Cross Reference: /src/sys/dev/pci

History log of /src/sys/dev/pci
Revision	Date	Author	Comments
1.16	15-Sep-2021	thorpej	Adjust the device_call() calling convention so as to provide type checking of the arguments passed to the call, using auto-generated argument structures and binding macros.
1.15	12-Dec-2018	maxv	Retire the LMC driver, and its associated lmcconfig tool. LMC has been mentioned repeatedly as a non-MP-safe driver that is hard to maintain, and no one is taking care of it. LMC was removed from OpenBSD three years ago, and from FreeBSD a few months ago.
1.14	28-Mar-2015	msaitoh	branches: 1.14.16; 1.14.18; Remove hdaudio subdir.
1.13	16-Sep-2009	sborrill	branches: 1.13.22; 1.13.40; Install hdaudio includes required by upcoming userland toolS
1.12	23-Jul-2006	bouyer	branches: 1.12.58; 1.12.70; 1.12.78; Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.11	06-Dec-2005	skrll	branches: 1.11.4; 1.11.8; Correct the includes for lmc.
1.10	25-Sep-2003	thorpej	branches: 1.10.14; 1.10.16; Install twereg.h and tweio.h
1.9	13-Sep-2001	thorpej	branches: 1.9.18; Install pcidevs.h, pcidevs_data.h, pciio.h, pcireg.h.
1.8	30-Jul-2001	ad	branches: 1.8.2; Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.7	11-Apr-2001	jdolecek	branches: 1.7.2; Only install headers which are actually used by our userland. This saves about 2.2MB under /usr/include/dev/. Discussed on tech-kern@ recently. I HOPE to get the list right. The headers I left in are ones used for MI tools and those whose usage I discovered by grep over tree sources. Feel free to put needed includes back in if you encounter anything which should not be removed from lists.
1.6	20-Jun-1999	pk	branches: 1.6.2; 1.6.12; 1.6.14; Remove dead file; alphabetize
1.5	20-Jun-1999	pk	Remove dead file
1.4	25-Mar-1999	explorer	branches: 1.4.4; install if_lmc*.h
1.3	16-Jun-1998	cgd	Point people to the correct Makefiles to build generated headers. (Suggested by Manuel Bouyer.)
1.2	12-Jun-1998	cgd	Rework the way kernel include files are installed. In the new method, as with user-land programs, include files are installed by each directory in the tree that has includes to install. (This allows more flexibility as to what gets installed, makes 'partial installs' easier, and gives us more options as to which machines' includes get installed at any given time.) The old SYS_INCLUDES={symlinks,copies} behaviours are _both_ still supported, though at least one bug in the 'symlinks' case is fixed by this change. Include files can't be build before installation, so directories that have includes as targets (e.g. dev/pci) have to move those targets into a different Makefile.
1.1	18-Jun-1995	cgd	learn about some PCI vendor and product numbers. not yet used
1.4.4.1	01-Jul-1999	thorpej	Sync w/ -current.
1.6.14.3	21-Sep-2001	nathanw	Catch up to -current.
1.6.14.2	24-Aug-2001	nathanw	Catch up with -current.
1.6.14.1	21-Jun-2001	nathanw	Catch up to -current.
1.6.12.2	26-Oct-2001	jhawk	Pull up via patch (requested by windsor): Move mlxio.h and mlxreg.h from dev/pci/Makefile to dev/ic/Makefile
1.6.12.1	25-Oct-2001	he	Apply patch (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.6.2.1	21-Apr-2001	bouyer	Sync with HEAD
1.7.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.2.1	03-Aug-2001	lukem	update to -current
1.8.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.9.18.4	11-Dec-2005	christos	Sync with head.
1.9.18.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.18.2	18-Sep-2004	skrll	Sync with HEAD.
1.9.18.1	03-Aug-2004	skrll	Sync with HEAD
1.10.16.2	30-Dec-2006	yamt	sync with head.
1.10.16.1	21-Jun-2006	yamt	sync with head.
1.10.14.1	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.11.8.1	11-Aug-2006	yamt	sync with head
1.11.4.1	09-Sep-2006	rpaulo	sync with head
1.12.78.1	21-Apr-2010	matt	sync to netbsd-5
1.12.70.1	18-Oct-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #1101): sys/dev/pci/Makefile: revision 1.13 sys/dev/pci/hdaudio/Makefile: revision 1.1 Install hdaudio includes required by userland tools
1.12.58.1	11-Mar-2010	yamt	sync with head
1.13.40.1	06-Apr-2015	skrll	Sync with HEAD
1.13.22.1	03-Dec-2017	jdolecek	update from HEAD
1.14.18.1	10-Jun-2019	christos	Sync with HEAD
1.14.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7	26-Oct-2016	pgoyette	Update the devlist2h.awk script to track the maximum lengths of vendor and product strings, and report the max values at end of the run. Update the Makefiles.{pci,usb,hdaudio}devs to point users at the places which might need to be updated if the maximum lengths get larger. Since this commit makes no changes to the generated files, we don't need to regenerate them now.
1.6	21-Sep-2014	christos	branches: 1.6.2; 1.6.4; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.5	19-Oct-2008	apb	branches: 1.5.38; Use ${TOOL_AWK} instead of ${AWK} or plain "awk" in make commands. Pass AWK=${TOOL_AWK:Q} to shell scripts that use awk.
1.4	11-Dec-2005	christos	branches: 1.4.70; 1.4.74; 1.4.80; merge ktrace-lwp.
1.3	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.2	16-Mar-1999	mjacob	branches: 1.2.42; Tron whacked my you-know-what because I didn't follow procedure. So I insered his description of the 'procedure' in the most likely place a developer will find it.
1.1	12-Jun-1998	cgd	Rework the way kernel include files are installed. In the new method, as with user-land programs, include files are installed by each directory in the tree that has includes to install. (This allows more flexibility as to what gets installed, makes 'partial installs' easier, and gives us more options as to which machines' includes get installed at any given time.) The old SYS_INCLUDES={symlinks,copies} behaviours are _both_ still supported, though at least one bug in the 'symlinks' case is fixed by this change. Include files can't be build before installation, so directories that have includes as targets (e.g. dev/pci) have to move those targets into a different Makefile.
1.2.42.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.42.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.42.1	03-Aug-2004	skrll	Sync with HEAD
1.4.80.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.4.74.1	04-May-2009	yamt	sync with head.
1.4.70.1	17-Jan-2009	mjf	Sync with HEAD.
1.5.38.1	03-Dec-2017	jdolecek	update from HEAD
1.6.4.1	04-Nov-2016	pgoyette	Sync with HEAD
1.6.2.1	05-Dec-2016	skrll	Sync with HEAD
1.3	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.2	27-Oct-1994	cgd	new RCS ID format.
1.1	09-Aug-1994	mycroft	Add PCI autoconfiguration support.
1.42	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.41	24-Apr-2021	thorpej	Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.40	09-Dec-2018	jdolecek	branches: 1.40.14; use pci_intr_establish_xname() everywhere
1.39	15-Oct-2018	uwe	Add Adaptec RAID 2445 and 2805. From Dima Veselov.
1.38	27-Sep-2016	pgoyette	branches: 1.38.8; 1.38.14; 1.38.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.37	07-Jul-2016	msaitoh	branches: 1.37.2; KNF. Remove extra spaces. No functional change.
1.36	29-Mar-2014	christos	branches: 1.36.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.35	27-Oct-2012	chs	branches: 1.35.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.34	23-Sep-2012	chs	match some more devices.
1.33	29-Sep-2011	is	branches: 1.33.2; 1.33.8; 1.33.12; Use symbolic constants for SUNs version of ADP2_ASR2200S.
1.32	18-Feb-2011	jmcneill	add Adaptec RAID 3405
1.31	07-May-2010	is	branches: 1.31.2; 1.31.4; SUN X4140's version of this board. From FreeBSD.
1.30	26-Nov-2009	njoly	branches: 1.30.2; 1.30.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.29	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.28	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.27	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.26	02-Jan-2009	briggs	branches: 1.26.2; Allocate enough space for the aac_pci_softc, not just the aac_softc.
1.25	08-May-2008	sborrill	branches: 1.25.6; 1.25.8; Add support for IBM ServeRAID 8k (and 8k-l)
1.24	28-Apr-2008	martin	branches: 1.24.2; Remove clause 3 and 4 from TNF licenses
1.23	10-Apr-2008	cegger	branches: 1.23.2; 1.23.4; use aprint_*_dev and device_xname
1.22	19-Oct-2007	ad	branches: 1.22.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.21	05-Jun-2007	briggs	branches: 1.21.6; 1.21.8; 1.21.12; Bring in changes from FreeBSD... * Allocate commands on-demand. * Update a bunch of constants and some structures. * Use __attribute__ ((__packed__)) instead of __packed to be consistent. * Support more commands for devices that can apparently handle them. * Support a "new comm. interface" present in more recent Adaptec firmware. This reduces the amount of PCI bus traffic in handling commands. * Support larger commands going to the adapter--if the adapter can support them. * Support 64-bit commands for archs where sizeof(bus_addr_t) > 4 and for adapters that advertise SGMAP64. * Handle the WINDOW4G option and NO4GB quirk by excluding 2G-4G window unless we have the WINDOW4G capability without the NO4GB quirk. * Ask the adapter more about its capabilities and try to use those if they seem sane. * Do our bus_dmamap_sync() inside dequeue_fib instead of following, since we have the information that we need there. * Provide access functions for some adapters that I haven't seen yet (MIPS-based "Rocket" adapters). Not yet used.
1.20	10-Mar-2007	christos	branches: 1.20.2; PR/35970: Todd Kover: Add support for HP ML110 G2 / Adaptec 2610SA
1.19	16-Nov-2006	christos	branches: 1.19.2; 1.19.4; 1.19.8; 1.19.10; __unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	28-Aug-2006	christos	branches: 1.17.2; 1.17.4; static comes first.
1.16	11-Dec-2005	christos	branches: 1.16.4; 1.16.8; merge ktrace-lwp.
1.15	28-Jun-2005	thorpej	branches: 1.15.2; Use ANSI function decls and static.
1.14	06-Apr-2005	martti	Added support for Dell CERC SATA RAID 1.5/6ch
1.13	19-Mar-2005	tron	Use PCI id symbol for the Adaptec ASR-2410SA instead of hardcoding the number.
1.12	01-Mar-2005	briggs	branches: 1.12.2; * Remove aif_get_mailboxstatus() from the aac_interface structure. Replace with aif_get_mailbox(). Make it return uint32_t instead of 'int'. * Add an AAC_GET_MAILBOX() macro and change AAC_GET_MAILBOXSTATUS() to use that. * Update the Dell PERC 2QC quirk code to use AAC_GET_MAILBOX instead of the StrongARM-specific code. While StrongARM access is correct for that card, it's a bad example of how to access the mailbox registers. * Add the GETINFO command and use it to get and display the card's supported options at a verbose level during attachment.
1.11	27-Feb-2005	perry	nuke trailing whitespace
1.10	10-May-2004	gendalia	branches: 1.10.4; 1.10.6; Add support for Adaptec AAR 2810SA raid controller.
1.9	06-May-2004	jdolecek	match Adaptec 2410SA XXX we should probably match all cards with product ID ASR2200S
1.8	26-Mar-2004	gendalia	branches: 1.8.2; Add support for Adaptec 2200S (Vulcan 2M) and Dell PERC 320/DC
1.7	31-Jan-2003	thorpej	branches: 1.7.2; Use aprint_*().
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	02-Aug-2002	ad	PR 17719 from Mark Davies: Additional Adaptec RAID controller types.
1.2	15-May-2002	augustss	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; Adapt to new bridge name.
1.1	26-Apr-2002	ad	Add a driver for Adaptec FSA RAID controllers, as often found in Dell servers. Based on the FreeBSD/OpenBSD versions.
1.2.8.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.8.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.2.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.8.1	15-May-2002	jdolecek	file aac_pci.c was added on branch kqueue on 2002-06-23 17:47:32 +0000
1.2.6.4	18-Oct-2002	nathanw	Catch up to -current.
1.2.6.3	13-Aug-2002	nathanw	Catch up to -current.
1.2.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.2.6.1	15-May-2002	nathanw	file aac_pci.c was added on branch nathanw_sa on 2002-06-20 03:45:19 +0000
1.2.4.5	01-Jun-2005	riz	Pull up revision 1.13 (requested by tron in ticket #5721): Use PCI id symbol for the Adaptec ASR-2410SA instead of hardcoding the number.
1.2.4.4	19-Mar-2005	tron	Pull up revision 1.10 (requested by he in ticket #1755): Add support for Adaptec AAR 2810SA raid controller.
1.2.4.3	19-Mar-2005	tron	Pull up revision 1.9 (requested by he in ticket #1755): match Adaptec 2410SA XXX we should probably match all cards with product ID ASR2200S
1.2.4.2	19-Mar-2005	tron	Pull up revision 1.8 (requested by he in ticket #1755): Add support for Adaptec 2200S (Vulcan 2M) and Dell PERC 320/DC
1.2.4.1	16-Nov-2002	he	Pull up revision 1.3 (requested by ad in ticket #646): Add additional Adaptec RAID controller types. Fixes PR#17719.
1.2.2.1	29-Aug-2002	gehenna	catch up with -current.
1.7.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.2.5	01-Apr-2005	skrll	Sync with HEAD.
1.7.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.2.2	08-Oct-2004	jmc	branches: 1.8.2.2.2; Pullup rev 1.10 (requested by he in ticket #896) Add support for Adaptec AAR 2810SA raid controller.
1.8.2.1	10-May-2004	tron	Pull up revision 1.9 (requested by jdolecek in ticket #274): match Adaptec 2410SA XXX we should probably match all cards with product ID ASR2200S
1.8.2.2.2.1	11-May-2005	riz	Pull up revision 1.14 (requested by martti in ticket #1402): Added support for Dell CERC SATA RAID 1.5/6ch
1.10.6.2	26-Mar-2005	yamt	sync with head.
1.10.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.10.4.1	29-Apr-2005	kent	sync with -current
1.12.2.2	07-Apr-2005	jmc	Pullup rev 1.13 (requested by tron in ticket #15) Add missing PCI id for Adaptec ASR-2410SA.
1.12.2.1	07-Apr-2005	tron	Pull up revision 1.14 (requested by martti in ticket #116): Added support for Dell CERC SATA RAID 1.5/6ch
1.15.2.3	27-Oct-2007	yamt	sync with head.
1.15.2.2	03-Sep-2007	yamt	sync with head.
1.15.2.1	30-Dec-2006	yamt	sync with head.
1.16.8.1	03-Sep-2006	yamt	sync with head.
1.16.4.1	09-Sep-2006	rpaulo	sync with head
1.17.4.2	10-Dec-2006	yamt	sync with head.
1.17.4.1	22-Oct-2006	yamt	sync with head
1.17.2.1	18-Nov-2006	ad	Sync with head.
1.19.10.3	03-Jun-2008	skrll	Sync with netbsd-4.
1.19.10.2	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.19.10.1	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.19.8.3	23-Oct-2007	ad	Sync with head.
1.19.8.2	09-Jun-2007	ad	Sync with head.
1.19.8.1	13-Mar-2007	ad	Sync with head.
1.19.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.19.2.4	16-Jan-2009	bouyer	Pull up following revision(s) (requested by briggs in ticket #1255): sys/dev/pci/aac_pci.c: revision 1.26 Allocate enough space for the aac_pci_softc, not just the aac_softc.
1.19.2.3	17-May-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1143): sys/dev/pci/aac_pci.c: revision 1.25 via patch share/man/man4/aac.4: revision 1.13 via patch sys/dev/pci/pcidevs: revision 1.939 via patch Add support for IBM ServeRAID 8k (and 8k-l)
1.19.2.2	25-Nov-2007	xtraeme	branches: 1.19.2.2.2; Pull up following revision(s) (requested by briggs in ticket #990): sys/dev/ic/aac_tables.h: revision 1.5 (patch) sys/dev/ic/aac.c: revision 1.31 (patch) sys/dev/ic/aac.c: revision 1.32 (patch) sys/dev/ic/aac.c: revision 1.33 (patch) sys/dev/ic/aac.c: revision 1.34 (patch) sys/arch/amd64/conf/INSTALL: revision 1.70 (patch) sys/dev/ic/aac.c: revision 1.37 (patch) sys/dev/ic/aacreg.h: revision 1.6 (patch) sys/dev/ic/aacreg.h: revision 1.7 (patch) sys/dev/ic/aacreg.h: revision 1.8 (patch) sys/dev/ic/aacreg.h: revision 1.9 (patch) sys/dev/pci/aac_pci.c: revision 1.21 (patch) sys/dev/ic/aacvar.h: revision 1.10 (patch) sys/dev/ic/aacvar.h: revision 1.9 (patch) sys/arch/amd64/conf/GENERIC: revision 1.144 (patch) sys/dev/ic/ld_aac.c: revision 1.14 (patch) sys/dev/ic/ld_aac.c: revision 1.15 (patch) Compile and run on amd64 and on >2GB RAM.
1.19.2.1	07-Oct-2007	xtraeme	Pull up following revision(s) (requested by christos in ticket #923): sys/dev/pci/aac_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.872 PR/35970: Todd Kover: Add support for HP ML110 G2 / Adaptec 2610SA
1.19.2.2.2.1	16-Jan-2009	bouyer	Pull up following revision(s) (requested by briggs in ticket #1255): sys/dev/pci/aac_pci.c: revision 1.26 Allocate enough space for the aac_pci_softc, not just the aac_softc.
1.20.2.1	11-Jul-2007	mjf	Sync with head.
1.21.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.21.8.1	06-Nov-2007	matt	sync with HEAD
1.21.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.22.16.2	17-Jan-2009	mjf	Sync with HEAD.
1.22.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.23.4.5	11-Aug-2010	yamt	sync with head.
1.23.4.4	11-Mar-2010	yamt	sync with head
1.23.4.3	16-May-2009	yamt	sync with head
1.23.4.2	04-May-2009	yamt	sync with head.
1.23.4.1	16-May-2008	yamt	sync with head.
1.23.2.1	18-May-2008	yamt	sync with head.
1.24.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.25.8.2	20-Nov-2010	riz	Pull up following revision(s) (requested by is in ticket #1394): sys/dev/pci/aac_pci.c: revision 1.31 SUN X4140's version of this board. From FreeBSD.
1.25.8.1	08-Jan-2009	snj	Pull up following revision(s) (requested by briggs in ticket #224): sys/dev/pci/aac_pci.c: revision 1.26 Allocate enough space for the aac_pci_softc, not just the aac_softc.
1.25.6.1	19-Jan-2009	skrll	Sync with HEAD.
1.26.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.30.4.2	05-Mar-2011	rmind	sync with head
1.30.4.1	30-May-2010	rmind	sync with head
1.30.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.31.4.1	05-Mar-2011	bouyer	Sync with HEAD
1.31.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.33.12.3	03-Dec-2017	jdolecek	update from HEAD
1.33.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.33.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.33.8.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #691): sys/dev/pci/aac_pci.c: revision 1.34 sys/dev/pci/if_sip.c: revision 1.155 sys/dev/pci/if_tlp_pci.c: revision 1.122 sys/dev/pci/if_an_pci.c: revision 1.34 match some more devices.
1.33.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.33.2.1	30-Oct-2012	yamt	sync with head
1.35.2.1	18-May-2014	rmind	sync with head
1.36.6.2	05-Oct-2016	skrll	Sync with HEAD
1.36.6.1	09-Jul-2016	skrll	Sync with HEAD
1.37.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.38.16.1	10-Jun-2019	christos	Sync with HEAD
1.38.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.38.14.1	20-Oct-2018	pgoyette	Sync with head
1.38.8.1	17-Oct-2018	martin	Pull up following revision(s) (requested by uwe in ticket #1064): sys/dev/pci/aac_pci.c: revision 1.39 sys/dev/pci/pcidevs: revision 1.1348 Add Adaptec RAID 2445 and 2805. From Dima Veselov.
1.40.14.1	28-Mar-2021	thorpej	Correct the names of the arguments passed to the rescan function.
1.32	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.31	07-Oct-2013	jakllsch	branches: 1.31.18; Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.30	31-Jul-2012	bouyer	branches: 1.30.2; 1.30.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.29	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.28	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.27	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.26	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.25	04-Apr-2011	dyoung	branches: 1.25.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.24	05-Nov-2010	jakllsch	branches: 1.24.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.23	14-May-2008	tsutsui	branches: 1.23.18; 1.23.20; Normalize my licenses.
1.22	18-Mar-2008	cube	branches: 1.22.2; 1.22.4; 1.22.6; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.21	09-Feb-2007	ad	branches: 1.21.20; 1.21.36; 1.21.40; Merge newlock2 to head.
1.20	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.19	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.18	11-Dec-2005	christos	branches: 1.18.20; 1.18.22; merge ktrace-lwp.
1.17	24-May-2005	tsutsui	branches: 1.17.2; ATP865 chips support Ultra133, so correct attach messages as well.
1.16	24-May-2005	lukem	add __KERNEL_RCSID
1.15	08-May-2005	tsutsui	ACARD ATP865-A (used on AEC-6280/6880) supports UltraDMA/133 too.
1.14	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.13	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.12	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.11	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.10	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.9	02-Aug-2004	bouyer	branches: 1.9.2; If acknowledge "bogus" interrupts (that is, interrupts that have happened according to the controller-dependant registers while wdc was not waiting for an interrupt). A reset can generate spurious interrupts (or even a flow of spurious interrupts on the promise) that needs to be properly acknowledged.
1.8	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.7	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.6	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.5	24-Oct-2003	tsutsui	Add support for Acard ATP865/A Ultra100 IDE Controllers. From Tsubai Masanari.
1.4	24-Oct-2003	tsutsui	Remove trailing whitespace.
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.9.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.9.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.9.2.2	03-Aug-2004	skrll	Sync with HEAD
1.9.2.1	02-Aug-2004	skrll	file acardide.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.17.2.3	24-Mar-2008	yamt	sync with head.
1.17.2.2	26-Feb-2007	yamt	sync with head.
1.17.2.1	30-Dec-2006	yamt	sync with head.
1.18.22.2	10-Dec-2006	yamt	sync with head.
1.18.22.1	22-Oct-2006	yamt	sync with head
1.18.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.18.20.1	18-Nov-2006	ad	Sync with head.
1.21.40.2	02-Jun-2008	mjf	Sync with HEAD.
1.21.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.21.36.1	24-Mar-2008	keiichi	sync with head.
1.21.20.1	23-Mar-2008	matt	sync with HEAD
1.22.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.22.4.1	16-May-2008	yamt	sync with head.
1.22.2.1	18-May-2008	yamt	sync with head.
1.23.20.2	21-Apr-2011	rmind	sync with head
1.23.20.1	05-Mar-2011	rmind	sync with head
1.23.18.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.24.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.25.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.25.4.1	30-Oct-2012	yamt	sync with head
1.30.4.1	18-May-2014	rmind	sync with head
1.30.2.3	03-Dec-2017	jdolecek	update from HEAD
1.30.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.30.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.31.18.1	05-Aug-2017	jdolecek	kill dead code
1.39	03-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "ali1543-ide-force-compat-mode".
1.38	20-Dec-2020	jdolecek	disable (U)DMA for ATAPI on aceride(4), the chip doesn't support it PR port-sparc64/55540
1.37	21-Jul-2017	nakayama	branches: 1.37.18; Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.36	07-Oct-2013	jakllsch	branches: 1.36.4; 1.36.6; 1.36.22; Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.35	31-Jul-2012	bouyer	branches: 1.35.2; 1.35.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.34	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.33	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.32	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.31	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.30	04-Apr-2011	dyoung	branches: 1.30.4; 1.30.10; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.29	05-Nov-2010	jakllsch	branches: 1.29.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.28	03-Nov-2010	nakayama	Add a workaround for 48-bit LBA DMA access bug on M5229 rev. <= 0xc4. - use DMA in 28-bit LBA addressing. - use PIO in 48-bit LBA addressing. Idea and message are from FreeBSD, and tested on M5229 rev. 0xc3 in my Sun Netra X1 with 160GB drive.
1.27	08-May-2010	nakayama	PCI-ISA Bridge paired with M5229 UDMA IDE Controller rev. 0xC3 is M1533. Follow the fix in rev 1.1004 of pcidevs.
1.26	19-Oct-2009	bouyer	branches: 1.26.2; 1.26.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.25	18-Mar-2008	cube	branches: 1.25.4; 1.25.14; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.24	01-Jan-2008	chris	branches: 1.24.2; 1.24.6; Add property to aceride to allow the forcing of compat mode. This allows cats to override the BIOS (aka ABLE) settings, which enable native mode, but doesn't route the interrupts correctly. Discussed last year on tech-kern and port-cats.
1.23	09-Feb-2007	ad	branches: 1.23.20; 1.23.26; 1.23.32; Merge newlock2 to head.
1.22	27-Nov-2006	christos	PR/35136: Gary Duzan: Enable Ultra/133 on newer aceride
1.21	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	11-Dec-2005	christos	branches: 1.19.20; 1.19.22; merge ktrace-lwp.
1.18	07-Aug-2005	bouyer	To disable a channel we need to clear bit 2 or 3, not all other bits but bit 2 or 3. Thanks to Takeshi Nakayama for catching it. (With this mistake the code was still working for the first channel, because the reset of the second channel would disable/enable the first one).
1.17	06-Aug-2005	bouyer	remove debug printf
1.16	06-Aug-2005	bouyer	Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.15	24-May-2005	lukem	branches: 1.15.2; add __KERNEL_RCSID
1.14	27-Feb-2005	perry	branches: 1.14.2; nuke trailing whitespace
1.13	21-Aug-2004	thorpej	branches: 1.13.4; 1.13.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.12	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.11	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.10	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.9	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.8	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.7	02-Aug-2004	bouyer	branches: 1.7.2; If acknowledge "bogus" interrupts (that is, interrupts that have happened according to the controller-dependant registers while wdc was not waiting for an interrupt). A reset can generate spurious interrupts (or even a flow of spurious interrupts on the promise) that needs to be properly acknowledged.
1.6	03-Jan-2004	thorpej	branches: 1.6.4; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.5	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.4	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.6.4.2	07-Aug-2005	riz	Pull up revision 1.18 (requested by bouyer in ticket #5555): To disable a channel we need to clear bit 2 or 3, not all other bits but bit 2 or 3. Thanks to Takeshi Nakayama for catching it. (With this mistake the code was still working for the first channel, because the reset of the second channel would disable/enable the first one).
1.6.4.1	07-Aug-2005	riz	Pull up revisions 1.16-1.17 via patch (requested by bouyer in ticket #5555): Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.7.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.7.2.2	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	02-Aug-2004	skrll	file aceride.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.13.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.4.1	29-Apr-2005	kent	sync with -current
1.14.2.3	18-Aug-2005	tron	Pull up revision 1.18 (requested by bouyer in ticket #653): To disable a channel we need to clear bit 2 or 3, not all other bits but bit 2 or 3. Thanks to Takeshi Nakayama for catching it. (With this mistake the code was still working for the first channel, because the reset of the second channel would disable/enable the first one).
1.14.2.2	18-Aug-2005	tron	Pull up revision 1.17 (requested by bouyer in ticket #653): remove debug printf
1.14.2.1	18-Aug-2005	tron	Pull up revision 1.16 (requested by bouyer in ticket #653): Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.15.2.5	24-Mar-2008	yamt	sync with head.
1.15.2.4	21-Jan-2008	yamt	sync with head
1.15.2.3	26-Feb-2007	yamt	sync with head.
1.15.2.2	30-Dec-2006	yamt	sync with head.
1.15.2.1	21-Jun-2006	yamt	sync with head.
1.19.22.2	10-Dec-2006	yamt	sync with head.
1.19.22.1	22-Oct-2006	yamt	sync with head
1.19.20.3	06-Feb-2007	ad	Quiten noisy boot messages.
1.19.20.2	12-Jan-2007	ad	Sync with head.
1.19.20.1	18-Nov-2006	ad	Sync with head.
1.23.32.1	02-Jan-2008	bouyer	Sync with HEAD
1.23.26.1	18-Feb-2008	mjf	Sync with HEAD.
1.23.20.2	23-Mar-2008	matt	sync with HEAD
1.23.20.1	09-Jan-2008	matt	sync with HEAD
1.24.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.24.2.1	24-Mar-2008	keiichi	sync with head.
1.25.14.1	21-Nov-2010	riz	Pull up following revision(s) (requested by nakayama in ticket #1471): sys/dev/pci/aceride.c: revision 1.28 Add a workaround for 48-bit LBA DMA access bug on M5229 rev. <= 0xc4. - use DMA in 28-bit LBA addressing. - use PIO in 48-bit LBA addressing. Idea and message are from FreeBSD, and tested on M5229 rev. 0xc3 in my Sun Netra X1 with 160GB drive.
1.25.4.2	11-Aug-2010	yamt	sync with head.
1.25.4.1	11-Mar-2010	yamt	sync with head
1.26.4.3	21-Apr-2011	rmind	sync with head
1.26.4.2	05-Mar-2011	rmind	sync with head
1.26.4.1	30-May-2010	rmind	sync with head
1.26.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.26.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.29.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.30.10.1	23-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #1463): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.30.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.30.4.1	30-Oct-2012	yamt	sync with head
1.35.4.1	18-May-2014	rmind	sync with head
1.35.2.3	03-Dec-2017	jdolecek	update from HEAD
1.35.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.35.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.36.22.1	25-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #153): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.36.6.1	28-Aug-2017	skrll	Sync with HEAD
1.36.4.1	26-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #1458): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.37.18.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.31	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.30	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.29	18-Oct-2014	snj	branches: 1.29.18; 1.29.20; src is too big these days to tolerate superfluous apostrophes. It's "its", people!
1.28	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.27	27-Oct-2012	chs	branches: 1.27.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.26	26-Nov-2009	njoly	branches: 1.26.12; 1.26.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.25	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.24	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.23	28-Apr-2008	martin	branches: 1.23.14; Remove clause 3 and 4 from TNF licenses
1.22	10-Apr-2008	cegger	branches: 1.22.2; 1.22.4; use aprint_*_dev and device_xname
1.21	19-Oct-2007	ad	branches: 1.21.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.20	16-Nov-2006	christos	branches: 1.20.8; 1.20.22; 1.20.24; 1.20.28; __unused removal on arguments; approved by core.
1.19	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.18	11-Dec-2005	christos	branches: 1.18.20; 1.18.22; merge ktrace-lwp.
1.17	28-Jun-2005	thorpej	branches: 1.17.2; Use ANSI function decls and static.
1.16	27-Feb-2005	perry	nuke trailing whitespace
1.15	04-Feb-2005	perry	de-__P
1.14	31-Jan-2003	thorpej	branches: 1.14.2; 1.14.10; 1.14.12; Use aprint_*().
1.13	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.9	13-Nov-2001	lukem	add RCSID
1.8	28-Dec-2000	sommerfeld	branches: 1.8.2; 1.8.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.7	12-Jun-1999	dante	branches: 1.7.2; Move chip version check from advlib.c to adv_pci.c
1.6	25-Feb-1999	dante	branches: 1.6.4; Fix AdvanSys Narrow driver to work on Alpha
1.5	26-Sep-1998	dante	Remove Wide boards references Update supported boards list
1.4	31-Aug-1998	dante	Remove __BROKEN_INDIRECT_CONFIG Correct currently known boards
1.3	29-Aug-1998	dante	Adjust copyright justification
1.2	29-Aug-1998	dante	Add rcs id
1.1	26-Aug-1998	dante	Add AdvanSys 930U and 940U SCSI controllers
1.6.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.7.2.1	05-Jan-2001	bouyer	Sync with HEAD
1.8.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.8.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.8.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.14.12.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.12.1	12-Feb-2005	yamt	sync with head.
1.14.10.1	29-Apr-2005	kent	sync with -current
1.14.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.17.2.2	27-Oct-2007	yamt	sync with head.
1.17.2.1	30-Dec-2006	yamt	sync with head.
1.18.22.2	10-Dec-2006	yamt	sync with head.
1.18.22.1	22-Oct-2006	yamt	sync with head
1.18.20.1	18-Nov-2006	ad	Sync with head.
1.20.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.20.24.1	06-Nov-2007	matt	sync with HEAD
1.20.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.20.8.1	23-Oct-2007	ad	Sync with head.
1.21.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.22.4.3	11-Mar-2010	yamt	sync with head
1.22.4.2	16-May-2009	yamt	sync with head
1.22.4.1	16-May-2008	yamt	sync with head.
1.22.2.1	18-May-2008	yamt	sync with head.
1.23.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.26.22.3	03-Dec-2017	jdolecek	update from HEAD
1.26.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.26.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.26.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.26.12.1	30-Oct-2012	yamt	sync with head
1.27.2.1	18-May-2014	rmind	sync with head
1.29.20.1	10-Jun-2019	christos	Sync with HEAD
1.29.18.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.30	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.29	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.28	18-Oct-2014	snj	branches: 1.28.18; 1.28.20; src is too big these days to tolerate superfluous apostrophes. It's "its", people!
1.27	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.26	27-Oct-2012	chs	branches: 1.26.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.25	26-Nov-2009	njoly	branches: 1.25.12; 1.25.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.24	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.23	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.22	28-Apr-2008	martin	branches: 1.22.14; Remove clause 3 and 4 from TNF licenses
1.21	10-Apr-2008	cegger	branches: 1.21.2; 1.21.4; use aprint_*_dev and device_xname
1.20	19-Oct-2007	ad	branches: 1.20.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.19	16-Nov-2006	christos	branches: 1.19.8; 1.19.22; 1.19.24; 1.19.28; __unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	11-Dec-2005	christos	branches: 1.17.20; 1.17.22; merge ktrace-lwp.
1.16	28-Jun-2005	thorpej	branches: 1.16.2; Use ANSI function decls and static.
1.15	04-Feb-2005	perry	de-__P
1.14	31-Jan-2003	thorpej	branches: 1.14.2; 1.14.10; 1.14.12; Use aprint_*().
1.13	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.9	13-Nov-2001	lukem	add RCSID
1.8	28-Dec-2000	sommerfeld	branches: 1.8.2; 1.8.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.7	26-May-2000	dante	Fix a bug introduced in last commit which caused a painc due to re-enqueueing an already free ccb. Prepare the background to have a unique initialization function for all of the supported host adapters. Rename a lot of #define in a more sane way. Move Carriers initialization and defines in adwmcode.{c,h} Don't lose Carrier nodes in case of a BUS/Chip reset explicitly invoked after a DMA failure. XXX - DMA failure still arise when AdvanSys U[2]W host adapters are used in conjunction with Intel 82443BX Host Bridge/Controller (rev. 0x03). !?!Have to understand why!?!
1.6	08-May-2000	dante	Minor changes to PCI configuration
1.5	30-Apr-2000	dante	Add new microcode for all of the supported boards. Fix minor glitches. Add preliminary support for ASB-3940U3W host adapters; it might work w/ autotermination on. ASB-3940U3W support is not yet tested due to lack of board.
1.4	04-Feb-2000	dante	Add a list of all the currently supported boards
1.3	03-Feb-2000	dante	Add AdvanSys U2W (LVD) boards support
1.2	26-Sep-1998	dante	branches: 1.2.12; Change sources to be KNF
1.1	26-Sep-1998	dante	Add support for AdvanSys Ultra Wide boards ABP-9xxUW
1.2.12.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.8.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.8.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.14.12.1	12-Feb-2005	yamt	sync with head.
1.14.10.1	29-Apr-2005	kent	sync with -current
1.14.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.16.2.2	27-Oct-2007	yamt	sync with head.
1.16.2.1	30-Dec-2006	yamt	sync with head.
1.17.22.2	10-Dec-2006	yamt	sync with head.
1.17.22.1	22-Oct-2006	yamt	sync with head
1.17.20.1	18-Nov-2006	ad	Sync with head.
1.19.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.19.24.1	06-Nov-2007	matt	sync with HEAD
1.19.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.19.8.1	23-Oct-2007	ad	Sync with head.
1.20.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.21.4.3	11-Mar-2010	yamt	sync with head
1.21.4.2	16-May-2009	yamt	sync with head
1.21.4.1	16-May-2008	yamt	sync with head.
1.21.2.1	18-May-2008	yamt	sync with head.
1.22.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.25.22.3	03-Dec-2017	jdolecek	update from HEAD
1.25.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.25.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.25.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.25.12.1	30-Oct-2012	yamt	sync with head
1.26.2.1	18-May-2014	rmind	sync with head
1.28.20.1	10-Jun-2019	christos	Sync with HEAD
1.28.18.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.88	22-May-2022	andvar	fix various small typos, mainly in comments.
1.87	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.86	15-Oct-2019	msaitoh	Use unsigned to avoid undefined behavior in agpattach(). Found by kUBSan.
1.85	27-Aug-2018	riastradh	branches: 1.85.6; Fill struct agp_info::ai_devid with the PCI id.
1.84	27-Feb-2017	msaitoh	branches: 1.84.12; 1.84.14; Decode AGP capability.
1.83	25-Jul-2014	dholland	branches: 1.83.4; 1.83.8; 1.83.12; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.82	12-Jun-2014	riastradh	Extend agp_generic_bind_memory with bounds to check. New routine agp_generic_bind_memory_bounded. Use it in agp_i810 to replace the pile of code I wrote as a replacement which is apparently broken.
1.81	16-Mar-2014	dholland	branches: 1.81.2; Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.80	06-Apr-2012	plunky	branches: 1.80.2; 1.80.4; device_pmf_is_registered() is not required
1.79	04-Apr-2011	dyoung	branches: 1.79.4; 1.79.8; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.78	24-Feb-2011	matt	Add Intel Pineview support
1.77	15-Feb-2011	jmcneill	add AGP v3 support to agp_generic_enable from FreeBSD
1.76	13-Nov-2010	uebayasi	branches: 1.76.2; 1.76.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.75	10-Oct-2010	christos	restore binary compatibility for amd64; requested by joerg.
1.74	02-Oct-2010	jym	Revert christos@ fix for the i386 paddr_t change and agp_allocate. Make 'physical' element of the struct agp_allocate an uint32_t. This is similar to what other kernels do (at least Solaris and Linux). This patch should make xserver work again within an i386 userland in a 64 bits paddr_t world. Thanks to Iain Hibbert for reporting the break, and providing initial hints. See http://mail-index.netbsd.org/current-users/2010/09/27/msg014418.html
1.73	28-Sep-2010	christos	make this compile.
1.72	28-Sep-2010	christos	- fixes from Jean-Yves Migeon - handle transitional paddr_t period for i386
1.71	27-Sep-2010	christos	move all the compat code here, and #ifdef __i386__ from mrg.
1.70	27-Sep-2010	christos	backwards compat code for paddr_t being 32 bits.
1.69	16-Jun-2010	riz	Add AGP support for a number of Intel onboard devices, including 82G41, 82B43, E7221, 82965GME, and "Iron Lake". Device types (i915, i965, G33, and G4X variants) from the Linux Intel AGP driver, and (for 82G41) from Henry Bent in PR#42906. There are a few more varieties that should be relatively low-hanging fruit ("Pineview" and "Sandy Bridge"), but will require a little bit of rejiggering of the "chiptype". OK mrg@
1.68	24-Feb-2010	dyoung	branches: 1.68.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.67	24-Feb-2010	jym	- Use ctob() instead of ptoa() to obtain physical addresses from frame numbers. Using ptoa() will cast to vaddr_t, which might no be adequate for architectures where sizeof(paddr_t) > sizeof(vaddr_t) (like i386 PAE). - small fix inside AGP heuristics to avoid masking high order bits for systems with more than 4GB. Reviewed by bouyer@. See also http://mail-index.netbsd.org/tech-kern/2010/02/22/msg007373.html
1.66	08-Jan-2010	dyoung	branches: 1.66.2; Expand PMF_FN_* macros.
1.65	27-Jan-2009	markd	Add some more Intel G4X class chipsets. Also fix gtt_size for G4X chipsets. Partially from FreeBSD. There is still an issue here as G4X chipsets report an aperture size of 0.
1.64	29-Nov-2008	christos	Add support for the Intel G45 AGP. From Arnaud Lacombe
1.63	08-Nov-2008	christos	Support for Intel G35 as found on Asus P5E-VM HDMI motherboard from Milos Negovanovic
1.62	17-Oct-2008	christos	branches: 1.62.2; 1.62.4; 1.62.8; add missing free.
1.61	22-Aug-2008	tnn	AGP support for Intel 945GME chipset, found on Acer Aspire One.
1.60	19-Aug-2008	matthias	Add agp support for Intel 946gz.
1.59	09-Jun-2008	freza	branches: 1.59.4; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.58	08-Jun-2008	tsutsui	Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.57	19-Apr-2008	njoly	branches: 1.57.2; 1.57.4; Add missing agp support for ALI M1689 chipset, for PR/35416.
1.56	10-Apr-2008	cegger	branches: 1.56.2; use aprint_*_dev and device_xname
1.55	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.54	09-Dec-2007	jmcneill	branches: 1.54.6; 1.54.10; Merge jmcneill-pm branch.
1.53	05-Dec-2007	ad	branches: 1.53.2; Match the docs: MUTEX_DRIVER/SPIN are now only for porting code written for Solaris.
1.52	24-Nov-2007	markd	branches: 1.52.2; Add agp support for Intel Q35/G33/Q33.
1.51	12-Nov-2007	joerg	Merge bunch of AGP changes from jmcneill-pm.
1.50	30-Oct-2007	jnemeth	branches: 1.50.2; PR/37201 - Yasushi Oshima -- Intel 82965G chipset support
1.49	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.48	15-Aug-2007	markd	branches: 1.48.2; 1.48.6; Add agp support for i965Q. Based on a patch for FreeBSD by Eric Anholt. OKed on tech-kern some months ago.
1.47	04-Aug-2007	kiyohara	branches: 1.47.2; Add support agp_amd64. Imported from FreeBSD.
1.46	06-Mar-2007	xtraeme	branches: 1.46.2; 1.46.10; 1.46.14; Use a mutex rather than lockmgr. Approved by Andrew Doran.
1.45	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.44	18-Dec-2006	christos	branches: 1.44.2; Don't cleanup memory mappings twice. From Blair Sadewitz
1.43	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.42	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.41	27-Aug-2006	christos	branches: 1.41.2; 1.41.4; complete initialization
1.40	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.39	30-Jul-2006	simonb	Add support for 82945GM and (untested) 82945P AGP. FreeBSD and Linux both treat the 945 series exactly the same as the 915 series for AGP, so just match a few extra device IDs.
1.38	27-Jun-2006	tron	Make this build with GCC 4.x.
1.37	16-Jan-2006	christos	branches: 1.37.2; 1.37.6; 1.37.14; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.36	11-Dec-2005	christos	branches: 1.36.2; merge ktrace-lwp.
1.35	28-Jun-2005	thorpej	branches: 1.35.2; Use ANSI function decls and static.
1.34	27-Feb-2005	perry	branches: 1.34.2; nuke trailing whitespace
1.33	30-Aug-2004	drochner	branches: 1.33.4; 1.33.6; Phase out the use of a string as first "attach args" member to control which bustype should be attached with a specific call to config_found() (from a "mainbus" or a bus bridge). Do it for isa/eisa/mca and pci/agp for now. These buses all attach to an mi interface attribute "isabus", "eisabus" etc., and the autoconf framework now allows to specify an interface attribute on config_found() and config_search(), which limits the search of matching config data to these which attach to that specific attribute. So we basically have to call config_found_ia(..., "foobus", ...) where such a bus is attached. As a consequence, where a "mainbus" or alike also attaches other devices (eg CPUs) which do not attach to a specific attribute yet, we need at least pass an attribute name (different from "foobus") so that the foo bus is not found at these places. This made some minor changes necessary which are not obviously related to the mentioned buses.
1.32	13-Feb-2004	wiz	Uppercase CPU, plural is CPUs.
1.31	26-Aug-2003	tron	Add support for i865G integrated graphics support to agp(4). Patches contributed by Michael D. Allen in PR port-i386/22600.
1.30	24-Aug-2003	tron	Add a missing blank line between two functions.
1.29	20-Jul-2003	hannken	Add support for i855 agp. From FreeBSD. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.28	29-Jun-2003	fvdl	branches: 1.28.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.27	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.26	24-Feb-2003	tron	Don't match a PCI graphics card in a system which supports an AGP. Patch supplied by Quentin Garnier in PR kern/19256.
1.25	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.24	31-Jan-2003	thorpej	Use aprint_*().
1.23	13-Dec-2002	scw	Add support for i830/i845 agp. From Shingo WATANABE (on tech-x11) and FreeBSD, with cleanup/KNF by me. Note: These chipsets are not well supported by the i810 driver in NetBSD's in-tree xsrc (based on XFree86 4.2.1 at this time). However, the driver works perfectly using bleeding-edge XFree86-current on my Omnibook's i830MG with these agp changes.
1.22	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.21	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.20	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.19	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.18	19-Sep-2002	nathanw	Slight wording cleanup in a comment.
1.17	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.16	11-Aug-2002	drochner	-avoid lock leak in error case -Free AGP memory on close, to avoid a memory leak in case the X server doesn't free it explicitely. (It appears that the X server never calls AGPIOC_DEALLOCATE.) Fixes PR kern/17869 by Emmanuel Dreyfus.
1.15	27-Jun-2002	drochner	Only allow the bus_dmamem_alloc() code to sleep if no physically contigous pages are insisted in. Atm, the pglistalloc code does the same automatically, but this might change.
1.14	22-Jan-2002	augustss	branches: 1.14.8; 1.14.10; Don't pretend we can handle i830; it needs a somewhat different AGP driver. (I added i830 to quickly.)
1.13	14-Jan-2002	augustss	Add i830M to agp chip tables.
1.12	13-Nov-2001	lukem	add RCSID
1.11	01-Oct-2001	fvdl	Don't map the aperture into KVA, it isn't needed.
1.10	16-Sep-2001	thorpej	branches: 1.10.2; 1.10.4; agp_generic_bind_memory(): Don't return NULL if we can't malloc memory, return ENOMEM.
1.9	15-Sep-2001	thorpej	agpopen(): check for NULL softc before dereferencing it.
1.8	15-Sep-2001	drochner	normalize error message if agp_map_aperture() fails
1.7	15-Sep-2001	thorpej	Give each AGP controller its own attribute, and let the "agpbus" device (rather, the device that carries that attribute) also carry one or more attributes indicating which type of controller it might be. This will allow systems that might have AGP, but would never have e.g. an Intel PCI-Host bridge, to trim out code that won't be used.
1.6	15-Sep-2001	thorpej	Use bus_space_mmap() to mmap the aperture.
1.5	15-Sep-2001	thorpej	Clean up the AGP match/attach code somewhat.
1.4	14-Sep-2001	drochner	-plug some resource leaks -don't clear memory ranges on "release" The X server survives a screen switch now.
1.3	13-Sep-2001	drochner	-allocate DMA maps as large as necessary in the worst case -allocate space for DMA segments as much as necessary -fix format warnings in a debug output -don't try to access AGP capability stuff in the PCI config header if it doesn't exist (as on the i810)
1.2	11-Sep-2001	fvdl	branches: 1.2.2; Fix i810/other intel chipsets attach mishap.
1.1	10-Sep-2001	fvdl	AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.2.2.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.2.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.2.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.2.2.1	11-Sep-2001	thorpej	file agp.c was added on branch kqueue on 2001-09-13 01:15:49 +0000
1.10.4.4	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.10.4.3	10-Oct-2001	fvdl	Convert all remaining devices.
1.10.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.10.4.1	16-Sep-2001	fvdl	file agp.c was added on branch thorpej-devvp on 2001-10-01 12:45:50 +0000
1.10.2.11	19-Dec-2002	thorpej	Sync with HEAD.
1.10.2.10	11-Nov-2002	nathanw	Catch up to -current
1.10.2.9	18-Oct-2002	nathanw	Catch up to -current.
1.10.2.8	17-Sep-2002	nathanw	Catch up to -current.
1.10.2.7	13-Aug-2002	nathanw	Catch up to -current.
1.10.2.6	01-Aug-2002	nathanw	Catch up to -current.
1.10.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.10.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.10.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.10.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.10.2.1	16-Sep-2001	nathanw	file agp.c was added on branch nathanw_sa on 2001-09-21 22:35:52 +0000
1.14.10.3	16-Jun-2003	grant	Pull up revision 1.26 (requested by tron in ticket #1181): Don't match a PCI graphics card in a system which supports an AGP. Patch supplied by Quentin Garnier in PR kern/19256.
1.14.10.2	28-Jan-2003	jmc	Pullup revisions 1.14-1.15 (requested by manu in ticket #1073. Missed in initial pullup) Only allow the bus_dmamem_alloc() code to sleep if no physically contigous pages are insisted in.
1.14.10.1	26-Jan-2003	jmc	Pullup revisions 1.15-1.16 (requested by manu in ticket #1073) -avoid lock leak in error case -Free AGP memory on close, to avoid a memory leak in case the X server doesn't free it explicitely. (It appears that the X server never calls AGPIOC_DEALLOCATE.)
1.14.8.3	29-Aug-2002	gehenna	catch up with -current.
1.14.8.2	15-Jul-2002	gehenna	catch up with -current.
1.14.8.1	16-May-2002	gehenna	Add the character device switch.
1.28.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.28.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.28.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.28.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.28.2.3	03-Sep-2004	skrll	Sync with HEAD
1.28.2.2	03-Aug-2004	skrll	Sync with HEAD
1.28.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.33.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.33.4.1	29-Apr-2005	kent	sync with -current
1.34.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.35.2.8	17-Mar-2008	yamt	sync with head.
1.35.2.7	21-Jan-2008	yamt	sync with head
1.35.2.6	07-Dec-2007	yamt	sync with head
1.35.2.5	15-Nov-2007	yamt	sync with head.
1.35.2.4	27-Oct-2007	yamt	sync with head.
1.35.2.3	03-Sep-2007	yamt	sync with head.
1.35.2.2	30-Dec-2006	yamt	sync with head.
1.35.2.1	21-Jun-2006	yamt	sync with head.
1.36.2.1	01-Feb-2006	yamt	sync with head.
1.37.14.1	13-Jul-2006	gdamore	Merge from HEAD.
1.37.6.2	03-Sep-2006	yamt	sync with head.
1.37.6.1	11-Aug-2006	yamt	sync with head
1.37.2.1	09-Sep-2006	rpaulo	sync with head
1.41.4.3	21-Dec-2006	yamt	sync with head.
1.41.4.2	10-Dec-2006	yamt	sync with head.
1.41.4.1	22-Oct-2006	yamt	sync with head
1.41.2.2	12-Jan-2007	ad	Sync with head.
1.41.2.1	18-Nov-2006	ad	Sync with head.
1.44.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.46.14.12	09-Dec-2007	jmcneill	Sync with HEAD.
1.46.14.11	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.46.14.10	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.46.14.9	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.46.14.8	31-Oct-2007	joerg	Sync with HEAD.
1.46.14.7	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.46.14.6	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.46.14.5	06-Sep-2007	jmcneill	agp(4) products list ordering fix, from Nicholas Joly on current-users.
1.46.14.4	27-Aug-2007	jmcneill	Add 965PM support.
1.46.14.3	16-Aug-2007	jmcneill	Sync with HEAD.
1.46.14.2	06-Aug-2007	jmcneill	Move agp_i810_power to agp, since it does nothing specific to the i810 hardware. Allow agp front-ends to override the default power handler if specific configuration needs to take place, and expose agp_power via agpvar.h so the front-ends can call back to it to do generic initialization.
1.46.14.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.46.10.1	15-Aug-2007	skrll	Sync with HEAD.
1.46.2.2	23-Oct-2007	ad	Sync with head.
1.46.2.1	20-Aug-2007	ad	Sync with HEAD.
1.47.2.2	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.47.2.1	04-Aug-2007	kiyohara	file agp.c was added on branch matt-mips64 on 2007-08-04 09:33:06 +0000
1.48.6.2	13-Nov-2007	bouyer	Sync with HEAD
1.48.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.48.2.3	23-Mar-2008	matt	sync with HEAD
1.48.2.2	09-Jan-2008	matt	sync with HEAD
1.48.2.1	06-Nov-2007	matt	sync with HEAD
1.50.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.50.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.50.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.52.2.2	26-Dec-2007	ad	Sync with head.
1.52.2.1	08-Dec-2007	ad	Sync with head.
1.53.2.1	11-Dec-2007	yamt	sync with head.
1.54.10.6	17-Jan-2009	mjf	Sync with HEAD.
1.54.10.5	28-Sep-2008	mjf	Sync with HEAD.
1.54.10.4	29-Jun-2008	mjf	Sync with HEAD.
1.54.10.3	02-Jun-2008	mjf	Sync with HEAD.
1.54.10.2	03-Apr-2008	mjf	Sync with HEAD.
1.54.10.1	29-Mar-2008	mjf	- etc/devfsd.conf: Add some rules to give nodes like /dev/tty and /dev/null better default modes, i.e. 0666. - sbin/init: Run devfsd -s before going to multiuser. - sys/arch: Provide arm32, i386, sparc with a mem_init() function to request device nodes for /dev/null, /dev/zero, etc. - sys/dev: Convert rnd, wd, agp, raid, cd, sd, wsdisplay, wskbd, wsmouse, wsmux, tty, bpf, swap to devfs New World Order. - sys/fs/devfs: Make the visibility attribute of device nodes configurable. Also provide a function to mount a devfs on boot. - sys/kern: Add a new boot flag, -n. This disables devfs support. Unless the -n flag is specified the kernel will mount a devfs file system on boot.
1.54.6.1	24-Mar-2008	keiichi	sync with head.
1.56.2.2	17-Jun-2008	yamt	sync with head.
1.56.2.1	18-May-2008	yamt	sync with head.
1.57.4.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.57.4.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.57.2.3	11-Aug-2010	yamt	sync with head.
1.57.2.2	11-Mar-2010	yamt	sync with head
1.57.2.1	04-May-2009	yamt	sync with head.
1.59.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.59.4.1	19-Oct-2008	haad	Sync with HEAD.
1.62.8.1	21-Apr-2010	matt	sync to netbsd-5
1.62.4.2	19-May-2011	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1560): sys/sys/agpio.h: revision 1.10 sys/dev/pci/agp_via.c: revision 1.21 sys/dev/pci/agp.c: revision 1.77 add AGP v3 support to agp_generic_enable from FreeBSD
1.62.4.1	05-May-2009	bouyer	Pull up following revision(s) (requested by snj in ticket #737): sys/arch/x86/pci/pchb.c: revisions 1.15 - 1.17 sys/dev/pci/agp.c: revisions 1.63 - 1.65 sys/dev/pci/agp_i810.c: revisions 1.57 - 1.64 sys/dev/pci/pcidevs: revisions 1.965, 1.967 via patch sys/dev/pci/agpreg.h: revision 1.20 Add AGP support for Intel G35, G45, and Q45.
1.62.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.62.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.66.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.66.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.66.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.68.2.3	21-Apr-2011	rmind	sync with head
1.68.2.2	05-Mar-2011	rmind	sync with head
1.68.2.1	03-Jul-2010	rmind	sync with head
1.76.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.76.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.76.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.79.8.1	29-Apr-2012	mrg	sync to latest -current.
1.79.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.79.4.1	17-Apr-2012	yamt	sync with head
1.80.4.1	18-May-2014	rmind	sync with head
1.80.2.2	03-Dec-2017	jdolecek	update from HEAD
1.80.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.81.2.1	10-Aug-2014	tls	Rebase.
1.83.12.1	21-Apr-2017	bouyer	Sync with HEAD
1.83.8.1	20-Mar-2017	pgoyette	Sync with HEAD
1.83.4.1	28-Aug-2017	skrll	Sync with HEAD
1.84.14.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.84.14.1	10-Jun-2019	christos	Sync with HEAD
1.84.12.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.85.6.1	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #672): sys/dev/ic/tulip.c: revision 1.198 sys/dev/pci/if_jme.c: revision 1.45 sys/dev/pci/agp.c: revision 1.86 sys/dev/pci/if_lii.c: revision 1.27 sys/dev/acpi/thinkpad_acpi.c: revision 1.47 sys/dev/scsipi/scsipi_base.c: revision 1.183 sys/dev/ic/aic6915reg.h: revision 1.6 Fix undefined behavior in thinkpad_mask_init(). Found by kUBSan. Use unsigned when initialize the transmit completion ring to avoid undefined behavior. Found by kUBSan. Avoid undefined behavior when setting multicast address. found by kUBSan. Use unsigned to avoid undefined behavior in agpattach(). Found by kUBSan. Avoid undefined behavior in jme_mii_write(). Found by kUBSan. Use unsigned to avoid undefined behavior in lii_setmulti(). Use unsigned to avoid undefined behavior in scsipi_{get,put}_tag(). Found by kUBSan.
1.17	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.16	13-Nov-2010	uebayasi	branches: 1.16.60; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.15	09-Jun-2008	freza	branches: 1.15.20; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.14	13-Apr-2008	njoly	branches: 1.14.2; 1.14.4; 1.14.6; Fix infinite loop with AGP on Ali M1689 (from PR/38269 and PR/35416), by disallowing aperture size values below 1MB.
1.13	04-Jan-2008	ad	branches: 1.13.6; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.12	19-Oct-2007	ad	branches: 1.12.2; 1.12.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	16-Nov-2006	christos	branches: 1.11.2; 1.11.8; 1.11.16; 1.11.22; 1.11.24; 1.11.28; __unused removal on arguments; approved by core.
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	16-Jan-2006	christos	branches: 1.9.18; 1.9.20; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.8	11-Dec-2005	christos	branches: 1.8.2; merge ktrace-lwp.
1.7	28-Jun-2005	thorpej	branches: 1.7.2; Use ANSI function decls and static.
1.6	27-Feb-2005	perry	branches: 1.6.2; nuke trailing whitespace
1.5	02-Jun-2004	christos	branches: 1.5.4; 1.5.6; Apply patch from "pandre" at netbsd.ports.i386, stops panic during attachment.
1.4	31-Jan-2003	thorpej	branches: 1.4.2; Use aprint_*().
1.3	13-Nov-2001	lukem	add RCSID
1.2	15-Sep-2001	thorpej	branches: 1.2.2; 1.2.4; Clean up the AGP match/attach code somewhat.
1.1	10-Sep-2001	fvdl	branches: 1.1.2; AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.1.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.2.1	10-Sep-2001	thorpej	file agp_ali.c was added on branch kqueue on 2001-09-13 01:15:50 +0000
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	15-Sep-2001	fvdl	file agp_ali.c was added on branch thorpej-devvp on 2001-10-01 12:45:50 +0000
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	15-Sep-2001	nathanw	file agp_ali.c was added on branch nathanw_sa on 2001-09-21 22:35:52 +0000
1.4.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.2.1	03-Aug-2004	skrll	Sync with HEAD
1.5.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.5.4.1	29-Apr-2005	kent	sync with -current
1.6.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.7.2.3	21-Jan-2008	yamt	sync with head
1.7.2.2	27-Oct-2007	yamt	sync with head.
1.7.2.1	21-Jun-2006	yamt	sync with head.
1.8.2.1	01-Feb-2006	yamt	sync with head.
1.9.20.2	10-Dec-2006	yamt	sync with head.
1.9.20.1	22-Oct-2006	yamt	sync with head
1.9.18.1	18-Nov-2006	ad	Sync with head.
1.11.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.24.2	09-Jan-2008	matt	sync with HEAD
1.11.24.1	06-Nov-2007	matt	sync with HEAD
1.11.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.16.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.11.8.1	23-Oct-2007	ad	Sync with head.
1.11.2.1	19-Apr-2008	bouyer	Pull up following revision(s) (requested by njoly in ticket #1125): sys/dev/pci/agp_ali.c: revision 1.14 Fix infinite loop with AGP on Ali M1689 (from PR/38269 and PR/35416), by disallowing aperture size values below 1MB.
1.12.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.12.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.13.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.13.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.14.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.14.4.1	04-May-2009	yamt	sync with head.
1.14.2.1	17-Jun-2008	yamt	sync with head.
1.15.20.1	05-Mar-2011	rmind	sync with head
1.16.60.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.22	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.21	13-Nov-2010	uebayasi	branches: 1.21.60; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.20	09-Jun-2008	freza	branches: 1.20.20; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.19	04-Jan-2008	ad	branches: 1.19.6; 1.19.8; 1.19.10; 1.19.12; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.18	04-Mar-2007	christos	branches: 1.18.16; 1.18.22; 1.18.28; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.17	16-Nov-2006	christos	branches: 1.17.4; __unused removal on arguments; approved by core.
1.16	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.15	11-Apr-2006	rpaulo	branches: 1.15.8; 1.15.10; Coverity ID 2313: free ``asc'' on error.
1.14	16-Jan-2006	christos	branches: 1.14.2; 1.14.4; 1.14.6; 1.14.8; 1.14.10; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.13	11-Dec-2005	christos	branches: 1.13.2; merge ktrace-lwp.
1.12	28-Jun-2005	thorpej	branches: 1.12.2; Use ANSI function decls and static.
1.11	27-Feb-2005	perry	branches: 1.11.2; nuke trailing whitespace
1.10	20-Oct-2004	xtraeme	branches: 1.10.4; 1.10.6; Add another case to match "PCI_PRODUCT_AMD_SC761_SC" (AMD761 System Controller) in agp_amd_match(). Ok'ed by christos@
1.9	31-Jan-2003	thorpej	branches: 1.9.2; 1.9.4; Use aprint_*().
1.8	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.7	13-Nov-2001	lukem	add RCSID
1.6	06-Oct-2001	thorpej	Remove and unnecessary cast, and avoid a memory leak in the event allocation of the GATT fails.
1.5	01-Oct-2001	fvdl	Don't map the aperture into KVA, it isn't needed.
1.4	01-Oct-2001	fvdl	Fix bogus vdir offset calculation.
1.3	16-Sep-2001	thorpej	branches: 1.3.2; 1.3.4; AGP registers are in memory space on the AMD-751.
1.2	15-Sep-2001	thorpej	Clean up the AGP match/attach code somewhat.
1.1	10-Sep-2001	fvdl	branches: 1.1.2; AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.1.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.2.1	10-Sep-2001	thorpej	file agp_amd.c was added on branch kqueue on 2001-09-13 01:15:50 +0000
1.3.4.3	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.3.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.3.4.1	16-Sep-2001	fvdl	file agp_amd.c was added on branch thorpej-devvp on 2001-10-01 12:45:50 +0000
1.3.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.5	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.4	22-Oct-2001	nathanw	Catch up to -current.
1.3.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.3.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.3.2.1	16-Sep-2001	nathanw	file agp_amd.c was added on branch nathanw_sa on 2001-09-21 22:35:52 +0000
1.9.4.1	12-Nov-2004	jmc	Pullup rev 1.10 (requested by xtraeme in ticket #948) Add another case to match "PCI_PRODUCT_AMD_SC761_SC" (AMD761 System Controller) in agp_amd_match().
1.9.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.9.2.1	02-Nov-2004	skrll	Sync with HEAD.
1.10.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.10.4.1	29-Apr-2005	kent	sync with -current
1.11.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.12.2.3	21-Jan-2008	yamt	sync with head
1.12.2.2	03-Sep-2007	yamt	sync with head.
1.12.2.1	21-Jun-2006	yamt	sync with head.
1.13.2.1	01-Feb-2006	yamt	sync with head.
1.14.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.14.8.1	19-Apr-2006	elad	sync with head.
1.14.6.1	24-May-2006	yamt	sync with head.
1.14.4.1	22-Apr-2006	simonb	Sync with head.
1.14.2.1	09-Sep-2006	rpaulo	sync with head
1.15.10.2	10-Dec-2006	yamt	sync with head.
1.15.10.1	22-Oct-2006	yamt	sync with head
1.15.8.1	18-Nov-2006	ad	Sync with head.
1.17.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.18.28.1	08-Jan-2008	bouyer	Sync with HEAD
1.18.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.18.16.1	09-Jan-2008	matt	sync with HEAD
1.19.12.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.19.10.1	04-May-2009	yamt	sync with head.
1.19.8.1	17-Jun-2008	yamt	sync with head.
1.19.6.1	29-Jun-2008	mjf	Sync with HEAD.
1.20.20.1	05-Mar-2011	rmind	sync with head
1.21.60.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.9	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.8	04-Apr-2015	riastradh	branches: 1.8.18; Fix error branches in agp_amd64.c. - agp_generic_detach always. - Free asc if it was allocated. (Found by Brainy, noted by maxv@.) - Free the GATT if it was allocated.
1.7	25-Feb-2012	tsutsui	branches: 1.7.2; 1.7.14; 1.7.16; - make agp_amd64_attach() also checks AMD64 Family 10h CPU's misc configuration devices - print proper error message if no misc configuration device is found Fixes kernel crash right after starting Xserver with radeondrm on ASRock AM2NF3-VSTA (AM2 + nForce3 250 AGP) with Athlon II X2 CPU.
1.6	13-Nov-2010	uebayasi	branches: 1.6.8; 1.6.12; 1.6.14; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.5	28-Mar-2008	kiyohara	branches: 1.5.26; Fix ugly format in aprint_normal{,_dev}.
1.4	27-Mar-2008	kiyohara	Split device_t/softc.
1.3	04-Jan-2008	ad	branches: 1.3.6; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.2	19-Oct-2007	ad	branches: 1.2.2; 1.2.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.1	04-Aug-2007	kiyohara	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.14; 1.1.16; Add support agp_amd64. Imported from FreeBSD.
1.1.16.2	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.1.16.1	04-Aug-2007	kiyohara	file agp_amd64.c was added on branch matt-mips64 on 2007-08-04 09:33:06 +0000
1.1.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.1.10.4	21-Jan-2008	yamt	sync with head
1.1.10.3	27-Oct-2007	yamt	sync with head.
1.1.10.2	03-Sep-2007	yamt	sync with head.
1.1.10.1	04-Aug-2007	yamt	file agp_amd64.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:17 +0000
1.1.8.2	09-Jan-2008	matt	sync with HEAD
1.1.8.1	06-Nov-2007	matt	sync with HEAD
1.1.6.3	23-Oct-2007	ad	Sync with head.
1.1.6.2	20-Aug-2007	ad	Sync with HEAD.
1.1.6.1	04-Aug-2007	ad	file agp_amd64.c was added on branch vmlocking on 2007-08-20 22:07:03 +0000
1.1.4.2	15-Aug-2007	skrll	Sync with HEAD.
1.1.4.1	04-Aug-2007	skrll	file agp_amd64.c was added on branch nick-csl-alignment on 2007-08-15 13:48:27 +0000
1.1.2.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.1.2.2	04-Aug-2007	jmcneill	Sync with HEAD.
1.1.2.1	04-Aug-2007	jmcneill	file agp_amd64.c was added on branch jmcneill-pm on 2007-08-04 12:33:09 +0000
1.2.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.2.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.3.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.5.26.1	05-Mar-2011	rmind	sync with head
1.6.14.1	08-Mar-2012	riz	Pull up following revision(s) (requested by tsutsui in ticket #95): sys/dev/pci/agp_amd64.c: revision 1.7 - make agp_amd64_attach() also checks AMD64 Family 10h CPU's misc configuration devices - print proper error message if no misc configuration device is found Fixes kernel crash right after starting Xserver with radeondrm on ASRock AM2NF3-VSTA (AM2 + nForce3 250 AGP) with Athlon II X2 CPU.
1.6.12.1	04-Mar-2012	mrg	sync to latest -current.
1.6.8.1	17-Apr-2012	yamt	sync with head
1.7.16.1	06-Apr-2015	skrll	Sync with HEAD
1.7.14.1	23-Apr-2015	snj	Pull up following revision(s) (requested by mrg in ticket #718): sys/arch/x86/include/pmap.h: revision 1.56 sys/arch/x86/x86/pmap.c: revision 1.188 sys/dev/pci/agp_amd64.c: revision 1.8 sys/dev/pci/agp_i810.c: revision 1.118 sys/external/bsd/drm2/dist/drm/i915/i915_dma.c: revision 1.16 sys/external/bsd/drm2/dist/drm/i915/i915_gem.c: revision 1.29 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_agp.c: revision 1.3 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_ttm.c: revision 1.4 sys/external/bsd/drm2/dist/drm/radeon/atombios_crtc.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_agp.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_display.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_crtc.c: revision 1.2 sys/external/bsd/drm2/dist/drm/radeon/radeon_object.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_ttm.c: revision 1.7 sys/external/bsd/drm2/dist/drm/ttm/ttm_bo.c: revisions 1.7-1.10 sys/external/bsd/drm2/dist/drm/ttm/ttm_bo_util.c: revision 1.5 sys/external/bsd/drm2/i915drm/intelfb.c: revision 1.13 sys/external/bsd/drm2/include/drm/drm_wait_netbsd.h: revisions 1.12, 1.13 sys/external/bsd/drm2/include/linux/mm.h: revision 1.5 sys/external/bsd/drm2/include/linux/pci.h: revisions 1.16, 1.17 sys/external/bsd/drm2/nouveau/nouveaufb.c: revision 1.2 sys/external/bsd/drm2/radeon/radeon_pci.c: revisions 1.8, 1.9 sys/uvm/uvm_init.c: revision 1.46 Hack against the blank console problem: Leave the CLUT alone on ancient cards. At least this leaves us with a semi working console (red and blue are flipped). Leave an example of what seems to be happening but disable it because colors are better than 444 bit greyscale. -- Initialize P->V tracking for unmanaged device pages in uvm_init. Conditional on __HAVE_PMAP_PV_TRACK until we add it to all pmaps. MI part of pmap_pv(9) change proposed on tech-kern: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html -- Implement pmap_pv(9) for x86 for P->V tracking of unmanaged pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html -- Use pmap_pv(9) to remove mappings of Intel graphics aperture pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html Further background at: https://mail-index.netbsd.org/tech-kern/2014/07/23/msg017392.html -- Use pmap_pv(9) to remove mappings of device pages in TTM. Adapt nouveau and radeon to do pmap_pv_track for their device pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html Further background at: https://mail-index.netbsd.org/tech-kern/2014/07/23/msg017392.html -- Fix error branches in agp_amd64.c. - agp_generic_detach always. - Free asc if it was allocated. (Found by Brainy, noted by maxv@.) - Free the GATT if it was allocated. -- pmf_device_register returns false on failure, not true -- In DRM_SPIN_WAIT_ON, don't stop after waiting only one tick. Continue the loop to recheck the condition and count the whole duration. -- Don't use the video BIOS memory as an i915 flush page! -- Don't let anyone else allocate the video BIOS either. -- Missed a zero: it's 0x100000, not 0x10000. -- Don't reserve if atomic -- caller must have pre-pinned the buffer. -- Don't reserve if atomic -- caller must have pre-pinned the buffer. -- almost add radeondrmkms suspend/resume support. it unfortunately doesn't work. -- Need the page's uvm object lock to do pmap_page_protect. -- Use KASSERTMSG to show bad base/offset. -- KASSERT about page-alignment on initialization too. -- Don't break when hardclock_ticks wraps around. Since we now only count time spent in wait, rather than determining the end time and checking whether we've passed it, timeouts might be marginally longer in effect. Unlikely to be an issue. -- Remove broken drm2 vm_mmap stub. Can't possibly have ever worked. -- apply some of the additional changes from Arto Huusko in PR#49645: - call pmf_device_deregister on detach. i've kept the "resume = true" for radeon_resume_kms() call as it seems to work for me (indeed, code inspection shows it is unused on netbsd :-) my old nforce4 box that can resume old drm (or could, last i tried several years ago) while X and GL apps were running, can at least survive a resume if X hasn't started. my one attempt so far with X exited, but having run, did not work. -- First attempt to make ttm_buffer_object_transfer less bogus. -- Make sure mem.bus.is_iomem is initialized. PR 49833
1.7.2.1	03-Dec-2017	jdolecek	update from HEAD
1.8.18.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.8	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.7	02-Nov-2014	christos	branches: 1.7.20; make the apple code compile again.
1.6	13-Nov-2010	uebayasi	branches: 1.6.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.5	09-Jun-2008	freza	branches: 1.5.20; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.4	29-Apr-2008	martin	branches: 1.4.2; Convert to new 2 clause license
1.3	04-Jan-2008	ad	branches: 1.3.6; 1.3.8; 1.3.10; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.2	19-Oct-2007	ad	branches: 1.2.2; 1.2.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.1	25-Mar-2007	macallan	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.14; 1.1.16; 1.1.18; 1.1.22; very experimental Apple UniNorth AGP support
1.1.22.1	25-Oct-2007	bouyer	Sync with HEAD.
1.1.18.4	21-Jan-2008	yamt	sync with head
1.1.18.3	27-Oct-2007	yamt	sync with head.
1.1.18.2	03-Sep-2007	yamt	sync with head.
1.1.18.1	25-Mar-2007	yamt	file agp_apple.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:18 +0000
1.1.16.2	09-Jan-2008	matt	sync with HEAD
1.1.16.1	06-Nov-2007	matt	sync with HEAD
1.1.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.1.8.2	11-Jul-2007	mjf	Sync with head.
1.1.8.1	25-Mar-2007	mjf	file agp_apple.c was added on branch mjf-ufs-trans on 2007-07-11 20:06:56 +0000
1.1.6.2	15-Apr-2007	yamt	sync with head.
1.1.6.1	25-Mar-2007	yamt	file agp_apple.c was added on branch yamt-idlelwp on 2007-04-15 16:03:23 +0000
1.1.2.3	23-Oct-2007	ad	Sync with head.
1.1.2.2	10-Apr-2007	ad	Sync with head.
1.1.2.1	25-Mar-2007	ad	file agp_apple.c was added on branch vmlocking on 2007-04-10 13:24:23 +0000
1.2.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.2.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.3.10.2	04-May-2009	yamt	sync with head.
1.3.10.1	16-May-2008	yamt	sync with head.
1.3.8.2	17-Jun-2008	yamt	sync with head.
1.3.8.1	18-May-2008	yamt	sync with head.
1.3.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.3.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.4.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.5.20.1	05-Mar-2011	rmind	sync with head
1.6.18.1	03-Dec-2017	jdolecek	update from HEAD
1.7.20.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.126	29-Jan-2024	riastradh	agp_i810(4): Use ipi(9) for chipset flush on all CPUs, not xcall(9). i915 now calls into this with a spin lock held, so we have to use ipi(9), which spin-waits for the other CPUs to complete, rather than xcall(9), which may sleep-wait. Fortunately, this is just to execute WBINVD on x86 (and if this code ever runs on other architectures, which it probably doesn't, it'll be a similar barrier instruction), so spinning to wait for that on all CPUs isn't too costly. PR kern/57878 XXX pullup-10
1.125	17-Jul-2022	riastradh	branches: 1.125.4; agp(4): Use bus_space_barrier, not membar_producer.
1.124	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.123	27-Aug-2018	riastradh	Restore gtt page table control register on resume. Additional subroutine agp_i810_reset is used by i915drmkms to make sure it happens early enough, since i915drmkms resumes before agp. XXX pullup-7 XXX pullup-8
1.122	01-May-2016	nonaka	branches: 1.122.10; 1.122.16; 1.122.18; kern/50453: Separate Pineview chipset from i915-family chipset support. Should use AGP_I965_IFPADDR instead AGP_I915_IFPADDR in Pineview, becuase AGP_I915_IFPADDR is used as PCIEXBAR.
1.121	21-Oct-2015	christos	Fix typo
1.120	21-Oct-2015	christos	CID 1328433: Anding with 0 and comparing != 0 will not work very will.
1.119	13-Oct-2015	riastradh	Fix mapping Intel graphics device registers. - Accept either 32-bit or 64-bit mappings for all devices. - Let the device always dictate size of the mapping. - Explain why we don't have a statically fixed mapping size. Fixes the main part of PR kern/50060. Still a display mode issue from one submitter, but it is almost certainly an unrelated issue.
1.118	05-Apr-2015	riastradh	Don't use the video BIOS memory as an i915 flush page!
1.117	06-Mar-2015	riastradh	Pass cache-related flags through to the GTT on pre-SNB devices. I had assumed for ages this would increase the amount of caching and thereby increase the chance of stale caches leading to rendering glitches. But apparently I was wrong, and failing to pass these through was causing all sorts of problems!
1.116	26-Feb-2015	riastradh	Include <sys/atomic.h> for membar_producer. (Why didn't this fail in my build?)
1.115	26-Feb-2015	riastradh	Issue a write barrier after updating the GTT. Linux never used to do this...until a month: https://bugs.freedesktop.org/show_bug.cgi?id=88191 commit 983d308cb8f602d1920a8c40196eb2ab6cc07bd2 Author: Chris Wilson <chris@chris-wilson.co.uk> Date: Mon Jan 26 10:47:10 2015 +0000 agp/intel: Serialise after GTT updates
1.114	24-Aug-2014	riastradh	branches: 1.114.2; Use the right BAR for MMIO register sizing on i810.
1.113	22-Aug-2014	riastradh	Revert to mapping the whole GTTMMADR region all at once for now. This was split up so we could map the GTT prefetchable without interfering with the MMIO registers, but it interferes with old drm which tries to map the whole region and which is still hooked up for the moment. Once we unhook old drm for good we can put this back.
1.112	25-Jul-2014	riastradh	branches: 1.112.2; Mark isc __diagused in agp_i810_unbind_memory.
1.111	23-Jul-2014	riastradh	Tweak style, make it compile...oops.
1.110	23-Jul-2014	riastradh	For the unbind code to work the bind code must set the offset.
1.109	23-Jul-2014	riastradh	Also unbind the right offset.
1.108	23-Jul-2014	riastradh	Bind i810 dcache pages at the requested offset, not at VA start. From John D Baker in PR xsrc/48344. XXX pullup to 6 (by patch)
1.107	01-Jul-2014	riastradh	Tweak debug printf directives.
1.106	27-Jun-2014	riastradh	Collect `#ifdef AGP_DEBUG printf(...)' into local DPRINTF macro.
1.105	25-Jun-2014	riastradh	No, that should be unsigned.
1.104	25-Jun-2014	riastradh	Fix integer type mismatch in debug printf.
1.103	12-Jun-2014	riastradh	Get GTT size from PGTBL_CTL on G4X chipsets.
1.102	12-Jun-2014	riastradh	Fix MMIO registers sizes for pre-Ironlake chipsets.
1.101	12-Jun-2014	riastradh	Extend agp_generic_bind_memory with bounds to check. New routine agp_generic_bind_memory_bounded. Use it in agp_i810 to replace the pile of code I wrote as a replacement which is apparently broken.
1.100	12-Jun-2014	riastradh	Fix type in cast too, after changing type of gtt_off.
1.99	12-Jun-2014	riastradh	Check bounds in agp_i810_borrow. Out of paranoia, do a bus_space_subregion in case the old drm code tries sizes that the agp_i810 code doesn't agree with.
1.98	12-Jun-2014	riastradh	Use correct type for gtt_off, an offset relative to a bus address.
1.97	12-Jun-2014	riastradh	Ensure we map all the MMIO registers. For i8xx chipsets, the GTT is in the middle of the MMIO space, so we can't map it separately. Use a subregion instead.
1.96	12-Jun-2014	christos	Giving a 0 gtt_size for G4X causes a panic when you try to bus map 0 size. Give it 256, like the default aperture size for it. Now it works again.
1.95	11-Jun-2014	riastradh	Fix fencepost.
1.94	11-Jun-2014	riastradh	Fix various mistakes in yesterday's commit. - Set mem->am_nseg, particularly before trying to use it. - Avoid error-prone local copies in agp_i810_bind_memory_main. - Omit missing DMA unload/free in agp_i810_unbind_memory.
1.93	11-Jun-2014	riastradh	Ignore result of agp_i810_unbind_page in agp_i810_unbind_memory.
1.92	11-Jun-2014	riastradh	Back out previous mistake. If you mixed `cvs diff' and `cvs commit' in your shell history, remember to hit C-p the right number of times before RET. Oops.
1.91	11-Jun-2014	riastradh	Add missing bus_dmamap_destroy.
1.90	11-Jun-2014	riastradh	Add missing bus_dmamap_destroy.
1.89	11-Jun-2014	riastradh	Check bounds more carefully to avoid integer overflow.
1.88	11-Jun-2014	riastradh	Don't trust the MSAC register after all; use the GMADR BAR size. Previously this read the MSAC register from the wrong PCI device. Experiments reveal that even on the right PCI device, per the datasheet, the GMADR BAR size is correct. Not clear that the driver has any business reading any of these GMCH registers to discern the aperture size -- seems to me they are meant to be written by the BIOS to determine the GMADR BAR size. Lacking a variety of hardware to test, though, I'll leave this as is for now...
1.87	11-Jun-2014	riastradh	Fix sense of conditional for GTTMMADR size.
1.86	10-Jun-2014	riastradh	Rework agp_i810 memory binding. Principal reason is that the GTT size is not necessarily the same as the aperture size: the GPU may have a bigger virtual address space than the CPU can see through the aperture. While here, factor the code a little more legibly and name some magic constants.
1.85	10-Jun-2014	riastradh	Another round of weed-whacking for agp_i810. - Make struct agp_i810_softc::gatt specific to i810 chipsets; use other members of struct agp_i810_softc for non-i810 chipsets. - agp_i810_init detects and sets isc->gtt_size. - Map GTT based on the GTT size detected by agp_i810_init. - Sprinkle some comments particularly about questionable calculations.
1.84	28-May-2014	riastradh	Cosmetic changes require testing too...oops.
1.83	28-May-2014	riastradh	Cosmetics: long lines, needless variables, PCI parameter verbiage.
1.82	28-May-2014	riastradh	Rationalize agp_i810 aprint output. Fix agp_i810_init error branches.
1.81	28-May-2014	riastradh	Oops -- gtt_off is relative to mmadr, not absolute.
1.80	27-May-2014	riastradh	Suppress uninitialized variable warnings in older and dumber GCC.
1.79	27-May-2014	riastradh	Rework agp_i810 attachment code a little. - Fix up error branches in agp_i810_attach. - Use a separate bus space handle for the GTT, whether it is in a separate BAR or a subregion of the MMIO device registers, so that (a) agp_i810_write_gtt_entry and agp_i810_post_gtt_entry are easier to follow, and (b) we can map the GTT prefetchable eventually.
1.78	26-May-2014	riastradh	Use correct address for 64-bit flush page config register.
1.77	26-May-2014	riastradh	Avoid xcall(9) while cold.
1.76	24-May-2014	riastradh	Handle 64-bit paddr registers delicately for non-PAE i386.
1.75	23-May-2014	riastradh	Implement Intel AGP chipset flush. While here, fix the use of the I915 PCI MSAC register: it lies on the bridge device (bus 0 dev 0 func 0), not the graphics device (bus 0 dev 2 func 0). I'm not sure we ever need to consult this register, really -- the PCI BARs should be sized. But I'll leave a rototill of this mess for another day.
1.74	18-Mar-2014	riastradh	branches: 1.74.2; Merge riastradh-drm2 to HEAD.
1.73	04-Apr-2011	dyoung	branches: 1.73.4; 1.73.10; 1.73.14; 1.73.16; 1.73.18; 1.73.24; 1.73.26; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.72	24-Feb-2011	matt	Add Intel Pineview support
1.71	30-Jan-2011	gsutre	Enable 36-bit addressing for chipsets that support it. While there, factorize offset computation. Inspired from OpenBSD and Intel docs. Note: agp_i810_bind/unbind_page will now fail with EINVAL if the physical address is too large for the chipset (instead of silently truncating it). ok jmcneill@
1.70	25-Jan-2011	gsutre	Typos, fix memory binding for memory type 2, and add consistency checks. From OpenBSD. ok jmcneill@
1.69	13-Nov-2010	uebayasi	branches: 1.69.2; 1.69.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.68	16-Jun-2010	riz	Add AGP support for a number of Intel onboard devices, including 82G41, 82B43, E7221, 82965GME, and "Iron Lake". Device types (i915, i965, G33, and G4X variants) from the Linux Intel AGP driver, and (for 82G41) from Henry Bent in PR#42906. There are a few more varieties that should be relatively low-hanging fruit ("Pineview" and "Sandy Bridge"), but will require a little bit of rejiggering of the "chiptype". OK mrg@
1.67	04-Apr-2010	jakllsch	Add (non-IGP) AGP port support for Intel 82855GM. agp_intel tested with radeondrm(4). agp_i810 tested with i915drm(4).
1.66	24-Feb-2010	dyoung	branches: 1.66.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.65	08-Jan-2010	dyoung	branches: 1.65.2; Expand PMF_FN_* macros.
1.64	04-May-2009	markd	Treat G4X like I915 and G33 for MSAC. Fixes aperture setting. Has been working successfully for me and snj for the last few months.
1.63	19-Feb-2009	markd	Add missing break.
1.62	27-Jan-2009	markd	branches: 1.62.2; Add some more Intel G4X class chipsets. Also fix gtt_size for G4X chipsets. Partially from FreeBSD. There is still an issue here as G4X chipsets report an aperture size of 0.
1.61	13-Dec-2008	sketch	Fix PGTBL sizes.
1.60	10-Dec-2008	christos	fix PGTBL size detection on the intel 965GM. From Anon Ymous.
1.59	04-Dec-2008	alc	Fix regression introducted by the previous commit. i915 and G33 Intel chipset family need special handling when writing an entry to the GTT. reported by: <tnn> [0] tested by: <tnn> [0]: http://mail-index.netbsd.org/tech-x11/2008/12/03/msg000276.html
1.58	29-Nov-2008	christos	Add support for the Intel G45 AGP. From Arnaud Lacombe
1.57	08-Nov-2008	christos	Support for Intel G35 as found on Asus P5E-VM HDMI motherboard from Milos Negovanovic
1.56	22-Aug-2008	tnn	branches: 1.56.2; 1.56.4; 1.56.8; AGP support for Intel 945GME chipset, found on Acer Aspire One.
1.55	19-Aug-2008	matthias	Add agp support for Intel 946gz.
1.54	09-Jun-2008	freza	branches: 1.54.4; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.53	10-May-2008	jmcneill	Don't use aprint_error in the success path.
1.52	10-Apr-2008	cegger	branches: 1.52.2; 1.52.4; 1.52.6; use aprint_*_dev and device_xname
1.51	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.50	26-Feb-2008	drochner	-Now that the address space conflict is solved, we can re-enable that old workaround for some VESA BIOS which overwrites PGTBL_CTL. Pointed out by Yorick Hardy. -make two variables introduced in the last commit static
1.49	22-Feb-2008	drochner	Add a horrible hack to allow the DRM driver to get at the bus_space_handle allocated by the AGP driver in the case of Intel chipset graphics. This is different from the patch by Yorick Hardy circulated a while ago in that it doesn't change the semantics of reference counting within the (3rd-party) DRM code -- it just bypasses it. Needless to say that it is uglier, but it eases future updates to the DRM code because the change there is just 2 LOC. Imo, a cleaner solution for all that would be to attach "agp" to "vga" in the intel chipset graphics case, which better reflects how the hardware is structured. This would still need a hack to the DRM code, but it would be confined to childs of the "vga" device, without need for global variables. Since there is a variety of intel chipsets with AGP and/or builtin graphics options, this would need a considerable testing effort.
1.48	04-Jan-2008	ad	branches: 1.48.2; 1.48.6; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.47	09-Dec-2007	jmcneill	branches: 1.47.2; Merge jmcneill-pm branch.
1.46	24-Nov-2007	markd	branches: 1.46.2; 1.46.4; Add agp support for Intel Q35/G33/Q33.
1.45	12-Nov-2007	joerg	Merge bunch of AGP changes from jmcneill-pm.
1.44	30-Oct-2007	jnemeth	branches: 1.44.2; PR/37201 - Yasushi Oshima -- Intel 82965G chipset support
1.43	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.42	15-Aug-2007	markd	branches: 1.42.2; 1.42.6; Add agp support for i965Q. Based on a patch for FreeBSD by Eric Anholt. OKed on tech-kern some months ago.
1.41	20-Jun-2007	sborrill	branches: 1.41.2; 1.41.6; Support more i915 stolen memory configurations (previous limit was 8MB, some configurations steal up to 64MB). OK christos@
1.40	24-Mar-2007	christos	comment out unmap; breaks X.
1.39	23-Mar-2007	drochner	simplify previous: no need to call pci_mapreg_info()
1.38	23-Mar-2007	jmcneill	Initialize mmadrsize in the non-i910 case, otherwise we end up passing an uninitialized variable to bus_space_unmap.
1.37	20-Mar-2007	drochner	Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11. Minor modifications by me: -use an mi device major number -(coarsly) divided into pci card specific and less specific parts, moved the latter to dev/drm -renamed autoconf attributes to reflect this Todo: -adapt all card frontends but i915 to drm include file location -review the mtrr change -make the change to agp_i810.c coexist with the fix for buggy VESA BIOSes which is commented out temporarily -RCS IDs etc style stuff -LKM support (rescan support for vga) -test
1.36	04-Mar-2007	christos	branches: 1.36.2; 1.36.4; 1.36.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.35	16-Nov-2006	christos	branches: 1.35.4; __unused removal on arguments; approved by core.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.32	30-Jul-2006	simonb	branches: 1.32.4; 1.32.6; Add support for 82945GM and (untested) 82945P AGP. FreeBSD and Linux both treat the 945 series exactly the same as the 915 series for AGP, so just match a few extra device IDs.
1.31	27-Jun-2006	tron	Make this build with GCC 4.x.
1.30	14-May-2006	elad	branches: 1.30.4; integrate kauth.
1.29	15-Apr-2006	rpaulo	s/#ifdef DEBUG/#ifdef AGP_DEBUG/
1.28	16-Jan-2006	christos	branches: 1.28.2; 1.28.4; 1.28.6; 1.28.8; 1.28.10; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.27	11-Dec-2005	christos	branches: 1.27.2; merge ktrace-lwp.
1.26	28-Jun-2005	thorpej	branches: 1.26.2; Use ANSI function decls and static.
1.25	27-Feb-2005	perry	branches: 1.25.2; nuke trailing whitespace
1.24	27-Jan-2005	jmcneill	Add agp_i810 powerhook, from Kentaro A. Kurahone
1.23	06-Nov-2004	xtraeme	branches: 1.23.4; 1.23.6; Add another case to match "PCI_PRODUCT_INTEL_82815_FULL_HUB" (Intel 82815 Hub) in agp_i810_attach(), closes PR 27863. Ok'ed christos@
1.22	20-Mar-2004	jdolecek	branches: 1.22.2; eliminate redundant aperture size printouts reported in PR kern/24859 by Nicolas Joly
1.21	23-Dec-2003	tron	Support i82845G/GL based motherboards without an Integrated Graphics Device. Patch supplied by Nicolas Joly in PR kern/23860.
1.20	07-Sep-2003	tron	Compile in the i845/i865 fallback attachment code only if generic AGP Intel support is enabled.
1.19	07-Sep-2003	tron	If we fail to find the Integrated Graphics Device of an Intel i840 or i865 chipset try the generic Intel AGP code because it is probably a motherboard without onboard graphics.
1.18	26-Aug-2003	tron	Add support for i865G integrated graphics support to agp(4). Patches contributed by Michael D. Allen in PR port-i386/22600.
1.17	20-Jul-2003	hannken	Add support for i855 agp. From FreeBSD. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.16	24-Mar-2003	drochner	branches: 1.16.2; Remove an aprint_normal() so that the output looks reasonable at least in the non-error case. (The aprint_* stuff makes it hard to keep track of newlines and device name tags, and it suffers from lack of rules.)
1.15	31-Jan-2003	thorpej	Use aprint_*().
1.14	13-Dec-2002	scw	Add support for i830/i845 agp. From Shingo WATANABE (on tech-x11) and FreeBSD, with cleanup/KNF by me. Note: These chipsets are not well supported by the i810 driver in NetBSD's in-tree xsrc (based on XFree86 4.2.1 at this time). However, the driver works perfectly using bleeding-edge XFree86-current on my Omnibook's i830MG with these agp changes.
1.13	11-Aug-2002	drochner	update the is_bound flag for local memory too
1.12	22-Jan-2002	augustss	branches: 1.12.8; 1.12.10; Don't pretend we can handle i830; it needs a somewhat different AGP driver. (I added i830 to quickly.)
1.11	14-Jan-2002	augustss	Add i830M to agp chip tables.
1.10	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.9	13-Nov-2001	lukem	add RCSID
1.8	20-Sep-2001	fvdl	branches: 1.8.2; 1.8.4; Add message to explain failure if the internal i810 config registers weren't found.
1.7	15-Sep-2001	drochner	normalize error message if agp_map_aperture() fails
1.6	15-Sep-2001	thorpej	Clean up the AGP match/attach code somewhat.
1.5	14-Sep-2001	drochner	make the hardware cursor work
1.4	13-Sep-2001	drochner	-allocate DMA maps as large as necessary in the worst case -Add a really bad hack: The X server (4.1.0 afaict) appearently overwrites the PGTBL_CTL which contains the base address of the pseudo-GATT. Restore the original value if we notice this happened. The X server works now, but only with soft cursor.
1.3	11-Sep-2001	fvdl	branches: 1.3.2; Fix i810/other intel chipsets attach mishap.
1.2	10-Sep-2001	fvdl	Fix reversed if condition in agp_i810_vgamatch.
1.1	10-Sep-2001	fvdl	AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.3.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.3.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.3.2.1	11-Sep-2001	thorpej	file agp_i810.c was added on branch kqueue on 2001-09-13 01:15:50 +0000
1.8.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.8.4.1	20-Sep-2001	fvdl	file agp_i810.c was added on branch thorpej-devvp on 2001-10-01 12:45:51 +0000
1.8.2.6	19-Dec-2002	thorpej	Sync with HEAD.
1.8.2.5	13-Aug-2002	nathanw	Catch up to -current.
1.8.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.8.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.8.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.8.2.1	20-Sep-2001	nathanw	file agp_i810.c was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.12.10.1	26-Jan-2003	jmc	Pullup revisions 1.12-1.13 (requested by manu in ticket #1073) update the is_bound flag for local memory too
1.12.8.1	29-Aug-2002	gehenna	catch up with -current.
1.16.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.16.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.16.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.16.2.4	14-Nov-2004	skrll	Sync with HEAD.
1.16.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.16.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.16.2.1	03-Aug-2004	skrll	Sync with HEAD
1.22.2.1	12-Nov-2004	jmc	Pullup rev 1.23 (requested by xtraeme in ticket #938) Add another case to match "PCI_PRODUCT_INTEL_82815_FULL_HUB" (Intel 82815 Hub) in agp_i810_attach(). PR#27863
1.23.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.23.6.1	12-Feb-2005	yamt	sync with head.
1.23.4.1	29-Apr-2005	kent	sync with -current
1.25.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.26.2.9	17-Mar-2008	yamt	sync with head.
1.26.2.8	27-Feb-2008	yamt	sync with head.
1.26.2.7	21-Jan-2008	yamt	sync with head
1.26.2.6	07-Dec-2007	yamt	sync with head
1.26.2.5	15-Nov-2007	yamt	sync with head.
1.26.2.4	27-Oct-2007	yamt	sync with head.
1.26.2.3	03-Sep-2007	yamt	sync with head.
1.26.2.2	30-Dec-2006	yamt	sync with head.
1.26.2.1	21-Jun-2006	yamt	sync with head.
1.27.2.1	01-Feb-2006	yamt	sync with head.
1.28.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.28.8.1	19-Apr-2006	elad	sync with head.
1.28.6.2	11-Aug-2006	yamt	sync with head
1.28.6.1	24-May-2006	yamt	sync with head.
1.28.4.1	22-Apr-2006	simonb	Sync with head.
1.28.2.1	09-Sep-2006	rpaulo	sync with head
1.30.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.32.6.2	10-Dec-2006	yamt	sync with head.
1.32.6.1	22-Oct-2006	yamt	sync with head
1.32.4.1	18-Nov-2006	ad	Sync with head.
1.35.4.3	15-Apr-2007	yamt	sync with head.
1.35.4.2	24-Mar-2007	yamt	sync with head.
1.35.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.36.6.1	29-Mar-2007	reinoud	Pullup to -current
1.36.4.1	11-Jul-2007	mjf	Sync with head.
1.36.2.4	23-Oct-2007	ad	Sync with head.
1.36.2.3	20-Aug-2007	ad	Sync with HEAD.
1.36.2.2	15-Jul-2007	ad	Sync with head.
1.36.2.1	10-Apr-2007	ad	Sync with head.
1.41.6.13	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.41.6.12	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.41.6.11	14-Nov-2007	joerg	Sync with HEAD.
1.41.6.10	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.41.6.9	31-Oct-2007	joerg	Sync with HEAD.
1.41.6.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.41.6.7	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.41.6.6	06-Sep-2007	jmcneill	Add Intel G33 AGP support, resolves PR36902, PR36903, PR36904, PR36906, PR36907, PR36908, PR36909, PR36910, PR36911, PR36912, PR36913, and PR36925.
1.41.6.5	27-Aug-2007	jmcneill	Add 965PM support.
1.41.6.4	16-Aug-2007	jmcneill	Sync with HEAD.
1.41.6.3	07-Aug-2007	jmcneill	Add i810 power handler.
1.41.6.2	06-Aug-2007	jmcneill	Move agp_i810_power to agp, since it does nothing specific to the i810 hardware. Allow agp front-ends to override the default power handler if specific configuration needs to take place, and expose agp_power via agpvar.h so the front-ends can call back to it to do generic initialization.
1.41.6.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.41.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.42.6.2	13-Nov-2007	bouyer	Sync with HEAD
1.42.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.42.2.3	23-Mar-2008	matt	sync with HEAD
1.42.2.2	09-Jan-2008	matt	sync with HEAD
1.42.2.1	06-Nov-2007	matt	sync with HEAD
1.44.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.44.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.44.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.44.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.46.4.1	11-Dec-2007	yamt	sync with head.
1.46.2.1	26-Dec-2007	ad	Sync with head.
1.47.2.1	08-Jan-2008	bouyer	Sync with HEAD
1.48.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.48.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.48.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.48.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.48.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.48.2.1	24-Mar-2008	keiichi	sync with head.
1.52.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.52.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.52.4.5	11-Aug-2010	yamt	sync with head.
1.52.4.4	11-Mar-2010	yamt	sync with head
1.52.4.3	16-May-2009	yamt	sync with head
1.52.4.2	04-May-2009	yamt	sync with head.
1.52.4.1	16-May-2008	yamt	sync with head.
1.52.2.2	17-Jun-2008	yamt	sync with head.
1.52.2.1	18-May-2008	yamt	sync with head.
1.54.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.54.4.1	19-Oct-2008	haad	Sync with HEAD.
1.56.8.1	21-Apr-2010	matt	sync to netbsd-5
1.56.4.2	20-Nov-2010	riz	Pull up following revision(s) (requested by jakllsch in ticket #1362): sys/dev/pci/pcidevs: revision 1.1026 sys/dev/pci/agp_intel.c: revision 1.35 sys/dev/pci/agp_i810.c: revision 1.67 Add Intel 82855GM Host-AGP Bridge. regen Add (non-IGP) AGP port support for Intel 82855GM. agp_intel tested with radeondrm(4). agp_i810 tested with i915drm(4).
1.56.4.1	05-May-2009	bouyer	Pull up following revision(s) (requested by snj in ticket #737): sys/arch/x86/pci/pchb.c: revisions 1.15 - 1.17 sys/dev/pci/agp.c: revisions 1.63 - 1.65 sys/dev/pci/agp_i810.c: revisions 1.57 - 1.64 sys/dev/pci/pcidevs: revisions 1.965, 1.967 via patch sys/dev/pci/agpreg.h: revision 1.20 Add AGP support for Intel G35, G45, and Q45.
1.56.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.56.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.62.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.65.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.65.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.66.2.4	21-Apr-2011	rmind	sync with head
1.66.2.3	05-Mar-2011	rmind	sync with head
1.66.2.2	03-Jul-2010	rmind	sync with head
1.66.2.1	30-May-2010	rmind	sync with head
1.69.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.69.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.69.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.73.26.2	05-Mar-2014	riastradh	Expose agp_i810_write_gtt_entry for drm2. Also add an agp_i810_post_gtt_entry to do a posting read for that GTT entry.
1.73.26.1	05-Mar-2014	riastradh	Expose the agp_i810 softc. This is an Intel-specific kludge so that i915drmkms can get at the GTT, graphics translation table, for pre-Sandy Bridge Intel graphics devices.
1.73.24.1	08-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1105): src/sys/dev/pci/agp_i810.c 1.108-1.110 via patch. Bind i810 dcache pages at the requested offset, not at VA start. From John D Baker in PR xsrc/48344.
1.73.18.1	18-May-2014	rmind	sync with head
1.73.16.1	08-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1105): src/sys/dev/pci/agp_i810.c 1.108-1.110 via patch. Bind i810 dcache pages at the requested offset, not at VA start. From John D Baker in PR xsrc/48344.
1.73.14.2	03-Dec-2017	jdolecek	update from HEAD
1.73.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.73.10.1	08-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1105): src/sys/dev/pci/agp_i810.c 1.108-1.110 via patch. Bind i810 dcache pages at the requested offset, not at VA start. From John D Baker in PR xsrc/48344.
1.73.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.74.2.1	10-Aug-2014	tls	Rebase.
1.112.2.5	27-Jan-2016	snj	Pull up following revision(s) (requested by riastradh in ticket #1078): sys/dev/pci/agp_i810.c: revision 1.120 sys/dev/pci/agp_i810.c: revision 1.121 CID 1328433: Anding with 0 and comparing != 0 will not work very will. Fix typo
1.112.2.4	06-Nov-2015	riz	Pull up following revision(s) (requested by riastradh in ticket #1000): sys/dev/pci/agp_i810.c: revision 1.119 Fix mapping Intel graphics device registers. - Accept either 32-bit or 64-bit mappings for all devices. - Let the device always dictate size of the mapping. - Explain why we don't have a statically fixed mapping size. Fixes the main part of PR kern/50060. Still a display mode issue from one submitter, but it is almost certainly an unrelated issue.
1.112.2.3	23-Apr-2015	snj	Pull up following revision(s) (requested by mrg in ticket #718): sys/arch/x86/include/pmap.h: revision 1.56 sys/arch/x86/x86/pmap.c: revision 1.188 sys/dev/pci/agp_amd64.c: revision 1.8 sys/dev/pci/agp_i810.c: revision 1.118 sys/external/bsd/drm2/dist/drm/i915/i915_dma.c: revision 1.16 sys/external/bsd/drm2/dist/drm/i915/i915_gem.c: revision 1.29 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_agp.c: revision 1.3 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_ttm.c: revision 1.4 sys/external/bsd/drm2/dist/drm/radeon/atombios_crtc.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_agp.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_display.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_crtc.c: revision 1.2 sys/external/bsd/drm2/dist/drm/radeon/radeon_object.c: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_ttm.c: revision 1.7 sys/external/bsd/drm2/dist/drm/ttm/ttm_bo.c: revisions 1.7-1.10 sys/external/bsd/drm2/dist/drm/ttm/ttm_bo_util.c: revision 1.5 sys/external/bsd/drm2/i915drm/intelfb.c: revision 1.13 sys/external/bsd/drm2/include/drm/drm_wait_netbsd.h: revisions 1.12, 1.13 sys/external/bsd/drm2/include/linux/mm.h: revision 1.5 sys/external/bsd/drm2/include/linux/pci.h: revisions 1.16, 1.17 sys/external/bsd/drm2/nouveau/nouveaufb.c: revision 1.2 sys/external/bsd/drm2/radeon/radeon_pci.c: revisions 1.8, 1.9 sys/uvm/uvm_init.c: revision 1.46 Hack against the blank console problem: Leave the CLUT alone on ancient cards. At least this leaves us with a semi working console (red and blue are flipped). Leave an example of what seems to be happening but disable it because colors are better than 444 bit greyscale. -- Initialize P->V tracking for unmanaged device pages in uvm_init. Conditional on __HAVE_PMAP_PV_TRACK until we add it to all pmaps. MI part of pmap_pv(9) change proposed on tech-kern: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html -- Implement pmap_pv(9) for x86 for P->V tracking of unmanaged pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html -- Use pmap_pv(9) to remove mappings of Intel graphics aperture pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html Further background at: https://mail-index.netbsd.org/tech-kern/2014/07/23/msg017392.html -- Use pmap_pv(9) to remove mappings of device pages in TTM. Adapt nouveau and radeon to do pmap_pv_track for their device pages. Proposed on tech-kern with no objections: https://mail-index.netbsd.org/tech-kern/2015/03/26/msg018561.html Further background at: https://mail-index.netbsd.org/tech-kern/2014/07/23/msg017392.html -- Fix error branches in agp_amd64.c. - agp_generic_detach always. - Free asc if it was allocated. (Found by Brainy, noted by maxv@.) - Free the GATT if it was allocated. -- pmf_device_register returns false on failure, not true -- In DRM_SPIN_WAIT_ON, don't stop after waiting only one tick. Continue the loop to recheck the condition and count the whole duration. -- Don't use the video BIOS memory as an i915 flush page! -- Don't let anyone else allocate the video BIOS either. -- Missed a zero: it's 0x100000, not 0x10000. -- Don't reserve if atomic -- caller must have pre-pinned the buffer. -- Don't reserve if atomic -- caller must have pre-pinned the buffer. -- almost add radeondrmkms suspend/resume support. it unfortunately doesn't work. -- Need the page's uvm object lock to do pmap_page_protect. -- Use KASSERTMSG to show bad base/offset. -- KASSERT about page-alignment on initialization too. -- Don't break when hardclock_ticks wraps around. Since we now only count time spent in wait, rather than determining the end time and checking whether we've passed it, timeouts might be marginally longer in effect. Unlikely to be an issue. -- Remove broken drm2 vm_mmap stub. Can't possibly have ever worked. -- apply some of the additional changes from Arto Huusko in PR#49645: - call pmf_device_deregister on detach. i've kept the "resume = true" for radeon_resume_kms() call as it seems to work for me (indeed, code inspection shows it is unused on netbsd :-) my old nforce4 box that can resume old drm (or could, last i tried several years ago) while X and GL apps were running, can at least survive a resume if X hasn't started. my one attempt so far with X exited, but having run, did not work. -- First attempt to make ttm_buffer_object_transfer less bogus. -- Make sure mem.bus.is_iomem is initialized. PR 49833
1.112.2.2	17-Mar-2015	riz	Pull up following revision(s) (requested by snj in ticket #590): sys/external/bsd/drm2/i915drm/intel_gtt.c: revision 1.5 sys/external/bsd/drm2/drm/drm_drv.c: revision 1.15 sys/external/bsd/drm2/dist/drm/i915/i915_gem.c: revision 1.28 sys/external/bsd/drm2/dist/drm/nouveau/core/subdev/clock/nouveau_subdev_clock_nv50.c: revision 1.3 sys/dev/pci/agp_i810.c: revision 1.115 sys/dev/pci/agp_i810.c: revision 1.116 sys/external/bsd/drm2/include/drm/intel-gtt.h: revision 1.5 sys/dev/pci/agp_i810.c: revision 1.117 sys/external/bsd/drm2/drm/drm_gem_vm.c: revision 1.6 sys/dev/pci/agp_i810var.h: revision 1.6 Issue a write barrier after updating the GTT. Linux never used to do this...until a month: <a rel="nofollow" href="https://bugs.freedesktop.org/show_bug.cgi?id=88191">https://bugs.freedesktop.org/show_bug.cgi?id=88191</a> commit 983d308cb8f602d1920a8c40196eb2ab6cc07bd2 Author: Chris Wilson <chris%chris-wilson.co.uk@localhost> Date: Mon Jan 26 10:47:10 2015 +0000 agp/intel: Serialise after GTT updates Include <sys/atomic.h> for membar_producer. (Why didn't this fail in my build?) Pass cache-related flags through to the GTT on pre-SNB devices. I had assumed for ages this would increase the amount of caching and thereby increase the chance of stale caches leading to rendering glitches. But apparently I was wrong, and failing to pass these through was causing all sorts of problems! Dedup the NetBSD portion of the code (ok Riastradh), no functional change. Don't return events that are too large and leave them in the list. Apply access control to gem mmap. fix gcc is stupid.
1.112.2.1	18-Sep-2014	martin	Pull up following revision(s) (requested by riastradh in ticket #94): sys/dev/pci/agp_i810.c: revision 1.113 sys/dev/pci/agp_i810.c: revision 1.114 Revert to mapping the whole GTTMMADR region all at once for now and use the right BAR for MMIO register sizing on i810.
1.114.2.3	29-May-2016	skrll	Sync with HEAD
1.114.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.114.2.1	06-Apr-2015	skrll	Sync with HEAD
1.122.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.122.18.1	10-Jun-2019	christos	Sync with HEAD
1.122.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.122.10.1	31-Aug-2018	martin	Pull up following revision(s) (requested by riastradh in ticket #998): sys/dev/pci/agp_i810var.h: revision 1.7 sys/external/bsd/drm2/i915drm/intel_gtt.c: revision 1.7 sys/dev/pci/agp_i810.c: revision 1.123 Restore gtt page table control register on resume. Additional subroutine agp_i810_reset is used by i915drmkms to make sure it happens early enough, since i915drmkms resumes before agp. XXX pullup-7 XXX pullup-8
1.125.4.1	03-Feb-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #574): sys/dev/pci/agp_i810.c: revision 1.126 agp_i810(4): Use ipi(9) for chipset flush on all CPUs, not xcall(9). i915 now calls into this with a spin lock held, so we have to use ipi(9), which spin-waits for the other CPUs to complete, rather than xcall(9), which may sleep-wait. Fortunately, this is just to execute WBINVD on x86 (and if this code ever runs on other architectures, which it probably doesn't, it'll be a similar barrier instruction), so spinning to wait for that on all CPUs isn't too costly. PR kern/57878
1.7	27-Aug-2018	riastradh	Restore gtt page table control register on resume. Additional subroutine agp_i810_reset is used by i915drmkms to make sure it happens early enough, since i915drmkms resumes before agp. XXX pullup-7 XXX pullup-8
1.6	06-Mar-2015	riastradh	branches: 1.6.10; 1.6.16; 1.6.18; Pass cache-related flags through to the GTT on pre-SNB devices. I had assumed for ages this would increase the amount of caching and thereby increase the chance of stale caches leading to rendering glitches. But apparently I was wrong, and failing to pass these through was causing all sorts of problems!
1.5	10-Jun-2014	riastradh	branches: 1.5.2; 1.5.4; 1.5.6; Another round of weed-whacking for agp_i810. - Make struct agp_i810_softc::gatt specific to i810 chipsets; use other members of struct agp_i810_softc for non-i810 chipsets. - agp_i810_init detects and sets isc->gtt_size. - Map GTT based on the GTT size detected by agp_i810_init. - Sprinkle some comments particularly about questionable calculations.
1.4	27-May-2014	riastradh	Rework agp_i810 attachment code a little. - Fix up error branches in agp_i810_attach. - Use a separate bus space handle for the GTT, whether it is in a separate BAR or a subregion of the MMIO device registers, so that (a) agp_i810_write_gtt_entry and agp_i810_post_gtt_entry are easier to follow, and (b) we can map the GTT prefetchable eventually.
1.3	23-May-2014	riastradh	Implement Intel AGP chipset flush. While here, fix the use of the I915 PCI MSAC register: it lies on the bridge device (bus 0 dev 0 func 0), not the graphics device (bus 0 dev 2 func 0). I'm not sure we ever need to consult this register, really -- the PCI BARs should be sized. But I'll leave a rototill of this mess for another day.
1.2	18-Mar-2014	riastradh	branches: 1.2.2; 1.2.4; 1.2.6; Merge riastradh-drm2 to HEAD.
1.1	05-Mar-2014	riastradh	branches: 1.1.2; file agp_i810var.h was initially added on branch riastradh-drm2.
1.1.2.2	05-Mar-2014	riastradh	Expose agp_i810_write_gtt_entry for drm2. Also add an agp_i810_post_gtt_entry to do a posting read for that GTT entry.
1.1.2.1	05-Mar-2014	riastradh	Expose the agp_i810 softc. This is an Intel-specific kludge so that i915drmkms can get at the GTT, graphics translation table, for pre-Sandy Bridge Intel graphics devices.
1.2.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.6.1	18-Mar-2014	yamt	file agp_i810var.h was added on branch yamt-pagecache on 2014-05-22 11:40:24 +0000
1.2.4.2	18-May-2014	rmind	sync with head
1.2.4.1	18-Mar-2014	rmind	file agp_i810var.h was added on branch rmind-smpnet on 2014-05-18 17:45:39 +0000
1.2.2.1	10-Aug-2014	tls	Rebase.
1.5.6.1	06-Apr-2015	skrll	Sync with HEAD
1.5.4.3	03-Dec-2017	jdolecek	update from HEAD
1.5.4.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.5.4.1	10-Jun-2014	tls	file agp_i810var.h was added on branch tls-maxphys on 2014-08-20 00:03:41 +0000
1.5.2.1	17-Mar-2015	riz	Pull up following revision(s) (requested by snj in ticket #590): sys/external/bsd/drm2/i915drm/intel_gtt.c: revision 1.5 sys/external/bsd/drm2/drm/drm_drv.c: revision 1.15 sys/external/bsd/drm2/dist/drm/i915/i915_gem.c: revision 1.28 sys/external/bsd/drm2/dist/drm/nouveau/core/subdev/clock/nouveau_subdev_clock_nv50.c: revision 1.3 sys/dev/pci/agp_i810.c: revision 1.115 sys/dev/pci/agp_i810.c: revision 1.116 sys/external/bsd/drm2/include/drm/intel-gtt.h: revision 1.5 sys/dev/pci/agp_i810.c: revision 1.117 sys/external/bsd/drm2/drm/drm_gem_vm.c: revision 1.6 sys/dev/pci/agp_i810var.h: revision 1.6 Issue a write barrier after updating the GTT. Linux never used to do this...until a month: <a rel="nofollow" href="https://bugs.freedesktop.org/show_bug.cgi?id=88191">https://bugs.freedesktop.org/show_bug.cgi?id=88191</a> commit 983d308cb8f602d1920a8c40196eb2ab6cc07bd2 Author: Chris Wilson <chris%chris-wilson.co.uk@localhost> Date: Mon Jan 26 10:47:10 2015 +0000 agp/intel: Serialise after GTT updates Include <sys/atomic.h> for membar_producer. (Why didn't this fail in my build?) Pass cache-related flags through to the GTT on pre-SNB devices. I had assumed for ages this would increase the amount of caching and thereby increase the chance of stale caches leading to rendering glitches. But apparently I was wrong, and failing to pass these through was causing all sorts of problems! Dedup the NetBSD portion of the code (ok Riastradh), no functional change. Don't return events that are too large and leave them in the list. Apply access control to gem mmap. fix gcc is stupid.
1.6.18.1	10-Jun-2019	christos	Sync with HEAD
1.6.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.6.10.1	31-Aug-2018	martin	Pull up following revision(s) (requested by riastradh in ticket #998): sys/dev/pci/agp_i810var.h: revision 1.7 sys/external/bsd/drm2/i915drm/intel_gtt.c: revision 1.7 sys/dev/pci/agp_i810.c: revision 1.123 Restore gtt page table control register on resume. Additional subroutine agp_i810_reset is used by i915drmkms to make sure it happens early enough, since i915drmkms resumes before agp. XXX pullup-7 XXX pullup-8
1.39	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.38	05-Feb-2019	mrg	fix a >15 year old code rework bug, and add a missing break;. the code in rev 1.7 looks like it does in this change, but rev 1.8 would write the I443 AGP_INTEL_AGPCTRL value, and then the other value.
1.37	04-Apr-2011	dyoung	branches: 1.37.48; 1.37.56; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.36	13-Nov-2010	uebayasi	branches: 1.36.2; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.35	04-Apr-2010	jakllsch	Add (non-IGP) AGP port support for Intel 82855GM. agp_intel tested with radeondrm(4). agp_i810 tested with i915drm(4).
1.34	24-Feb-2010	dyoung	branches: 1.34.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.33	08-Jan-2010	dyoung	branches: 1.33.2; Expand PMF_FN_* macros.
1.32	09-Jun-2008	freza	branches: 1.32.8; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.31	10-Apr-2008	cegger	branches: 1.31.2; 1.31.4; 1.31.6; use aprint_*_dev and device_xname
1.30	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.29	21-Feb-2008	drochner	clean up the error bit clear code: -use an aligned pci config space address as everyone else -- I'm sorry about that because I like gson's way a lot -- it keeps the address offset due to alignment visually close to the data shift --, but since aligned addresses are used everywhere else in definitions, this causes confusion -the mask applied to the data didn't make much sense, a look at the FreeBSD code where this code originates from suggests that there was just a mistake -- one trailing zero missing -- anyway, the datasheet tells that the error bits are all write-one- to-clear, so just write back the value and we are done
1.28	15-Feb-2008	gson	branches: 1.28.2; 1.28.6; Fix unaligned write to PCI configuration space. Should fix PR kern/37824.
1.27	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.26	15-Dec-2007	perry	__FUNCTION__ -> __func__
1.25	09-Dec-2007	jmcneill	branches: 1.25.2; Merge jmcneill-pm branch.
1.24	12-Nov-2007	joerg	branches: 1.24.2; 1.24.4; Merge bunch of AGP changes from jmcneill-pm.
1.23	19-Oct-2007	ad	branches: 1.23.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.22	26-Mar-2007	hubertf	branches: 1.22.8; 1.22.10; 1.22.14; Remove duplicate include of <sys/agpio.h> From: Slava Semushin <php-coder@altlinux.ru>
1.21	16-Nov-2006	christos	branches: 1.21.4; 1.21.8; 1.21.10; 1.21.12; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.18	11-Mar-2006	jmcneill	branches: 1.18.10; 1.18.12; Add Intel AGP powerhook. Tested on an 82855PM.
1.17	16-Jan-2006	christos	branches: 1.17.2; 1.17.4; 1.17.6; 1.17.8; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.16	11-Dec-2005	christos	branches: 1.16.2; merge ktrace-lwp.
1.15	28-Jun-2005	thorpej	branches: 1.15.2; Use ANSI function decls and static.
1.14	26-Aug-2003	tron	branches: 1.14.14; Rework support for i845, i865 and i875P. agp(4) now doesn't "kill" the i82547EI on my i865PE motherboard any longer.
1.13	26-Aug-2003	tron	Replaces space with tabs.
1.12	22-Jul-2003	simonb	Clean up autoconfiguration output a little.
1.11	06-Jul-2003	tron	Handle i865 and i875P chipsets like the i845. This works good enough to make XFree86 work with agp(4) enabled on an i865PE motherboard.
1.10	25-Jun-2003	ichiro	branches: 1.10.2; The problem which stops in the middle of booting was solved. (kern/21843 Brian Stark <bpstark@pacbell.net>)
1.9	14-Jun-2003	ichiro	make compile
1.8	14-Jun-2003	ichiro	fix some bugs
1.7	09-Jun-2003	ichiro	fix bug report - xsrc/17810, kern/16898 add AGP configuration registers for i840/855PM, i840/i850E, 82443LX/BX/GX Although there chips are well alike, but initialization and so on procedures differ.
1.6	31-Jan-2003	thorpej	Use aprint_*().
1.5	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.4	13-Nov-2001	lukem	add RCSID
1.3	15-Sep-2001	thorpej	branches: 1.3.2; 1.3.4; Clean up the AGP match/attach code somewhat.
1.2	11-Sep-2001	fvdl	branches: 1.2.2; Don't forget to install the pointer to the chip-specific chipc.
1.1	10-Sep-2001	fvdl	AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.2.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.2.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.2.2.1	11-Sep-2001	thorpej	file agp_intel.c was added on branch kqueue on 2001-09-13 01:15:51 +0000
1.3.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.3.4.1	15-Sep-2001	fvdl	file agp_intel.c was added on branch thorpej-devvp on 2001-10-01 12:45:51 +0000
1.3.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.3.2.1	15-Sep-2001	nathanw	file agp_intel.c was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.10.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.14.14.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.15.2.8	17-Mar-2008	yamt	sync with head.
1.15.2.7	27-Feb-2008	yamt	sync with head.
1.15.2.6	21-Jan-2008	yamt	sync with head
1.15.2.5	15-Nov-2007	yamt	sync with head.
1.15.2.4	27-Oct-2007	yamt	sync with head.
1.15.2.3	03-Sep-2007	yamt	sync with head.
1.15.2.2	30-Dec-2006	yamt	sync with head.
1.15.2.1	21-Jun-2006	yamt	sync with head.
1.16.2.1	01-Feb-2006	yamt	sync with head.
1.17.8.1	19-Apr-2006	elad	sync with head.
1.17.6.1	13-Mar-2006	yamt	sync with head.
1.17.4.1	22-Apr-2006	simonb	Sync with head.
1.17.2.1	09-Sep-2006	rpaulo	sync with head
1.18.12.2	10-Dec-2006	yamt	sync with head.
1.18.12.1	22-Oct-2006	yamt	sync with head
1.18.10.1	18-Nov-2006	ad	Sync with head.
1.21.12.1	29-Mar-2007	reinoud	Pullup to -current
1.21.10.1	11-Jul-2007	mjf	Sync with head.
1.21.8.2	23-Oct-2007	ad	Sync with head.
1.21.8.1	10-Apr-2007	ad	Sync with head.
1.21.4.1	15-Apr-2007	yamt	sync with head.
1.22.14.2	13-Nov-2007	bouyer	Sync with HEAD
1.22.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.22.10.3	23-Mar-2008	matt	sync with HEAD
1.22.10.2	09-Jan-2008	matt	sync with HEAD
1.22.10.1	06-Nov-2007	matt	sync with HEAD
1.22.8.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.22.8.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.22.8.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.22.8.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.22.8.1	08-Aug-2007	jmcneill	Add Intel (older version) AGP power management support.
1.23.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.23.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.23.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.24.4.1	11-Dec-2007	yamt	sync with head.
1.24.2.1	26-Dec-2007	ad	Sync with head.
1.25.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.25.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.28.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.28.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.28.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.28.2.1	24-Mar-2008	keiichi	sync with head.
1.31.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.31.4.3	11-Aug-2010	yamt	sync with head.
1.31.4.2	11-Mar-2010	yamt	sync with head
1.31.4.1	04-May-2009	yamt	sync with head.
1.31.2.1	17-Jun-2008	yamt	sync with head.
1.32.8.1	20-Nov-2010	riz	Pull up following revision(s) (requested by jakllsch in ticket #1362): sys/dev/pci/pcidevs: revision 1.1026 sys/dev/pci/agp_intel.c: revision 1.35 sys/dev/pci/agp_i810.c: revision 1.67 Add Intel 82855GM Host-AGP Bridge. regen Add (non-IGP) AGP port support for Intel 82855GM. agp_intel tested with radeondrm(4). agp_i810 tested with i915drm(4).
1.33.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.34.2.3	21-Apr-2011	rmind	sync with head
1.34.2.2	05-Mar-2011	rmind	sync with head
1.34.2.1	30-May-2010	rmind	sync with head
1.36.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.37.56.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.37.56.1	10-Jun-2019	christos	Sync with HEAD
1.37.48.1	01-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1311): sys/dev/pci/agp_intel.c: revision 1.38 fix a >15 year old code rework bug, and add a missing break;. the code in rev 1.7 looks like it does in this change, but rev 1.8 would write the I443 AGP_INTEL_AGPCTRL value, and then the other value.
1.15	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.14	13-Nov-2010	uebayasi	branches: 1.14.60; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.13	09-Jun-2008	freza	branches: 1.13.20; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.12	04-Jan-2008	ad	branches: 1.12.6; 1.12.8; 1.12.10; 1.12.12; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.11	19-Oct-2007	ad	branches: 1.11.2; 1.11.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.10	16-Nov-2006	christos	branches: 1.10.8; 1.10.22; 1.10.24; 1.10.28; __unused removal on arguments; approved by core.
1.9	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.8	16-Jan-2006	christos	branches: 1.8.18; 1.8.20; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.7	11-Dec-2005	christos	branches: 1.7.2; merge ktrace-lwp.
1.6	28-Jun-2005	thorpej	branches: 1.6.2; Use ANSI function decls and static.
1.5	27-Feb-2005	perry	branches: 1.5.2; nuke trailing whitespace
1.4	31-Jan-2003	thorpej	branches: 1.4.2; 1.4.10; 1.4.12; Use aprint_*().
1.3	13-Nov-2001	lukem	add RCSID
1.2	15-Sep-2001	thorpej	branches: 1.2.2; 1.2.4; Clean up the AGP match/attach code somewhat.
1.1	10-Sep-2001	fvdl	branches: 1.1.2; AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.1.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.2.1	10-Sep-2001	thorpej	file agp_sis.c was added on branch kqueue on 2001-09-13 01:15:51 +0000
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	15-Sep-2001	fvdl	file agp_sis.c was added on branch thorpej-devvp on 2001-10-01 12:45:51 +0000
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	15-Sep-2001	nathanw	file agp_sis.c was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.4.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.10.1	29-Apr-2005	kent	sync with -current
1.4.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.6.2.3	21-Jan-2008	yamt	sync with head
1.6.2.2	27-Oct-2007	yamt	sync with head.
1.6.2.1	21-Jun-2006	yamt	sync with head.
1.7.2.1	01-Feb-2006	yamt	sync with head.
1.8.20.2	10-Dec-2006	yamt	sync with head.
1.8.20.1	22-Oct-2006	yamt	sync with head
1.8.18.1	18-Nov-2006	ad	Sync with head.
1.10.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.10.24.2	09-Jan-2008	matt	sync with HEAD
1.10.24.1	06-Nov-2007	matt	sync with HEAD
1.10.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.10.8.1	23-Oct-2007	ad	Sync with head.
1.11.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.11.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.12.12.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.12.10.1	04-May-2009	yamt	sync with head.
1.12.8.1	17-Jun-2008	yamt	sync with head.
1.12.6.1	29-Jun-2008	mjf	Sync with HEAD.
1.13.20.1	05-Mar-2011	rmind	sync with head
1.14.60.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.22	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.21	19-Feb-2011	jmcneill	branches: 1.21.56; more AGP v3 fixes from FreeBSD: - fix a typo when enabling the aperture (read wrong register) - get_aperture/set_aperture: handle AGP v3 APSIZE register layout
1.20	15-Feb-2011	jmcneill	print AGP mode (v3/v2compat/v2) with aprint_debug instead of aprint_normal
1.19	13-Nov-2010	uebayasi	branches: 1.19.2; 1.19.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.18	31-Jan-2010	hubertf	branches: 1.18.4; Replace more printfs with aprint_normal / aprint_verbose Makes "boot -z" go mostly silent for me.
1.17	09-Jun-2008	freza	branches: 1.17.8; 1.17.16; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.16	11-Mar-2008	joerg	branches: 1.16.2; 1.16.4; 1.16.6; Avoid unaligned pci_conf_read.
1.15	04-Jan-2008	ad	branches: 1.15.2; 1.15.6; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.14	19-Oct-2007	ad	branches: 1.14.2; 1.14.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	27-Mar-2007	jmcneill	branches: 1.13.8; 1.13.10; 1.13.14; Instead of having a lookup table of devices that support AGP V3, use the capabilities register to determine the chipset's supported AGP mode. We still fallback to V2 compatibility mode when necessary.
1.12	27-Mar-2007	jmcneill	Add AGPv3 support to VIA AGP driver from FreeBSD. I can now use DRI w/ AGP enabled on my r300.
1.11	16-Nov-2006	christos	branches: 1.11.4; 1.11.8; 1.11.10; 1.11.12; __unused removal on arguments; approved by core.
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	16-Jan-2006	christos	branches: 1.9.18; 1.9.20; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.8	11-Dec-2005	christos	branches: 1.8.2; merge ktrace-lwp.
1.7	28-Jun-2005	thorpej	branches: 1.7.2; Use ANSI function decls and static.
1.6	27-Feb-2005	perry	branches: 1.6.2; nuke trailing whitespace
1.5	31-Jan-2003	thorpej	branches: 1.5.2; 1.5.10; 1.5.12; Use aprint_*().
1.4	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.3	13-Nov-2001	lukem	add RCSID
1.2	15-Sep-2001	thorpej	branches: 1.2.2; 1.2.4; Clean up the AGP match/attach code somewhat.
1.1	10-Sep-2001	fvdl	branches: 1.1.2; AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.1.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.2.1	10-Sep-2001	thorpej	file agp_via.c was added on branch kqueue on 2001-09-13 01:15:51 +0000
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	15-Sep-2001	fvdl	file agp_via.c was added on branch thorpej-devvp on 2001-10-01 12:45:52 +0000
1.2.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	15-Sep-2001	nathanw	file agp_via.c was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.5.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.5.10.1	29-Apr-2005	kent	sync with -current
1.5.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.7.2.5	17-Mar-2008	yamt	sync with head.
1.7.2.4	21-Jan-2008	yamt	sync with head
1.7.2.3	27-Oct-2007	yamt	sync with head.
1.7.2.2	03-Sep-2007	yamt	sync with head.
1.7.2.1	21-Jun-2006	yamt	sync with head.
1.8.2.1	01-Feb-2006	yamt	sync with head.
1.9.20.2	10-Dec-2006	yamt	sync with head.
1.9.20.1	22-Oct-2006	yamt	sync with head
1.9.18.1	18-Nov-2006	ad	Sync with head.
1.11.12.1	29-Mar-2007	reinoud	Pullup to -current
1.11.10.1	11-Jul-2007	mjf	Sync with head.
1.11.8.2	23-Oct-2007	ad	Sync with head.
1.11.8.1	10-Apr-2007	ad	Sync with head.
1.11.4.1	15-Apr-2007	yamt	sync with head.
1.13.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.10.3	23-Mar-2008	matt	sync with HEAD
1.13.10.2	09-Jan-2008	matt	sync with HEAD
1.13.10.1	06-Nov-2007	matt	sync with HEAD
1.13.8.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.14.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.14.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.15.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.15.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.2.1	24-Mar-2008	keiichi	sync with head.
1.16.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.16.4.2	11-Mar-2010	yamt	sync with head
1.16.4.1	04-May-2009	yamt	sync with head.
1.16.2.1	17-Jun-2008	yamt	sync with head.
1.17.16.1	21-Apr-2010	matt	sync to netbsd-5
1.17.8.2	19-May-2011	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1560): sys/sys/agpio.h: revision 1.10 sys/dev/pci/agp_via.c: revision 1.21 sys/dev/pci/agp.c: revision 1.77 add AGP v3 support to agp_generic_enable from FreeBSD
1.17.8.1	14-Feb-2010	bouyer	Pull up following revision(s) (requested by hubertf in ticket #1290): sys/kern/kern_ksyms.c: revision 1.53 sys/dev/pci/agp_via.c: revision 1.18 sys/netipsec/key.c: revision 1.63 sys/arch/x86/x86/x86_autoconf.c: revision 1.49 sys/kern/init_main.c: revision 1.415 sys/kern/cnmagic.c: revision 1.11 sys/netipsec/ipsec.c: revision 1.47 sys/arch/x86/x86/pmap.c: revision 1.100 sys/netkey/key.c: revision 1.176 Replace more printfs with aprint_normal / aprint_verbose Makes "boot -z" go mostly silent for me.
1.18.4.1	05-Mar-2011	rmind	sync with head
1.19.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.19.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.19.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.21.56.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.5	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.4	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.3	11-Dec-2005	christos	branches: 1.3.20; 1.3.22; merge ktrace-lwp.
1.2	24-May-2005	lukem	branches: 1.2.2; add __KERNEL_RCSID
1.1	30-Aug-2004	drochner	branches: 1.1.2; add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.1.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	03-Sep-2004	skrll	Sync with HEAD
1.1.2.1	30-Aug-2004	skrll	file agpbusprint.c was added on branch ktrace-lwp on 2004-09-03 12:45:27 +0000
1.2.2.1	21-Jun-2006	yamt	sync with head.
1.3.22.2	10-Dec-2006	yamt	sync with head.
1.3.22.1	22-Oct-2006	yamt	sync with head
1.3.20.1	18-Nov-2006	ad	Sync with head.
1.24	27-Feb-2017	msaitoh	Decode AGP capability.
1.23	01-May-2016	nonaka	branches: 1.23.2; 1.23.4; kern/50453: Separate Pineview chipset from i915-family chipset support. Should use AGP_I965_IFPADDR instead AGP_I915_IFPADDR in Pineview, becuase AGP_I915_IFPADDR is used as PCIEXBAR.
1.22	12-Jun-2014	riastradh	branches: 1.22.4; Get GTT size from PGTBL_CTL on G4X chipsets.
1.21	23-May-2014	riastradh	Implement Intel AGP chipset flush. While here, fix the use of the I915 PCI MSAC register: it lies on the bridge device (bus 0 dev 0 func 0), not the graphics device (bus 0 dev 2 func 0). I'm not sure we ever need to consult this register, really -- the PCI BARs should be sized. But I'll leave a rototill of this mess for another day.
1.20	29-Nov-2008	christos	branches: 1.20.26; 1.20.40; Add support for the Intel G45 AGP. From Arnaud Lacombe
1.19	11-Mar-2008	joerg	branches: 1.19.4; 1.19.10; 1.19.12; 1.19.14; 1.19.18; Avoid unaligned pci_conf_read.
1.18	21-Feb-2008	drochner	clean up the error bit clear code: -use an aligned pci config space address as everyone else -- I'm sorry about that because I like gson's way a lot -- it keeps the address offset due to alignment visually close to the data shift --, but since aligned addresses are used everywhere else in definitions, this causes confusion -the mask applied to the data didn't make much sense, a look at the FreeBSD code where this code originates from suggests that there was just a mistake -- one trailing zero missing -- anyway, the datasheet tells that the error bits are all write-one- to-clear, so just write back the value and we are done
1.17	15-Jan-2008	drochner	branches: 1.17.2; 1.17.6; fix some PCI config space offsets to be 32-bit aligned, the code did already do the right bit shifting
1.16	24-Nov-2007	markd	branches: 1.16.6; Add agp support for Intel Q35/G33/Q33.
1.15	12-Nov-2007	joerg	Merge bunch of AGP changes from jmcneill-pm.
1.14	15-Aug-2007	markd	branches: 1.14.2; 1.14.6; 1.14.8; Add agp support for i965Q. Based on a patch for FreeBSD by Eric Anholt. OKed on tech-kern some months ago.
1.13	04-Aug-2007	kiyohara	branches: 1.13.2; Add support agp_amd64. Imported from FreeBSD.
1.12	20-Jun-2007	sborrill	branches: 1.12.2; 1.12.6; Support more i915 stolen memory configurations (previous limit was 8MB, some configurations steal up to 64MB). OK christos@
1.11	27-Mar-2007	jmcneill	Add AGPv3 support to VIA AGP driver from FreeBSD. I can now use DRI w/ AGP enabled on my r300.
1.10	16-Jan-2006	christos	branches: 1.10.24; 1.10.28; 1.10.30; 1.10.32; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.9	11-Dec-2005	christos	branches: 1.9.2; merge ktrace-lwp.
1.8	27-Feb-2005	perry	branches: 1.8.2; 1.8.4; nuke trailing whitespace
1.7	04-Feb-2004	soren	branches: 1.7.8; 1.7.10; Use the right bits for the AGP version.
1.6	20-Jul-2003	hannken	Add support for i855 agp. From FreeBSD. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.5	25-Jun-2003	ichiro	branches: 1.5.2; The problem which stops in the middle of booting was solved. (kern/21843 Brian Stark <bpstark@pacbell.net>)
1.4	14-Jun-2003	ichiro	fix some bugs
1.3	09-Jun-2003	ichiro	fix bug report - xsrc/17810, kern/16898 add AGP configuration registers for i840/855PM, i840/i850E, 82443LX/BX/GX Although there chips are well alike, but initialization and so on procedures differ.
1.2	13-Dec-2002	scw	Add support for i830/i845 agp. From Shingo WATANABE (on tech-x11) and FreeBSD, with cleanup/KNF by me. Note: These chipsets are not well supported by the i810 driver in NetBSD's in-tree xsrc (based on XFree86 4.2.1 at this time). However, the driver works perfectly using bleeding-edge XFree86-current on my Omnibook's i830MG with these agp changes.
1.1	10-Sep-2001	fvdl	branches: 1.1.2; 1.1.4; 1.1.6; AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.1.6.2	01-Oct-2001	fvdl	Catch up with -current.
1.1.6.1	10-Sep-2001	fvdl	file agpreg.h was added on branch thorpej-devvp on 2001-10-01 12:45:52 +0000
1.1.4.3	19-Dec-2002	thorpej	Sync with HEAD.
1.1.4.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.4.1	10-Sep-2001	nathanw	file agpreg.h was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.1.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.2.1	10-Sep-2001	thorpej	file agpreg.h was added on branch kqueue on 2001-09-13 01:15:51 +0000
1.5.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.8.4.7	17-Mar-2008	yamt	sync with head.
1.8.4.6	27-Feb-2008	yamt	sync with head.
1.8.4.5	21-Jan-2008	yamt	sync with head
1.8.4.4	07-Dec-2007	yamt	sync with head
1.8.4.3	15-Nov-2007	yamt	sync with head.
1.8.4.2	03-Sep-2007	yamt	sync with head.
1.8.4.1	21-Jun-2006	yamt	sync with head.
1.8.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.9.2.1	01-Feb-2006	yamt	sync with head.
1.10.32.1	29-Mar-2007	reinoud	Pullup to -current
1.10.30.1	11-Jul-2007	mjf	Sync with head.
1.10.28.3	20-Aug-2007	ad	Sync with HEAD.
1.10.28.2	15-Jul-2007	ad	Sync with head.
1.10.28.1	10-Apr-2007	ad	Sync with head.
1.10.24.1	15-Apr-2007	yamt	sync with head.
1.12.6.4	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.12.6.3	06-Sep-2007	jmcneill	Add Intel G33 AGP support, resolves PR36902, PR36903, PR36904, PR36906, PR36907, PR36908, PR36909, PR36910, PR36911, PR36912, PR36913, and PR36925.
1.12.6.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.12.6.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.12.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.13.2.2	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.13.2.1	04-Aug-2007	kiyohara	file agpreg.h was added on branch matt-mips64 on 2007-08-04 09:33:06 +0000
1.14.8.3	18-Feb-2008	mjf	Sync with HEAD.
1.14.8.2	08-Dec-2007	mjf	Sync with HEAD.
1.14.8.1	19-Nov-2007	mjf	Sync with HEAD.
1.14.6.1	13-Nov-2007	bouyer	Sync with HEAD
1.14.2.2	23-Mar-2008	matt	sync with HEAD
1.14.2.1	09-Jan-2008	matt	sync with HEAD
1.16.6.1	19-Jan-2008	bouyer	Sync with HEAD
1.17.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.17.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.17.2.1	24-Mar-2008	keiichi	sync with head.
1.19.18.1	21-Apr-2010	matt	sync to netbsd-5
1.19.14.1	05-May-2009	bouyer	Pull up following revision(s) (requested by snj in ticket #737): sys/arch/x86/pci/pchb.c: revisions 1.15 - 1.17 sys/dev/pci/agp.c: revisions 1.63 - 1.65 sys/dev/pci/agp_i810.c: revisions 1.57 - 1.64 sys/dev/pci/pcidevs: revisions 1.965, 1.967 via patch sys/dev/pci/agpreg.h: revision 1.20 Add AGP support for Intel G35, G45, and Q45.
1.19.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.19.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.19.4.1	04-May-2009	yamt	sync with head.
1.20.40.1	10-Aug-2014	tls	Rebase.
1.20.26.2	03-Dec-2017	jdolecek	update from HEAD
1.20.26.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.22.4.2	28-Aug-2017	skrll	Sync with HEAD
1.22.4.1	29-May-2016	skrll	Sync with HEAD
1.23.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.23.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.22	10-Apr-2022	andvar	fix various typos in comments and output/log messages.
1.21	02-Nov-2014	christos	make the apple code compile again.
1.20	12-Jun-2014	riastradh	Extend agp_generic_bind_memory with bounds to check. New routine agp_generic_bind_memory_bounded. Use it in agp_i810 to replace the pile of code I wrote as a replacement which is apparently broken.
1.19	12-Jun-2014	riastradh	Check bounds in agp_i810_borrow. Out of paranoia, do a bus_space_subregion in case the old drm code tries sizes that the agp_i810 code doesn't agree with.
1.18	06-May-2009	cegger	branches: 1.18.22; 1.18.36; struct device * -> device_t, no functional changes intended.
1.17	09-Jun-2008	freza	branches: 1.17.12; o Split device_t/softc for agp(4). o agp_ali.c: remove unused 'agp' member from agp_ali_softc. o drm: agp_find_device() returns 'void *', not a device_t. o Use device_t, cfdata_t instead of struct pointers. o Don't cast void pointers in assignments.
1.16	22-Feb-2008	drochner	branches: 1.16.2; 1.16.4; 1.16.6; Add a horrible hack to allow the DRM driver to get at the bus_space_handle allocated by the AGP driver in the case of Intel chipset graphics. This is different from the patch by Yorick Hardy circulated a while ago in that it doesn't change the semantics of reference counting within the (3rd-party) DRM code -- it just bypasses it. Needless to say that it is uglier, but it eases future updates to the DRM code because the change there is just 2 LOC. Imo, a cleaner solution for all that would be to attach "agp" to "vga" in the intel chipset graphics case, which better reflects how the hardware is structured. This would still need a hack to the DRM code, but it would be confined to childs of the "vga" device, without need for global variables. Since there is a variety of intel chipsets with AGP and/or builtin graphics options, this would need a considerable testing effort.
1.15	04-Aug-2007	kiyohara	branches: 1.15.2; 1.15.18; 1.15.22; 1.15.24; Add support agp_amd64. Imported from FreeBSD.
1.14	06-Mar-2007	xtraeme	branches: 1.14.2; 1.14.10; 1.14.14; Use a mutex rather than lockmgr. Approved by Andrew Doran.
1.13	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.12	17-Aug-2006	christos	branches: 1.12.8; Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.11	16-Jan-2006	christos	branches: 1.11.2; 1.11.6; PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.10	11-Dec-2005	christos	branches: 1.10.2; merge ktrace-lwp.
1.9	27-Feb-2005	perry	branches: 1.9.2; 1.9.4; nuke trailing whitespace
1.8	30-Aug-2004	drochner	branches: 1.8.4; 1.8.6; Phase out the use of a string as first "attach args" member to control which bustype should be attached with a specific call to config_found() (from a "mainbus" or a bus bridge). Do it for isa/eisa/mca and pci/agp for now. These buses all attach to an mi interface attribute "isabus", "eisabus" etc., and the autoconf framework now allows to specify an interface attribute on config_found() and config_search(), which limits the search of matching config data to these which attach to that specific attribute. So we basically have to call config_found_ia(..., "foobus", ...) where such a bus is attached. As a consequence, where a "mainbus" or alike also attaches other devices (eg CPUs) which do not attach to a specific attribute yet, we need at least pass an attribute name (different from "foobus") so that the foo bus is not found at these places. This made some minor changes necessary which are not obviously related to the mentioned buses.
1.7	30-Aug-2004	drochner	add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.6	08-Jul-2003	itojun	function prototype must not have variable name
1.5	01-Feb-2003	thorpej	branches: 1.5.2; Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.4	01-Oct-2001	fvdl	Don't map the aperture into KVA, it isn't needed.
1.3	15-Sep-2001	thorpej	branches: 1.3.2; 1.3.4; Clean up the AGP match/attach code somewhat.
1.2	11-Sep-2001	fvdl	branches: 1.2.2; Fix i810/other intel chipsets attach mishap.
1.1	10-Sep-2001	fvdl	AGP GART support code. Originally written by Doug Rabson for FreeBSD, modifications to fit it into the NetBSD device/config structure and to use bus_dma by me.
1.2.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.2.2.1	11-Sep-2001	thorpej	file agpvar.h was added on branch kqueue on 2001-09-13 01:15:52 +0000
1.3.4.3	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.3.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.3.4.1	15-Sep-2001	fvdl	file agpvar.h was added on branch thorpej-devvp on 2001-10-01 12:45:52 +0000
1.3.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.3.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.3.2.1	15-Sep-2001	nathanw	file agpvar.h was added on branch nathanw_sa on 2001-09-21 22:35:53 +0000
1.5.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.2	03-Sep-2004	skrll	Sync with HEAD
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.9.4.4	27-Feb-2008	yamt	sync with head.
1.9.4.3	03-Sep-2007	yamt	sync with head.
1.9.4.2	30-Dec-2006	yamt	sync with head.
1.9.4.1	21-Jun-2006	yamt	sync with head.
1.9.2.1	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.10.2.1	01-Feb-2006	yamt	sync with head.
1.11.6.1	03-Sep-2006	yamt	sync with head.
1.11.2.1	09-Sep-2006	rpaulo	sync with head
1.12.8.1	12-Mar-2007	rmind	Sync with HEAD.
1.14.14.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.14.14.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.14.14.2	06-Aug-2007	jmcneill	Move agp_i810_power to agp, since it does nothing specific to the i810 hardware. Allow agp front-ends to override the default power handler if specific configuration needs to take place, and expose agp_power via agpvar.h so the front-ends can call back to it to do generic initialization.
1.14.14.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.14.10.1	15-Aug-2007	skrll	Sync with HEAD.
1.14.2.1	20-Aug-2007	ad	Sync with HEAD.
1.15.24.2	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.15.24.1	04-Aug-2007	kiyohara	file agpvar.h was added on branch matt-mips64 on 2007-08-04 09:33:06 +0000
1.15.22.2	29-Jun-2008	mjf	Sync with HEAD.
1.15.22.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.18.1	24-Mar-2008	keiichi	sync with head.
1.15.2.1	23-Mar-2008	matt	sync with HEAD
1.16.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.16.4.2	16-May-2009	yamt	sync with head
1.16.4.1	04-May-2009	yamt	sync with head.
1.16.2.1	17-Jun-2008	yamt	sync with head.
1.17.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.36.1	10-Aug-2014	tls	Rebase.
1.18.22.2	03-Dec-2017	jdolecek	update from HEAD
1.18.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.77	03-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "aic7xxx-use-target-defaults" and "aic7xxx-override-ultra".
1.76	27-Jun-2025	andvar	Grammar and spelling fixes, mainly in comments. A few in documentation, logging, test description, and SCSI ASC/ASCQ assignment descriptions.
1.75	09-Feb-2024	andvar	branches: 1.75.2; fix spelling mistakes, mainly in comments and log messages.
1.74	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.73	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.72	03-Jun-2018	maxv	branches: 1.72.2; Constify ahc_pci_ident_table[] so that it lands in .rodata (1488 bytes).
1.71	14-Jul-2016	msaitoh	branches: 1.71.16; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.70	29-Mar-2014	christos	branches: 1.70.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.69	18-Feb-2014	macallan	if MMIO is allowed and available, prefer it over PIO now my ADPT,2930CU works again
1.68	26-Nov-2009	njoly	branches: 1.68.12; 1.68.22; 1.68.26; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.67	02-Sep-2009	tsutsui	Call pmf_device_register(9) in ahc_attahc() rather than ahc_pci_attach() since pmf_device_deregister(9) is called from ahc_detach() so that cardbus backend also gets proper pmf(9) calls. PCI backend is tested on on O2, but cardbus is untested.
1.66	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.65	05-May-2009	cegger	device_t/softc split
1.64	03-Jan-2009	yamt	branches: 1.64.2; remove extra semicolons.
1.63	22-Feb-2008	dyoung	branches: 1.63.4; 1.63.12; Use device_t and accessors. Use aprint__dev(). Use PMF_FN_.
1.62	28-Jan-2008	macallan	branches: 1.62.2; 1.62.6; get rid of the shutdown hook, add suspend and resume handlers instead XXX: ahc_suspend panics here so suspend calls ahc_shutdown and resume does nothing. This fixes a hangup and subsequent panic which occured when the shutdown hooks were called and /afterwards/ the sd suspend handler tried to flush caches through an already shut down ahc.
1.61	15-Dec-2007	tsutsui	Rename a device property name: "override_ultra" -> "aic7xxx-override-ultra"
1.60	19-Oct-2007	ad	branches: 1.60.2; 1.60.4; 1.60.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.59	26-Sep-2007	macallan	branches: 1.59.2; allow to override the check for an external precision resistor via device property so we can have 40MB/s on SGI O2
1.58	16-Nov-2006	christos	branches: 1.58.8; 1.58.22; 1.58.24; 1.58.26; __unused removal on arguments; approved by core.
1.57	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.56	10-Jul-2006	thorpej	branches: 1.56.4; 1.56.6; Put appropriate prefixes on property names to reduce chances of name collisions.
1.55	05-May-2006	thorpej	branches: 1.55.4; Remove the devprop API and switch everthing over to the new proplib. Add a new device_properties() accessor for device_t that returns the device's property dictionary.
1.54	15-Apr-2006	tsutsui	Add AHC_USETARGETDEFAULTS to ahc_flag, which enables to use default target (but not for host) device settings on ahc(4) interface without its SEEPROM (mostly on-board one). This flag could be useful if host's BIOS (firmware) initializes ahc(4) chip with some conservative (async, no tagged queuing etc.) target settings for minimum support to load its primary loader etc., and it will be enabled in each attachment according to MD device properties(9). Currently this property is set only on sgimips for O2 on-board aic7880 adapters. No objection on tech-kern (and netbsd-bugs), and closes PR kern/23276.
1.53	24-Dec-2005	perry	branches: 1.53.4; 1.53.6; 1.53.8; 1.53.10; 1.53.12; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.52	11-Dec-2005	christos	merge ktrace-lwp.
1.51	28-Jun-2005	thorpej	branches: 1.51.2; Use ANSI function decls and static.
1.50	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.49	27-Feb-2005	perry	nuke trailing whitespace
1.48	04-Feb-2005	perry	de-__P
1.47	17-Oct-2004	christos	branches: 1.47.4; 1.47.6; Bring back ALLOW_MEMIO change that was lost between 1.38 and 1.39. This change reportedly makes the driver work on macppc. There must be a better way of doing this (based on the chip revision?).
1.46	16-Oct-2004	christos	Make a commented out print statement compile and avoid wrapped lines. No functional change.
1.45	16-Mar-2004	simonb	branches: 1.45.2; Fix an assigned-to-but-not-used variable in the non AHC_ALLOW_MEMIO case.
1.44	13-Jan-2004	augustss	Print the device description in the attach routine like all other drivers do.
1.43	18-Aug-2003	taca	make ahc(4) to recognize Adaptec 2915LP Ultra160 SCSI adapter.
1.42	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.41	04-Jun-2003	pk	branches: 1.41.2; ahc_pci_probe: avoid placeholder entries in the `identities' table.
1.40	25-Apr-2003	fvdl	On the first read of the DEVCONFIG PCI config register, sparc64 systems seem to return all 1s. The second read is ok, though. So, as a workaround, read it twice. From Manuel Bouyer.
1.39	19-Apr-2003	fvdl	Newer ahc driver, port from Justin Gibbs' driver in FreeBSD. Ported by Pascal Renauld at Network Storage Solutions, Inc, with some changes by me.
1.38	31-Jan-2003	thorpej	Use aprint_*().
1.37	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.36	20-Jan-2003	simonb	Remove variable that is only assigned too but not referenced.
1.35	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.34	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.33	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.32	21-Nov-2001	wiz	branches: 1.32.10; Presence looks better with a 'c' instead of an 's'.
1.31	13-Nov-2001	lukem	add RCSID
1.30	24-Mar-2001	christos	branches: 1.30.2; There was a stray debugging message "OptionMode = %x\n". Prefix it with the device name, and enable it only in debugging mode.
1.29	18-Jan-2001	jdolecek	branches: 1.29.2; constify
1.28	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.27	24-Sep-2000	jdolecek	use new generic bootverbose instead of local definition
1.26	04-Sep-2000	soren	Fix 'RAID unsupported' printf.
1.25	04-Sep-2000	soren	Add support for the AAA-131 RAID controller by treating it as a 7890. Taken from rev 1.32 of the FreeBSD ahc_pci.c.
1.24	10-May-2000	thorpej	branches: 1.24.4; Support Adaptec U2W cards with a 64-bit PCI memory range (such cards are available with Power Macintosh G3 and G4 systems).
1.23	16-Mar-2000	fvdl	Set bus_intr to ahc_pci_intr.
1.22	15-Mar-2000	fvdl	PCI frontend for new ahc driver.
1.21	26-Jan-2000	thorpej	Remove the FreeBSD support from this file. It hasn't worked in FreeBSD for some time, and only made the code harder to read.
1.20	26-Jan-2000	thorpej	Split the code that reads the SEEPROM into its own file.
1.19	12-Oct-1999	hannken	branches: 1.19.2; Add support for `Adaptec AHA 2940UW Pro'. Behaves like the `Adaptec AHA 2940UW'. Device info: product id = 0x8778, Chip = AIC7880P.
1.18	08-Jun-1998	thorpej	branches: 1.18.14; Nuke __BROKEN_INDIRECT_CONFIG.
1.17	16-Mar-1998	leo	Use bus-dma functions.
1.16	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.15	08-May-1997	thorpej	branches: 1.15.2; 1.15.4; Garbage-collect and uneeded #ifdef.
1.14	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.13	13-Apr-1997	cgd	use pci_map_register(). map both mem and I/O spaces, preferring mem space if it's usable, but falling back to I/O space if mem isn't usable.
1.12	10-Apr-1997	cgd	clean these up (mostly rename bus_space_{tag,handle}_t variables) so that this code makes equal sense for memory and I/O space, prefer to map the PCI front end via memory space (conditionalized on a patchable kernel variable), and do a bit of other random NetBSD-specific cleanup. (These changes were sent to Justin Gibbs on March 28.)
1.11	13-Mar-1997	cgd	clean up the __BROKEN_INDIRECT_CONFIG chunks of the last change
1.10	13-Mar-1997	cgd	fixes from Matt Jacob so that these can compile and run on the Alpha. (aic7xxx has been tested and works on the Alpha, bha has not yet been tested on thne Alpha.)
1.9	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.8	13-Oct-1996	christos	backout kprintf changes
1.7	10-Oct-1996	christos	- move a variable declaration that is unused under NetBSD into the ifdef for FreeBSD - printf -> kprintf, sprintf -> ksprintf
1.6	08-Oct-1996	gibbs	dev/microcode/aic7xxx/aic7xxx.seq, dev/microcode/aic7xxx_seq.h, dev/ic/aic7xxxreg.h: Remove intrinsic knowledge about SDTR and WDTR messages and replace it with a generic message system that allows the kernel driver to handle SDTR, WDTR and any other type of extended message it chooses too. This makes the sequencer code much simpler, makes extended message handling debuggable since the bulk of the work is in the kernel driver, and saves lots of instruction space. Regen microcode header file. dev/ic/aic7xxx.c, dev/ic/aic7xxxvar.h: Add code to handle WDTR and SDTR negotiation in light of the changes in the message interface to the sequencer. Don't reject targets that negotiate async by sending an SDTR with a 0 offset. Use an sdtr message with 0,0 to negotiate async when a target suggests a period that is too long for us to handle. Some tape and cdrom drives don't like us doing the message reject that we did in the past. Fix a problem with handing the QUEUE FULL condition. Fix a race condition (most likely the cause of the SCB paging problems) that might allow the sequencer to get unpaused before the condition that caused it to be paused (a SEQINT) was handled. Race condition pointed out by Doug Ledford <dledford@dialnet.net> and by "Dan Willis" <dan@plutotech.com>. dev/pci/ahc_pci.c: Add support for the 2940AU, an aic7860 based controller. dev/pci/pcidevs.h, dev/pci/pcidevs_data.h: Add product IDs for the 2940AU, aic7860 and aic7855. Regen data file. scsi/scsi_message.h: Add MSG_EXT_SDTR_LEN and MSG_EXT_WDTR_LEN - the length of bytes in these extended messages. Thanks to Chuck Cranor <chuck@maria.wustl.edu> for testing these changes out for me.
1.5	28-Aug-1996	thorpej	Quiet the ahc driver down a bit by only enabling some of less useful messages #if defined(DEBUG) in the NetBSD case.
1.4	14-Jul-1996	cgd	#ifdefs are not allowed in macro invocations.
1.3	10-Jul-1996	explorer	Updates to aic7xxx driver ; from pr port-i386/2600
1.2	20-May-1996	thorpej	branches: 1.2.4; RCS Id police.
1.1	16-May-1996	mycroft	New version, with changes from Justin Gibbs and Noriyuki Soda.
1.2.4.2	04-Mar-1997	mycroft	Pull up latest ahc driver. Fixes several bugs.
1.2.4.1	18-Jul-1996	jtc	Pulled up from rev 1.4, ahc fixes
1.15.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.15.2.1	01-Jul-1997	bouyer	Updates for new scsipi subsystem. Actally known to work on i386 and sparc.
1.18.14.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.19.2.4	27-Mar-2001	bouyer	Sync with HEAD.
1.19.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.19.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.19.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.24.4.2	06-Feb-2002	he	Pull up revision 1.30 (requested by jlam): Move stray debugging message to only occur in debug mode.
1.24.4.1	05-Sep-2000	soren	Pull up revs 1.25-1.26: > Add support for the AAA-131 RAID controller by treating it as a 7890. > Taken from rev 1.32 of the FreeBSD ahc_pci.c. > Fix 'RAID unsupported' printf.
1.29.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.29.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.29.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.29.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.30.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.30.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.32.10.1	23-Aug-2003	tron	Pull up revision 1.43 via patch (requested by taca in ticket #1430): make ahc(4) to recognize Adaptec 2915LP Ultra160 SCSI adapter.
1.41.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.41.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.41.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.41.2.4	19-Oct-2004	skrll	Sync with HEAD
1.41.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.41.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.41.2.1	03-Aug-2004	skrll	Sync with HEAD
1.45.2.1	12-Nov-2004	jmc	Pullup rev 1.46-1.47 (requested by christos in ticket #930) Bring back ALLOW_MEMIO change that was lost between 1.38 and 1.39.
1.47.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.47.6.1	12-Feb-2005	yamt	sync with head.
1.47.4.1	29-Apr-2005	kent	sync with -current
1.51.2.6	27-Feb-2008	yamt	sync with head.
1.51.2.5	04-Feb-2008	yamt	sync with head.
1.51.2.4	21-Jan-2008	yamt	sync with head
1.51.2.3	27-Oct-2007	yamt	sync with head.
1.51.2.2	30-Dec-2006	yamt	sync with head.
1.51.2.1	21-Jun-2006	yamt	sync with head.
1.53.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.53.10.2	11-May-2006	elad	sync with head
1.53.10.1	19-Apr-2006	elad	sync with head.
1.53.8.2	11-Aug-2006	yamt	sync with head
1.53.8.1	24-May-2006	yamt	sync with head.
1.53.6.2	01-Jun-2006	kardel	Sync with head.
1.53.6.1	22-Apr-2006	simonb	Sync with head.
1.53.4.1	09-Sep-2006	rpaulo	sync with head
1.55.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.56.6.2	10-Dec-2006	yamt	sync with head.
1.56.6.1	22-Oct-2006	yamt	sync with head
1.56.4.1	18-Nov-2006	ad	Sync with head.
1.58.26.1	06-Oct-2007	yamt	sync with head.
1.58.24.3	23-Mar-2008	matt	sync with HEAD
1.58.24.2	09-Jan-2008	matt	sync with HEAD
1.58.24.1	06-Nov-2007	matt	sync with HEAD
1.58.22.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.58.22.1	02-Oct-2007	joerg	Sync with HEAD.
1.58.8.2	23-Oct-2007	ad	Sync with head.
1.58.8.1	09-Oct-2007	ad	Sync with head.
1.59.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.60.8.1	02-Jan-2008	bouyer	Sync with HEAD
1.60.4.1	26-Dec-2007	ad	Sync with head.
1.60.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.62.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.62.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.62.2.1	24-Mar-2008	keiichi	sync with head.
1.63.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.63.4.4	11-Mar-2010	yamt	sync with head
1.63.4.3	16-Sep-2009	yamt	sync with head
1.63.4.2	16-May-2009	yamt	sync with head
1.63.4.1	04-May-2009	yamt	sync with head.
1.64.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.68.26.1	18-May-2014	rmind	sync with head
1.68.22.2	03-Dec-2017	jdolecek	update from HEAD
1.68.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.68.12.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.70.6.1	05-Oct-2016	skrll	Sync with HEAD
1.71.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.71.16.1	25-Jun-2018	pgoyette	Sync with HEAD
1.72.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.72.2.1	10-Jun-2019	christos	Sync with HEAD
1.75.2.1	02-Aug-2025	perseant	Sync with HEAD
1.5	12-May-2007	bouyer	Split the ahcisata driver in pci front-end and bus-independant back-end.
1.4	04-Mar-2007	christos	branches: 1.4.2; 1.4.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	03-Jan-2007	bouyer	branches: 1.3.2; 1.3.4; Fix cut'n'paste, don't print version 1.0 when we have a version 1.1 controller. Pointed out by Craig Metz.
1.2	02-Dec-2006	bouyer	branches: 1.2.2; 1.2.4; Don't call the done callback twice if a special action is waiting on the queue. Don't forget to free the xfer in the normal bio path.
1.1	30-Nov-2006	bouyer	branches: 1.1.2; Add ahcisata(4), a driver for AHCI 1.0 and 1.1 controllers. Tested on the sata ports of a Intel 63xxESB chipset. Does not support NCQ yet.
1.1.2.3	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.1.2.2	04-Jan-2007	tron	branches: 1.1.2.2.2; Pull up following revision(s) (requested by bouyer in ticket #329): sys/dev/pci/ahcisata.c: revision 1.3 Fix cut'n'paste, don't print version 1.0 when we have a version 1.1 controller. Pointed out by Craig Metz.
1.1.2.1	03-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #242): sys/dev/pci/ahcisata.c: revision 1.2 Don't call the done callback twice if a special action is waiting on the queue. Don't forget to free the xfer in the normal bio path.
1.1.2.2.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.2.4.4	03-Sep-2007	yamt	sync with head.
1.2.4.3	26-Feb-2007	yamt	sync with head.
1.2.4.2	30-Dec-2006	yamt	sync with head.
1.2.4.1	02-Dec-2006	yamt	file ahcisata.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:41 +0000
1.2.2.2	10-Dec-2006	yamt	sync with head.
1.2.2.1	02-Dec-2006	yamt	file ahcisata.c was added on branch yamt-splraiseipl on 2006-12-10 07:17:41 +0000
1.3.4.2	17-May-2007	yamt	sync with head.
1.3.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.3.2.2	12-Jan-2007	ad	Sync with head.
1.3.2.1	03-Jan-2007	ad	file ahcisata.c was added on branch newlock2 on 2007-01-12 00:57:40 +0000
1.4.4.1	11-Jul-2007	mjf	Sync with head.
1.4.2.2	27-May-2007	ad	Sync with head.
1.4.2.1	05-Apr-2007	ad	Compile fixes.
1.72	20-Apr-2025	andvar	viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad.
1.71	10-Oct-2023	abs	branches: 1.71.6; Add support for ASMEDIA 0x0624 ... an AHCI SATA adaptor which identifies itself as an IDE storage device
1.70	10-Sep-2023	abs	Rework AHCISATA_EXTRA_DELAY for kern/56737 - Remove AHCI_QUIRK_EXTRA_DELAY as issue appears to be drive and not controller related - Replace AHCISATA_EXTRA_DELAY with AHCISATA_REMOVE_EXTRA_DELAY, so defaulting to enabling the extra delay, as the downside of slower probing on systems which do not need it is less than having other systems intermittently fail to probe and attach drives - Also allow disabling extra delay with AHCISATA_EXTRA_DELAY_MS = 0 We should return to this code to work out which of the extra delays are needed, and how long they need to be. It may be that faster systems are more likely to trigger the issue (I've only seen it on a 13th gen i7-13700, though only tested on a limited set) XXX pullup -10
1.69	31-Jul-2023	tnn	ahcisata: ASM106x needs EXTRA_DELAY quirk
1.68	12-Oct-2022	macallan	branches: 1.68.2; set AHCI_QUIRK_EXTRA_DELAY for ASMedia ASM1061 - now it reliably attaches old 1.5GBit/s only disks
1.67	02-Oct-2022	martin	Add extra delay quirk for AMD FCH SATA Controller D.
1.66	02-Oct-2022	martin	regen (AMD FCH SATA D added)
1.65	31-May-2022	andvar	fix various typos in comments, documentation and messages.
1.64	29-May-2022	rin	Apply extra delay quirk for - Intel 7 Series (mobile) SATA Controller (AHCI): rev. 0x04 sometimes fails to probe Seagate ST500LT012. Reported by RVP. Thanks!
1.63	29-May-2022	rin	Apply extra delay quirk for followings: - Intel Braswell AHCI: rev. 0x35 fails for Seagate ST2000LX001. - Intel 8 Series (desktop) SATA Controller (AHCI): rev. 0x04 fails for Seagate ST1000LX015 and WD WD10JPVX, whereas it works without the quirk for some SSD models... Reported by Matthias Petermann. Thanks!
1.62	14-May-2022	rin	Apply extra-delay quirk to "Intel 9 Series SATA Controller (AHCI)"; without the quirk, the controller fails to probe some HDD models, at least "Seagate ST2000DM008". Info and patch provided by Tiago Seco, thanks!
1.61	19-Nov-2021	rin	ahcisata(4): Introduce AHCI_QUIRK_EXTRA_DELAY quirk for devices that need extra delays as done by AHCISATA_EXTRA_DELAY option. Enable this quirk for "C600/X79 AHCI". Also add commented out quirk entries for "Bay Trail SATA (AHCI)" and "Mobile AHCI SATA Controller", for which non-reproducible failures worked around by extra delays have been reported. 500 ms of delays inserted by these option/quirk may be too much. Add AHCISATA_EXTRA_DELAY_MS option to adjust number of delays in ms, like: ---- options AHCISATA_EXTRA_DELAY_MS=200 ---- Thanks prlw1@ and jun@ for testing!
1.60	12-Nov-2021	skrll	Fix the conditional for failed interrupt establishment
1.59	08-Nov-2021	rin	Group quirks by vendors. No functional changes.
1.58	28-Dec-2020	jmcneill	Remove duplicate opt_ahcisata_pci.h include
1.57	18-Jan-2020	simonb	branches: 1.57.6; Add new AHCI_QUIRK_BADNCQ quick for controllers that have issues with NCQ on (some) drives. Enable this quirk for ATI (AMD) SB600/SB700 controllers. Alternate fix for kern/54790 and kern/54855. ok jdolecek@, tested on my SB700 chipset and tsutsui's SB600 chipset.
1.56	18-Oct-2019	tnn	branches: 1.56.2; ahcisata: make sure bus mastering and memory space are actually enabled This makes the "ROCKPro64 PCI-e to Dual SATA-II Interface Card" work.
1.55	27-Jan-2019	pgoyette	branches: 1.55.4; Merge the [pgoyette-compat] branch
1.54	22-Jan-2019	jdolecek	fix pasto; still PR kern/53307
1.53	22-Jan-2019	jdolecek	add BADPMP quirk also for AMD Hudson AHCI SATA Controller, seems in some configurations/BIOS versions the PMP reset works, in others not; it's better to have the common configuration working related to PR kern/53307
1.52	22-Jan-2019	jdolecek	add BADPMP quirk for Intel 82801JI SATA Controller (AHCI mode), which apparently has problem with the PMP reset part of PR kern/53307 - Christos machine
1.51	14-Jan-2019	jdolecek	nothing handles AHCI_QUIRK_BADPMPRESET quirk, remove and change all entries using it (ATI SB600/SB700) to instead use AHCI_QUIRK_BADPMP
1.50	08-Dec-2018	jdolecek	support multi-vector MSI-X tested and confirmed working by Masanobu SAITOH on C3000
1.49	04-Dec-2018	jdolecek	disable MSI-X for time being since it's not working; allow only MSI and INTx which both are confirmed working
1.48	30-Nov-2018	jdolecek	simplify intr establish code - rely on pci_intr_alloc() to allow also MSI-X, and to return interrupt types which are possible for pci_intr_establish(); remove fallbacks to retry with MSI/MSI-X explicitly disabled discussed on tech-kern@ https://mail-index.netbsd.org/tech-kern/2018/11/27/msg024240.html
1.47	26-Nov-2018	jdolecek	add AHCI_PCI_QUIRK_FORCE for apu2 to be recognized; patch by Mike Pumford on current-users@
1.46	24-Nov-2018	skrll	Handle Cavium AHCI/RAID controller BAR weirdness. Support 64bit BARs as well.
1.45	20-Nov-2018	prlw1	Trivial build fix
1.44	20-Nov-2018	skrll	Add support for MSI/MSI-X to ahcisata at pci. The options AHCISATA_DISABLE_MSI and AHCISATA_DISABLE_MSIX are available if required.
1.43	18-Nov-2018	skrll	Trailing whitespace
1.42	25-Oct-2018	jdolecek	release intr if pci_intr_establish_xname() fails
1.41	24-Oct-2018	jdolecek	detach the controller itself on shutdown; adjust to not detach already detached atabus/channel
1.40	22-Oct-2018	jdolecek	first disestablish interrupt, then release
1.39	22-Oct-2018	jdolecek	enable MSI support where available
1.38	13-Oct-2016	jdolecek	branches: 1.38.8; 1.38.14; 1.38.16; provide intr xname
1.37	23-Aug-2016	msaitoh	Sync with pcidevs rev. 1.1263. No functional change.
1.36	29-Mar-2014	christos	branches: 1.36.6; 1.36.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.35	21-Jan-2014	msaitoh	Rename 88SE91XX.
1.34	10-Jan-2014	msaitoh	Sync with pcidev's change. Rename PCI_PRODUCT_MARVELL2_88SE9128 to PCI_PRODUCT_MARVELL2_88SE912Z. No functional change.
1.33	09-Jan-2014	christos	revert unintended changes
1.32	09-Jan-2014	htodd	Fix build. Only compile tested.
1.31	21-Oct-2012	matt	branches: 1.31.2; Add FORCE quirks for ASMEDIA ASM1061
1.30	20-Aug-2012	bouyer	branches: 1.30.2; Fix typo, pointed out by Markus W Kilbinger
1.29	10-Aug-2012	bouyer	Remove leftover comment.
1.28	10-Aug-2012	bouyer	Work around some SATA PMP issues in some AHCI controllers by either disabling PMP entirely, or special handling in the reset function. Controller list from linux and FreeBSD.
1.27	30-Jan-2012	drochner	branches: 1.27.2; 1.27.4; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.26	06-Aug-2011	jakllsch	branches: 1.26.2; 1.26.6; Marvell 88SE9128 needs AHCI_PCI_QUIRK_FORCE.
1.25	27-May-2011	matt	Add quirk for ALI M5228
1.24	03-Apr-2011	dyoung	Deregister in ahci_pci_detach() the PMF hooks registered in ahci_pci_attach().
1.23	13-Nov-2010	uebayasi	branches: 1.23.2; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.22	07-Aug-2010	jmcneill	Add NVIDIA MCP77 to the quirk table, so ahcisata will attach when the BIOS is set to 'RAID' mode.
1.21	27-Jul-2010	jakllsch	Use use 64-bit DMA tag (where available and supported) for ahcisata. Avoid the possible ATI SB600 64-bit DMA problem.
1.20	27-Jul-2010	jakllsch	Support detachment of ahcisata(4). Use use 64-bit DMA tag (where available) for ahcisata(4) at jmide(4). Beginnings of detach/resume support for jmide(4). Sprinkle static. Misc. little changes.
1.19	24-Feb-2010	dyoung	branches: 1.19.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.18	08-Jan-2010	dyoung	branches: 1.18.2; Expand PMF_FN_* macros.
1.17	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.16	19-Aug-2009	pgoyette	Use __arraycount() for readability
1.15	11-Jun-2009	cegger	Attach Marvel 88SE6121 SATA II controller on ahci. W/o this, it attaches on pciide.
1.14	27-Mar-2009	dholland	Add the other three PCI ids for the nvidia MCP65 SATA controller to the quirk table. This way they attach as ahcisata and work, instead of attaching as viaide and failing miserably. Fixes PR kern/37826. XXX: There are a bunch of additional MCP67 and MCP73 PCI ids that XXX: should probably be added also, but I don't want to do that without XXX: further investigation and testing.
1.13	08-Dec-2008	tron	branches: 1.13.2; Force the nVidia nForce 630i MCP73 to attach to ahcisata(4) as well. It would otherwise attach to pciide(4) and use PIO mode. Patch submitted by Bernhard Moellemann in PR kern/40134.
1.12	01-Aug-2008	dillo	branches: 1.12.2; 1.12.4; Force ahcisata to attach to NVIDIA MCP65 and MCP67, as they don't work with viaide. Reviewed by bouyer.
1.11	20-Mar-2008	cube	branches: 1.11.4; 1.11.6; 1.11.10; Fix previous. Sigh
1.10	20-Mar-2008	cube	Don't forget to se the device_t back pointer in the softc.
1.9	18-Mar-2008	cube	Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.8	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.7	23-Feb-2008	mlelstv	make this less verbose
1.6	23-Feb-2008	mlelstv	Always attach AHCI_SATA devices, even when not configured for AHCI mode. Fixes PR kern/38081.
1.5	19-Feb-2008	jnemeth	branches: 1.5.2; 1.5.6; KNF
1.4	19-Feb-2008	jnemeth	PR/38056 - Simon Burge -- In ahci_pci_match, return the correct value so that we only match SATA RAID controllers in AHCI mode. Tested by: simonb@ Approved by: bouyer@
1.3	11-Feb-2008	xtraeme	Attach ahcisata(4) to any RAID controller that supports AHCI, and make it pass the ATAC_CAP_RAID capability for them. ok'ed by bouyer@.
1.2	09-Dec-2007	jmcneill	Merge jmcneill-pm branch.
1.1	12-May-2007	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.12; 1.1.14; 1.1.16; 1.1.18; 1.1.20; 1.1.26; 1.1.28; 1.1.30; Split the ahcisata driver in pci front-end and bus-independant back-end.
1.1.30.1	11-Dec-2007	yamt	sync with head.
1.1.28.1	26-Dec-2007	ad	Sync with head.
1.1.26.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.20.7	24-Mar-2008	yamt	sync with head.
1.1.20.6	17-Mar-2008	yamt	sync with head.
1.1.20.5	27-Feb-2008	yamt	sync with head.
1.1.20.4	11-Feb-2008	yamt	sync with head.
1.1.20.3	21-Jan-2008	yamt	sync with head
1.1.20.2	03-Sep-2007	yamt	sync with head.
1.1.20.1	12-May-2007	yamt	file ahcisata_pci.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:20 +0000
1.1.18.2	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.1.18.1	12-May-2007	wrstuden	file ahcisata_pci.c was added on branch wrstuden-fixsa on 2007-09-03 07:04:27 +0000
1.1.16.2	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.1.16.1	12-May-2007	pavel	file ahcisata_pci.c was added on branch netbsd-4 on 2007-08-31 20:09:25 +0000
1.1.14.2	23-Mar-2008	matt	sync with HEAD
1.1.14.1	09-Jan-2008	matt	sync with HEAD
1.1.12.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.1.12.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.1.12.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.1.12.1	04-Aug-2007	he	Factor out the hardware initialization code in ahcisata_core.c, and provide them to the pci frontend, which has now grown its own softc and a power management handler. There may be one too many re-initializations done on resume (witness the multiple kernel messages from ahcisata), but at least this is sufficient to get the Lenovo T60 to come out of resume with a working disk controller and disk.
1.1.6.2	11-Jul-2007	mjf	Sync with head.
1.1.6.1	12-May-2007	mjf	file ahcisata_pci.c was added on branch mjf-ufs-trans on 2007-07-11 20:06:58 +0000
1.1.4.2	09-Jun-2007	ad	Sync with head.
1.1.4.1	12-May-2007	ad	file ahcisata_pci.c was added on branch vmlocking on 2007-06-09 21:37:23 +0000
1.1.2.2	17-May-2007	yamt	sync with head.
1.1.2.1	12-May-2007	yamt	file ahcisata_pci.c was added on branch yamt-idlelwp on 2007-05-17 13:41:29 +0000
1.5.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.5.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.5.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.5.2.1	24-Mar-2008	keiichi	sync with head.
1.11.10.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.11.10.1	19-Oct-2008	haad	Sync with HEAD.
1.11.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.11.4.5	11-Aug-2010	yamt	sync with head.
1.11.4.4	11-Mar-2010	yamt	sync with head
1.11.4.3	16-Sep-2009	yamt	sync with head
1.11.4.2	20-Jun-2009	yamt	sync with head
1.11.4.1	04-May-2009	yamt	sync with head.
1.12.4.3	21-Nov-2010	riz	branches: 1.12.4.3.2; Pull up following revision(s) (requested by jmcneill in ticket #1437): sys/dev/pci/ahcisata_pci.c: revision 1.22 Add NVIDIA MCP77 to the quirk table, so ahcisata will attach when the BIOS is set to 'RAID' mode.
1.12.4.2	30-Mar-2009	snj	branches: 1.12.4.2.4; Pull up following revision(s) (requested by dholland in ticket #616): sys/dev/pci/ahcisata_pci.c: revision 1.14 Add the other three PCI ids for the nvidia MCP65 SATA controller to the quirk table. This way they attach as ahcisata and work, instead of attaching as viaide and failing miserably. Fixes PR kern/37826. XXX: There are a bunch of additional MCP67 and MCP73 PCI ids that XXX: should probably be added also, but I don't want to do that without XXX: further investigation and testing.
1.12.4.1	10-Dec-2008	snj	Pull up following revision(s) (requested by tron in ticket #170): sys/dev/pci/ahcisata_pci.c: revision 1.13 Force the nVidia nForce 630i MCP73 to attach to ahcisata(4) as well. It would otherwise attach to pciide(4) and use PIO mode. Patch submitted by Bernhard Moellemann in PR kern/40134.
1.12.4.3.2.1	14-Oct-2011	matt	revision 1.25 date: 2011/05/27 15:11:59; author: matt; state: Exp; lines: +4 -2 Add quirk for ALI M5228
1.12.4.2.4.2	26-Mar-2014	matt	Use ahci_resume
1.12.4.2.4.1	05-Nov-2013	matt	Support a 64-bit BAR for AHCI
1.12.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.12.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.13.2.2	23-Jul-2009	jym	Sync with HEAD.
1.13.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.18.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.19.2.3	31-May-2011	rmind	sync with head
1.19.2.2	21-Apr-2011	rmind	sync with head
1.19.2.1	05-Mar-2011	rmind	sync with head
1.23.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.26.6.1	18-Feb-2012	mrg	merge to -current.
1.26.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.26.2.2	30-Oct-2012	yamt	sync with head
1.26.2.1	17-Apr-2012	yamt	sync with head
1.27.4.1	18-Nov-2012	matt	Merge tickets 658 and 659.
1.27.2.1	18-Nov-2012	msaitoh	Pull up following revision(s) (requested by matt in ticket #659): sys/dev/pci/ahcisata_pci.c: revision 1.31 Add FORCE quirks for ASMEDIA ASM1061
1.30.2.4	03-Dec-2017	jdolecek	update from HEAD
1.30.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.30.2.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.30.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.31.2.1	18-May-2014	rmind	sync with head
1.36.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.36.6.2	05-Dec-2016	skrll	Sync with HEAD
1.36.6.1	05-Oct-2016	skrll	Sync with HEAD
1.38.16.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.38.16.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.38.16.1	10-Jun-2019	christos	Sync with HEAD
1.38.14.4	26-Jan-2019	pgoyette	Sync with HEAD
1.38.14.3	18-Jan-2019	pgoyette	Synch with HEAD
1.38.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.38.14.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.38.8.2	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #1911): sys/dev/pci/ahcisata_pci.c: revision 1.71 Add support for ASMEDIA 0x0624 ... an AHCI SATA adaptor which identifies itself as an IDE storage device
1.38.8.1	18-Dec-2018	martin	Pull up following revision(s) (requested by is in ticket #1139): sys/dev/pci/ahcisata_pci.c: revision 1.47 add AHCI_PCI_QUIRK_FORCE for apu2 to be recognized; patch by Mike Pumford on current-users@
1.55.4.4	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1944): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.55.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #1748): sys/dev/pci/ahcisata_pci.c: revision 1.71 Add support for ASMEDIA 0x0624 ... an AHCI SATA adaptor which identifies itself as an IDE storage device
1.55.4.2	21-Jan-2020	martin	Pull up following revision(s) (requested by simonb in ticket #630): sys/dev/ic/ahcisatavar.h: revision 1.24 sys/dev/pci/ahcisata_pci.c: revision 1.57 sys/dev/ic/ahcisata_core.c: revision 1.81 Add new AHCI_QUIRK_BADNCQ quick for controllers that have issues with NCQ on (some) drives. Enable this quirk for ATI (AMD) SB600/SB700 controllers. Alternate fix for kern/54790 and kern/54855. ok jdolecek@, tested on my SB700 chipset and tsutsui's SB600 chipset.
1.55.4.1	23-Oct-2019	martin	Pull up following revision(s) (requested by tnn in ticket #358): sys/dev/pci/ahcisata_pci.c: revision 1.56 ahcisata: make sure bus mastering and memory space are actually enabled This makes the "ROCKPro64 PCI-e to Dual SATA-II Interface Card" work.
1.56.2.1	25-Jan-2020	ad	Sync with head.
1.57.6.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.68.2.4	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1101): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.68.2.3	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #403): sys/dev/pci/ahcisata_pci.c: revision 1.71 Add support for ASMEDIA 0x0624 ... an AHCI SATA adaptor which identifies itself as an IDE storage device
1.68.2.2	11-Sep-2023	martin	Pull up following revision(s) (requested by abs in ticket #366): sys/dev/pci/ahcisata_pci.c: revision 1.70 sys/dev/ic/ahcisata_core.c: revision 1.108 sys/dev/ic/ahcisatavar.h: revision 1.28 sys/conf/files: revision 1.1309 Rework AHCISATA_EXTRA_DELAY for kern/56737 - Remove AHCI_QUIRK_EXTRA_DELAY as issue appears to be drive and not controller related - Replace AHCISATA_EXTRA_DELAY with AHCISATA_REMOVE_EXTRA_DELAY, so defaulting to enabling the extra delay, as the downside of slower probing on systems which do not need it is less than having other systems intermittently fail to probe and attach drives - Also allow disabling extra delay with AHCISATA_EXTRA_DELAY_MS = 0 We should return to this code to work out which of the extra delays are needed, and how long they need to be. It may be that faster systems are more likely to trigger the issue (I've only seen it on a 13th gen i7-13700, though only tested on a limited set)
1.68.2.1	22-Aug-2023	martin	Pull up following revision(s) (requested by tnn in ticket #334): sys/dev/pci/ahcisata_pci.c: revision 1.69 ahcisata: ASM106x needs EXTRA_DELAY quirk
1.71.6.1	02-Aug-2025	perseant	Sync with HEAD
1.2	12-May-2007	bouyer	Split the ahcisata driver in pci front-end and bus-independant back-end.
1.1	30-Nov-2006	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.14; 1.1.16; 1.1.22; Add ahcisata(4), a driver for AHCI 1.0 and 1.1 controllers. Tested on the sata ports of a Intel 63xxESB chipset. Does not support NCQ yet.
1.1.22.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.1.16.1	11-Jul-2007	mjf	Sync with head.
1.1.14.2	09-Jun-2007	ad	Sync with head.
1.1.14.1	27-May-2007	ad	Sync with head.
1.1.10.1	17-May-2007	yamt	sync with head.
1.1.8.2	12-Jan-2007	ad	Sync with head.
1.1.8.1	30-Nov-2006	ad	file ahcisatareg.h was added on branch newlock2 on 2007-01-12 00:57:40 +0000
1.1.6.3	03-Sep-2007	yamt	sync with head.
1.1.6.2	30-Dec-2006	yamt	sync with head.
1.1.6.1	30-Nov-2006	yamt	file ahcisatareg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:41 +0000
1.1.4.2	10-Dec-2006	yamt	sync with head.
1.1.4.1	30-Nov-2006	yamt	file ahcisatareg.h was added on branch yamt-splraiseipl on 2006-12-10 07:17:41 +0000
1.1.2.1	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.40	28-May-2022	andvar	fix various typos, mainly in comments.
1.39	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.38	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.37	23-Jun-2018	maxv	branches: 1.37.2; constify
1.36	15-Jan-2018	maya	branches: 1.36.2; malloc+memset zero to malloc(, \|M_ZERO); ok rkujawa
1.35	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.34	21-Apr-2013	msaitoh	branches: 1.34.4; Delete "PCI_" from PCIX and PICE capability registers.
1.33	30-Dec-2011	christos	branches: 1.33.6; PR/45760: Henning Petersen: Add Adaptec29329LPE to ahd_pci.c
1.32	21-Apr-2010	dyoung	branches: 1.32.8; 1.32.12; Don't printf bus_space_tag_t's in a MI driver.
1.31	26-Sep-2009	tsutsui	branches: 1.31.2; 1.31.4; Split device_t/softc. Tested on aic7901A.
1.30	05-Sep-2009	tsutsui	Use device_t self arg to get a device name rather than through &ahd->sc_dev which needs to be changed on future device_t/softc split.
1.29	05-Sep-2009	tsutsui	Wrap long lines.
1.28	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.27	21-Mar-2008	dyoung	branches: 1.27.4; 1.27.18; Use aprint_*_dev(), and device_xname(). Constify a bit.
1.26	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.25	16-Nov-2006	christos	branches: 1.25.24; 1.25.44; 1.25.48; __unused removal on arguments; approved by core.
1.24	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.23	17-Jun-2006	christos	branches: 1.23.4; 1.23.6; re-factor the pci powestate api. reviewed by gimpy
1.22	08-Mar-2006	lukem	branches: 1.22.4; 1.22.6; Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.21	24-Dec-2005	perry	branches: 1.21.4; 1.21.6; 1.21.8; 1.21.10; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.20	11-Dec-2005	christos	merge ktrace-lwp.
1.19	28-Jun-2005	thorpej	branches: 1.19.2; Use ANSI function decls and static.
1.18	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.17	25-Apr-2005	bad	Recognise the Dell OEM version of the ASR-29320B as found in recent e.g. Dell SC1420 boxen. This adapter can't be put out of HostRaid mode, so recognise it in that mode regardless.
1.16	25-Apr-2005	bad	Bring in some changes from rev. 1.17 of FreeBSD aic79xx_pci.c: Fix the _IROC_ masks so that the correct bits get masked out for the HostRaid features. The 29320 and 29320B really are 7902 based, not 7901A based. Use the approriate chip setup routine. Recognise generic 7901 based boards by matching against ID_9005_GENERIC_MASK.
1.15	27-Feb-2005	perry	branches: 1.15.2; nuke trailing whitespace
1.14	04-Feb-2005	perry	de-__P
1.13	24-Feb-2004	wiz	branches: 1.13.6; 1.13.8; 1.13.10; parameter with two es. From Peter Postma.
1.12	18-Nov-2003	briggs	Use aprint_* instead of printf.
1.11	10-Oct-2003	fvdl	Add a missing chip reset in attach. Makes things work for an on-board 7902B on a SuperMicro board.
1.10	10-Oct-2003	fvdl	Set some termination defaults correctly. Test register access like in the original FreeBSD driver.
1.9	10-Oct-2003	thorpej	Fix some errors in device mapping and errata work-arounds, and generally clean up mapping code.
1.8	09-Oct-2003	fvdl	Don't treat a 66-100Mhz PCI-X bus as plain PCI. From yamt (the code was already like this in the FreeBSD driver).
1.7	02-Sep-2003	fvdl	Make this less chatty on probe/attach, to match the output of other controllers.
1.6	29-Aug-2003	thorpej	Apply the following change from 06/28/2003 04:39:49 by gibbs to the FreeBSD ahd driver: aic79xx.h: Add a 7901A specific feature definition. aic79xx_pci.c: Split out the general aic790X setup into it's own setup handler that works on single and dual controllers. Adjust all other PCI setup handlers to initialize the chips basic features and type before calling the generic handler. Turn off a few Rev B workarounds that are not required on the 7901B.
1.5	29-Aug-2003	thorpej	Apply the following change from 5/26/2003 21:15:52 made by gibbs to the FreeBSD ahd driver: Add 7901B support. Sort IDs based on chip type. Remove IROC IDs. We'll switch to using the IROC masks if/when we want to start attaching to IROC controllers. Approved by: RE
1.4	29-Aug-2003	thorpej	Apply the following change checked in 2003/05/04 00:20:07 by gibbs to the FreeBSD ahd driver: Correct spelling errors. Switch to handling bad SCSI status as a sequencer interrupt instead of having the kernel proccess these failures via the completion queue. This is done because: o The old scheme required us to pause the sequencer and clear critical sections for each SCB. It seems that these pause actions, if coincident with a sequencer FIFO interrupt, would result in a FIFO interrupt getting lost or directing to the wrong FIFO. This caused hangs when the driver was stressed under high "queue full" loads. o The completion code assumed that it was always called with the sequencer running. This may not be the case in timeout processing where completions occur manually via ahd_pause_and_flushwork(). o With this scheme, the extra expense of clearing critical sections is avoided since the sequencer will only self pause once all pending selections have cleared and it is not in a critical section. aic79xx.c Add code to handle the new BAD_SCB_STATUS sequencer interrupt code. This just redirects the SCB through the already existing ahd_complete_scb() code path. Remove code in ahd_handle_scsi_status() that paused the sequencer, made sure that no selections where pending, and cleared critical sections. Bad status SCBs are now only processed when all of these conditions are true. aic79xx.reg: Add the BAD_SCB_STATUS sequencer interrupt code. aic79xx.seq: When completing an SCB upload to the host, if we are doing this because the SCB contains non-zero SCSI status, defer completing the SCB until there are no pending selection events. When completing these SCBs, use the new BAD_SCB_STATUS sequencer interrupt. For all other uploaded SCBs (currently only for underruns), the SCB is completed via the normal done queue. Additionally, keep the SCB that is currently being uploaded on the COMPLETE_DMA_SCB list until the dma is completed, not just until the DMA is started. This ensures that the DMA is restarted properly should the host disable the DMA transfer for some reason. In our RevA workaround for Maxtor drives, guard against the host pausing us while trying to pause I/O until the first data-valid REQ by clearing the current snapshot so that we can tell if the transfer has completed prior to us noticing the REQINIT status. In cfg4data_intr, shave off an instruction before getting the data path running by adding an entrypoint to the overrun handler to also increment the FIFO use count. In the overrun handler, be sure to clear our LONGJMP address in both exit paths. Perform a few sequencer optimizations. aic79xx.c: Print the full path from the SCB when a packetized status overrun occurs. Remove references to LONGJMP_SCB which is being removed from firmware usage. Print the new SCB_FIFO_USE_COUNT field in the per-SCB section of ahd_dump_card_state(). The SCB_TAG field is now re-used by the sequencer, so it no longer makes sense to reference this field in the kernel driver. aic79xx.h: Re-arrange fields in the hardware SCB from largest size type to smallest. This makes it easier to move fields without changing field alignment. The hardware scb tag field is now down near the "spare" portion of the SCB to facilitate reuse by the sequencer. aic79xx.reg: Remove LONGJMP_ADDR. Rearrange SCB fields to match aic79xx.h. Add SCB_FIFO_USE_COUNT as the first byte of the SCB_TAG field. aic79xx.seq: Add a per-SCB "Fifos in use count" field and use it to determine when it is safe (all data posted) to deliver status back to the host. The old method involved polling one or both FIFOs to verify that the current task did not have pending data. This makes running down the GSFIFO very cheap, so we will empty the GSFIFO in one idle loop pass in all cases. Use this simplification of the completion process to prune down the data FIFO teardown sequencer for packetized transfers. Much more code is now shared between the data residual and transfer complete cases. Correct some issues in the packetized status handler. It used to be possible to CLRCHN our FIFO before status had fully transferred to the host. We also failed to handle NONPACKREQ phases that could occur should a CRC error occur during transmission of the status data packet. Correct a few big endian issues: aic79xx.c: aic79xx_inline.h: aic79xx_pci.c: aic79xx_osm.c: o Always get the SCB's tag via the SCB_GET_TAG acccessor o Add missing use of byte swapping macros when touching hscb fields. o Don't double swap SEEPROM data when it is printed. Correct a big-endian bug. We cannot assign a o When assigning a 32bit LE variable to a 64bit LE variable, we must be explict about how the words of the 64bit LE variable are initialized. Cast to (uint32_t*) to do this. aic79xx.c: In ahd_clear_critical_section(), hit CRLSCSIINT after restoring the interrupt masks to avoid what appears to be a glitch on SCSIINT. Any real SCSIINT status will be persistent and will immidiately reset SCSIINT. This clear should only get rid of spurious SCSIINTs. This glitch was the cause of the "Unexpected PKT busfree" status that occurred under high queue full loads Call ahd_fini_scbdata() after shutdown so that any ahd_chip_init() routine that might access SCB data will not access free'd memory. Reset the bus on an IOERR since the chip doesn't seem to reset to the new voltage level without this. Change offset calculation for scatter gather maps so that the calculation is correct if an integral multiple of sg lists does not fit in the allocation size. Adjust bus dma tag for data buffers based on 39BIT addressing flag in our softc. Use the QFREEZE count to simplify ahd_pause_and_flushworkd(). We can thus rely on the sequencer eventually clearing ENSELO. In ahd_abort_scbs(), fix a bug that could potentially corrupt sequencer state. The saved SCB was being restored in the SCSI mode instead of the saved mode. It turns out that the SCB did not need to be saved at all as the scbptr is already restored by all subroutines called during this function that modify that register. aic79xx.c: aic79xx.h: aic79xx_pci.c: Add support for parsing the seeprom vital product data. The VPD data are currently unused. aic79xx.h: aic79xx.seq: aic79xx_pci.c: Add a firmware workaround to make the LED blink brighter during packetized operations on the H2A. aic79xx_inline.h: The host does not use timer interrupts, so don't gate our decision on whether or not to unpause the sequencer on whether or not a timer interrupt is pending.
1.3	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.2	14-May-2003	wiz	branches: 1.2.2; Correct spelling of coalesce. (Will forward to Justin Gibbs.)
1.1	21-Apr-2003	fvdl	ahd, driver for Adaptec 790x U320 controllers. Originally by Justin Gibbs for FreeBSD, ported to NetBSD by Pascal Renauld, Network Storage Solutions, Inc, plus a bunch of changes by me. This driver is as yet untested in this final form, it will be added to config files when it has been tested.
1.2.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.2.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.1	03-Aug-2004	skrll	Sync with HEAD
1.13.10.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.10.1	12-Feb-2005	yamt	sync with head.
1.13.8.1	29-Apr-2005	kent	sync with -current
1.13.6.1	13-May-2005	riz	Pull up revisions 1.16, 1.17 (requested by bad in ticket #1480): Bring in some changes from rev. 1.17 of FreeBSD aic79xx_pci.c: Fix the _IROC_ masks so that the correct bits get masked out for the HostRaid features. The 29320 and 29320B really are 7902 based, not 7901A based. Use the approriate chip setup routine. Recognise generic 7901 based boards by matching against ID_9005_GENERIC_MASK. Recognise the Dell OEM version of the ASR-29320B as found in recent e.g. Dell SC1420 boxen. This adapter can't be put out of HostRaid mode, so recognise it in that mode regardless.
1.15.2.2	28-Apr-2005	tron	Pull up revision 1.17 (requested by bad in ticket #204): Recognise the Dell OEM version of the ASR-29320B as found in recent e.g. Dell SC1420 boxen. This adapter can't be put out of HostRaid mode, so recognise it in that mode regardless.
1.15.2.1	28-Apr-2005	tron	Pull up revision 1.16 (requested by bad in ticket #204): Bring in some changes from rev. 1.17 of FreeBSD aic79xx_pci.c: Fix the _IROC_ masks so that the correct bits get masked out for the HostRaid features. The 29320 and 29320B really are 7902 based, not 7901A based. Use the approriate chip setup routine. Recognise generic 7901 based boards by matching against ID_9005_GENERIC_MASK.
1.19.2.3	24-Mar-2008	yamt	sync with head.
1.19.2.2	30-Dec-2006	yamt	sync with head.
1.19.2.1	21-Jun-2006	yamt	sync with head.
1.21.10.1	19-Apr-2006	elad	sync with head.
1.21.8.2	26-Jun-2006	yamt	sync with head.
1.21.8.1	13-Mar-2006	yamt	sync with head.
1.21.6.1	22-Apr-2006	simonb	Sync with head.
1.21.4.1	09-Sep-2006	rpaulo	sync with head
1.22.6.1	13-Jul-2006	gdamore	Merge from HEAD.
1.22.4.1	19-Jun-2006	chap	Sync with head.
1.23.6.2	10-Dec-2006	yamt	sync with head.
1.23.6.1	22-Oct-2006	yamt	sync with head
1.23.4.1	18-Nov-2006	ad	Sync with head.
1.25.48.1	03-Apr-2008	mjf	Sync with HEAD.
1.25.44.1	24-Mar-2008	keiichi	sync with head.
1.25.24.1	23-Mar-2008	matt	sync with HEAD
1.27.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.27.4.4	11-Aug-2010	yamt	sync with head.
1.27.4.3	11-Mar-2010	yamt	sync with head
1.27.4.2	16-Sep-2009	yamt	sync with head
1.27.4.1	16-May-2009	yamt	sync with head
1.31.4.1	30-May-2010	rmind	sync with head
1.31.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.32.12.1	18-Feb-2012	mrg	merge to -current.
1.32.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.32.8.1	17-Apr-2012	yamt	sync with head
1.33.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.33.6.1	23-Jun-2013	tls	resync from head
1.34.4.1	18-May-2014	rmind	sync with head
1.36.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.36.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.37.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.37.2.1	10-Jun-2019	christos	Sync with HEAD
1.10	16-May-1996	mycroft	New version, with changes from Justin Gibbs and Noriyuki Soda.
1.9	13-May-1996	mycroft	Use intr.h.
1.8	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.7	04-Mar-1996	cgd	reorganize mapping register definitions
1.6	31-Jan-1996	cgd	update for new version of 'pcidevs'
1.5	28-Dec-1995	thorpej	Squish a couple of rogue old-style RCS ids.
1.4	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.3	12-Nov-1995	christos	Don't use the softc at probe time on pci devices because it has not been allocated yet. Thanks Charles...
1.2	10-Nov-1995	christos	- pcidevs: add 3c590 and adaptec ultra - aic7xxx.c: fix the probe routines to use pcidevs.h
1.1	09-Oct-1995	mycroft	Add an Adaptec 2940 driver, by John Aycock and Justin Gibbs, ported to NetBSD by Stefan Grefen, with several bug fixes by me.
1.14	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.13	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.12	24-Apr-2021	thorpej	branches: 1.12.2; 1.12.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.11	22-Dec-2019	thorpej	branches: 1.11.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.10	18-Mar-2012	martin	branches: 1.10.40; Split device_t
1.9	11-Mar-2012	pgoyette	Since there are no RW_READERs for the device's rw_lock, convert it to a mutex and reduce the overhead. Also, even if we're polling, we need to use the lock to control access to the bus, otherwise we get anomolous results. (These same changes were made several weeks ago to other drivers; I missed this one.)
1.8	10-Feb-2011	hannken	branches: 1.8.4; 1.8.8; Make the SMbus Send/Receive functions work. If cmdlen = 0 and len = 1 use the XXX_HC_CMD_BYTE protocol, send data through XXX_SMB_HCMD or receive data from XXX_SMB_HD0. Ok: Paul Goyette
1.7	11-Mar-2010	mrg	branches: 1.7.2; 1.7.4; 1.7.6; various aprint_* fixes.
1.6	12-May-2009	cegger	branches: 1.6.2; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.5	11-May-2009	cegger	use device_xname()
1.4	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.3	18-Mar-2009	cegger	branches: 1.3.2; bzero -> memset
1.2	03-Feb-2009	pgoyette	branches: 1.2.2; Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.1	29-Oct-2008	jkunz	branches: 1.1.6; 1.1.8; 1.1.10; Ported alipm(4) and admtemp(4) from OpenBSD.
1.1.10.4	28-Apr-2009	skrll	Sync with HEAD.
1.1.10.3	03-Mar-2009	skrll	Sync with HEAD.
1.1.10.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.10.1	29-Oct-2008	skrll	file alipm.c was added on branch nick-hppapmap on 2009-01-19 13:18:24 +0000
1.1.8.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.8.1	29-Oct-2008	mjf	file alipm.c was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.1.6.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.1.6.1	29-Oct-2008	haad	file alipm.c was added on branch haad-dm on 2008-12-13 01:14:34 +0000
1.2.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.3.2.4	11-Aug-2010	yamt	sync with head.
1.3.2.3	16-May-2009	yamt	sync with head
1.3.2.2	04-May-2009	yamt	sync with head.
1.3.2.1	18-Mar-2009	yamt	file alipm.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:54 +0000
1.6.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.7.6.1	17-Feb-2011	bouyer	Sync with HEAD
1.7.4.1	06-Jun-2011	jruoho	Sync with HEAD.
1.7.2.1	05-Mar-2011	rmind	sync with head
1.8.8.1	05-Apr-2012	mrg	sync to latest -current.
1.8.4.1	17-Apr-2012	yamt	sync with head
1.10.40.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.11.10.1	23-Mar-2021	thorpej	Convert config_found_ia() call sites where the device only carries a single interface attribute to bare config_found() calls.
1.12.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.12.2.1	14-May-2021	thorpej	Pass along our devhandle to the iic bus instance.
1.4	18-Dec-2022	reinoud	Add amdsmn(4) and amdccp(4) power management stubs.
1.3	26-Oct-2022	msaitoh	branches: 1.3.2; Add AMD F17/1x and F17/9x CCP(PSP) support to amdccp(4).
1.2	25-Jun-2020	thorpej	Include all necessary letters in the word "Cryptographic".
1.1	24-Jun-2020	thorpej	Add a PCI front-end for the "amdccp" (AMD Cryptographic Coprocessor) driver.
1.3.2.1	19-Dec-2022	martin	Pull up following revision(s) (requested by reinoud in ticket #3): sys/dev/pci/amdccp_pci.c: revision 1.4 sys/arch/x86/pci/amdsmn.c: revision 1.15 sys/dev/acpi/amdccp_acpi.c: revision 1.6 sys/dev/fdt/amdccp_fdt.c: revision 1.7 Add amdsmn(4) and amdccp(4) power management stubs.
1.43	19-Mar-2022	riastradh	rnd(9): Adjust IPL of locks used by rndsource callbacks. These no longer ever run from hard interrupt context or with a spin lock held, so there is no longer any need to have them at IPL_VM to block hard interrupts. Instead, lower them to IPL_SOFTSERIAL.
1.42	30-May-2020	jdolecek	remove #if 0'ed pci_conf_print() call
1.41	30-Apr-2020	riastradh	rnd_attach_source calls the callback itself now. No need for every driver to explicitly call it to prime the pool. Eliminate now-unused <sys/rndpool.h>.
1.40	30-Apr-2020	riastradh	Don't attach rndsource until it's actually ready to run.
1.39	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.38	10-Aug-2014	tls	branches: 1.38.4; Merge tls-earlyentropy branch into HEAD.
1.37	13-Jun-2013	tls	branches: 1.37.6; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.36	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.35	30-Jan-2012	drochner	branches: 1.35.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.34	07-Feb-2010	pgoyette	branches: 1.34.12; 1.34.16; Recognize the i2c bus on the AMD768 PCI-ISA/LPC Bridge Resolves PR/42759 - thanks Michael Stapelberg
1.33	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.32	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.31	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.30	28-Apr-2008	martin	branches: 1.30.14; Remove clause 3 and 4 from TNF licenses
1.29	10-Apr-2008	cegger	branches: 1.29.2; 1.29.4; use aprint_*_dev and device_xname
1.28	19-Oct-2007	ad	branches: 1.28.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.27	26-Aug-2007	xtraeme	branches: 1.27.2; 1.27.6; There's no need to add timecounter code inside of __HAVE_TIMECOUNTER ifdefs.
1.26	09-Jul-2007	ad	branches: 1.26.2; 1.26.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.25	05-Feb-2007	jmcneill	branches: 1.25.6; 1.25.8; Disgusting Xbox hack; The "pci0 dev 1 function 2" aka "System Management" on the nForce chipset isn't detected on the Xbox. If XBOX is defined and arch_i386_is_xbox is true, force intrline to 12 and map the system management IO space at 0x8000. We can now call iic_smbus_intr when a system management interrupt takes place. While we're here, make amdpm_smbus honour the I2C_F_POLL flag. The Xbox no longer resets when you press the eject button.
1.24	06-Jan-2007	jmcneill	Make sure that we don't use the amdpm timecounter on Xbox.
1.23	06-Jan-2007	jmcneill	nForce uses a slightly different register offset for the smbus controls; compensate for this.
1.22	06-Jan-2007	jmcneill	Add nVidia nForce support, tested on Xbox.
1.21	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	27-Aug-2006	christos	branches: 1.19.2; 1.19.4; restructure ifdefs
1.18	11-Jul-2006	drochner	include machine/bus.h explicitely, fixes build problem
1.17	11-Jul-2006	drochner	use timecounter code in dev/ic/acpipmtimer.*, kill interface attribute in device declaration
1.16	10-Jul-2006	xtraeme	Attach amdpm timer on AMD8111 too, as was suggested by Nicolas Joly.
1.15	10-Jul-2006	christos	PR/33968: Nicolas Joly: amdpm(4) timer, fix and AMD8111 support Refactor the code to use different variable names to hold each conf register and avoid reading/witing the conf space multiple times.
1.14	21-Jun-2006	explorer	formatting nit, newline, print device name
1.13	21-Jun-2006	explorer	Actually set amdpm_timecounter.tc_priv before using it later.
1.12	17-Jun-2006	xtraeme	Support for the 24/32bit timer in the AMD 768 Power Management Controller, to try it use `sysctl -w kern.timecounter.hardware=amdpm`. From OpenBSD.
1.11	19-Feb-2006	tls	branches: 1.11.2; 1.11.8; 1.11.10; Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.10	28-Jan-2006	kleink	branches: 1.10.2; 1.10.4; Add support for the AMD8111 RNG; from Nicolas Joly in PR kern/32284.
1.9	04-Dec-2005	christos	branches: 1.9.2; PR/32220: Murray Armfield: Enable AMD768 random no generator.
1.8	28-Jun-2005	thorpej	branches: 1.8.2; Use ANSI function decls and static.
1.7	01-Sep-2003	tls	branches: 1.7.14; Revert previous change, which was wrong. As Enami points out, we really do want the NO_ESTIMATE flag here, because if the source doesn't have it, rnd_add_data will use the time the data was added directly as entropy, which is not a good idea either for a hardware RNG that works like this one (periodically polled by the driver) or, really, like any other that I've seen, because even for "interrupt when ready" RNGs the data rate (and thus interrupt time) will vary only slightly for a given RNG configuration over time, if at all. The data we add _is_ counted into the pool estimage because we directly pass in an entropy estimate. The overloading of the NO_ESTIMATE flag with this meaning seems, to me, like a design bug (albeit easily fixed) in the rng interface; but until that's changed, we need to use it here.
1.6	12-Aug-2003	tls	Remove RND_FLAG_NO_ESTIMATE and comment claiming we can't estimate (which means "count", in this context) the entropy we're adding because we poll the device periodically. Sure, we poll it periodically, but it's a hardware RNG -- the data returned should be random no matter when we read the register!
1.5	31-Jan-2003	thorpej	branches: 1.5.2; Use aprint_*().
1.4	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	02-Jun-2002	enami	branches: 1.1.2; 1.1.4; 1.1.6; Collect random number from AMD 768MPX power management controller.
1.1.6.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.6.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.6.1	02-Jun-2002	jdolecek	file amdpm.c was added on branch kqueue on 2002-06-23 17:47:32 +0000
1.1.4.2	20-Jun-2002	gehenna	catch up with -current.
1.1.4.1	02-Jun-2002	gehenna	file amdpm.c was added on branch gehenna-devsw on 2002-06-20 16:33:29 +0000
1.1.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.2	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.1	02-Jun-2002	nathanw	file amdpm.c was added on branch nathanw_sa on 2002-06-20 03:45:19 +0000
1.5.2.5	11-Dec-2005	christos	Sync with head.
1.5.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.14.3	20-Feb-2006	tron	Pull up following revision(s) (requested by tls in ticket #1179): sys/dev/pci/amdpm.c: revision 1.11 sys/dev/i2c/files.i2c: revision 1.7 sys/dev/i2c/adt7463.c: revision 1.1 sys/dev/pci/files.pci: revision 1.242 via patch sys/dev/pci/amdpmvar.h: revision 1.1 sys/dev/i2c/adt7463reg.h: revision 1.1 sys/dev/DEVNAMES: revision 1.192 via patch sys/dev/pci/amdpm_smbus.c: revision 1.1 sys/dev/pci/amdpm_smbusreg.h: revision 1.1 Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.7.14.2	05-Feb-2006	riz	Pull up following revision(s) (requested by tls in ticket #1159): sys/dev/pci/amdpm.c: revision 1.10 share/man/man4/amdpm.4: revision 1.4 Add support for the AMD8111 RNG; from Nicolas Joly in PR kern/32284.
1.7.14.1	05-Feb-2006	riz	Pull up following revision(s) (requested by tls in ticket #1158): sys/dev/pci/amdpm.c: revision 1.9 PR/32220: Murray Armfield: Enable AMD768 random no generator.
1.8.2.5	27-Oct-2007	yamt	sync with head.
1.8.2.4	03-Sep-2007	yamt	sync with head.
1.8.2.3	26-Feb-2007	yamt	sync with head.
1.8.2.2	30-Dec-2006	yamt	sync with head.
1.8.2.1	21-Jun-2006	yamt	sync with head.
1.9.2.2	01-Mar-2006	yamt	sync with head.
1.9.2.1	01-Feb-2006	yamt	sync with head.
1.10.4.1	22-Apr-2006	simonb	Sync with head.
1.10.2.1	09-Sep-2006	rpaulo	sync with head
1.11.10.1	13-Jul-2006	gdamore	Merge from HEAD.
1.11.8.1	19-Jun-2006	chap	Sync with head.
1.11.2.3	03-Sep-2006	yamt	sync with head.
1.11.2.2	11-Aug-2006	yamt	sync with head
1.11.2.1	26-Jun-2006	yamt	sync with head.
1.19.4.2	10-Dec-2006	yamt	sync with head.
1.19.4.1	22-Oct-2006	yamt	sync with head
1.19.2.3	09-Feb-2007	ad	Sync with HEAD.
1.19.2.2	12-Jan-2007	ad	Sync with head.
1.19.2.1	18-Nov-2006	ad	Sync with head.
1.25.8.1	11-Jul-2007	mjf	Sync with head.
1.25.6.3	23-Oct-2007	ad	Sync with head.
1.25.6.2	09-Oct-2007	ad	Sync with head.
1.25.6.1	01-Jul-2007	ad	Adapt to callout API change.
1.26.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.26.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.26.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.27.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.27.2.1	06-Nov-2007	matt	sync with HEAD
1.28.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.29.4.3	11-Mar-2010	yamt	sync with head
1.29.4.2	16-May-2009	yamt	sync with head
1.29.4.1	16-May-2008	yamt	sync with head.
1.29.2.1	18-May-2008	yamt	sync with head.
1.30.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.34.16.1	18-Feb-2012	mrg	merge to -current.
1.34.12.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.34.12.2	30-Oct-2012	yamt	sync with head
1.34.12.1	17-Apr-2012	yamt	sync with head
1.35.6.4	03-Dec-2017	jdolecek	update from HEAD
1.35.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.35.6.2	23-Jun-2013	tls	resync from head
1.35.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.37.6.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.38.4.1	06-Jun-2015	skrll	Sync with HEAD
1.26	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.25	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.24	24-Apr-2021	thorpej	branches: 1.24.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.23	22-Dec-2019	thorpej	branches: 1.23.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.22	14-Feb-2016	chs	branches: 1.22.18; zero the i2c_attach_args structure before filling it in. fixes occasional crashes in iic_attach().
1.21	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.20	13-Jun-2013	tls	branches: 1.20.10; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.19	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.18	14-Feb-2012	pgoyette	branches: 1.18.6; Replace the xxx_acquire()/xxx_release() rwlocks with mutexes. There are only RW_WRITERs for these, and no RW_READERs, so no need to incur the extra overhead of allowing for both. As discussed on tech-kern. For piixpm and ichsmb, the acquire/release protocol needs to be used, even if the request is I2C_F_POLL'd (or if the device supports only polled mode). Otherwise multiple requests can be running at the same time, and they stomp on each other and create anomolous results. Part 2 addresses my PR kern/45889 3 ACKs from releng
1.17	19-Nov-2011	christos	branches: 1.17.4; bye xbox
1.16	03-Feb-2009	pgoyette	branches: 1.16.14; Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.15	10-Apr-2008	cegger	branches: 1.15.4; 1.15.12; use aprint_*_dev and device_xname
1.14	27-Aug-2007	xtraeme	branches: 1.14.22; Use rwlock(9) rather than lockmgr(9).
1.13	27-Aug-2007	xtraeme	Apply some KNF.
1.12	06-Feb-2007	jmcneill	branches: 1.12.6; 1.12.14; 1.12.18; Fix build failure for !XBOX, pointed out by Paul Goyette on current-users
1.11	06-Feb-2007	jmcneill	Another Xbox hack; If we're handling system management interrupts ourselves, the system locks solid when the ACPI timer SCI fires. Disable this interrupt so we can have an uptime of more than ~10 minutes.
1.10	05-Feb-2007	jmcneill	Disgusting Xbox hack; The "pci0 dev 1 function 2" aka "System Management" on the nForce chipset isn't detected on the Xbox. If XBOX is defined and arch_i386_is_xbox is true, force intrline to 12 and map the system management IO space at 0x8000. We can now call iic_smbus_intr when a system management interrupt takes place. While we're here, make amdpm_smbus honour the I2C_F_POLL flag. The Xbox no longer resets when you press the eject button.
1.9	06-Jan-2007	jmcneill	There is absolutely no way that this code could have worked for reading values from the smbus before. amdpm_smbus_exec was using the values read from the registers and using them as return values instead of filling in the caller's buffer.
1.8	06-Jan-2007	jmcneill	nForce uses a slightly different register offset for the smbus controls; compensate for this.
1.7	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.6	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.5	24-Sep-2006	jmcneill	Use /* ... */ for comment instead of // ...
1.4	26-Jun-2006	drochner	branches: 1.4.4; 1.4.6; 1.4.8; use the "i2cbus" interface attribute rather than putting a string name into the i2cbus attach args
1.3	30-Apr-2006	xtraeme	branches: 1.3.4; 1.3.6; PR kern/32913: amdpm(4) smbus/i2cbus needs locking. Apply patch from Nicolas Joly.
1.2	19-Feb-2006	xtraeme	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12; Add __KERNEL_RCSID.
1.1	19-Feb-2006	tls	Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.2.12.3	01-Jun-2006	kardel	Sync with head.
1.2.12.2	22-Apr-2006	simonb	Sync with head.
1.2.12.1	19-Feb-2006	simonb	file amdpm_smbus.c was added on branch simonb-timecounters on 2006-04-22 11:39:13 +0000
1.2.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.2.8.1	11-May-2006	elad	sync with head
1.2.6.2	11-Aug-2006	yamt	sync with head
1.2.6.1	24-May-2006	yamt	sync with head.
1.2.4.2	01-Mar-2006	yamt	sync with head.
1.2.4.1	19-Feb-2006	yamt	file amdpm_smbus.c was added on branch yamt-uio_vmspace on 2006-03-01 09:28:21 +0000
1.2.2.2	20-Feb-2006	tron	Pull up following revision(s) (requested by tls in ticket #1179): sys/dev/pci/amdpm.c: revision 1.11 sys/dev/i2c/files.i2c: revision 1.7 sys/dev/i2c/adt7463.c: revision 1.1 sys/dev/pci/files.pci: revision 1.242 via patch sys/dev/pci/amdpmvar.h: revision 1.1 sys/dev/i2c/adt7463reg.h: revision 1.1 sys/dev/DEVNAMES: revision 1.192 via patch sys/dev/pci/amdpm_smbus.c: revision 1.1 sys/dev/pci/amdpm_smbusreg.h: revision 1.1 Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.2.2.1	19-Feb-2006	tron	file amdpm_smbus.c was added on branch netbsd-3 on 2006-02-20 23:00:27 +0000
1.3.6.5	03-Sep-2007	yamt	sync with head.
1.3.6.4	26-Feb-2007	yamt	sync with head.
1.3.6.3	30-Dec-2006	yamt	sync with head.
1.3.6.2	21-Jun-2006	yamt	sync with head.
1.3.6.1	30-Apr-2006	yamt	file amdpm_smbus.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.3.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.4.8.2	10-Dec-2006	yamt	sync with head.
1.4.8.1	22-Oct-2006	yamt	sync with head
1.4.6.2	09-Sep-2006	rpaulo	sync with head
1.4.6.1	26-Jun-2006	rpaulo	file amdpm_smbus.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:16 +0000
1.4.4.3	09-Feb-2007	ad	Sync with HEAD.
1.4.4.2	12-Jan-2007	ad	Sync with head.
1.4.4.1	18-Nov-2006	ad	Sync with head.
1.12.18.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.12.14.1	03-Sep-2007	skrll	Sync with HEAD.
1.12.6.1	09-Oct-2007	ad	Sync with head.
1.14.22.1	02-Jun-2008	mjf	Sync with HEAD.
1.15.12.1	03-Mar-2009	skrll	Sync with HEAD.
1.15.4.1	04-May-2009	yamt	sync with head.
1.16.14.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.16.14.2	30-Oct-2012	yamt	sync with head
1.16.14.1	17-Apr-2012	yamt	sync with head
1.17.4.1	18-Feb-2012	mrg	merge to -current.
1.18.6.3	03-Dec-2017	jdolecek	update from HEAD
1.18.6.2	23-Jun-2013	tls	resync from head
1.18.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.20.10.2	19-Mar-2016	skrll	Sync with HEAD
1.20.10.1	06-Jun-2015	skrll	Sync with HEAD
1.22.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.23.10.1	23-Mar-2021	thorpej	Convert config_found_ia() call sites where the device only carries a single interface attribute to bare config_found() calls.
1.24.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.3	16-Sep-2021	andvar	fix typos in word "successfully", mainly s/succesfully/successfully/.
1.2	03-Feb-2009	pgoyette	Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.1	19-Feb-2006	tls	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.18; 1.1.24; 1.1.78; 1.1.88; Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.1.88.1	03-Mar-2009	skrll	Sync with HEAD.
1.1.78.1	04-May-2009	yamt	sync with head.
1.1.24.2	09-Sep-2006	rpaulo	sync with head
1.1.24.1	19-Feb-2006	rpaulo	file amdpm_smbusreg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:16 +0000
1.1.18.2	21-Jun-2006	yamt	sync with head.
1.1.18.1	19-Feb-2006	yamt	file amdpm_smbusreg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.1.12.2	22-Apr-2006	simonb	Sync with head.
1.1.12.1	19-Feb-2006	simonb	file amdpm_smbusreg.h was added on branch simonb-timecounters on 2006-04-22 11:39:13 +0000
1.1.4.2	01-Mar-2006	yamt	sync with head.
1.1.4.1	19-Feb-2006	yamt	file amdpm_smbusreg.h was added on branch yamt-uio_vmspace on 2006-03-01 09:28:21 +0000
1.1.2.2	20-Feb-2006	tron	Pull up following revision(s) (requested by tls in ticket #1179): sys/dev/pci/amdpm.c: revision 1.11 sys/dev/i2c/files.i2c: revision 1.7 sys/dev/i2c/adt7463.c: revision 1.1 sys/dev/pci/files.pci: revision 1.242 via patch sys/dev/pci/amdpmvar.h: revision 1.1 sys/dev/i2c/adt7463reg.h: revision 1.1 sys/dev/DEVNAMES: revision 1.192 via patch sys/dev/pci/amdpm_smbus.c: revision 1.1 sys/dev/pci/amdpm_smbusreg.h: revision 1.1 Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.1.2.1	19-Feb-2006	tron	file amdpm_smbusreg.h was added on branch netbsd-3 on 2006-02-20 23:00:27 +0000
1.4	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.3	06-Jan-2007	jmcneill	branches: 1.3.40; 1.3.42; 1.3.44; Add nVidia nForce support, tested on Xbox.
1.2	17-Jun-2006	xtraeme	branches: 1.2.4; Support for the 24/32bit timer in the AMD 768 Power Management Controller, to try it use `sysctl -w kern.timecounter.hardware=amdpm`. From OpenBSD.
1.1	02-Jun-2002	enami	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.28; 1.1.42; 1.1.46; 1.1.52; 1.1.54; Collect random number from AMD 768MPX power management controller.
1.1.54.1	13-Jul-2006	gdamore	Merge from HEAD.
1.1.52.1	19-Jun-2006	chap	Sync with head.
1.1.46.1	26-Jun-2006	yamt	sync with head.
1.1.42.1	09-Sep-2006	rpaulo	sync with head
1.1.28.2	26-Feb-2007	yamt	sync with head.
1.1.28.1	21-Jun-2006	yamt	sync with head.
1.1.6.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.6.1	02-Jun-2002	jdolecek	file amdpmreg.h was added on branch kqueue on 2002-06-23 17:47:32 +0000
1.1.4.2	20-Jun-2002	gehenna	catch up with -current.
1.1.4.1	02-Jun-2002	gehenna	file amdpmreg.h was added on branch gehenna-devsw on 2002-06-20 16:33:29 +0000
1.1.2.2	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.1	02-Jun-2002	nathanw	file amdpmreg.h was added on branch nathanw_sa on 2002-06-20 03:45:19 +0000
1.2.4.1	12-Jan-2007	ad	Sync with head.
1.3.44.1	16-May-2008	yamt	sync with head.
1.3.42.1	18-May-2008	yamt	sync with head.
1.3.40.1	02-Jun-2008	mjf	Sync with HEAD.
1.12	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.11	13-Apr-2015	riastradh	branches: 1.11.18; Convert sys/dev to use <sys/rndsource.h>.
1.10	13-Jun-2013	tls	branches: 1.10.10; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.9	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.8	14-Feb-2012	pgoyette	branches: 1.8.6; Replace the xxx_acquire()/xxx_release() rwlocks with mutexes. There are only RW_WRITERs for these, and no RW_READERs, so no need to incur the extra overhead of allowing for both. As discussed on tech-kern. For piixpm and ichsmb, the acquire/release protocol needs to be used, even if the request is I2C_F_POLL'd (or if the device supports only polled mode). Otherwise multiple requests can be running at the same time, and they stomp on each other and create anomolous results. Part 2 addresses my PR kern/45889 3 ACKs from releng
1.7	19-Nov-2011	tls	branches: 1.7.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.6	28-Apr-2008	martin	branches: 1.6.34; Remove clause 3 and 4 from TNF licenses
1.5	27-Aug-2007	xtraeme	branches: 1.5.22; 1.5.24; 1.5.26; Use rwlock(9) rather than lockmgr(9).
1.4	05-Feb-2007	jmcneill	branches: 1.4.6; 1.4.14; 1.4.18; Disgusting Xbox hack; The "pci0 dev 1 function 2" aka "System Management" on the nForce chipset isn't detected on the Xbox. If XBOX is defined and arch_i386_is_xbox is true, force intrline to 12 and map the system management IO space at 0x8000. We can now call iic_smbus_intr when a system management interrupt takes place. While we're here, make amdpm_smbus honour the I2C_F_POLL flag. The Xbox no longer resets when you press the eject button.
1.3	06-Jan-2007	jmcneill	Add nVidia nForce support, tested on Xbox.
1.2	30-Apr-2006	xtraeme	branches: 1.2.6; 1.2.10; 1.2.12; PR kern/32913: amdpm(4) smbus/i2cbus needs locking. Apply patch from Nicolas Joly.
1.1	19-Feb-2006	tls	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.12; Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.1.12.3	01-Jun-2006	kardel	Sync with head.
1.1.12.2	22-Apr-2006	simonb	Sync with head.
1.1.12.1	19-Feb-2006	simonb	file amdpmvar.h was added on branch simonb-timecounters on 2006-04-22 11:39:13 +0000
1.1.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.1.8.1	11-May-2006	elad	sync with head
1.1.6.1	24-May-2006	yamt	sync with head.
1.1.4.2	01-Mar-2006	yamt	sync with head.
1.1.4.1	19-Feb-2006	yamt	file amdpmvar.h was added on branch yamt-uio_vmspace on 2006-03-01 09:28:21 +0000
1.1.2.2	20-Feb-2006	tron	Pull up following revision(s) (requested by tls in ticket #1179): sys/dev/pci/amdpm.c: revision 1.11 sys/dev/i2c/files.i2c: revision 1.7 sys/dev/i2c/adt7463.c: revision 1.1 sys/dev/pci/files.pci: revision 1.242 via patch sys/dev/pci/amdpmvar.h: revision 1.1 sys/dev/i2c/adt7463reg.h: revision 1.1 sys/dev/DEVNAMES: revision 1.192 via patch sys/dev/pci/amdpm_smbus.c: revision 1.1 sys/dev/pci/amdpm_smbusreg.h: revision 1.1 Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.1.2.1	19-Feb-2006	tron	file amdpmvar.h was added on branch netbsd-3 on 2006-02-20 23:00:27 +0000
1.2.12.2	09-Sep-2006	rpaulo	sync with head
1.2.12.1	30-Apr-2006	rpaulo	file amdpmvar.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:16 +0000
1.2.10.2	09-Feb-2007	ad	Sync with HEAD.
1.2.10.1	12-Jan-2007	ad	Sync with head.
1.2.6.4	03-Sep-2007	yamt	sync with head.
1.2.6.3	26-Feb-2007	yamt	sync with head.
1.2.6.2	21-Jun-2006	yamt	sync with head.
1.2.6.1	30-Apr-2006	yamt	file amdpmvar.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.4.18.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.4.14.1	03-Sep-2007	skrll	Sync with HEAD.
1.4.6.1	09-Oct-2007	ad	Sync with head.
1.5.26.1	16-May-2008	yamt	sync with head.
1.5.24.1	18-May-2008	yamt	sync with head.
1.5.22.1	02-Jun-2008	mjf	Sync with HEAD.
1.6.34.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6.34.2	30-Oct-2012	yamt	sync with head
1.6.34.1	17-Apr-2012	yamt	sync with head
1.7.2.1	18-Feb-2012	mrg	merge to -current.
1.8.6.3	03-Dec-2017	jdolecek	update from HEAD
1.8.6.2	23-Jun-2013	tls	resync from head
1.8.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.10.10.1	06-Jun-2015	skrll	Sync with HEAD
1.11.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.68	02-Feb-2024	andvar	fix various typos in comments.
1.67	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.66	24-Apr-2021	thorpej	branches: 1.66.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.65	10-Nov-2019	chs	branches: 1.65.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.64	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.63	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.62	27-Sep-2016	pgoyette	branches: 1.62.14; 1.62.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.61	14-Jul-2016	msaitoh	branches: 1.61.2; KNF. No functional change.
1.60	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.59	02-Mar-2015	christos	From Tom Ivar Helbekkmo: - Use mutexes and condvars instead of spl() calls. - Add a couple of bugfixes from FreeBSD (the easily observed ones). - Add a simple watchdog thread for the newer versions of the hardware, that tickles the controller at task submission time if it seems not ready (from FreeBSD). - Add a buffer allocation size hack to the ioctl handler, to work around a firmware bug in some versions of the controller (also from FreeBSD). - Rearrange the bus_dmamap_sync() calls to make the use match the documentation.
1.58	25-Jul-2014	dholland	branches: 1.58.4; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.57	29-Mar-2014	christos	branches: 1.57.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.55	27-Jul-2012	jakllsch	branches: 1.55.2; 1.55.4; Convert amr(4) to device_t (CFATTACH_DECL_NEW).
1.54	13-Nov-2010	uebayasi	branches: 1.54.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.53	26-Nov-2009	njoly	branches: 1.53.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.52	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.51	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.50	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.49	08-Jun-2008	tsutsui	branches: 1.49.12; Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.48	28-Apr-2008	martin	branches: 1.48.2; Remove clause 3 and 4 from TNF licenses
1.47	10-Apr-2008	cegger	branches: 1.47.2; 1.47.4; use aprint_*_dev and device_xname
1.46	19-Oct-2007	ad	branches: 1.46.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.45	09-Jul-2007	ad	branches: 1.45.6; 1.45.8; 1.45.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.44	04-Mar-2007	christos	branches: 1.44.2; 1.44.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.43	02-Dec-2006	elad	branches: 1.43.2; Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.42	25-Nov-2006	christos	branches: 1.42.2; spell precede; from Zafer
1.41	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.40	08-Nov-2006	elad	Replace securelevel checks with kauth(9) calls.
1.39	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.38	27-Aug-2006	christos	branches: 1.38.2; 1.38.4; - static needs to be first - complete the cdevsw
1.37	23-Aug-2006	christos	Acting as Elad's keyboard: Place the securelevel checks in their logical locations. This will be clearer in the future when code changes to use kauth(9) calls. input and okay ad@
1.36	23-Jul-2006	bouyer	Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.35	07-Jun-2006	kardel	merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.34	17-Apr-2006	elad	branches: 1.34.2; Fix typos.
1.33	15-Apr-2006	he	The type returned by __arraycount() is a size_t, not an int, so print it with %zd and not %d.
1.32	14-Apr-2006	christos	Coverity CID 1535: Avoid reading past the end of arrays because we don't pay proper attention to the number of logical drives each controller can support.
1.31	11-Dec-2005	jonathan	branches: 1.31.4; 1.31.6; 1.31.8; 1.31.10; 1.31.12; Add pcidevs entries and support to amr(4) for newer LSI MegaRAID devices, including the LSI MegaRAID SCSI 320-0x, 320-2x, 320-4x, 320-1E, 320-2E, the LSI MegaRAID SATA 300-6x and 300-8x. Tested on an LSI MegaRAID SATA 300-8x, which confirms private advice that after PCI-IDs to our amr(4) driver, the newer LSI MegaRAID cards will Just Work The Linux megaraid2 driver accepts various Intel cards with the same pci-device IDs as the LSI MegaRAID cards, treating them exaclty as the above-listed LSI cards. Rework our amr(4) driver to also match and attach those device-ID if seen with an Intel vendor-ID. Add the Intel-vendor PCI decice-ids to pcidevs, so that PCI_VERBOSE will correctly identify the Intel cards. Update src/share/man/man4/amr.4 to list the newly-supported cards. (NB: I don't have the LSI SCSI or any Intel cards to test, just the LSI SATA, which works fine for me.)
1.30	11-Dec-2005	christos	merge ktrace-lwp.
1.29	26-Aug-2005	drochner	kill some more simple submatch() functions, use config_stdsubmatch()
1.28	25-Aug-2005	drochner	replace the "locdesc_t" structure carrying the number of locators explicitely by a plain integer array the length in now known to all relevant parties, so this avoids duplication of information, and we can allocate that thing in drivers without hacks
1.27	28-Jun-2005	thorpej	branches: 1.27.2; Use ANSI function decls and static.
1.26	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.25	27-Feb-2005	perry	branches: 1.25.2; nuke trailing whitespace
1.24	28-Oct-2004	martti	branches: 1.24.4; 1.24.6; Added Dell PERC 4/SC
1.23	20-Oct-2004	martti	Added support for DELL PERC 4e/Si RAID controller
1.22	13-Sep-2004	drochner	a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.21	02-Sep-2004	he	Make this driver recognize LSI Logic MegaRAID 320-2 RAID controllers. Fixes PR#26834, patch from that PR.
1.20	22-Apr-2004	itojun	sprintf -> snprintf
1.19	05-Nov-2003	fvdl	branches: 1.19.2; Apparently, Dell PERC 4/Di controllers can have two memory BARs, with the second one empty. The previous revision made it pick the empty one. Reinstate the BAR size check.
1.18	29-Oct-2003	mycroft	Rework to remove bogus initializers. Also fixes a potential problem with amr_teardown() unmapping too much.
1.17	25-Oct-2003	christos	Fix completely bogus printf that used uninitialized variables.
1.16	25-Oct-2003	christos	Fix uninitialized variable warnings
1.15	21-Oct-2003	fvdl	Don't pass NULL as an integer value.
1.14	14-Oct-2003	martti	Added another Dell PERC 4/Di
1.13	13-Oct-2003	ad	Make this work with the newer 'Quartz' controllers.
1.12	26-Sep-2003	matt	Add DELL Perc 4/di (from Freebsd).
1.11	15-May-2003	fvdl	branches: 1.11.2; Fix reversed arguments to bus_space_unmap.
1.10	14-May-2003	ad	Fix thinko.
1.9	04-May-2003	ad	- Fix the mailbox protocol. Not tested with newer 'Quartz' controllers. - Handle failure during initalisation more gracefully. - Create a watchdog thread to poke the board and check for status changes. - Make ready for > 64kB transfers.
1.8	31-Jan-2003	thorpej	Use aprint_*().
1.7	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	27-Sep-2002	thorpej	Introduce a new routine, config_match(), which invokes the cfattach->ca_match function in behalf of the caller. Use it rather than invoking cfattach->ca_match directly.
1.2	01-Jun-2002	lukem	SIMPLEQ rototill: - implement SIMPLEQ_REMOVE(head, elm, type, field). whilst it's O(n), this mirrors the functionality of SLIST_REMOVE() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE() - remove the unnecessary elm arg from SIMPLEQ_REMOVE_HEAD(). this mirrors the functionality of SLIST_REMOVE_HEAD() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE_HEAD() - remove notes about SIMPLEQ not supporting arbitrary element removal - use SIMPLEQ_FOREACH() instead of home-grown for loops - use SIMPLEQ_EMPTY() appropriately - use SIMPLEQ_*() instead of accessing sqh_first,sqh_last,sqe_next directly - reorder manual page; be consistent about how the types are listed - other minor cleanups
1.1	30-Jan-2002	ad	branches: 1.1.2; 1.1.4; 1.1.12; Bare-bones driver for AMI RAID. Parts taken from FreeBSD. This was tried a good while ago and it had problems under load. Changes were made to address that, but I don't have the ability to test them. So, I'm committing it before it rots.
1.1.12.1	20-Jun-2002	gehenna	catch up with -current.
1.1.4.5	03-Jan-2003	thorpej	Sync with HEAD.
1.1.4.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.3	20-Jun-2002	nathanw	Catch up to -current.
1.1.4.2	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.1	30-Jan-2002	nathanw	file amr.c was added on branch nathanw_sa on 2002-02-28 04:13:57 +0000
1.1.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.1	30-Jan-2002	jdolecek	file amr.c was added on branch kqueue on 2002-02-11 20:09:55 +0000
1.11.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.11.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.11.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.11.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.11.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.11.2.2	03-Sep-2004	skrll	Sync with HEAD
1.11.2.1	03-Aug-2004	skrll	Sync with HEAD
1.19.2.3	12-Nov-2004	jmc	Pullup rev 1.24 (requested by martti in ticket #953) Added Dell PERC 4/SC
1.19.2.2	12-Nov-2004	jmc	Pullup rev 1.23 (requested by martti in ticket #947) Added support for DELL PERC 4e/Si RAID controller
1.19.2.1	11-Sep-2004	tron	Pull up revision 1.21 (requested by he in ticket #826): Make this driver recognize LSI Logic MegaRAID 320-2 RAID controllers. Fixes PR#26834, patch from that PR.
1.24.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.24.4.1	29-Apr-2005	kent	sync with -current
1.25.2.2	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.25.2.1	15-Dec-2005	tron	Pull up following revision(s) (requested by jonathan in ticket #1047): share/man/man4/amr.4: revision 1.11 sys/dev/pci/amr.c: revision 1.31 sys/dev/pci/pcidevs: revision 1.746 via patch Add pcidevs entries and support to amr(4) for newer LSI MegaRAID devices, including the LSI MegaRAID SCSI 320-0x, 320-2x, 320-4x, 320-1E, 320-2E, the LSI MegaRAID SATA 300-6x and 300-8x. Tested on an LSI MegaRAID SATA 300-8x, which confirms private advice that after PCI-IDs to our amr(4) driver, the newer LSI MegaRAID cards will Just Work The Linux megaraid2 driver accepts various Intel cards with the same pci-device IDs as the LSI MegaRAID cards, treating them exaclty as the above-listed LSI cards. Rework our amr(4) driver to also match and attach those device-ID if seen with an Intel vendor-ID. Add the Intel-vendor PCI decice-ids to pcidevs, so that PCI_VERBOSE will correctly identify the Intel cards. Update src/share/man/man4/amr.4 to list the newly-supported cards. (NB: I don't have the LSI SCSI or any Intel cards to test, just the LSI SATA, which works fine for me.)
1.27.2.4	27-Oct-2007	yamt	sync with head.
1.27.2.3	03-Sep-2007	yamt	sync with head.
1.27.2.2	30-Dec-2006	yamt	sync with head.
1.27.2.1	21-Jun-2006	yamt	sync with head.
1.31.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.31.10.1	19-Apr-2006	elad	sync with head.
1.31.8.4	03-Sep-2006	yamt	sync with head.
1.31.8.3	11-Aug-2006	yamt	sync with head
1.31.8.2	26-Jun-2006	yamt	sync with head.
1.31.8.1	24-May-2006	yamt	sync with head.
1.31.6.2	22-Apr-2006	simonb	Sync with head.
1.31.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.31.4.1	09-Sep-2006	rpaulo	sync with head
1.34.2.1	19-Jun-2006	chap	Sync with head.
1.38.4.2	10-Dec-2006	yamt	sync with head.
1.38.4.1	22-Oct-2006	yamt	sync with head
1.38.2.2	12-Jan-2007	ad	Sync with head.
1.38.2.1	18-Nov-2006	ad	Sync with head.
1.42.2.1	04-Dec-2006	tron	Pull up following revision(s) (requested by elad in ticket #247): sys/dev/ic/dpt.c: revision 1.55 sys/dev/pci/amr.c: revision 1.43 sys/secmodel/bsd44/secmodel_bsd44_securelevel.c: revision 1.19 sys/dev/pci/mly.c: revision 1.33 share/man/man9/kauth.9: revision 1.37 sys/dev/ic/mlx.c: revision 1.49 sys/dev/ic/icp_ioctl.c: revision 1.14 sys/dev/i2o/iop.c: revision 1.62 sys/dev/pci/twe.c: revision 1.82 sys/sys/kauth.h: revision 1.25 sys/dev/i2o/dpti.c: revision 1.31 sys/kern/kern_auth.c: revision 1.33 sys/dev/tc/stic.c: revision 1.37 Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.43.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.44.4.1	11-Jul-2007	mjf	Sync with head.
1.44.2.4	23-Oct-2007	ad	Sync with head.
1.44.2.3	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.44.2.2	10-Apr-2007	ad	Nuke the deferred kthread creation stuff, as it's no longer needed. Pointed out by thorpej@.
1.44.2.1	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.45.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.45.8.1	06-Nov-2007	matt	sync with HEAD
1.45.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.46.16.3	29-Jun-2008	mjf	Sync with HEAD.
1.46.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.46.16.1	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.47.4.4	11-Mar-2010	yamt	sync with head
1.47.4.3	16-May-2009	yamt	sync with head
1.47.4.2	04-May-2009	yamt	sync with head.
1.47.4.1	16-May-2008	yamt	sync with head.
1.47.2.2	17-Jun-2008	yamt	sync with head.
1.47.2.1	18-May-2008	yamt	sync with head.
1.48.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.49.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.53.4.1	05-Mar-2011	rmind	sync with head
1.54.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.54.8.1	30-Oct-2012	yamt	sync with head
1.55.4.1	18-May-2014	rmind	sync with head
1.55.2.3	03-Dec-2017	jdolecek	update from HEAD
1.55.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.2.1	12-Sep-2012	tls	Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for physio (I/O to raw devices); needs more doing to get it going with the filesystems, but it shouldn't damage data. All work's been done on amd64 so far. Not hard to add support to other ports. If others want to pitch in, one very helpful thing would be to sort out when and how IDE disks can do 128K or larger transfers, and adjust the various PCI IDE (or at least ahcisata) drivers and wd.c accordingly -- it would make testing much easier. Another very helpful thing would be to implement a smart minphys() for RAIDframe along the lines detailed in the MAXPHYS-NOTES file.
1.57.2.1	10-Aug-2014	tls	Rebase.
1.58.4.3	05-Oct-2016	skrll	Sync with HEAD
1.58.4.2	09-Jul-2016	skrll	Sync with HEAD
1.58.4.1	06-Apr-2015	skrll	Sync with HEAD
1.61.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.62.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.62.16.1	10-Jun-2019	christos	Sync with HEAD
1.62.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.62.14.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.65.10.3	28-Mar-2021	thorpej	Correct the names of the arguments passed to the rescan function.
1.65.10.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.65.10.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.66.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	23-Jul-2006	bouyer	branches: 1.1.2; 1.1.6; 1.1.10; 1.1.16; Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.1.16.2	30-Dec-2006	yamt	sync with head.
1.1.16.1	23-Jul-2006	yamt	file amrio.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:41 +0000
1.1.10.2	09-Sep-2006	rpaulo	sync with head
1.1.10.1	23-Jul-2006	rpaulo	file amrio.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:16 +0000
1.1.6.2	11-Aug-2006	yamt	sync with head
1.1.6.1	23-Jul-2006	yamt	file amrio.h was added on branch yamt-pdpolicy on 2006-08-11 15:44:25 +0000
1.1.2.2	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.1.2.1	23-Jul-2006	tron	file amrio.h was added on branch netbsd-3 on 2006-07-30 16:38:59 +0000
1.6	15-Aug-2023	mrg	amr(4): fix the size of the the drive format array. now this is actually 1024 bytes long, like it wants to be. found by GCC 12.
1.5	08-Sep-2008	gmcgarry	Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.4	28-Apr-2008	martin	branches: 1.4.2; 1.4.6; Remove clause 3 and 4 from TNF licenses
1.3	23-Jul-2006	bouyer	branches: 1.3.54; 1.3.56; 1.3.58; Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.2	04-May-2003	ad	branches: 1.2.16; 1.2.18; 1.2.32; 1.2.36; - Fix the mailbox protocol. Not tested with newer 'Quartz' controllers. - Handle failure during initalisation more gracefully. - Create a watchdog thread to poke the board and check for status changes. - Make ready for > 64kB transfers.
1.1	30-Jan-2002	ad	branches: 1.1.2; 1.1.4; Bare-bones driver for AMI RAID. Parts taken from FreeBSD. This was tried a good while ago and it had problems under load. Changes were made to address that, but I don't have the ability to test them. So, I'm committing it before it rots.
1.1.4.2	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.1	30-Jan-2002	nathanw	file amrreg.h was added on branch nathanw_sa on 2002-02-28 04:13:57 +0000
1.1.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.1	30-Jan-2002	jdolecek	file amrreg.h was added on branch kqueue on 2002-02-11 20:09:55 +0000
1.2.36.1	11-Aug-2006	yamt	sync with head
1.2.32.1	09-Sep-2006	rpaulo	sync with head
1.2.18.1	30-Dec-2006	yamt	sync with head.
1.2.16.1	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.3.58.2	04-May-2009	yamt	sync with head.
1.3.58.1	16-May-2008	yamt	sync with head.
1.3.56.1	18-May-2008	yamt	sync with head.
1.3.54.2	28-Sep-2008	mjf	Sync with HEAD.
1.3.54.1	02-Jun-2008	mjf	Sync with HEAD.
1.4.6.1	19-Oct-2008	haad	Sync with HEAD.
1.4.2.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.10	02-Mar-2015	christos	From Tom Ivar Helbekkmo: - Use mutexes and condvars instead of spl() calls. - Add a couple of bugfixes from FreeBSD (the easily observed ones). - Add a simple watchdog thread for the newer versions of the hardware, that tickles the controller at task submission time if it seems not ready (from FreeBSD). - Add a buffer allocation size hack to the ioctl handler, to work around a firmware bug in some versions of the controller (also from FreeBSD). - Rearrange the bus_dmamap_sync() calls to make the use match the documentation.
1.9	27-Jul-2012	jakllsch	branches: 1.9.2; 1.9.16; Convert amr(4) to device_t (CFATTACH_DECL_NEW).
1.8	28-Apr-2008	martin	branches: 1.8.34; Remove clause 3 and 4 from TNF licenses
1.7	09-Jul-2007	ad	branches: 1.7.28; 1.7.30; 1.7.32; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.6	23-Jul-2006	bouyer	branches: 1.6.14; 1.6.16; Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.5	11-Dec-2005	christos	branches: 1.5.4; 1.5.8; merge ktrace-lwp.
1.4	13-Sep-2004	drochner	branches: 1.4.10; 1.4.12; a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.3	14-May-2003	ad	branches: 1.3.2; Fix thinko.
1.2	04-May-2003	ad	- Fix the mailbox protocol. Not tested with newer 'Quartz' controllers. - Handle failure during initalisation more gracefully. - Create a watchdog thread to poke the board and check for status changes. - Make ready for > 64kB transfers.
1.1	30-Jan-2002	ad	branches: 1.1.2; 1.1.4; Bare-bones driver for AMI RAID. Parts taken from FreeBSD. This was tried a good while ago and it had problems under load. Changes were made to address that, but I don't have the ability to test them. So, I'm committing it before it rots.
1.1.4.2	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.1	30-Jan-2002	nathanw	file amrvar.h was added on branch nathanw_sa on 2002-02-28 04:13:57 +0000
1.1.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.1	30-Jan-2002	jdolecek	file amrvar.h was added on branch kqueue on 2002-02-11 20:09:55 +0000
1.3.2.2	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.2.1	18-Sep-2004	skrll	Sync with HEAD.
1.4.12.2	03-Sep-2007	yamt	sync with head.
1.4.12.1	30-Dec-2006	yamt	sync with head.
1.4.10.1	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.5.8.1	11-Aug-2006	yamt	sync with head
1.5.4.1	09-Sep-2006	rpaulo	sync with head
1.6.16.1	11-Jul-2007	mjf	Sync with head.
1.6.14.1	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.7.32.1	16-May-2008	yamt	sync with head.
1.7.30.1	18-May-2008	yamt	sync with head.
1.7.28.1	02-Jun-2008	mjf	Sync with HEAD.
1.8.34.1	30-Oct-2012	yamt	sync with head
1.9.16.1	06-Apr-2015	skrll	Sync with HEAD
1.9.2.1	03-Dec-2017	jdolecek	update from HEAD
1.3	05-Dec-2007	xtraeme	branches: 1.3.8; Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.2	04-Dec-2007	xtraeme	Use a rwlock(4) for arc_lock/unlock. arc_msgbuf() is too fragile and we want to be sure that only one consumer is using it at a time.
1.1	04-Dec-2007	xtraeme	Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD. arc0 at pci2 dev 14 function 0: interrupting at ioapic0 pin 18 (irq 5) arc0: Areca ARC-1210 Host Adapter RAID controller arc0: 4 ports, 256MB SDRAM, firmware <V1.43 2007-4-17> scsibus0 at arc0: 16 targets, 8 luns per target [...] scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <Areca, ARC-1210-VOL#00, R001> disk fixed sd0: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors sd1 at scsibus0 target 0 lun 1: <Areca, ARC-1210-VOL#01, R001> disk fixed sd1: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors bioctl(4) output with two RAID0 volumes: Volume Status Size Device arc0 0 Online 466G ARC-1210-VOL#00 RAID0 0 Online 234G 0:0.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:1.0 noencl <WDC WD2500YS-01SHB1 20.06C06> arc0 1 Online 466G ARC-1210-VOL#01 RAID0 0 Online 234G 0:2.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:3.0 noencl <WDC WD2500YS-01SHB1 20.06C06> The driver still needs changes related to locking and talking to the firmware, which sometimes is unable to answer... Raid card donated by Areca Technology Corporation via Trent George. Disks used for testing were contributed by TNF. Thank you very much.
1.3.8.2	05-Dec-2007	xtraeme	Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.3.8.1	05-Dec-2007	xtraeme	file arc.c was added on branch mjf-devfs on 2007-12-05 00:18:08 +0000
1.45	09-Feb-2024	andvar	s/firwmare/firmware/ in comments.
1.44	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.43	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.42	13-Jun-2021	mlelstv	branches: 1.42.2; Fix race when freeing sensors.
1.41	24-Apr-2021	thorpej	branches: 1.41.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.40	01-Oct-2019	chs	branches: 1.40.10; in many device attach paths, allocate memory with KM_SLEEP instead of KM_NOSLEEP and remove code to handle failures that can no longer happen.
1.39	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.38	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.37	12-Aug-2017	mlelstv	branches: 1.37.2; 1.37.4; tag arcmsr scsipi adapter and pci interrupt as MPSAFE. I/O is now done without taking kernel lock.
1.36	19-Jun-2016	dholland	more NBIO > 0 (not sure why test-compiling yesterday didn't expose these, probably improper cflags)
1.35	19-Jun-2016	dholland	Broaden the #if NBIO > 0 block. Should fix broken evbppc build.
1.34	12-Jun-2016	christos	Make internal functions static to avoid conflicts with arc_* from zfs.
1.33	02-May-2016	christos	move scsipi_strvis -> libkern:strnvisx() change the prototype to match userland fix sizes of strings passed to it
1.32	12-Mar-2015	christos	Dedup the conversion of bioc_disk and bioc_vol to envsys_data_t
1.31	29-Mar-2014	christos	branches: 1.31.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.30	20-Jun-2011	pgoyette	branches: 1.30.2; 1.30.12; 1.30.16; Initialize current value for ENVSYS_DRIVE sensors
1.29	20-Jun-2011	pgoyette	Initialize sensors states before registering.
1.28	20-Jun-2011	pgoyette	Maintain our own storage for volume and disk ID info rather than (ab)using envstat's storage. XXX Compile-tested only.
1.27	04-Jun-2011	pgoyette	Update to use the new name for the sensor's private data. XXX This driver still uses the sensor's value_min for an unrelated XXX purpose, but changing that will have to wait for another day.
1.26	13-Nov-2010	uebayasi	branches: 1.26.2; 1.26.6; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.25	03-Apr-2010	jruoho	Only unregister the sensors if these were successfully installed; sysmon_envsys_unregister() does not check for NULL pointers gracefully.
1.24	03-Apr-2010	pgoyette	If we're detaching the device, make sure we unregister the sensors.
1.23	14-Mar-2010	pgoyette	branches: 1.23.2; Remove setting of edata->monitor since that member no longer exists.
1.22	23-Sep-2008	christos	branches: 1.22.14; PR/39583: Brad du Plessis: acrmsr(4) driver doesn't report number of volumes correctly to bioctl(8) PR/39584: Juan RP: arcmsr(4) driver disk state values are incorrect (for ARC-1220)
1.21	24-Jun-2008	gmcgarry	branches: 1.21.2; Replace gcc variadic macros with c99 variadic macros.
1.20	03-Apr-2008	xtraeme	branches: 1.20.4; 1.20.6; 1.20.8; 1.20.10; Add ARC_FW_DISK_INITIALIZED flag for device_state and do not treat it as unknown; in this state the disk is correctly working so set it as online. Also remove some bogus htole32() for diskinfo->device_state. This should fix PR kern/38299 by Hiroyuki Bessho.
1.19	05-Mar-2008	xtraeme	Split device_t/softc, and other related cosmetic changes.
1.18	03-Mar-2008	xtraeme	Detect and initialize correctly RAID 1+0 levels.
1.17	01-Mar-2008	xtraeme	RAID 1/1+0 volumes may only be created with foreground initialization, use the quick_init member in the ARC_FW_CREATE_VOLUME command code. You can now create RAID 1/1+0 volumes through bioctl(8), but only one volume will be initialized; the other ones that need initialization will wait until the first is done.
1.16	01-Mar-2008	xtraeme	arc_create_sensors: if there are no valid volumes (all passthru volumes), just exit from the kthread.
1.15	29-Feb-2008	xtraeme	arc_bio_inq: ask the firmware only once for the ARC_FW_SYSINFO command code and use the softc values on the next bio requests.
1.14	29-Feb-2008	xtraeme	If a disk has been disconnected in a volume set, mark it as offline and put it on another bus and print "disk missing"... the OpenBSD driver does that but I forgot to re-add it. Also save some values provided by the firmware in the softc and use them in the bio(4) ioctls, this speeds up considerably bioctl(8) and avoids so many firmware commands every time we want the current status.
1.13	29-Feb-2008	xtraeme	arc_refresh_sensors: show rebuilding status.
1.12	29-Feb-2008	xtraeme	arc_bio_disk_novol: don't forget to free diskinfo if arc_msgbuf returns an error.
1.11	28-Feb-2008	xtraeme	If pmf_device_register1() fails for some reason, make it panic(9) rather than printing an error. The shutdown handler is mandatory on this driver to flush controller's cache or stopping background rebuilds.
1.10	28-Feb-2008	xtraeme	Register a sensor per disk connected to a valid volume, so that we can get more details if state on a disk or volume has changed, previously only state about the volume was available. You'll see something like this with envstat(8) now: $ envstat -darcmsr0 RAID 5 volume0 (sd0): building disk0 volume0 (sd0): online disk1 volume0 (sd0): online disk2 volume0 (sd0): online disk3 volume0 (sd0): online $ Convert the driver to use pmf(9) and register a shutdown power handler via pmf_device_register1() that was added recently. Misc cosmetics tweaks while I'm here.
1.9	02-Jan-2008	xtraeme	branches: 1.9.2; 1.9.4; 1.9.6; 1.9.10; New functionality for arcmsr(4) native management via bioctl(8): Added support to create/remove hot-spare, pass-through disks. Added support to create/remove volume sets. Added support to show information about physical disks, even if they are marked unused, hot-spares or pass-through. sd(4) devices are attached/detached automagically when a pass-through disk or volume set is created/removed... thanks scsipi(9) and cube@ for hints.
1.8	07-Dec-2007	xtraeme	branches: 1.8.2; 1.8.4; 1.8.6; 1.8.10; 1.8.12; Add BIOC_SVMIGRATING to bio(4) and bioctl(8) to report if a volume is migrating currently showing the percentage. Update arcmsr(4) to report this, like: $ sudo ./bioctl -h arcmsr0 Volume Status Size Device arcmsr0 0 Migrating 698G ARC-1210-VOL#00 RAID 5 7% done 0 Online 234G 0:0.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:1.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 2 Online 234G 0:2.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 3 Online 234G 0:3.0 noencl <WDC WD2500YS-01SHB1 20.06C06> $
1.7	07-Dec-2007	xtraeme	- Improve the locking based in comments from ad@. - malloc(9) -> kmem(9) now that most of the bio code doesn't need to run on interrupt context. - Reduce code that runs in interrupt context to a small part in arc_msgbuf() and is protected by arc_lock()/arc_unlock(). Reviewed and help by ad@.
1.6	05-Dec-2007	gmcgarry	Fix locking botch.
1.5	05-Dec-2007	xtraeme	Boh! I spoke too soon before, without the rwlock(9) sometimes arc_wait() waits forever, so put it back until I figure any other way to remove the rwlock(9).
1.4	05-Dec-2007	xtraeme	Use SCSIPI_CHAN_NOSETTLE... the driver doesn't have to wait.
1.3	05-Dec-2007	xtraeme	Make it use mutex(9) and condvar(9), bye bye spl(9) and tsleep(9). Fully stable with all debugging options turned on, unless someday any problem appears :-)
1.2	05-Dec-2007	xtraeme	- arc_msgbuf: free wbuf and rbuf with the correct type (M_TEMP vs M_DEVBUF) this fixes a panic with debugging options. - Do not use a callout to refresh sensor data, and make it available every time someone requests it. - Enable ENVSYS_FMONSTCHANGED for notifications in the volumes.
1.1	05-Dec-2007	xtraeme	Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.8.12.1	08-Jan-2008	bouyer	Sync with HEAD
1.8.10.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.8.10.1	07-Dec-2007	jmcneill	file arcmsr.c was added on branch jmcneill-pm on 2007-12-09 19:37:52 +0000
1.8.6.3	18-Feb-2008	mjf	Sync with HEAD.
1.8.6.2	08-Dec-2007	mjf	Sync with HEAD.
1.8.6.1	07-Dec-2007	mjf	file arcmsr.c was added on branch mjf-devfs on 2007-12-08 18:19:41 +0000
1.8.4.2	08-Dec-2007	ad	Sync with head.
1.8.4.1	07-Dec-2007	ad	file arcmsr.c was added on branch vmlocking2 on 2007-12-08 17:57:25 +0000
1.8.2.4	17-Mar-2008	yamt	sync with head.
1.8.2.3	21-Jan-2008	yamt	sync with head
1.8.2.2	07-Dec-2007	yamt	sync with head
1.8.2.1	07-Dec-2007	yamt	file arcmsr.c was added on branch yamt-lazymbuf on 2007-12-07 17:30:22 +0000
1.9.10.4	28-Sep-2008	mjf	Sync with HEAD.
1.9.10.3	29-Jun-2008	mjf	Sync with HEAD.
1.9.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.9.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.9.6.1	24-Mar-2008	keiichi	sync with head.
1.9.4.4	24-Sep-2008	bouyer	Pull up following revision(s) (requested by christos in ticket #1205): sys/dev/pci/arcmsrvar.h: revision 1.13 sys/dev/pci/arcmsr.c: revision 1.22 PR/39583: Brad du Plessis: acrmsr(4) driver doesn't report number of volumes correctly to bioctl(8) PR/39584: Juan RP: arcmsr(4) driver disk state values are incorrect (for ARC-1220)
1.9.4.3	29-Aug-2008	bouyer	Apply patch, requested by christos in tickets #1175: sys/dev/bio.c patch sys/dev/biovar.h patch sys/dev/pci/arcmsr.c patch sys/dev/pci/arcmsrvar.h patch sbin/bioctl/Makefile patch sbin/bioctl/bioctl.8 patch sbin/bioctl/bioctl.c patch sbin/bioctl/dehumanize_number.c patch sbin/bioctl/strtonum.c patch sbin/bioctl/strtonum.h patch Port bioctl(8) and arcmsr(4) from current: bioctl(8): Rewritten to handle new features like creating/removing hot-spare, pass-through disks and RAID volumes, start/stop consistency checks in volumes and others. arcmsr(4): Added support to create/remove hot-spare, pass-through disks and RAID volumes, start/stop consistency checks in volumes as well as showing information about physical disks (even if they are marked as hot-spare, pass-through or unused).
1.9.4.2	11-Jan-2008	xtraeme	Pull up following revision(s) (requested by christos in ticket #1037): sys/dev/pci/arcmsr.c: patch sys/dev/pci/arcmsrvar.h: patch sys/dev/pci/pcidevs: patch sys/dev/pci/files.pci: patch sys/arch/i386/conf/GENERIC: patch sys/arch/i386/conf/XEN2_DOM0: patch sys/arch/i386/conf/INSTALL_LARGE: patch sys/arch/amd64/conf/GENERIC: patch sys/arch/amd64/conf/INSTALL: patch share/man/man4/Makefile: patch share/man/man4/arcmsr.4: patch distrib/sets/lists/man/mi: patch Add the Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD.
1.9.4.1	02-Jan-2008	xtraeme	file arcmsr.c was added on branch netbsd-4 on 2008-01-11 17:03:16 +0000
1.9.2.3	23-Mar-2008	matt	sync with HEAD
1.9.2.2	09-Jan-2008	matt	sync with HEAD
1.9.2.1	02-Jan-2008	matt	file arcmsr.c was added on branch matt-armv6 on 2008-01-09 01:53:32 +0000
1.20.10.1	27-Jun-2008	simonb	Sync with head.
1.20.8.3	04-Sep-2008	skrll	Sync with netbsd-4.
1.20.8.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.20.8.1	03-Apr-2008	skrll	file arcmsr.c was added on branch wrstuden-fixsa on 2008-06-03 20:47:23 +0000
1.20.6.2	10-Oct-2008	skrll	Sync with HEAD.
1.20.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.20.4.2	11-Aug-2010	yamt	sync with head.
1.20.4.1	04-May-2009	yamt	sync with head.
1.21.2.1	19-Oct-2008	haad	Sync with HEAD.
1.22.14.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.23.2.3	12-Jun-2011	rmind	sync with head
1.23.2.2	05-Mar-2011	rmind	sync with head
1.23.2.1	30-May-2010	rmind	sync with head
1.26.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.26.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.30.16.1	18-May-2014	rmind	sync with head
1.30.12.3	03-Dec-2017	jdolecek	update from HEAD
1.30.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.30.12.1	02-Dec-2012	tls	Areca controller maximum-transfer size limits -- looked up in FreeBSD driver. Actually exporting the per-volume transfer size for this driver will be quite hard; the controller firmware shows the host the configured volumes when the host probes the controller's "scsibus", and we can't really associate them with particular RAID volumes until later. I think we will have to intercept the SCSI inquiry commands -- yuck.
1.30.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.31.6.4	28-Aug-2017	skrll	Sync with HEAD
1.31.6.3	09-Jul-2016	skrll	Sync with HEAD
1.31.6.2	29-May-2016	skrll	Sync with HEAD
1.31.6.1	06-Apr-2015	skrll	Sync with HEAD
1.37.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.37.4.1	10-Jun-2019	christos	Sync with HEAD
1.37.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.37.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.40.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.41.2.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.42.2.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.16	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.15	12-Jun-2016	christos	Make internal functions static to avoid conflicts with arc_* from zfs.
1.14	20-Jun-2011	pgoyette	branches: 1.14.12; 1.14.30; Maintain our own storage for volume and disk ID info rather than (ab)using envstat's storage. XXX Compile-tested only.
1.13	23-Sep-2008	christos	branches: 1.13.26; PR/39583: Brad du Plessis: acrmsr(4) driver doesn't report number of volumes correctly to bioctl(8) PR/39584: Juan RP: arcmsr(4) driver disk state values are incorrect (for ARC-1220)
1.12	03-Apr-2008	xtraeme	branches: 1.12.4; 1.12.6; 1.12.8; 1.12.12; Add ARC_FW_DISK_INITIALIZED flag for device_state and do not treat it as unknown; in this state the disk is correctly working so set it as online. Also remove some bogus htole32() for diskinfo->device_state. This should fix PR kern/38299 by Hiroyuki Bessho.
1.11	05-Mar-2008	xtraeme	Split device_t/softc, and other related cosmetic changes.
1.10	29-Feb-2008	xtraeme	If a disk has been disconnected in a volume set, mark it as offline and put it on another bus and print "disk missing"... the OpenBSD driver does that but I forgot to re-add it. Also save some values provided by the firmware in the softc and use them in the bio(4) ioctls, this speeds up considerably bioctl(8) and avoids so many firmware commands every time we want the current status.
1.9	28-Feb-2008	xtraeme	Register a sensor per disk connected to a valid volume, so that we can get more details if state on a disk or volume has changed, previously only state about the volume was available. You'll see something like this with envstat(8) now: $ envstat -darcmsr0 RAID 5 volume0 (sd0): building disk0 volume0 (sd0): online disk1 volume0 (sd0): online disk2 volume0 (sd0): online disk3 volume0 (sd0): online $ Convert the driver to use pmf(9) and register a shutdown power handler via pmf_device_register1() that was added recently. Misc cosmetics tweaks while I'm here.
1.8	03-Jan-2008	xtraeme	branches: 1.8.2; 1.8.4; 1.8.6; 1.8.10; Fix indentation.
1.7	02-Jan-2008	xtraeme	New functionality for arcmsr(4) native management via bioctl(8): Added support to create/remove hot-spare, pass-through disks. Added support to create/remove volume sets. Added support to show information about physical disks, even if they are marked unused, hot-spares or pass-through. sd(4) devices are attached/detached automagically when a pass-through disk or volume set is created/removed... thanks scsipi(9) and cube@ for hints.
1.6	09-Dec-2007	xtraeme	branches: 1.6.4; 1.6.6; Erm fix some defines in previous...
1.5	08-Dec-2007	xtraeme	branches: 1.5.2; 1.5.4; Add some more firmware command/status codes from the freebsd driver.
1.4	05-Dec-2007	xtraeme	branches: 1.4.2; Boh! I spoke too soon before, without the rwlock(9) sometimes arc_wait() waits forever, so put it back until I figure any other way to remove the rwlock(9).
1.3	05-Dec-2007	xtraeme	Make it use mutex(9) and condvar(9), bye bye spl(9) and tsleep(9). Fully stable with all debugging options turned on, unless someday any problem appears :-)
1.2	05-Dec-2007	xtraeme	- arc_msgbuf: free wbuf and rbuf with the correct type (M_TEMP vs M_DEVBUF) this fixes a panic with debugging options. - Do not use a callout to refresh sensor data, and make it available every time someone requests it. - Enable ENVSYS_FMONSTCHANGED for notifications in the volumes.
1.1	05-Dec-2007	xtraeme	Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.4.2.4	17-Mar-2008	yamt	sync with head.
1.4.2.3	21-Jan-2008	yamt	sync with head
1.4.2.2	07-Dec-2007	yamt	sync with head
1.4.2.1	05-Dec-2007	yamt	file arcmsrvar.h was added on branch yamt-lazymbuf on 2007-12-07 17:30:23 +0000
1.5.4.4	18-Feb-2008	mjf	Sync with HEAD.
1.5.4.3	27-Dec-2007	mjf	Sync with HEAD.
1.5.4.2	08-Dec-2007	mjf	Sync with HEAD.
1.5.4.1	08-Dec-2007	mjf	file arcmsrvar.h was added on branch mjf-devfs on 2007-12-08 18:19:41 +0000
1.5.2.3	26-Dec-2007	ad	Sync with head.
1.5.2.2	08-Dec-2007	ad	Sync with head.
1.5.2.1	08-Dec-2007	ad	file arcmsrvar.h was added on branch vmlocking2 on 2007-12-08 17:57:26 +0000
1.6.6.1	08-Jan-2008	bouyer	Sync with HEAD
1.6.4.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.6.4.1	09-Dec-2007	jmcneill	file arcmsrvar.h was added on branch jmcneill-pm on 2007-12-09 19:37:53 +0000
1.8.10.3	28-Sep-2008	mjf	Sync with HEAD.
1.8.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.8.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.8.6.1	24-Mar-2008	keiichi	sync with head.
1.8.4.4	24-Sep-2008	bouyer	Pull up following revision(s) (requested by christos in ticket #1205): sys/dev/pci/arcmsrvar.h: revision 1.13 sys/dev/pci/arcmsr.c: revision 1.22 PR/39583: Brad du Plessis: acrmsr(4) driver doesn't report number of volumes correctly to bioctl(8) PR/39584: Juan RP: arcmsr(4) driver disk state values are incorrect (for ARC-1220)
1.8.4.3	29-Aug-2008	bouyer	Apply patch, requested by christos in tickets #1175: sys/dev/bio.c patch sys/dev/biovar.h patch sys/dev/pci/arcmsr.c patch sys/dev/pci/arcmsrvar.h patch sbin/bioctl/Makefile patch sbin/bioctl/bioctl.8 patch sbin/bioctl/bioctl.c patch sbin/bioctl/dehumanize_number.c patch sbin/bioctl/strtonum.c patch sbin/bioctl/strtonum.h patch Port bioctl(8) and arcmsr(4) from current: bioctl(8): Rewritten to handle new features like creating/removing hot-spare, pass-through disks and RAID volumes, start/stop consistency checks in volumes and others. arcmsr(4): Added support to create/remove hot-spare, pass-through disks and RAID volumes, start/stop consistency checks in volumes as well as showing information about physical disks (even if they are marked as hot-spare, pass-through or unused).
1.8.4.2	11-Jan-2008	xtraeme	Pull up following revision(s) (requested by christos in ticket #1037): sys/dev/pci/arcmsr.c: patch sys/dev/pci/arcmsrvar.h: patch sys/dev/pci/pcidevs: patch sys/dev/pci/files.pci: patch sys/arch/i386/conf/GENERIC: patch sys/arch/i386/conf/XEN2_DOM0: patch sys/arch/i386/conf/INSTALL_LARGE: patch sys/arch/amd64/conf/GENERIC: patch sys/arch/amd64/conf/INSTALL: patch share/man/man4/Makefile: patch share/man/man4/arcmsr.4: patch distrib/sets/lists/man/mi: patch Add the Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD.
1.8.4.1	03-Jan-2008	xtraeme	file arcmsrvar.h was added on branch netbsd-4 on 2008-01-11 17:03:17 +0000
1.8.2.3	23-Mar-2008	matt	sync with HEAD
1.8.2.2	09-Jan-2008	matt	sync with HEAD
1.8.2.1	03-Jan-2008	matt	file arcmsrvar.h was added on branch matt-armv6 on 2008-01-09 01:53:33 +0000
1.12.12.1	19-Oct-2008	haad	Sync with HEAD.
1.12.8.3	04-Sep-2008	skrll	Sync with netbsd-4.
1.12.8.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.12.8.1	03-Apr-2008	skrll	file arcmsrvar.h was added on branch wrstuden-fixsa on 2008-06-03 20:47:24 +0000
1.12.6.1	10-Oct-2008	skrll	Sync with HEAD.
1.12.4.1	04-May-2009	yamt	sync with head.
1.13.26.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.14.30.1	09-Jul-2016	skrll	Sync with HEAD
1.14.12.2	03-Dec-2017	jdolecek	update from HEAD
1.14.12.1	02-Dec-2012	tls	Areca controller maximum-transfer size limits -- looked up in FreeBSD driver. Actually exporting the per-volume transfer size for this driver will be quite hard; the controller firmware shows the host the configured volumes when the host probes the controller's "scsibus", and we can't really associate them with particular RAID volumes until later. I think we will have to intercept the SCSI inquiry commands -- yuck.
1.3	05-Dec-2007	xtraeme	branches: 1.3.8; Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.2	04-Dec-2007	xtraeme	Use a rwlock(4) for arc_lock/unlock. arc_msgbuf() is too fragile and we want to be sure that only one consumer is using it at a time.
1.1	04-Dec-2007	xtraeme	Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD. arc0 at pci2 dev 14 function 0: interrupting at ioapic0 pin 18 (irq 5) arc0: Areca ARC-1210 Host Adapter RAID controller arc0: 4 ports, 256MB SDRAM, firmware <V1.43 2007-4-17> scsibus0 at arc0: 16 targets, 8 luns per target [...] scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <Areca, ARC-1210-VOL#00, R001> disk fixed sd0: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors sd1 at scsibus0 target 0 lun 1: <Areca, ARC-1210-VOL#01, R001> disk fixed sd1: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors bioctl(4) output with two RAID0 volumes: Volume Status Size Device arc0 0 Online 466G ARC-1210-VOL#00 RAID0 0 Online 234G 0:0.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:1.0 noencl <WDC WD2500YS-01SHB1 20.06C06> arc0 1 Online 466G ARC-1210-VOL#01 RAID0 0 Online 234G 0:2.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:3.0 noencl <WDC WD2500YS-01SHB1 20.06C06> The driver still needs changes related to locking and talking to the firmware, which sometimes is unable to answer... Raid card donated by Areca Technology Corporation via Trent George. Disks used for testing were contributed by TNF. Thank you very much.
1.3.8.2	05-Dec-2007	xtraeme	Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.3.8.1	05-Dec-2007	xtraeme	file arcvar.h was added on branch mjf-devfs on 2007-12-05 00:18:08 +0000
1.30	23-Jan-2023	andvar	s/cyliner/cylinder/ and s/specturm/spectrum/ in messages.
1.29	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.28	20-Oct-2017	jdolecek	branches: 1.28.2; 1.28.4; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.27	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.26	29-Mar-2014	christos	branches: 1.26.18; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.25	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.24	31-Jul-2012	bouyer	branches: 1.24.2; 1.24.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.23	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.22	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.21	04-Apr-2011	dyoung	branches: 1.21.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.20	05-Nov-2010	jakllsch	branches: 1.20.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.19	26-Nov-2009	njoly	branches: 1.19.2; 1.19.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.18	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.17	18-Mar-2008	cube	branches: 1.17.2; 1.17.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.16	19-Jul-2007	dsl	branches: 1.16.6; 1.16.22; 1.16.26; 1.16.28; include cdefs.h before any opt_xxx.h
1.15	09-Feb-2007	ad	branches: 1.15.6; 1.15.14; Merge newlock2 to head.
1.14	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.13	25-Oct-2006	bouyer	Use common wdc_sataprobe() where possible; explain why we can't in satalink.c
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	26-Jun-2006	xtraeme	branches: 1.11.4; 1.11.6; Add missing __KERNEL_RCSID()
1.10	29-Mar-2006	thorpej	branches: 1.10.4; Use device_cfdata().
1.9	16-Jan-2006	bouyer	branches: 1.9.2; 1.9.4; 1.9.6; 1.9.8; 1.9.10; properly use ata_channel->ch_ndrive: - initialize it properly in the bus front-ends (all 2, exept in wdc_pcmcia.c for the "Sandisk CompactFlash Card" where it's set to 1) - remplace hardcoded '2' by ata_channel->ch_ndrive in MI IDE drivers. From Christos Zoulas in kern/32501.
1.8	11-Dec-2005	christos	branches: 1.8.2; merge ktrace-lwp.
1.7	27-Feb-2005	perry	branches: 1.7.4; nuke trailing whitespace
1.6	12-Feb-2005	rearnsha	Change PCIIDE_I31244_ENABLEDMA to PCIIDE_I31244_DISABLEDMA as suggested by briggs. Put the flag in opt_pciide.h.
1.5	11-Feb-2005	rearnsha	Add support for the Artisea device operating in DPA mode. Approved by briggs.
1.4	20-Aug-2004	thorpej	branches: 1.4.4; 1.4.6; Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.3	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.2	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.1	14-Dec-2003	thorpej	branches: 1.1.4; Split the Intel i31244 SATA controller ("Artisea") driver out of piixide into its own driver (artsata).
1.1.4.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.6	15-Feb-2005	skrll	Sync with HEAD.
1.1.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	14-Dec-2003	skrll	file artsata.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.4.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.6.1	12-Feb-2005	yamt	sync with head.
1.4.4.1	29-Apr-2005	kent	sync with -current
1.7.4.5	24-Mar-2008	yamt	sync with head.
1.7.4.4	03-Sep-2007	yamt	sync with head.
1.7.4.3	26-Feb-2007	yamt	sync with head.
1.7.4.2	30-Dec-2006	yamt	sync with head.
1.7.4.1	21-Jun-2006	yamt	sync with head.
1.8.2.1	01-Feb-2006	yamt	sync with head.
1.9.10.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.9.8.1	19-Apr-2006	elad	sync with head.
1.9.6.2	11-Aug-2006	yamt	sync with head
1.9.6.1	01-Apr-2006	yamt	sync with head.
1.9.4.1	22-Apr-2006	simonb	Sync with head.
1.9.2.1	09-Sep-2006	rpaulo	sync with head
1.10.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.11.6.2	10-Dec-2006	yamt	sync with head.
1.11.6.1	22-Oct-2006	yamt	sync with head
1.11.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.11.4.1	18-Nov-2006	ad	Sync with head.
1.15.14.1	15-Aug-2007	skrll	Sync with HEAD.
1.15.6.1	12-Oct-2007	ad	Fix merge errors.
1.16.28.2	19-Jul-2007	dsl	include cdefs.h before any opt_xxx.h
1.16.28.1	19-Jul-2007	dsl	file artsata.c was added on branch matt-mips64 on 2007-07-19 21:53:16 +0000
1.16.26.2	02-Jun-2008	mjf	Sync with HEAD.
1.16.26.1	03-Apr-2008	mjf	Sync with HEAD.
1.16.22.1	24-Mar-2008	keiichi	sync with head.
1.16.6.1	23-Mar-2008	matt	sync with HEAD
1.17.4.2	11-Mar-2010	yamt	sync with head
1.17.4.1	16-May-2008	yamt	sync with head.
1.17.2.1	18-May-2008	yamt	sync with head.
1.19.4.2	21-Apr-2011	rmind	sync with head
1.19.4.1	05-Mar-2011	rmind	sync with head
1.19.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.20.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.21.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.21.4.1	30-Oct-2012	yamt	sync with head
1.24.4.1	18-May-2014	rmind	sync with head
1.24.2.3	03-Dec-2017	jdolecek	update from HEAD
1.24.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.24.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.26.18.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.28.4.1	10-Jun-2019	christos	Sync with HEAD
1.28.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.16	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.15	30-Nov-2018	jmcneill	Add support for polling com devices when no interrupt is available.
1.14	29-Mar-2014	christos	branches: 1.14.28; 1.14.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.13	27-Oct-2012	chs	branches: 1.13.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.12	13-Nov-2010	uebayasi	branches: 1.12.8; 1.12.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.11	26-Nov-2009	njoly	branches: 1.11.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.10	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.9	16-Apr-2008	cegger	branches: 1.9.2; 1.9.4; device_t / softc split. From Hans Rosenfeld.
1.8	15-Apr-2008	cegger	device_t / softc split. From Hans Rosenfeld
1.7	10-Apr-2008	cegger	use aprint_*_dev and device_xname
1.6	19-Oct-2007	ad	branches: 1.6.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	04-Mar-2007	christos	branches: 1.5.2; 1.5.14; 1.5.16; 1.5.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.4	11-Dec-2005	christos	branches: 1.4.26; merge ktrace-lwp.
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	10-Feb-2004	jdolecek	branches: 1.2.4; 1.2.10; 1.2.12; checkpoint DMA support; still disabled, not tested yet
1.1	25-Jan-2004	jdolecek	add puc(4) atppc(4) attachment; probes correctly, but needs further work to support DMA, and some actual transfer testing
1.2.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.10.1	29-Apr-2005	kent	sync with -current
1.2.4.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.4.2	03-Aug-2004	skrll	Sync with HEAD
1.2.4.1	10-Feb-2004	skrll	file atppc_puc.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.3.4.2	27-Oct-2007	yamt	sync with head.
1.3.4.1	03-Sep-2007	yamt	sync with head.
1.4.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.5.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.16.1	06-Nov-2007	matt	sync with HEAD
1.5.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.2.1	23-Oct-2007	ad	Sync with head.
1.6.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.9.4.2	11-Mar-2010	yamt	sync with head
1.9.4.1	16-May-2008	yamt	sync with head.
1.9.2.1	18-May-2008	yamt	sync with head.
1.11.4.1	05-Mar-2011	rmind	sync with head
1.12.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.12.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.12.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.8.1	30-Oct-2012	yamt	sync with head
1.13.2.1	18-May-2014	rmind	sync with head
1.14.30.1	10-Jun-2019	christos	Sync with HEAD
1.14.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.39	29-Feb-2020	isaki	round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.38	08-Jun-2019	isaki	branches: 1.38.2; 1.38.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.37	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.36	16-Mar-2019	isaki	branches: 1.36.2; Use C99 style struct initializer to audio_hw_if.
1.35	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.34	23-Jun-2018	maxv	branches: 1.34.2; constify
1.33	01-Jun-2017	chs	branches: 1.33.8; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.32	29-Mar-2014	christos	branches: 1.32.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.31	27-Oct-2012	chs	branches: 1.31.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.30	24-Nov-2011	mrg	branches: 1.30.8; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.29	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.28	13-Nov-2010	uebayasi	branches: 1.28.8; 1.28.10; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.27	24-Feb-2010	dyoung	branches: 1.27.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.26	08-Jan-2010	dyoung	branches: 1.26.2; Expand PMF_FN_* macros.
1.25	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.24	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.23	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.22	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.21	28-Apr-2008	martin	branches: 1.21.12; 1.21.14; Remove clause 3 and 4 from TNF licenses
1.20	10-Apr-2008	cegger	branches: 1.20.2; 1.20.4; use aprint_*_dev and device_xname
1.19	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.18	09-Dec-2007	jmcneill	branches: 1.18.6; 1.18.10; Merge jmcneill-pm branch.
1.17	19-Oct-2007	ad	branches: 1.17.4; 1.17.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	04-Mar-2007	christos	branches: 1.16.2; 1.16.14; 1.16.16; 1.16.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.15	16-Nov-2006	christos	branches: 1.15.4; 1.15.6; __unused removal on arguments; approved by core.
1.14	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.13	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.12	27-Aug-2006	christos	branches: 1.12.2; 1.12.4; add missing initializer
1.11	11-Dec-2005	christos	branches: 1.11.4; 1.11.8; merge ktrace-lwp.
1.10	28-Jun-2005	thorpej	branches: 1.10.2; Use ANSI function decls and static.
1.9	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.8	17-Jan-2005	augustss	Make it compile with debug.
1.7	15-Jan-2005	kent	branches: 1.7.2; ansify and KNF
1.6	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.5	16-Nov-2004	augustss	branches: 1.5.2; Add RECORD to the conversion table, otherwise we panic on boot. The fact that we cannot record (yet) is found out later anyway.
1.4	13-Nov-2004	kent	adopt auconv_set_converter()
1.3	10-Nov-2004	kent	move IS_FIXED_RATE(), SUPPORTS_4CH(), and SUPPORTS_6CH to ac97var.h
1.2	09-Nov-2004	kent	remove unnecessary ac97-related code
1.1	10-Oct-2004	augustss	branches: 1.1.2; Add a driver for the Acer Labs M5455 audio controller. This controller is integrated in some ALi Southbridges. XXX This driver is incomplete and slightly buggy, but it works enough to enable me to listen to music on my Sharp MM20.
1.1.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.6	24-Jan-2005	skrll	Sync with HEAD.
1.1.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.1.2.4	29-Nov-2004	skrll	Sync with HEAD.
1.1.2.3	14-Nov-2004	skrll	Sync with HEAD.
1.1.2.2	19-Oct-2004	skrll	Sync with HEAD
1.1.2.1	10-Oct-2004	skrll	file auacer.c was added on branch ktrace-lwp on 2004-10-19 15:56:57 +0000
1.5.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.5.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.7.2.1	29-Apr-2005	kent	sync with -current
1.10.2.5	17-Mar-2008	yamt	sync with head.
1.10.2.4	21-Jan-2008	yamt	sync with head
1.10.2.3	27-Oct-2007	yamt	sync with head.
1.10.2.2	03-Sep-2007	yamt	sync with head.
1.10.2.1	30-Dec-2006	yamt	sync with head.
1.11.8.1	03-Sep-2006	yamt	sync with head.
1.11.4.1	09-Sep-2006	rpaulo	sync with head
1.12.4.2	10-Dec-2006	yamt	sync with head.
1.12.4.1	22-Oct-2006	yamt	sync with head
1.12.2.1	18-Nov-2006	ad	Sync with head.
1.15.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.15.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.16.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.16.3	23-Mar-2008	matt	sync with HEAD
1.16.16.2	09-Jan-2008	matt	sync with HEAD
1.16.16.1	06-Nov-2007	matt	sync with HEAD
1.16.14.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.16.14.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.16.14.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.14.1	04-Oct-2007	joerg	Convert to generic PCI power handler.
1.16.2.1	23-Oct-2007	ad	Sync with head.
1.17.6.1	11-Dec-2007	yamt	sync with head.
1.17.4.1	26-Dec-2007	ad	Sync with head.
1.18.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.18.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.18.6.1	24-Mar-2008	keiichi	sync with head.
1.20.4.3	11-Mar-2010	yamt	sync with head
1.20.4.2	16-May-2009	yamt	sync with head
1.20.4.1	16-May-2008	yamt	sync with head.
1.20.2.1	18-May-2008	yamt	sync with head.
1.21.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.21.12.3	13-Dec-2008	ad	Fix locking for ac97.
1.21.12.2	12-Dec-2008	ad	Checkpoint work in progress.
1.21.12.1	08-Dec-2008	ad	Convert some more drivers.
1.26.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.27.2.1	05-Mar-2011	rmind	sync with head
1.28.10.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.28.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.28.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.28.8.2	30-Oct-2012	yamt	sync with head
1.28.8.1	17-Apr-2012	yamt	sync with head
1.30.8.3	03-Dec-2017	jdolecek	update from HEAD
1.30.8.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.30.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.31.2.1	18-May-2014	rmind	sync with head
1.32.6.1	28-Aug-2017	skrll	Sync with HEAD
1.33.8.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.33.8.1	25-Jun-2018	pgoyette	Sync with HEAD
1.34.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.34.2.1	10-Jun-2019	christos	Sync with HEAD
1.36.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.36.2.2	27-Apr-2019	isaki	Adapt to audio2.
1.36.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.38.4.1	29-Feb-2020	ad	Sync with head.
1.38.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by isaki in ticket #792): sys/arch/amiga/dev/aucc.c: revision 1.48 sys/dev/pci/auixp.c: revision 1.49 sys/dev/pci/gcscaudio.c: revision 1.19 sys/dev/pci/auich.c: revision 1.159 sys/dev/sbus/dbri.c: revision 1.42 sys/dev/pci/auvia.c: revision 1.85 sys/dev/pci/auacer.c: revision 1.39 Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017. round_blocksize must return a multiple of the framesize. aucc(4) supports 3 channels mode. round_blocksize must return a multiple of the framesize even if passed blocksize is greater than the upper limit. round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4. round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.5	08-Feb-2018	dholland	Typos.
1.4	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.3	11-Dec-2005	christos	branches: 1.3.70; 1.3.72; 1.3.74; merge ktrace-lwp.
1.2	15-Jan-2005	kent	ansify and KNF
1.1	10-Oct-2004	augustss	branches: 1.1.2; Add a driver for the Acer Labs M5455 audio controller. This controller is integrated in some ALi Southbridges. XXX This driver is incomplete and slightly buggy, but it works enough to enable me to listen to music on my Sharp MM20.
1.1.2.3	17-Jan-2005	skrll	Sync with HEAD.
1.1.2.2	19-Oct-2004	skrll	Sync with HEAD
1.1.2.1	10-Oct-2004	skrll	file auacerreg.h was added on branch ktrace-lwp on 2004-10-19 15:56:57 +0000
1.3.74.1	16-May-2008	yamt	sync with head.
1.3.72.1	18-May-2008	yamt	sync with head.
1.3.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.161	10-May-2023	riastradh	auich(4): Use config_detach_children.
1.160	06-Feb-2021	isaki	Hold sc_intr_lock in open()/close(). vtbl->{lock,unlock} (in fact ac97_{lock,unlock}) don't seem to need releasing an interrupt lock. Confirmed on auich.
1.159	29-Feb-2020	isaki	branches: 1.159.6; round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.158	06-Dec-2019	maxv	branches: 1.158.2; Minor changes, reported by the LGTM bot.
1.157	08-Jun-2019	isaki	branches: 1.157.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.156	11-May-2019	christos	Don't return the raw audio formats; return the ones after we disabled some of them because the device does not support them. Fixes panic in virtualbox.
1.155	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.154	18-Apr-2019	isaki	branches: 1.154.2; White space and indent fix.
1.153	16-Mar-2019	isaki	Use C99 style struct initializer to audio_hw_if.
1.152	24-Mar-2018	jdolecek	branches: 1.152.2; use pci_intr_establish_xname()
1.151	01-Jun-2017	chs	branches: 1.151.8; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.150	25-Jan-2017	martin	auich_round_blocksize: never return 0 as blocksize
1.149	07-Jul-2016	msaitoh	branches: 1.149.2; 1.149.4; KNF. Remove extra spaces. No functional change.
1.148	24-May-2015	christos	PR/49926: Robert Millan: Avoid deadlock condition in auich_read_codec() and auich_write_codec() when running in userspase.
1.147	24-May-2015	christos	PR/49927: Robert Millan: Avoid zerodivide on emulated hardware.
1.146	24-May-2015	christos	use aprint consistently; note that we use aprint_normal_dev in debugging statements instead of aprint_debug_dev because those are controlled by the driver debugging, not the centralized one.
1.145	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.144	10-Aug-2014	tls	branches: 1.144.4; Merge tls-earlyentropy branch into HEAD.
1.143	29-Mar-2014	christos	branches: 1.143.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.142	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.141	02-Jun-2012	dsl	branches: 1.141.2; 1.141.4; Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.140	24-Nov-2011	mrg	put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.139	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.138	13-Nov-2010	uebayasi	branches: 1.138.8; 1.138.10; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.137	24-Feb-2010	dyoung	branches: 1.137.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.136	01-Feb-2010	njoly	branches: 1.136.2; Do not print ac97 link rate message unless verbose boot is requested.
1.135	13-Jan-2010	jakllsch	Revert code I added in 1.133 that touches the PCI_COMMAND_IO_ENABLE and PCI_COMMAND_MEM_ENABLE bits in PCI_COMMAND_STATUS_REG. This causes problems where the BAR is 0. 'options PCI_ADDR_FIXUP' is more appropriate when these bits are not set correctly by the BIOS.
1.134	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.133	30-Dec-2009	jakllsch	Ensure that the memory or i/o space we need is accessible at attach and resume. Rework register mapping code path, using 'goto', so duplicate code is reduced. I've tested this on nvidia and sis auich(4)-compatible controllers. In addition this change is reported to fix PR/39652.
1.132	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.131	12-Nov-2009	dyoung	Remove superfluous activation hook.
1.130	03-Sep-2009	sborrill	SiS 7012 uses different multichannel control bits compared to other codecs (data from linux via OpenBSD).
1.129	17-Mar-2009	dyoung	Handle child-detachment by NULL'ing the child pointer so that auich_detach() does not subsequently dereference a dangling pointer.
1.128	08-Nov-2008	dyoung	branches: 1.128.2; 1.128.4; To finish the device_t/softc split, convert some casts to struct auich_softc * to device_private() calls.
1.127	04-May-2008	xtraeme	branches: 1.127.6; 1.127.8; 1.127.10; 1.127.14; device_t/softc split and other related cosmetic changes.
1.126	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.125	10-Apr-2008	cegger	branches: 1.125.2; 1.125.4; use aprint_*_dev and device_xname
1.124	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.123	06-Jan-2008	kent	branches: 1.123.2; 1.123.6; fix a bug of S/PDIF audio format selection PR/36493 by Arcady Antipin
1.122	13-Dec-2007	jmcneill	Make this compile again.
1.121	13-Dec-2007	jmcneill	Restore the native mode PCI conf register on select chipsets on resume, in case it was lost when power was removed.
1.120	09-Dec-2007	jmcneill	branches: 1.120.2; Merge jmcneill-pm branch.
1.119	19-Oct-2007	ad	branches: 1.119.2; 1.119.4; 1.119.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.118	02-Sep-2007	xtraeme	branches: 1.118.4; Attach to the ICH6 AC-97 Modem function.
1.117	04-Mar-2007	christos	branches: 1.117.2; 1.117.10; 1.117.14; 1.117.16; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.116	21-Feb-2007	thorpej	branches: 1.116.2; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.115	16-Nov-2006	christos	branches: 1.115.4; __unused removal on arguments; approved by core.
1.114	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.113	25-Sep-2006	jmcneill	No need to call pci_set_powerstate from our powerhook, the bus handles this for us now.
1.112	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.111	03-Aug-2006	bsh	branches: 1.111.4; 1.111.6; Device 10de:026c is not AC-97 controller but nForce430 HD Audio controller. Pointed out by Brad. Thanks.
1.110	01-Aug-2006	bsh	Support nForce430's audio. (I have tested this only with nFroce410, not with nFroce430)
1.109	28-Jul-2006	kent	add support for S/PDIF.
1.108	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.107	27-Apr-2006	jmcneill	branches: 1.107.2; 1.107.4; Disable my aggressive power saving changes; they're causing too many problems.
1.106	16-Apr-2006	jmcneill	Use AC97_HOST_INVERTED_EAMP on the Sony VAIO VGN-B1XP. From PR#33272 by Dieter Roelants.
1.105	16-Apr-2006	christos	Deal with machines that have inverted eamp power. Data from the FreeBSD driver. Finally sound works on my vaio vgn-t250p.
1.104	14-Apr-2006	rpaulo	s/#ifdef DEBUG/#ifdef AUICH_DEBUG/
1.103	07-Mar-2006	jmcneill	branches: 1.103.2; Don't try to suspend while already suspended.
1.102	07-Mar-2006	jmcneill	branches: 1.102.2; Put the device in D3 state when it's not in use to save power.
1.101	04-Dec-2005	xtraeme	branches: 1.101.4; 1.101.6; 1.101.8; Add support for Intel 6300ESB.
1.100	02-Dec-2005	xtraeme	Add support for ICH7. Tested by gendalia@.
1.99	22-Aug-2005	rpaulo	Use #ifdef AUICH_DEBUG instead of #if 0/#if 1. Suggested by perry@.
1.98	21-Aug-2005	rpaulo	PR#31030: debug printf in auich_halt_pipe. From lmneto@gmail.com. It was under #if 1, so I just replaced the if by 0.
1.97	28-Jun-2005	thorpej	branches: 1.97.2; Use ANSI function decls and static.
1.96	20-Jun-2005	atatat	Change the rest of the sysctl subsystem to use const consistently. The __UNCONST macro is now used only where necessary and the RW macros are gone. Most of the changes here are consumers of the sysctl_createv(9) interface that now takes a pair of const pointers which used not to be.
1.95	12-Apr-2005	jmcneill	Some broken codecs don't release the CAS on reads to AC97_REG_GPIO_STATUS. In auich_read_codec, do a dummy read to the RESET register if we're reading GPIO_STATUS to be safe.
1.94	11-Apr-2005	jmcneill	i82801CA is an ICH3, not ICH2. Fix typo in auich_modem_devices. Pointed out by Nicolas Joly.
1.93	08-Apr-2005	jmcneill	Add identifiers for Intel ICH3 AC'97 modems and CXT33 modem codecs, from Nicolas Joly <njoly _at_ pasteur _dot_ fr>.
1.92	07-Apr-2005	jmcneill	More modem fixes: * Ignore ICH_PCM246_MASK for modem devices * Set LINE1 rate and level in set_params * Enable mmap for modem devices. This was previously disallowed due to the ext_id field not advertising that the device supported variable rates. Since modems don't use ext_id (they use ext_mid instead), enable mmap. Multiple rates are supported on AC'97 modems.
1.91	04-Apr-2005	jmcneill	Work-in-progress of AC'97 modem support. Currently disabled by default (can be enabled by options AUICH_ATTACH_MODEM), and untested. Shouldn't have any impact on auich audio capabilities.
1.90	11-Mar-2005	mycroft	branches: 1.90.2; Three changes: 1) After setting an input or output channel to pause, watch for DCH (DMA halted) to be set before resetting the registers. The manual says this is required. This may fix random auich lockups. 2) Add auich_{trigger,intr,halt}_pipe(), which can operate on any of the input and output channels. This reduces code duplication quite a bit. While I'm at it, fill in the mic interrupt handling. 3) The mic in interrupt is MINT, not MIINT. (The latter is modem input.)
1.89	27-Feb-2005	perry	nuke trailing whitespace
1.88	27-Jan-2005	jdolecek	match nForce2 Ultra 400 audio
1.87	27-Jan-2005	kent	support nForce 4
1.86	26-Jan-2005	jmcneill	Add pci_conf_{capture,restore} to auich(4) powerhook
1.85	15-Jan-2005	kent	branches: 1.85.2; 1.85.4; ansify and KNF
1.84	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.83	11-Dec-2004	cube	Some SiS 7012 chips come muted by default, so un-mute all SiS 7012 chips at attach time the way Linux does it. Reported and tested by Jonathan Schleifer, I checked it didn't break my own 7012 device which doesn't need that manipulation.
1.82	17-Nov-2004	kent	branches: 1.82.2; implement auich_activate() and auich_detach()
1.81	17-Nov-2004	kent	make functions static as possible
1.80	13-Nov-2004	kent	adopt auconv_set_converter()
1.79	11-Nov-2004	kent	device matching by pcireg_t, not a pair of vendor ID and product ID. This change reduces 170 Byte of i386 kernel size.
1.78	10-Nov-2004	cube	Add support for the ICH6 AC97 controller.
1.77	10-Nov-2004	cube	Return an error when a read operation fails, instead of pretending it succeeded and transmitting a bogus value.
1.76	10-Nov-2004	cube	In auich_calibrate, set sc_ac97_clock to a default value of 48000 first. That way, if the calibration fails or times out, sc_ac97_clock won't be 0 and there won't be a division by zero later in ac97_set_rate.
1.75	10-Nov-2004	kent	move IS_FIXED_RATE(), SUPPORTS_4CH(), and SUPPORTS_6CH to ac97var.h
1.74	06-Nov-2004	xtraeme	Add support for "nForce3 250 MCP-T AC-97 Audio", from Kouichirou Hiratsuka in PR 27843. Ok'ed christos@
1.73	01-Nov-2004	mycroft	Some miscellaneous cleanup: * Encapsulate all the ring-specific variables into a substructure. (This will help with some potential code sharing.) * Don't bother with "last buffer complete" interrupts -- we should be taking an interrupt anyway because we set IOC on every buffer. * Likewise, ignore CELV; it's meaningless to us. * Get rid of the FIFO error counters, since this doesn't actually happen any more.
1.72	31-Oct-2004	mycroft	Copyright maintenance.
1.71	31-Oct-2004	mycroft	More code reduction. Also, allow any sample rate between 8000 and 48000Hz. (This range should probably be larger, but this is what's known to work.) Another 180 bytes shaved.
1.70	31-Oct-2004	mycroft	Shave another 112 bytes by reducing some divisions to shifts.
1.69	31-Oct-2004	mycroft	Fix a regression -- my previous changes made the kernel slightly larger.
1.68	31-Oct-2004	mycroft	In the trigger routines, start all the blocks, not just the first one. This was causing some "fifo under/overrun" errors during the initial trigger. Also fixes a ring synchronization problem introduced in the earlier changes. This should completely fix the FIFO errors with auich.
1.67	31-Oct-2004	mycroft	Reorder some register writes just to be safe.
1.66	31-Oct-2004	mycroft	Attempt to mitigate "fifo underrun" errors in two ways: 1) Update the queue pointers any time we get a block completion interrupt, not just when we hit the last block. 2) Set the "back-to-back enable" bit. In addition: 3) Make sure we ack every block we transfer; there was a bug that could cause the audio layer to get out of sync.
1.65	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.64	27-Oct-2004	kent	add "hw.auichN.ac97rate" sysctl MIB to specify AC'97 link rate. It is needed because automatic calibration by auich_calibrate() is not so precise. - Why not ioctl? It is not good idea to add a new ioctl operation to MI audio for a specific audio device. - Why not mixerctl item? AC'97 linke rate is not related to audio mixing.
1.63	17-Oct-2004	kent	no need to restore AC97_REG_EXT_AUDIO_CTRL value on resuming because ac97_codec_if_vtbl::restore_ports() does it.
1.62	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.61	07-Aug-2004	soren	- Don't attempt to clear the CELV bit by setting it high like the FIFOE/BCIS/LVBCI bits; it's is cleared automatically by the hardware. - Separate AUICH_DEBUG out from AUDIO_DEBUG and make it possible to debug interrupt handling separately. - A little KNF.
1.60	09-Jul-2004	mycroft	Clean up. Attempt to make these safer against stray interrupts.
1.59	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.58	13-Jan-2004	kent	branches: 1.58.2; remove/modify confusing error messsges.
1.57	03-Jan-2004	soren	Only report codec readiness if DEBUG to avoid printing it on every resume.
1.56	28-Dec-2003	kent	ICH4/ICH5: When native i/o mapping fails, try compatible i/o mapping.
1.55	22-Nov-2003	kent	For ICH4/ICH5, use native BARs instead of compatible BARs. The patch was provided by Lars Heidieker.
1.54	06-Nov-2003	mycroft	Force the codec into a known state before trying to determine its speed.
1.53	31-Oct-2003	kent	auich_calibrate(): - make the buffer longer to improve the precision of measurement - round off the result to the nearest multiple of 1000.
1.52	30-Oct-2003	kent	For ICH4/ICH5, set IOSE flag to make them compatible to old ICHs.
1.51	28-Oct-2003	mycroft	Eliminate bogus initializer (the code was technically wrong).
1.50	25-Oct-2003	christos	Fix uninitialized variable warnings
1.49	24-Oct-2003	kent	auich_reset_codec: wait for readiness of one of all codecs.
1.48	23-Oct-2003	kent	Change the short names of nForce: nForce-MCP -> nForce nForce-MCP-T -> nForce2 nForce-MCP-T -> nForce3 Use snprintf() and strlcpy() instead of sprintf() and strcpy().
1.47	23-Oct-2003	kent	- remove unused code - print diagnostic messages if the primary codec does not get ready in auich_reset_codec().
1.46	23-Oct-2003	kent	No need to check the codec ready flag in read_codec() and write_codec(). According to the ICH datasheets, the flag is unchanged unless resetting the codec. The flag is useless after enabling bus mastering. This behavior is common to all ICH chips. The quirk handling code is removed.
1.45	22-Oct-2003	manu	Patch from http://mail-index.netbsd.org/current-users/2003/09/04/0018.html to fix problems with a sound card
1.44	22-Oct-2003	fvdl	Match the nForce3 AC-97 chip.
1.43	21-Oct-2003	fvdl	Add missing field initializer.
1.42	02-Oct-2003	mycroft	Cleanup: * auich_calibrate() assumes that the AC97 part is in its reset state. To ensure this, call audio_attach_mi() after auich_calibrate(). * Explicitly support 12000Hz and 24000Hz. (Why is there a discrete list at all?) * Fix an obvious recording bug -- we were acking the wrong interrupt. * Ensure that we don't get an interrupt during the AC97 speed probe by clearing the "interrupt on completion" bit in the DMA setup.
1.41	28-Sep-2003	kent	Support for 4ch/6ch audio playback.
1.40	19-Aug-2003	erh	Add entry for Intel 82801EB AC-97 audio. Audio sounds fine with this, although the mixer settings are a bit odd: outputs.master doesn't do anything, but outputs.surround does.
1.39	13-Jun-2003	kent	branches: 1.39.2; Fix a regression of the last commit. ICH4 needs IGNORE_CODEC_READY_MAYBE.
1.38	13-Jun-2003	kent	Not only ICH4 but also ICH0 and ICH2 might set no ICH_PCR. The patch was provided by lha. PR: kern/21031
1.37	03-May-2003	wiz	DMA, not dma nor Dma.
1.36	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.35	31-Jan-2003	thorpej	Use aprint_*().
1.34	28-Jan-2003	kent	Apply BUS_DMA_NOCACHE for the buffers in the case of 440MX. The code was provided by DHOYASHIKI Shinichi in kern/19919, and address kern/19435.
1.33	21-Jan-2003	kent	Add DIAGNOSTIC code provided in PR kern/19919 by DHOYASHIKI Shinichi.
1.32	16-Jan-2003	gendalia	add support for nForce2 MCP-T
1.31	15-Jan-2003	kent	Fixed a problem that /dev/sound can not be opened in some cases. auich_set_rate() modified audio_params::sample_rate to actual sample rate though auich_set_params() rejected sample rates other than some specific rates. For example, setting 8000Hz with auich_set_params() modified sample_rate to 7998Hz in the case of overclocked AC97 codec, and calling auich_set_params() again returned with EINVAL because sample_rate was 7998Hz and auich_set_params() rejected 7998Hz. For now, auich_set_params() never modify audio_params::sample_rate.
1.30	12-Oct-2002	wiz	Fix typo in comment.
1.29	11-Oct-2002	kent	Kick auich_calibrate() by config_interrupts(). auichi_calibrate(): If the measured rate is less than 48k, use 48k.
1.28	08-Oct-2002	kent	auich_set_params(): Don't return in the for loop.
1.27	08-Oct-2002	kent	auich_softc: Remove sc_fixed_rate and sc_ac97rate. ac97_softc has such information. auich_attach(): Remove VRA/VRM enabling code. ac97_attach() does it. auich_set_rate(): Use ac97_codec_if::set_rate(). auich_get_props(): Return no AUDIO_PROP_MMAP if the codec is fixed-rate. auichreg.h: Remove unnecessary ICH_PM_*.
1.26	06-Oct-2002	kent	Add some constant symbols for AC97 registers and values based on OpenBSD. Rename some symbols to sync with OpenBSD: AC97_REG_EXTENDED_ID -> AC97_REG_EXT_AUDIO_ID AC97_REG_EXTENDED_STATUS -> AC97_REG_EXT_AUDIO_CTRL AC97_CODEC_DOES_VRA/AC97_ENAB_VRA -> AC97_EXT_AUDIO_VRA AC97_CODEC_DOES_MICVRA/AC97_ENAB_MICVRA -> AC97_EXT_AUDIO_VRM ac97feature: "mic channel" -> "dedicated mic channel" ac97_attach(): print extended capabilities; VRA, DRA, SPDIF, VRM, CDAC, SDAC, LDAC.
1.25	04-Oct-2002	kent	Improve nForce MCP support. Submitted by: FUKAUMI Naoki <naoki at fukaumi dot org> PR: kern/18524
1.24	02-Oct-2002	kent	Remove a line added accidentally(?) at revision 1.21.
1.23	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.22	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.21	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.20	28-Aug-2002	dsainty	Fix playback at 48kHz sample rate (integer overflow bug introduced in 1.18)
1.19	16-Aug-2002	kent	Fix division-by-zero when auich_set_params() is called before auich_calibrate().
1.18	07-Aug-2002	kent	- Port clock calibration code (auich_calibrate) from ich.c of FreeBSD to solve kern/16373. - SiS 7012 support from kern/17716 submitted by Quentin Garnier. - ICH4 support - Experimental support for nForce MCP, AMD768 and AMD8111. - Rename the short name for ICH3: "i830M" -> "ICH3"
1.17	11-Apr-2002	augustss	branches: 1.17.2; 1.17.4; When the DAC claims to support variable rate it doesn't seem to always do this, so do some extra chechs and fall back on fixed rate.
1.16	23-Mar-2002	kent	whitespace and folding fix.
1.15	21-Mar-2002	kent	Apply a patch by Mark Davies. http://mail-index.netbsd.org/port-i386/2002/03/18/0000.html Check codec status in auich_read/write_codec().
1.14	15-Mar-2002	tacha	Use appropriate emulation functions and set ADC rate(not DAC date!) for recording. Now /dev/audio with auich becomes usable.
1.13	10-Mar-2002	kent	auich_set_params() sets 2 to hw_channels whenever monaural format is requested.
1.12	07-Mar-2002	kent	Sampling rate and mono-stereo conversion described in http://mail-index.netbsd.org/tech-kern/2002/03/04/0005.html auconv.c: Add conversion functions audio.c: Sample alignment, calling conversion functions, etc. audio_if.h: Add four hw_* members to "struct audio_params" audiovar.h: Add conversion buffers, etc. auich and uaudio: Add conversion request code to *_set_params().
1.11	14-Feb-2002	augustss	Oops, I messed up the last commit, try again.
1.10	14-Feb-2002	augustss	Make fixed rate codecs work. From Tsuyoshi Iguchi in kern/15606.
1.9	02-Feb-2002	augustss	Make auich support variable rate codec, and suspend/resume. From URA Hiroshi in PR kern/15431 (partly from OpenBSD).
1.8	14-Jan-2002	augustss	Add i830M to the chip list.
1.7	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.6	12-Jan-2002	enami	Skip not yet emulated 8bit ulinear encoding correctly.
1.5	13-Nov-2001	lukem	add RCSID
1.4	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.3	28-Dec-2000	sommerfeld	branches: 1.3.2; 1.3.4; 1.3.6; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.2	28-Nov-2000	thorpej	branches: 1.2.2; Use AC97_REG_* constants rather than defining them ourselves.
1.1	28-Nov-2000	thorpej	Device driver for the built-in audio on Intel ICH, ICH0, ICH2, and i440MX chipsets. Based on a driver by Michael Shalayeff, modified somewhat by me to use bus_dma properly, and fix some audio encoding emulation bugs. Thanks to Manuel Bouyer for testing and feedback.
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.2.2.1	28-Nov-2000	bouyer	file auich.c was added on branch thorpej_scsipi on 2000-12-08 09:12:30 +0000
1.3.6.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.3.4.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.4.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.4.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.4.3	16-Mar-2002	jdolecek	Catch up with -current.
1.3.4.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.3.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.11	17-Jan-2003	thorpej	Sync with HEAD.
1.3.2.10	15-Jan-2003	thorpej	Sync with HEAD.
1.3.2.9	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.8	17-Sep-2002	nathanw	Catch up to -current.
1.3.2.7	27-Aug-2002	nathanw	Catch up to -current.
1.3.2.6	13-Aug-2002	nathanw	Catch up to -current.
1.3.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.3.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.3.2.3	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.1	08-Oct-2001	nathanw	Catch up to -current.
1.17.4.1	15-Mar-2004	jmc	Pullup patch (requested by kent in ticket #1612) Add support for ICH3, ICH4, ICH5, SiS 7012, nForce MCP, nForce2 MCP-T, and nForce3 MCP-T to auich driver
1.17.2.1	29-Aug-2002	gehenna	catch up with -current.
1.39.2.16	11-Dec-2005	christos	Sync with head.
1.39.2.15	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.39.2.14	01-Apr-2005	skrll	Sync with HEAD.
1.39.2.13	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.39.2.12	04-Feb-2005	skrll	Sync with HEAD.
1.39.2.11	17-Jan-2005	skrll	Sync with HEAD.
1.39.2.10	18-Dec-2004	skrll	Sync with HEAD.
1.39.2.9	29-Nov-2004	skrll	Sync with HEAD.
1.39.2.8	14-Nov-2004	skrll	Sync with HEAD.
1.39.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.39.2.6	19-Oct-2004	skrll	Sync with HEAD
1.39.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.39.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.39.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.39.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.39.2.1	03-Aug-2004	skrll	Sync with HEAD
1.58.2.1	22-Sep-2004	jmc	branches: 1.58.2.1.2; Pullup rev 1.62 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.58.2.1.2.2	19-May-2005	tron	Pull up revision 1.78 (requested by riz in ticket #1533): Add support for the ICH6 AC97 controller.
1.58.2.1.2.1	07-Apr-2005	tron	Pull up revision 1.74 (requested by hira in ticket #1103): Add support for "nForce3 250 MCP-T AC-97 Audio", from Kouichirou Hiratsuka in PR 27843. Ok'ed christos@
1.82.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.82.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.85.4.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.85.4.1	12-Feb-2005	yamt	sync with head.
1.85.2.1	29-Apr-2005	kent	sync with -current
1.90.2.2	06-Dec-2005	riz	Apply patch (requested by xtraeme in ticket #1027): sys/dev/pci/auich.c: patch sys/dev/pci/auichreg.h: patch Add support for Intel ICH7 and 6300ESB.
1.90.2.1	30-Oct-2005	jmc	Pullup revs 1.98-1.99 (requested by rpaulo in ticket #920) Fix debug printf in auich_halt_pipe. PR#31030
1.97.2.7	17-Mar-2008	yamt	sync with head.
1.97.2.6	21-Jan-2008	yamt	sync with head
1.97.2.5	27-Oct-2007	yamt	sync with head.
1.97.2.4	03-Sep-2007	yamt	sync with head.
1.97.2.3	26-Feb-2007	yamt	sync with head.
1.97.2.2	30-Dec-2006	yamt	sync with head.
1.97.2.1	21-Jun-2006	yamt	sync with head.
1.101.8.4	11-Aug-2006	yamt	sync with head
1.101.8.3	26-Jun-2006	yamt	sync with head.
1.101.8.2	24-May-2006	yamt	sync with head.
1.101.8.1	13-Mar-2006	yamt	sync with head.
1.101.6.2	01-Jun-2006	kardel	Sync with head.
1.101.6.1	22-Apr-2006	simonb	Sync with head.
1.101.4.1	09-Sep-2006	rpaulo	sync with head
1.102.2.3	11-May-2006	elad	sync with head
1.102.2.2	19-Apr-2006	elad	sync with head.
1.102.2.1	07-Mar-2006	elad	file auich.c was added on branch elad-kernelauth on 2006-04-19 03:25:33 +0000
1.103.2.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.107.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.107.2.1	19-Jun-2006	chap	Sync with head.
1.111.6.2	10-Dec-2006	yamt	sync with head.
1.111.6.1	22-Oct-2006	yamt	sync with head
1.111.4.1	18-Nov-2006	ad	Sync with head.
1.115.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.115.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.116.2.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.117.16.3	23-Mar-2008	matt	sync with HEAD
1.117.16.2	09-Jan-2008	matt	sync with HEAD
1.117.16.1	06-Nov-2007	matt	sync with HEAD
1.117.14.7	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.117.14.6	11-Nov-2007	joerg	In auich_intr, check if device is currently powered down.
1.117.14.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.117.14.4	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.117.14.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.117.14.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.117.14.1	05-Aug-2007	jmcneill	Convert auich powerhook to pnp power handler (untested).
1.117.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.117.2.2	23-Oct-2007	ad	Sync with head.
1.117.2.1	09-Oct-2007	ad	Sync with head.
1.118.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.119.6.1	11-Dec-2007	yamt	sync with head.
1.119.4.1	26-Dec-2007	ad	Sync with head.
1.119.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.120.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.120.2.1	13-Dec-2007	bouyer	Sync with HEAD
1.123.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.123.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.123.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.123.2.1	24-Mar-2008	keiichi	sync with head.
1.125.4.4	11-Mar-2010	yamt	sync with head
1.125.4.3	16-Sep-2009	yamt	sync with head
1.125.4.2	04-May-2009	yamt	sync with head.
1.125.4.1	16-May-2008	yamt	sync with head.
1.125.2.1	18-May-2008	yamt	sync with head.
1.127.14.1	21-Apr-2010	matt	sync to netbsd-5
1.127.10.2	15-Jan-2010	snj	Revert ticket 1218.
1.127.10.1	09-Jan-2010	snj	Pull up following revision(s) (requested by jakllsch in ticket #1218): sys/dev/pci/auich.c: revision 1.133 Ensure that the memory or i/o space we need is accessible at attach and resume. Rework register mapping code path, using 'goto', so duplicate code is reduced. I've tested this on nvidia and sis auich(4)-compatible controllers. In addition this change is reported to fix PR/39652.
1.127.8.2	28-Apr-2009	skrll	Sync with HEAD.
1.127.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.127.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.128.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.128.2.3	13-Dec-2008	ad	Fix locking for ac97.
1.128.2.2	12-Dec-2008	ad	Checkpoint work in progress.
1.128.2.1	08-Dec-2008	ad	Convert some more drivers.
1.136.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.137.2.1	05-Mar-2011	rmind	sync with head
1.138.10.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.138.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.138.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.138.8.2	30-Oct-2012	yamt	sync with head
1.138.8.1	17-Apr-2012	yamt	sync with head
1.141.4.1	18-May-2014	rmind	sync with head
1.141.2.2	03-Dec-2017	jdolecek	update from HEAD
1.141.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.143.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.144.4.4	28-Aug-2017	skrll	Sync with HEAD
1.144.4.3	05-Feb-2017	skrll	Sync with HEAD
1.144.4.2	09-Jul-2016	skrll	Sync with HEAD
1.144.4.1	06-Jun-2015	skrll	Sync with HEAD
1.149.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.149.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.151.8.1	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.152.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.152.2.1	10-Jun-2019	christos	Sync with HEAD
1.154.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.154.2.2	27-Apr-2019	isaki	Adapt to audio2. - Correct frequency_type in modem_audio_format. - {spdif,modem}_audio_format in my previous commit (in this branch) was also wrong. :(
1.154.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.157.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by isaki in ticket #792): sys/arch/amiga/dev/aucc.c: revision 1.48 sys/dev/pci/auixp.c: revision 1.49 sys/dev/pci/gcscaudio.c: revision 1.19 sys/dev/pci/auich.c: revision 1.159 sys/dev/sbus/dbri.c: revision 1.42 sys/dev/pci/auvia.c: revision 1.85 sys/dev/pci/auacer.c: revision 1.39 Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017. round_blocksize must return a multiple of the framesize. aucc(4) supports 3 channels mode. round_blocksize must return a multiple of the framesize even if passed blocksize is greater than the upper limit. round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4. round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.158.2.1	29-Feb-2020	ad	Sync with head.
1.159.6.1	03-Apr-2021	thorpej	Sync with HEAD.
1.13	08-Feb-2018	dholland	Typos.
1.12	03-Sep-2009	sborrill	SiS 7012 uses different multichannel control bits compared to other codecs (data from linux via OpenBSD).
1.11	14-Jan-2008	martin	branches: 1.11.10; Make sure pci config space access only happens on 32bit aligned addresses.
1.10	02-Dec-2005	xtraeme	branches: 1.10.46; 1.10.52; 1.10.60; Add support for ICH7. Tested by gendalia@.
1.9	11-Apr-2005	jmcneill	branches: 1.9.2; Add ICH_CODEC_OFFSET definition
1.8	11-Dec-2004	cube	branches: 1.8.2; 1.8.8; Some SiS 7012 chips come muted by default, so un-mute all SiS 7012 chips at attach time the way Linux does it. Reported and tested by Jonathan Schleifer, I checked it didn't break my own 7012 device which doesn't need that manipulation.
1.7	22-Nov-2003	kent	For ICH4/ICH5, use native BARs instead of compatible BARs. The patch was provided by Lars Heidieker.
1.6	30-Oct-2003	kent	For ICH4/ICH5, set IOSE flag to make them compatible to old ICHs.
1.5	03-May-2003	wiz	branches: 1.5.2; DMA, not dma nor Dma.
1.4	08-Oct-2002	kent	auich_softc: Remove sc_fixed_rate and sc_ac97rate. ac97_softc has such information. auich_attach(): Remove VRA/VRM enabling code. ac97_attach() does it. auich_set_rate(): Use ac97_codec_if::set_rate(). auich_get_props(): Return no AUDIO_PROP_MMAP if the codec is fixed-rate. auichreg.h: Remove unnecessary ICH_PM_*.
1.3	07-Aug-2002	kent	- Port clock calibration code (auich_calibrate) from ich.c of FreeBSD to solve kern/16373. - SiS 7012 support from kern/17716 submitted by Quentin Garnier. - ICH4 support - Experimental support for nForce MCP, AMD768 and AMD8111. - Rename the short name for ICH3: "i830M" -> "ICH3"
1.2	28-Nov-2000	thorpej	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.18; 1.2.20; Use AC97_REG_* constants rather than defining them ourselves.
1.1	28-Nov-2000	thorpej	Device driver for the built-in audio on Intel ICH, ICH0, ICH2, and i440MX chipsets. Based on a driver by Michael Shalayeff, modified somewhat by me to use bus_dma properly, and fix some audio encoding emulation bugs. Thanks to Manuel Bouyer for testing and feedback.
1.2.20.1	15-Mar-2004	jmc	Pullup patch (requested by kent in ticket #1612) Add support for ICH3, ICH4, ICH5, SiS 7012, nForce MCP, nForce2 MCP-T, and nForce3 MCP-T to auich driver
1.2.18.1	29-Aug-2002	gehenna	catch up with -current.
1.2.6.1	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.2.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.4.1	13-Aug-2002	nathanw	Catch up to -current.
1.2.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.2.2.1	28-Nov-2000	bouyer	file auichreg.h was added on branch thorpej_scsipi on 2000-12-08 09:12:30 +0000
1.5.2.6	11-Dec-2005	christos	Sync with head.
1.5.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.4	18-Dec-2004	skrll	Sync with HEAD.
1.5.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.8.1	06-Dec-2005	riz	Apply patch (requested by xtraeme in ticket #1027): sys/dev/pci/auich.c: patch sys/dev/pci/auichreg.h: patch Add support for Intel ICH7 and 6300ESB.
1.8.2.1	29-Apr-2005	kent	sync with -current
1.9.2.2	21-Jan-2008	yamt	sync with head
1.9.2.1	21-Jun-2006	yamt	sync with head.
1.10.60.1	19-Jan-2008	bouyer	Sync with HEAD
1.10.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.10.46.1	23-Mar-2008	matt	sync with HEAD
1.11.10.1	16-Sep-2009	yamt	sync with head
1.55	08-Feb-2024	andvar	s/inlucde/include/, s/implemetation/implementation/ in comments and error msg.
1.54	27-Aug-2022	skrll	Correct the abbreviation of approximately to 'approx.'
1.53	25-Jan-2022	andvar	fix various typos in comments.
1.52	21-Aug-2021	andvar	s/helt/held+s/eroneously/erroneously/+s/splitted/split/+s/recommented/recommended/
1.51	20-Aug-2021	andvar	fix various typos in comments and log messages.
1.50	03-Jul-2020	isaki	Fix an argument passes to auixp_intr(). This is rest of rev 1.39 (split device_t/softc) in 2012. Problem reported and tested by Riccardo Mottola.
1.49	29-Feb-2020	isaki	round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.48	16-Oct-2019	maya	branches: 1.48.2; Switch files copyright Reinoud Zandijk from BSD 4 clause to BSD 2 clause. OK'd by reinoud in email (from 4 Feb 2019).
1.47	08-Jun-2019	isaki	branches: 1.47.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.46	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.45	16-Mar-2019	isaki	branches: 1.45.2; Use C99 style struct initializer to audio_hw_if.
1.44	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.43	01-Jun-2017	chs	branches: 1.43.8; 1.43.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.42	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.41	18-Oct-2014	snj	branches: 1.41.2; src is too big these days to tolerate superfluous apostrophes. It's "its", people!
1.40	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.39	27-Oct-2012	chs	branches: 1.39.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.38	30-Jan-2012	drochner	branches: 1.38.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.37	02-Dec-2011	jmcneill	branches: 1.37.2; audio cookie is a struct auixp_codec, not auixp_softc
1.36	24-Nov-2011	mrg	put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.35	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.34	24-Feb-2010	dyoung	branches: 1.34.10; 1.34.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.33	08-Jan-2010	dyoung	branches: 1.33.2; Expand PMF_FN_* macros.
1.32	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.31	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.30	18-Mar-2009	cegger	bzero -> memset
1.29	05-Mar-2009	msaitoh	Fix typo(s/interupt/interrupt/)
1.28	10-Apr-2008	cegger	branches: 1.28.4; 1.28.12; 1.28.16; 1.28.18; use aprint_*_dev and device_xname
1.27	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.26	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.25	09-Dec-2007	jmcneill	branches: 1.25.6; 1.25.10; Merge jmcneill-pm branch.
1.24	19-Oct-2007	ad	branches: 1.24.4; 1.24.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.23	29-Apr-2007	msaitoh	branches: 1.23.6; 1.23.8; 1.23.12; fix typos
1.22	04-Mar-2007	christos	branches: 1.22.2; 1.22.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.21	16-Nov-2006	christos	branches: 1.21.4; 1.21.6; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.18	24-Sep-2006	jmcneill	Don't use // for comments.
1.17	18-Sep-2006	reinoud	Since SPDIF support fo the auixp driver is still half done disable it for now since some chipset revisions will freak out on the aparent half-initialisation. Even on my machine i can't seem to get the SPDIF led to light up so something is wrong. Also delay the setting of the DMA bits until after the codec detection but before the enabling of interrupts. Note that the dma has to be explicitly started when the device is opened.
1.16	27-Aug-2006	christos	branches: 1.16.2; 1.16.4; Add missing initializers.
1.15	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.14	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.13	14-May-2006	elad	branches: 1.13.2; 1.13.4; integrate kauth.
1.12	11-Apr-2006	rpaulo	Coverity ID 2311: in auixp_allocate_dma_chain() free dma variable if we are unable to allocate hardware dma descriptors.
1.11	24-Jan-2006	reinoud	branches: 1.11.2; 1.11.4; 1.11.6; 1.11.8; 1.11.10; Duh! If askes to halt the input dont halt the output instead. This also fixes the timeout on draining when closing.
1.10	11-Dec-2005	christos	branches: 1.10.2; merge ktrace-lwp.
1.9	27-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static. Fix a memory leak uncovered by doing so (auixp_freemem() was never called).
1.8	15-Jun-2005	reinoud	Cosmetic change giving better feedback on what's going on. It would allways tell the hardware reset timed out when it didn't. Also cleaned up code comments.
1.7	12-Jun-2005	reinoud	Implement auixp AC-link reset
1.6	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.5	26-Jan-2005	fvdl	branches: 1.5.4; Don't print a message when there's no interrupt to be handled, since this can happen when one is shared. Also, don't claim it in that case. Make some printfs more consistent.
1.4	13-Jan-2005	simonb	branches: 1.4.2; 1.4.4; 1.4.6; More KNF nits.
1.3	12-Jan-2005	kent	remove empty auixp_open() and auixp_close(). They are optional. style fixes * Don't associate a name with the types * The function type must be declared on a line by itself preceding the function. * DO NOT initialize variables in the declarations. * No parentheses are needed around the return value.
1.2	12-Jan-2005	reinoud	Update copyright messages
1.1	12-Jan-2005	reinoud	Initial commit of auixp(4) driver. It's a driver for the audio part of ATI IXP-200 and family motherboard chips. Its functional for both recording and playback. Note that quadraphonic and Dolby 5.1 audio are not tested by me but ought to work fine. I can set the number of channels etc. OK but i can't listen to the result. Any feedback on this performance would be greatly apreciated.
1.4.6.1	12-Feb-2005	yamt	sync with head.
1.4.4.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.4.3	04-Feb-2005	skrll	Sync with HEAD.
1.4.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.4.4.1	13-Jan-2005	skrll	file auixp.c was added on branch ktrace-lwp on 2005-01-17 19:31:23 +0000
1.4.2.1	29-Apr-2005	kent	sync with -current
1.5.4.1	28-Jan-2006	tron	Pull up following revision(s) (requested by reinoud in ticket #1138): sys/dev/pci/auixp.c: revision 1.11 Duh! If askes to halt the input dont halt the output instead. This also fixes the timeout on draining when closing.
1.9.2.7	24-Mar-2008	yamt	sync with head.
1.9.2.6	17-Mar-2008	yamt	sync with head.
1.9.2.5	21-Jan-2008	yamt	sync with head
1.9.2.4	27-Oct-2007	yamt	sync with head.
1.9.2.3	03-Sep-2007	yamt	sync with head.
1.9.2.2	30-Dec-2006	yamt	sync with head.
1.9.2.1	21-Jun-2006	yamt	sync with head.
1.10.2.1	01-Feb-2006	yamt	sync with head.
1.11.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.11.8.1	19-Apr-2006	elad	sync with head.
1.11.6.3	03-Sep-2006	yamt	sync with head.
1.11.6.2	26-Jun-2006	yamt	sync with head.
1.11.6.1	24-May-2006	yamt	sync with head.
1.11.4.1	22-Apr-2006	simonb	Sync with head.
1.11.2.1	09-Sep-2006	rpaulo	sync with head
1.13.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.13.2.1	19-Jun-2006	chap	Sync with head.
1.16.4.2	10-Dec-2006	yamt	sync with head.
1.16.4.1	22-Oct-2006	yamt	sync with head
1.16.2.1	18-Nov-2006	ad	Sync with head.
1.21.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.21.4.2	07-May-2007	yamt	sync with head.
1.21.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.22.4.1	11-Jul-2007	mjf	Sync with head.
1.22.2.2	23-Oct-2007	ad	Sync with head.
1.22.2.1	27-May-2007	ad	Sync with head.
1.23.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.23.8.3	23-Mar-2008	matt	sync with HEAD
1.23.8.2	09-Jan-2008	matt	sync with HEAD
1.23.8.1	06-Nov-2007	matt	sync with HEAD
1.23.6.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.23.6.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.23.6.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.23.6.2	04-Oct-2007	joerg	Also deregister PNP handler at detach time.
1.23.6.1	04-Oct-2007	joerg	Add blindly codec power management support.
1.24.6.1	11-Dec-2007	yamt	sync with head.
1.24.4.1	26-Dec-2007	ad	Sync with head.
1.25.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.25.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.25.6.1	24-Mar-2008	keiichi	sync with head.
1.28.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.28.16.3	13-Dec-2008	ad	Fix locking for ac97.
1.28.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.28.16.1	08-Dec-2008	ad	Convert some more drivers.
1.28.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.28.4.3	11-Mar-2010	yamt	sync with head
1.28.4.2	16-May-2009	yamt	sync with head
1.28.4.1	04-May-2009	yamt	sync with head.
1.33.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.34.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.34.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.34.10.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.34.10.2	30-Oct-2012	yamt	sync with head
1.34.10.1	17-Apr-2012	yamt	sync with head
1.37.2.1	18-Feb-2012	mrg	merge to -current.
1.38.6.3	03-Dec-2017	jdolecek	update from HEAD
1.38.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.38.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.39.2.1	18-May-2014	rmind	sync with head
1.41.2.2	28-Aug-2017	skrll	Sync with HEAD
1.41.2.1	09-Jul-2016	skrll	Sync with HEAD
1.43.10.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.43.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.43.10.1	10-Jun-2019	christos	Sync with HEAD
1.43.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.45.2.2	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.45.2.1	27-Apr-2019	isaki	Adapt to audio2. - Move auixp_formats' definition from header to source. - Make compilable even with DEBUG_AUIXP.
1.47.2.2	07-Jul-2020	martin	Pull up following revision(s) (requested by isaki in ticket #983): sys/dev/pci/auixp.c: revision 1.50 Fix an argument passes to auixp_intr(). This is rest of rev 1.39 (split device_t/softc) in 2012. Problem reported and tested by Riccardo Mottola.
1.47.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by isaki in ticket #792): sys/arch/amiga/dev/aucc.c: revision 1.48 sys/dev/pci/auixp.c: revision 1.49 sys/dev/pci/gcscaudio.c: revision 1.19 sys/dev/pci/auich.c: revision 1.159 sys/dev/sbus/dbri.c: revision 1.42 sys/dev/pci/auvia.c: revision 1.85 sys/dev/pci/auacer.c: revision 1.39 Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017. round_blocksize must return a multiple of the framesize. aucc(4) supports 3 channels mode. round_blocksize must return a multiple of the framesize even if passed blocksize is greater than the upper limit. round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4. round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.48.2.1	29-Feb-2020	ad	Sync with head.
1.4	16-Oct-2019	maya	Switch files copyright Reinoud Zandijk from BSD 4 clause to BSD 2 clause. OK'd by reinoud in email (from 4 Feb 2019).
1.3	11-Dec-2005	christos	branches: 1.3.164; merge ktrace-lwp.
1.2	12-Jan-2005	reinoud	branches: 1.2.4; Update copyright messages
1.1	12-Jan-2005	reinoud	Initial commit of auixp(4) driver. It's a driver for the audio part of ATI IXP-200 and family motherboard chips. Its functional for both recording and playback. Note that quadraphonic and Dolby 5.1 audio are not tested by me but ought to work fine. I can set the number of channels etc. OK but i can't listen to the result. Any feedback on this performance would be greatly apreciated.
1.2.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.2.4.1	12-Jan-2005	skrll	file auixpreg.h was added on branch ktrace-lwp on 2005-01-17 19:31:23 +0000
1.3.164.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.11	20-Aug-2021	andvar	fix various typos in comments and log messages.
1.10	16-Oct-2019	maya	Switch files copyright Reinoud Zandijk from BSD 4 clause to BSD 2 clause. OK'd by reinoud in email (from 4 Feb 2019).
1.9	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.8	27-Oct-2012	chs	branches: 1.8.38; 1.8.42; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.7	23-Nov-2011	jmcneill	branches: 1.7.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.6	09-Dec-2007	jmcneill	branches: 1.6.26; 1.6.48; 1.6.50; Merge jmcneill-pm branch.
1.5	04-Mar-2007	christos	branches: 1.5.14; 1.5.16; 1.5.24; 1.5.26; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.4	11-Dec-2005	christos	branches: 1.4.26; 1.4.28; merge ktrace-lwp.
1.3	12-Jan-2005	kent	branches: 1.3.4; 1.3.12; remove empty auixp_open() and auixp_close(). They are optional. style fixes * Don't associate a name with the types * The function type must be declared on a line by itself preceding the function. * DO NOT initialize variables in the declarations. * No parentheses are needed around the return value.
1.2	12-Jan-2005	reinoud	Update copyright messages
1.1	12-Jan-2005	reinoud	Initial commit of auixp(4) driver. It's a driver for the audio part of ATI IXP-200 and family motherboard chips. Its functional for both recording and playback. Note that quadraphonic and Dolby 5.1 audio are not tested by me but ought to work fine. I can set the number of channels etc. OK but i can't listen to the result. Any feedback on this performance would be greatly apreciated.
1.3.12.2	21-Jan-2008	yamt	sync with head
1.3.12.1	03-Sep-2007	yamt	sync with head.
1.3.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.3.4.1	12-Jan-2005	skrll	file auixpvar.h was added on branch ktrace-lwp on 2005-01-17 19:31:23 +0000
1.4.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.4.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.5.26.1	11-Dec-2007	yamt	sync with head.
1.5.24.1	26-Dec-2007	ad	Sync with head.
1.5.16.1	09-Jan-2008	matt	sync with HEAD
1.5.14.1	04-Oct-2007	joerg	Add blindly codec power management support.
1.6.50.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.6.48.2	30-Oct-2012	yamt	sync with head
1.6.48.1	17-Apr-2012	yamt	sync with head
1.6.26.1	08-Dec-2008	ad	Convert some more drivers.
1.7.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.8.42.3	05-May-2019	isaki	Remove sc_encodings. (I forgot it)
1.8.42.2	27-Apr-2019	isaki	Adapt to audio2. - Move auixp_formats' definition from header to source. - Make compilable even with DEBUG_AUIXP.
1.8.42.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.8.38.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.8.38.1	10-Jun-2019	christos	Sync with HEAD
1.60	25-Jan-2022	andvar	fix various typos in comments.
1.59	29-Feb-2020	isaki	Remove rounding by 4 bytes on round_blocksize(). For drivers which supports only 16bit * 2channels sampling, rounding by 4 bytes no longer meaningful.
1.58	08-Jun-2019	isaki	branches: 1.58.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.57	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.56	16-Mar-2019	isaki	branches: 1.56.2; Use C99 style struct initializer to audio_hw_if.
1.55	01-Jun-2017	chs	branches: 1.55.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.54	10-May-2017	msaitoh	- Use pci_intr_establish_xname(). - Style change.
1.53	07-Jul-2016	msaitoh	branches: 1.53.8; KNF. Remove extra spaces. No functional change.
1.52	01-Aug-2014	joerg	branches: 1.52.4; Byte access is only used in the midi code.
1.51	29-Mar-2014	christos	branches: 1.51.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.50	17-Mar-2012	martin	branches: 1.50.2; 1.50.4; Split device_t
1.49	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.48	24-Nov-2011	mrg	branches: 1.48.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.47	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.46	24-Feb-2010	dyoung	branches: 1.46.10; 1.46.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.45	08-Jan-2010	dyoung	branches: 1.45.2; Expand PMF_FN_* macros.
1.44	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.43	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.42	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.41	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.40	03-Jul-2008	gson	branches: 1.40.8; 1.40.10; Revert the change of revision 1.37 - a driver must not blindly set the mem/io enable bits of a PCI device as the BARs may not be correctly configured. Fixes PR kern/38857.
1.39	10-Apr-2008	cegger	branches: 1.39.4; 1.39.6; 1.39.8; use aprint_*_dev and device_xname
1.38	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.37	27-Jan-2008	jmcneill	branches: 1.37.2; 1.37.6; Fixup PCI_COMMAND_STATUS_REG on attach.
1.36	09-Dec-2007	jmcneill	Merge jmcneill-pm branch.
1.35	19-Oct-2007	ad	branches: 1.35.2; 1.35.4; 1.35.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.34	16-Nov-2006	christos	branches: 1.34.6; 1.34.8; 1.34.22; 1.34.24; 1.34.28; __unused removal on arguments; approved by core.
1.33	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.32	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.31	28-Aug-2006	christos	branches: 1.31.2; 1.31.4; add missing initializer
1.30	08-Mar-2006	lukem	Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.29	22-Feb-2006	garbled	branches: 1.29.2; 1.29.4; IBM makes a pcn card that shows up as: Trident Microsystems 4DWAVE DX (ethernet network, revision 0x25) Unfortunately, this is an autri sound card. Special case the match routines for both drivers so if_pcn properly picks it up, and autri doesn't.
1.28	24-Dec-2005	perry	branches: 1.28.2; 1.28.4; 1.28.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.27	11-Dec-2005	christos	merge ktrace-lwp.
1.26	28-Jun-2005	thorpej	branches: 1.26.2; Use ANSI function decls and static.
1.25	02-Feb-2005	gson	Generate an interrupt only after processing all the samples in a block rather than at the time of the last sample, to eliminate a click at the end of each block when recording. Fixes kern/18066.
1.24	15-Jan-2005	kent	branches: 1.24.2; 1.24.4; ansify and KNF
1.23	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.22	09-Nov-2004	kent	branches: 1.22.2; remove mixer setting code in *_attach() because ac97_attach() does it.
1.21	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.20	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.19	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.18	29-Oct-2003	mycroft	branches: 1.18.2; Bogus initializers suck.
1.17	25-Oct-2003	christos	Fix uninitialized variable warnings
1.16	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.15	04-Apr-2003	perry	branches: 1.15.2; fix from OpenBSD, contributed by Gabriel Gonzalez in PR kern/20999
1.14	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.13	31-Jan-2003	thorpej	Use aprint_*().
1.12	06-Jan-2003	wiz	interrupt with two rs.
1.11	07-Nov-2002	someya	Fix recording problem on SiS7018.
1.10	20-Oct-2002	someya	Bit shifts for interrupt channel exceeded u_int32_t width.
1.9	12-Oct-2002	someya	add autri_flags_codec() from OpenBSD, reported by grendel@zeitbombe.org
1.8	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.7	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.6	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.5	30-May-2002	someya	Fix playback problem on SiS7018. - mute the volume for interrupt channel - add some improvements for device initialization but recording does not work properly. Reported and tested by Jan Wagner <waja@cyconet.org>. Thanks.
1.4	05-Feb-2002	eeh	branches: 1.4.8; Fix a blatant inconsistency in PCI_PRODUCT handling.
1.3	18-Nov-2001	augustss	branches: 1.3.2; 1.3.4; ANSIfy driver.
1.2	18-Nov-2001	augustss	Don't use a static variable in the power hook. It doesn't work with more than one device.
1.1	18-Nov-2001	someya	add Trident 4DWAVE based (4DWAVE DX/NX,SiS 7018,ALi M5451) PCI audio driver
1.3.4.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.4.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.4.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.3.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.4.1	18-Nov-2001	thorpej	file autri.c was added on branch kqueue on 2002-01-10 19:56:26 +0000
1.3.2.7	07-Jan-2003	thorpej	Sync with HEAD.
1.3.2.6	11-Nov-2002	nathanw	Catch up to -current
1.3.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.3.2.3	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.3.2.1	18-Nov-2001	nathanw	file autri.c was added on branch nathanw_sa on 2002-01-08 00:30:58 +0000
1.4.8.1	20-Jun-2002	gehenna	catch up with -current.
1.15.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.15.2.8	04-Feb-2005	skrll	Sync with HEAD.
1.15.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.15.2.6	14-Nov-2004	skrll	Sync with HEAD.
1.15.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.15.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.15.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.15.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.15.2.1	03-Aug-2004	skrll	Sync with HEAD
1.18.2.1	22-Sep-2004	jmc	Pullup rev 1.20 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.22.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.22.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.24.4.1	12-Feb-2005	yamt	sync with head.
1.24.2.1	29-Apr-2005	kent	sync with -current
1.26.2.6	17-Mar-2008	yamt	sync with head.
1.26.2.5	04-Feb-2008	yamt	sync with head.
1.26.2.4	21-Jan-2008	yamt	sync with head
1.26.2.3	27-Oct-2007	yamt	sync with head.
1.26.2.2	30-Dec-2006	yamt	sync with head.
1.26.2.1	21-Jun-2006	yamt	sync with head.
1.28.6.1	22-Apr-2006	simonb	Sync with head.
1.28.4.1	09-Sep-2006	rpaulo	sync with head
1.28.2.1	01-Mar-2006	yamt	sync with head.
1.29.4.1	19-Apr-2006	elad	sync with head.
1.29.2.2	03-Sep-2006	yamt	sync with head.
1.29.2.1	13-Mar-2006	yamt	sync with head.
1.31.4.2	10-Dec-2006	yamt	sync with head.
1.31.4.1	22-Oct-2006	yamt	sync with head
1.31.2.1	18-Nov-2006	ad	Sync with head.
1.34.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.34.24.3	23-Mar-2008	matt	sync with HEAD
1.34.24.2	09-Jan-2008	matt	sync with HEAD
1.34.24.1	06-Nov-2007	matt	sync with HEAD
1.34.22.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.34.22.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.34.22.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.34.22.1	04-Oct-2007	joerg	Convert autri(4) to PNP power handling.
1.34.8.1	23-Oct-2007	ad	Sync with head.
1.34.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.35.6.1	11-Dec-2007	yamt	sync with head.
1.35.4.1	26-Dec-2007	ad	Sync with head.
1.35.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.37.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.37.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.37.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.37.2.1	24-Mar-2008	keiichi	sync with head.
1.39.8.1	03-Jul-2008	simonb	Sync with head.
1.39.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.39.4.3	11-Mar-2010	yamt	sync with head
1.39.4.2	16-May-2009	yamt	sync with head
1.39.4.1	04-May-2009	yamt	sync with head.
1.40.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.40.8.2	12-Dec-2008	ad	Checkpoint work in progress.
1.40.8.1	08-Dec-2008	ad	Convert some more drivers.
1.45.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.46.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.46.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.46.10.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.46.10.1	17-Apr-2012	yamt	sync with head
1.48.2.2	05-Apr-2012	mrg	sync to latest -current.
1.48.2.1	18-Feb-2012	mrg	merge to -current.
1.50.4.1	18-May-2014	rmind	sync with head
1.50.2.2	03-Dec-2017	jdolecek	update from HEAD
1.50.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.51.2.1	10-Aug-2014	tls	Rebase.
1.52.4.2	28-Aug-2017	skrll	Sync with HEAD
1.52.4.1	09-Jul-2016	skrll	Sync with HEAD
1.53.8.1	11-May-2017	pgoyette	Sync with HEAD
1.55.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.55.10.1	10-Jun-2019	christos	Sync with HEAD
1.56.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.56.2.3	27-Apr-2019	isaki	More adapt to audio2. - Shrink autri_formats[]. HW actually supports frequencies other than 48kHz but they may include (ignorable) rounding error in thier calculation. So accept only 48kHz. It's enough. - Remove obsoleted and empty methods.
1.56.2.2	27-Apr-2019	isaki	Adapt to audio2. Tested on sparc64 by macallan@. Thank you.
1.56.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.58.4.1	29-Feb-2020	ad	Sync with head.
1.3	07-Nov-2002	someya	Fix recording problem on SiS7018.
1.2	30-May-2002	someya	Fix playback problem on SiS7018. - mute the volume for interrupt channel - add some improvements for device initialization but recording does not work properly. Reported and tested by Jan Wagner <waja@cyconet.org>. Thanks.
1.1	18-Nov-2001	someya	branches: 1.1.2; 1.1.4; 1.1.12; add Trident 4DWAVE based (4DWAVE DX/NX,SiS 7018,ALi M5451) PCI audio driver
1.1.12.1	20-Jun-2002	gehenna	catch up with -current.
1.1.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.1	18-Nov-2001	thorpej	file autrireg.h was added on branch kqueue on 2002-01-10 19:56:26 +0000
1.1.2.4	11-Nov-2002	nathanw	Catch up to -current
1.1.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.1	18-Nov-2001	nathanw	file autrireg.h was added on branch nathanw_sa on 2002-01-08 00:30:59 +0000
1.6	17-Mar-2012	martin	Split device_t
1.5	23-Nov-2011	jmcneill	branches: 1.5.2; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.4	09-Dec-2007	jmcneill	branches: 1.4.26; 1.4.48; 1.4.50; Merge jmcneill-pm branch.
1.3	04-Mar-2007	christos	branches: 1.3.14; 1.3.16; 1.3.24; 1.3.26; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.2	18-Nov-2001	augustss	branches: 1.2.2; 1.2.4; 1.2.36; 1.2.74; 1.2.76; ANSIfy driver.
1.1	18-Nov-2001	someya	add Trident 4DWAVE based (4DWAVE DX/NX,SiS 7018,ALi M5451) PCI audio driver
1.2.76.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.2.74.1	12-Mar-2007	rmind	Sync with HEAD.
1.2.36.2	21-Jan-2008	yamt	sync with head
1.2.36.1	03-Sep-2007	yamt	sync with head.
1.2.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.4.1	18-Nov-2001	thorpej	file autrivar.h was added on branch kqueue on 2002-01-10 19:56:27 +0000
1.2.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.2.2.1	18-Nov-2001	nathanw	file autrivar.h was added on branch nathanw_sa on 2002-01-08 00:30:59 +0000
1.3.26.1	11-Dec-2007	yamt	sync with head.
1.3.24.1	26-Dec-2007	ad	Sync with head.
1.3.16.1	09-Jan-2008	matt	sync with HEAD
1.3.14.1	04-Oct-2007	joerg	Convert autri(4) to PNP power handling.
1.4.50.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.4.48.1	17-Apr-2012	yamt	sync with head
1.4.26.1	08-Dec-2008	ad	Convert some more drivers.
1.5.2.1	05-Apr-2012	mrg	sync to latest -current.
1.87	06-Feb-2021	isaki	Hold sc_intr_lock in open()/close(). vtbl->{lock,unlock} (in fact ac97_{lock,unlock}) don't seem to need releasing an interrupt lock. Confirmed on auich.
1.86	19-Apr-2020	isaki	branches: 1.86.2; Fix round_blocksize not to return 0.
1.85	28-Feb-2020	isaki	branches: 1.85.4; Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017.
1.84	08-Jun-2019	isaki	branches: 1.84.2; 1.84.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.83	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.82	16-Mar-2019	isaki	branches: 1.82.2; Use C99 style struct initializer to audio_hw_if.
1.81	16-Mar-2019	isaki	Fix (probably) typo. SLINEAR_LE -> ULINEAR_LE.
1.80	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.79	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.78	01-Jun-2017	chs	branches: 1.78.8; 1.78.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.77	29-Mar-2014	christos	branches: 1.77.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.76	16-Oct-2013	christos	remove unused variables.
1.75	24-Nov-2011	mrg	branches: 1.75.8; 1.75.12; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.74	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.73	13-Nov-2010	uebayasi	branches: 1.73.8; 1.73.10; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.72	24-Feb-2010	dyoung	branches: 1.72.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.71	08-Jan-2010	dyoung	branches: 1.71.2; Expand PMF_FN_* macros.
1.70	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.69	06-May-2009	cegger	device_t/softc split tested by jakllsch@
1.68	05-May-2009	cegger	struct cfdata * -> cfdata_t
1.67	11-Oct-2008	dholland	branches: 1.67.6; 1.67.8; Print a message on two previously silent error paths in auvia_attach.
1.66	01-Jul-2008	jmcneill	branches: 1.66.2; VT8233+ uses a different register base for capturing, tested by sborrill
1.65	28-Apr-2008	martin	branches: 1.65.2; 1.65.4; Remove clause 3 and 4 from TNF licenses
1.64	10-Apr-2008	cegger	branches: 1.64.2; 1.64.4; use aprint_*_dev and device_xname
1.63	23-Feb-2008	dyoung	Detach self and children. Use device_t and accessors. Use PMF_FN_*.
1.62	14-Jan-2008	tsutsui	branches: 1.62.2; 1.62.6; Use htole32() to write parameters to DMA descriptors for big endian machines. Tested on Pegasos by mrg@ on port-ofppc.
1.61	09-Dec-2007	jmcneill	branches: 1.61.2; Merge jmcneill-pm branch.
1.60	04-Mar-2007	christos	branches: 1.60.14; 1.60.16; 1.60.22; 1.60.24; 1.60.26; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.59	21-Feb-2007	thorpej	branches: 1.59.2; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.58	16-Nov-2006	christos	branches: 1.58.2; 1.58.4; 1.58.6; __unused removal on arguments; approved by core.
1.57	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.56	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.55	28-Aug-2006	christos	branches: 1.55.2; 1.55.4; add missing initializer
1.54	27-Jul-2006	kent	add support for AC'97 S/PDIF
1.53	28-Nov-2005	rpaulo	branches: 1.53.4; 1.53.8; PR 32178: Nicolas Joly: VT8233 AC'97 Audio revision 0x10 small display fix. I just changed the comment a bit.
1.52	28-Jun-2005	thorpej	branches: 1.52.2; 1.52.8; Use ANSI function decls and static.
1.51	15-Jan-2005	kent	ansify and KNF
1.50	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.49	16-Nov-2004	xtraeme	branches: 1.49.2; Fix typo.
1.48	13-Nov-2004	kent	adopt auconv_set_converter()
1.47	10-Nov-2004	kent	remove unused variable
1.46	10-Nov-2004	kent	move IS_FIXED_RATE(), SUPPORTS_4CH(), and SUPPORTS_6CH to ac97var.h
1.45	08-Nov-2004	kent	print dmesg correctly for VT8237
1.44	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.43	17-Oct-2004	kent	add support for suspending/resuming. patch provided by Rhialto in PR#26432
1.42	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.41	09-Jul-2004	mycroft	Clean up. Attempt to make these safer against stray interrupts.
1.40	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.39	22-Apr-2004	itojun	sprintf -> snprintf
1.38	12-Apr-2004	jmmv	Be consistent with other messages: output "VIA Technologies" instead of only "VIA". Also add a missing "Audio" word to a message. Closes PR kern/25152 by Nicolas Joly.
1.37	31-Mar-2004	jmcneill	Don't allow block sizes < 288 bytes on VT823x chips (patch from kent@). Fixes kern/22788.
1.36	25-Mar-2004	xtraeme	branches: 1.36.2; err fix previous (return -> break) in case statement.
1.35	25-Mar-2004	xtraeme	There are two revisions of the VT8235 audio chipset (0x50 and 0x60), so print any of them as VT8235, approved by martin@.
1.34	25-Mar-2004	martin	Better identify some revisions of via 823* audio chips. From PR kern/24906.
1.33	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.32	14-Sep-2003	jmmv	Enable auvia_reset_codec's code and wait until the card is available. Fixes 'invalid codec' problems (at least with AD1980). Ok'ed by tsarna@.
1.31	03-May-2003	wiz	branches: 1.31.2; DMA, not dma nor Dma.
1.30	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.29	31-Jan-2003	thorpej	Use aprint_*().
1.28	04-Nov-2002	kent	auvia_set_params: For a fixed-rate codec, when AUMODE_RECORD and AUDIO_ENCODING_MULAW or AUDIO_ENCODING_ALAW, use AUDIO_ENCODING_SLINEAR_LE/16bit for native encoding instead of AUDIO_ENCODING_ULINEAR/8bit because aurateconv does not support sampling rate conversion for 8 bit PCM. This change fixes PR kern/18834.
1.27	16-Oct-2002	kent	Support for 4ch/6ch audio playback with VT8233/VT8235.
1.26	08-Oct-2002	kent	- Enable in-kernel sampling rate conversion for fixed-rate codec - Remove VRA/VRM enabling in auvia_attach() - Use ac97_codec_if::set_rate() - Return no AUDIO_PROPS_MMAP for fixed-rate codec - Remove unused variables in auvia_softc
1.25	07-Oct-2002	kent	Revive VIA8233_RP_DXS_[LR]VOL in auvia_trigger_*().
1.24	06-Oct-2002	kent	- auvia_trigger_output(): correct register; AUVIA_RP_CONTROL -> AUVIA_RP_MODE - auvia_trigger_output(), auvia_trigger_input(): Remove DXS_[LR]VOL setting. - whitespace
1.23	06-Oct-2002	kent	Add some constant symbols for AC97 registers and values based on OpenBSD. Rename some symbols to sync with OpenBSD: AC97_REG_EXTENDED_ID -> AC97_REG_EXT_AUDIO_ID AC97_REG_EXTENDED_STATUS -> AC97_REG_EXT_AUDIO_CTRL AC97_CODEC_DOES_VRA/AC97_ENAB_VRA -> AC97_EXT_AUDIO_VRA AC97_CODEC_DOES_MICVRA/AC97_ENAB_MICVRA -> AC97_EXT_AUDIO_VRM ac97feature: "mic channel" -> "dedicated mic channel" ac97_attach(): print extended capabilities; VRA, DRA, SPDIF, VRM, CDAC, SDAC, LDAC.
1.22	05-Oct-2002	kent	Add support for VIA VT8233/VT8235. The patch was provided by Ola Eriksson <ola at mreriksson dot net> in PR kern/18375. It is based on auvia of OpenBSD.
1.21	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.20	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.19	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.18	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.17	02-Apr-2002	fvdl	Return 0 when no interrupt was handled.
1.16	16-Mar-2002	jmcneill	Backout addition of the VIA VT8233.
1.15	16-Mar-2002	jmcneill	Add support for the VT8233 AC'97.
1.14	13-Nov-2001	lukem	add RCSID
1.13	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.12	04-Aug-2001	jdolecek	branches: 1.12.2; It seems the fixed rate versions are actually fixed at 48 kHz, not 44.1 kHz.
1.11	19-Feb-2001	fvdl	branches: 1.11.2; 1.11.4; Switch line in mute off by default.
1.10	05-Jan-2001	augustss	Don't make local defines for stuff that's already in ac97reg.h.
1.9	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.8	10-Dec-2000	jdolecek	do not try to get/set sample rate on codecs which don't support variable rate audio - always report/use fixed rate of 44100 for them, so that the other audio code DTRT
1.7	15-Nov-2000	jdolecek	branches: 1.7.2; if AC97_CODEC_DOES_VRA is not supported, only print warning than hw Variable Rate Audio is not supported and continue attaching the driver. This is I believe more correct than what was done in rev 1.5
1.6	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.5	01-Sep-2000	jdolecek	even through the integrated audio on my MSI 694D Pro does not set AC97_CODEC_DOES_VRA in AC97_REG_EXT_AUDIO_ID request, it works okay; only require the bit set for revision < 'H'
1.4	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.3	15-May-2000	thorpej	branches: 1.3.4; ac97.h -> ac97var.h
1.2	08-Apr-2000	tsarna	Fix auvia_free, clean up more on close, and be more careful in interrupt handler.
1.1	31-Mar-2000	tsarna	Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.3.4.2	01-Sep-2000	jdolecek	pullup rev. 1.5 (approved by thorpej): rev. H apparently doesn't set AC97_CODEC_DOES_VRA bit but works ok with this driver, so only require the bit set for revisions < H
1.3.4.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.7.2.6	12-Mar-2001	bouyer	Sync with HEAD.
1.7.2.5	05-Jan-2001	bouyer	Sync with HEAD
1.7.2.4	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.7.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.7.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.1	15-Nov-2000	bouyer	file auvia.c was added on branch thorpej_scsipi on 2000-11-20 11:42:14 +0000
1.11.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.11.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.11.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.11.4.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.11.2.7	11-Nov-2002	nathanw	Catch up to -current
1.11.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.11.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.11.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.11.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.11.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.11.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.12.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.31.2.11	11-Dec-2005	christos	Sync with head.
1.31.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.31.2.9	17-Jan-2005	skrll	Sync with HEAD.
1.31.2.8	29-Nov-2004	skrll	Sync with HEAD.
1.31.2.7	14-Nov-2004	skrll	Sync with HEAD.
1.31.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.31.2.5	19-Oct-2004	skrll	Sync with HEAD
1.31.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.31.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.31.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.31.2.1	03-Aug-2004	skrll	Sync with HEAD
1.36.2.2	22-Sep-2004	jmc	branches: 1.36.2.2.2; Pullup rev 1.42 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.36.2.1	01-Apr-2004	jmc	Pullup rev 1.37 (requested by jmcneill in ticket #36) Don't allow block sizes < 288 bytes on VT823x chips. PR#22788
1.36.2.2.2.1	30-Jan-2005	he	Pull up revision 1.43 (requested by kent in ticket #949): Add support for suspending/resuming. Fixes PR#26432.
1.49.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.49.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.52.8.1	29-Nov-2005	yamt	sync with head.
1.52.2.6	27-Feb-2008	yamt	sync with head.
1.52.2.5	21-Jan-2008	yamt	sync with head
1.52.2.4	03-Sep-2007	yamt	sync with head.
1.52.2.3	26-Feb-2007	yamt	sync with head.
1.52.2.2	30-Dec-2006	yamt	sync with head.
1.52.2.1	21-Jun-2006	yamt	sync with head.
1.53.8.2	03-Sep-2006	yamt	sync with head.
1.53.8.1	11-Aug-2006	yamt	sync with head
1.53.4.1	09-Sep-2006	rpaulo	sync with head
1.55.4.2	10-Dec-2006	yamt	sync with head.
1.55.4.1	22-Oct-2006	yamt	sync with head
1.55.2.1	18-Nov-2006	ad	Sync with head.
1.58.6.1	04-Sep-2008	skrll	Sync with netbsd-4.
1.58.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.58.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.58.2.1	06-Jul-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1161): sys/dev/pci/auvia.c: revision 1.66 VT8233+ uses a different register base for capturing, tested by sborrill
1.59.2.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.60.26.1	11-Dec-2007	yamt	sync with head.
1.60.24.1	26-Dec-2007	ad	Sync with head.
1.60.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.60.16.2	23-Mar-2008	matt	sync with HEAD
1.60.16.1	09-Jan-2008	matt	sync with HEAD
1.60.14.3	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.60.14.2	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.60.14.1	04-Oct-2007	joerg	Convert auvia(4) to PNP power management.
1.61.2.1	19-Jan-2008	bouyer	Sync with HEAD
1.62.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.62.6.3	02-Jul-2008	mjf	Sync with HEAD.
1.62.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.62.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.62.2.1	24-Mar-2008	keiichi	sync with head.
1.64.4.4	11-Mar-2010	yamt	sync with head
1.64.4.3	16-May-2009	yamt	sync with head
1.64.4.2	04-May-2009	yamt	sync with head.
1.64.4.1	16-May-2008	yamt	sync with head.
1.64.2.1	18-May-2008	yamt	sync with head.
1.65.4.1	03-Jul-2008	simonb	Sync with head.
1.65.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.66.2.1	19-Oct-2008	haad	Sync with HEAD.
1.67.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.67.6.3	13-Dec-2008	ad	Fix locking for ac97.
1.67.6.2	11-Dec-2008	ad	Checkpoint work in progress.
1.67.6.1	08-Dec-2008	ad	Convert some more drivers.
1.71.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.72.2.1	05-Mar-2011	rmind	sync with head
1.73.10.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.73.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.73.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.73.8.1	17-Apr-2012	yamt	sync with head
1.75.12.1	18-May-2014	rmind	sync with head
1.75.8.2	03-Dec-2017	jdolecek	update from HEAD
1.75.8.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.77.6.1	28-Aug-2017	skrll	Sync with HEAD
1.78.10.3	21-Apr-2020	martin	Sync with HEAD
1.78.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.78.10.1	10-Jun-2019	christos	Sync with HEAD
1.78.8.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.78.8.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.82.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.82.2.3	03-May-2019	isaki	Remove unnecessary header file.
1.82.2.2	27-Apr-2019	isaki	Adapt to audio2. - auvia_set_params_sub is also renamed because caller auvia_set_params is renamed to auvia_set_format.
1.82.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.84.4.1	29-Feb-2020	ad	Sync with head.
1.84.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by isaki in ticket #792): sys/arch/amiga/dev/aucc.c: revision 1.48 sys/dev/pci/auixp.c: revision 1.49 sys/dev/pci/gcscaudio.c: revision 1.19 sys/dev/pci/auich.c: revision 1.159 sys/dev/sbus/dbri.c: revision 1.42 sys/dev/pci/auvia.c: revision 1.85 sys/dev/pci/auacer.c: revision 1.39 Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017. round_blocksize must return a multiple of the framesize. aucc(4) supports 3 channels mode. round_blocksize must return a multiple of the framesize even if passed blocksize is greater than the upper limit. round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4. round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.85.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.86.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.17	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	23-Nov-2011	jmcneill	branches: 1.16.50; 1.16.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.15	06-May-2009	cegger	branches: 1.15.12; 1.15.14; device_t/softc split tested by jakllsch@
1.14	28-Apr-2008	martin	branches: 1.14.12; 1.14.14; Remove clause 3 and 4 from TNF licenses
1.13	23-Feb-2008	dyoung	branches: 1.13.2; 1.13.4; Detach self and children. Use device_t and accessors. Use PMF_FN_*.
1.12	09-Dec-2007	jmcneill	branches: 1.12.6; 1.12.10; Merge jmcneill-pm branch.
1.11	21-Feb-2007	thorpej	branches: 1.11.2; 1.11.16; 1.11.18; 1.11.26; 1.11.28; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.10	27-Jul-2006	kent	branches: 1.10.10; add support for AC'97 S/PDIF
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.8; merge ktrace-lwp.
1.8	13-Nov-2004	kent	branches: 1.8.12; adopt auconv_set_converter()
1.7	10-Nov-2004	kent	move IS_FIXED_RATE(), SUPPORTS_4CH(), and SUPPORTS_6CH to ac97var.h
1.6	17-Oct-2004	kent	add support for suspending/resuming. patch provided by Rhialto in PR#26432
1.5	16-Oct-2002	kent	branches: 1.5.6; 1.5.10; Support for 4ch/6ch audio playback with VT8233/VT8235.
1.4	08-Oct-2002	kent	- Enable in-kernel sampling rate conversion for fixed-rate codec - Remove VRA/VRM enabling in auvia_attach() - Use ac97_codec_if::set_rate() - Return no AUDIO_PROPS_MMAP for fixed-rate codec - Remove unused variables in auvia_softc
1.3	05-Oct-2002	kent	Add support for VIA VT8233/VT8235. The patch was provided by Ola Eriksson <ola at mreriksson dot net> in PR kern/18375. It is based on auvia of OpenBSD.
1.2	10-Dec-2000	jdolecek	branches: 1.2.2; do not try to get/set sample rate on codecs which don't support variable rate audio - always report/use fixed rate of 44100 for them, so that the other audio code DTRT
1.1	31-Mar-2000	tsarna	branches: 1.1.6; Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.1.6.3	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	31-Mar-2000	bouyer	file auviavar.h was added on branch thorpej_scsipi on 2000-11-20 11:42:14 +0000
1.2.2.1	18-Oct-2002	nathanw	Catch up to -current.
1.5.10.1	30-Jan-2005	he	Pull up revision 1.6 (requested by kent in ticket #949): Add support for suspending/resuming. Fixes PR#26432.
1.5.6.3	29-Nov-2004	skrll	Sync with HEAD.
1.5.6.2	14-Nov-2004	skrll	Sync with HEAD.
1.5.6.1	19-Oct-2004	skrll	Sync with HEAD
1.8.12.4	27-Feb-2008	yamt	sync with head.
1.8.12.3	21-Jan-2008	yamt	sync with head
1.8.12.2	26-Feb-2007	yamt	sync with head.
1.8.12.1	30-Dec-2006	yamt	sync with head.
1.9.8.1	11-Aug-2006	yamt	sync with head
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.10.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.11.28.1	11-Dec-2007	yamt	sync with head.
1.11.26.1	26-Dec-2007	ad	Sync with head.
1.11.18.2	23-Mar-2008	matt	sync with HEAD
1.11.18.1	09-Jan-2008	matt	sync with HEAD
1.11.16.1	04-Oct-2007	joerg	Convert auvia(4) to PNP power management.
1.11.2.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.12.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.6.1	24-Mar-2008	keiichi	sync with head.
1.13.4.2	16-May-2009	yamt	sync with head
1.13.4.1	16-May-2008	yamt	sync with head.
1.13.2.1	18-May-2008	yamt	sync with head.
1.14.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.14.12.2	11-Dec-2008	ad	Checkpoint work in progress.
1.14.12.1	08-Dec-2008	ad	Convert some more drivers.
1.15.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.15.12.1	17-Apr-2012	yamt	sync with head
1.16.54.1	27-Apr-2019	isaki	Adapt to audio2. - auvia_set_params_sub is also renamed because caller auvia_set_params is renamed to auvia_set_format.
1.16.50.1	10-Jun-2019	christos	Sync with HEAD
1.89	09-Feb-2020	jmcneill	Retire azalia(4).
1.88	08-Jun-2019	isaki	branches: 1.88.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.87	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.86	16-Mar-2019	isaki	branches: 1.86.2; Use C99 style struct initializer to audio_hw_if.
1.85	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.84	01-Jun-2017	chs	branches: 1.84.8; 1.84.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.83	09-Nov-2014	nonaka	branches: 1.83.2; Wait for the RIRB DMA engine to come ready. Needed on the audio device integrated in the Vortex86EX SoC. From OpenBSD.
1.82	21-Sep-2014	christos	Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.81	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.80	20-Oct-2013	christos	move debugging code inside ifdef debug
1.79	24-Nov-2011	mrg	branches: 1.79.8; 1.79.12; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.78	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.77	29-Aug-2011	jmcneill	branches: 1.77.2; 1.77.4; have pci register itself as a module, and make PCI driver modules depend on it
1.76	28-Aug-2011	jmcneill	set default WARNS for modules to 3 -- the only one that needs < 3 now is pf
1.75	25-May-2010	pgoyette	Revert earlier changes to callers of pci_find{vendor,product} since the original routine entrypoint name has been restored. Requested by dyoung@ and mrg@
1.74	24-May-2010	pgoyette	Update all callers of the pci_find{vendor,product} routines to now call these routines through their global pointers.
1.73	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.72	20-Jan-2010	tonnerre	branches: 1.72.2; Signedness bug/crash in azalia and hdaudio. Fixes PR 42604. By Pierre Pronchery.
1.71	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.70	12-Nov-2009	dyoung	Remove superfluous activation hook.
1.69	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.68	02-Apr-2009	dyoung	Cosmetic changes, only, to clarify and to save a couple of lines: return a constant, 0, instead of returning ret when it is always 0. Wait to initialize ret until we really need to.
1.67	25-Feb-2009	jmcneill	PR# port-i386/40143: Viewing an mpeg transport stream with mplayer causes crash Not directly related to the PR but this bug was discovered while testing. In azalia_stream_halt, invalidate the intr ptr to prevent a race condition where azalia_stream_intr can jump off into lala land during stream shutdown.
1.66	17-Dec-2008	cegger	branches: 1.66.2; make this build with AZALIA_DEBUG
1.65	16-Dec-2008	christos	replace bitmask_snprintf(9) with snprintb(3)
1.64	26-Jun-2008	kent	branches: 1.64.4; 1.64.6; 1.64.8; * azalia_rirb_intr() fix a problem that normal responses are skipped unintentionally. * azalia_stream_start() access HDA_INTCTL with AZ_READ_4()/AZ_WRITE_4(). access HDA_SD_CTL with STR_READ_1()/AZ_WRITE_1(). * azalia_modcmd() enclose config_cfdata_attach() with splaudio().
1.63	08-Jun-2008	cegger	branches: 1.63.2; make this build with AZALIA_DEBUG get rid of XNAME macro
1.62	07-Jun-2008	freza	Split device_t/softc. Remove unused macros.
1.61	02-Jun-2008	jmcneill	Backout previous; causes a panic on attach/detach/reattach for some unknown reason.
1.60	02-Jun-2008	jmcneill	Halt streams before deleting them on detach.
1.59	02-Jun-2008	jmcneill	Fix MODULE_CMD_FINI case, azalia can be unloaded now.
1.58	02-Jun-2008	jmcneill	Allow azalia to be built as a module.
1.57	13-May-2008	mjf	Do not dereference this->intr if it's NULL.
1.56	10-May-2008	jmcneill	Cleanup aprint messages, remove unnecessary usage of aprint_error.
1.55	28-Apr-2008	martin	branches: 1.55.2; Remove clause 3 and 4 from TNF licenses
1.54	10-Apr-2008	cegger	branches: 1.54.2; 1.54.4; use aprint_*_dev and device_xname
1.53	23-Feb-2008	dyoung	Add a method to detach children. Use device_t and accessors. Use PMF_FN_*.
1.52	09-Dec-2007	jmcneill	branches: 1.52.6; 1.52.10; Merge jmcneill-pm branch.
1.51	14-Aug-2007	kent	branches: 1.51.2; 1.51.10; 1.51.12; support for playback-only devices such as Radeon HD 2x00
1.50	13-May-2007	kent	branches: 1.50.2; 1.50.6; Add PCI_SUBCLASS_MULTIMEDIA_HDAUDIO to pcireg.h and azalia(4) uses it
1.49	08-May-2007	kent	* azalia_code_connect_stream() set stream#0 for unused converters in a group PR#35658 by kikuma norimitsu
1.48	08-May-2007	kent	Some codecs need explicit initialization even for uni-directional pins. Pointed out by Deanna Phillips
1.47	11-Mar-2007	kent	branches: 1.47.2; don't start dmesg lines with \t to do grep easily
1.46	10-Mar-2007	kent	branches: 1.46.2; add diagnostic code in azalia_codec_construct_format()
1.45	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.44	21-Feb-2007	thorpej	branches: 1.44.2; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.43	05-Feb-2007	kent	branches: 1.43.2; If both of AZALIA_DEBUG and AZALIA_DEBUG_DOT are defined, a DOT file for Graphviz representing the widget structure is printed.
1.42	16-Nov-2006	christos	branches: 1.42.2; 1.42.4; __unused removal on arguments; approved by core.
1.41	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.40	24-Sep-2006	jmcneill	Don't use // for comments.
1.39	28-Aug-2006	christos	branches: 1.39.2; 1.39.4; add missing initializer.
1.38	21-Jul-2006	kent	* azalia_init_corb() No need to confirm CORBRPRST is set even with ICH6/ICH7
1.37	13-Jul-2006	kent	For MCP55, don't check CORBRPRST flag a part of PR#33741
1.36	27-Jun-2006	kent	* azalia_pci_attach() print the vendor name of the device * azalia_pci_detach() add some debug messages * azalia_codec_delete() fix a possible NULL dereference
1.35	25-Jun-2006	kent	* disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.34	25-Jun-2006	kent	ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.33	25-Jun-2006	kent	printf() -> DPRINTF(())
1.32	25-Jun-2006	kent	avoid infinite recursive call of azalia_rirb_kick_unsol_events()
1.31	25-Jun-2006	kent	handle unsolicited responses. A codec can receive unsolicited responses by codec_t::unsol_event()
1.30	24-Jun-2006	kent	If the encoding of an audio converter is 0, refers the audio function node. CMI9880 0.2 has such ADCs. PR#33787
1.29	16-Jun-2006	kent	check return values of azalia_init_corb() and azalia_init_rirb() a part of PR#33741
1.28	15-Jun-2006	kent	support for a converter group of which members have non-identical bits_rates.
1.27	13-Jun-2006	kent	branches: 1.27.2; - always prints the codec index in dmesg - add missing \n
1.26	11-Jun-2006	kent	fix audio_format allocation for recording
1.25	11-Jun-2006	kent	support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.24	11-Jun-2006	kent	If the codec supports for modem, print that we support for no modem explicitly.
1.23	09-Jun-2006	kent	- print audio format only when the format is changed - print the name of STAC9220 codec
1.22	09-Jun-2006	kent	shorten audio format message
1.21	07-Jun-2006	kent	* move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.20	07-May-2006	kent	branches: 1.20.2; Don't assume the device always has PCI_MAPREG_MEM_TYPE_64BIT. This change might help nForce 410/430 support.
1.19	16-Jan-2006	kent	branches: 1.19.2; 1.19.4; 1.19.6; 1.19.8; 1.19.10; Don't crash when a connection list has invalid NIDs. PR#32485 by Samuel Tran
1.18	08-Jan-2006	kent	add some trace messages in azalia_mixer_init().
1.17	03-Jan-2006	kent	- make this compile with AZALIA_DEBUG Thanks to Tomokatsu HAYAKAWA. - introduce WIDGET_CHANNELS macro
1.16	11-Dec-2005	christos	branches: 1.16.2; merge ktrace-lwp.
1.15	29-Sep-2005	kent	branches: 1.15.6; * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.14	28-Sep-2005	kent	add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.13	28-Sep-2005	kent	split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.12	26-Sep-2005	kent	* azalia_attach_intr() ignore errors of codec initializations if at least one codec is initialized successfully * azalia_alloc_dmamem() fail if the HDA controller does not support 64 bit addressing and buf_dmamap_* allocate a 64 bit address. Pointed out by Brett Lymn
1.11	02-Aug-2005	kent	* azalia_mixer_init() set AUDIO_MAX_GAIN/2 to all of volume items.
1.10	02-Aug-2005	kent	- handle widgets without `format override' or `amp param override' - support for volume knob
1.9	31-Jul-2005	kent	- make sure the power state of a node is D0 - ignore unsoliticited responses
1.8	29-Jul-2005	kent	make this compile with AZALIA_DEBUG. thanks to cube@
1.7	25-Jun-2005	kent	branches: 1.7.2; 1.7.4; - free all of allocated resources when an error occurs - code cleanup
1.6	21-Jun-2005	kent	support for recording
1.5	20-Jun-2005	kent	- print the controller name if possible - make the max values of mixers AUDIO_MAX_GAIN - reset the stream in trigger_output()
1.4	19-Jun-2005	kent	- add capabilities to specify a DAC group or an ADC. - support for 4,6,8 channels playback
1.3	19-Jun-2005	kent	- add support for headphone-boost mixer controls - fix a bug of creation of output mute - unmute all controls by azalia_set_mixer() - if a bidirectional pin is green, orange, gray, or black, set it as "output" direction by default.
1.2	19-Jun-2005	kent	partial support for mixer. Still we have no ways to specify DACs or ADCs.
1.1	18-Jun-2005	kent	A driver for High Definition Audio. The driver is highly experimental at this moment: - limited support for playback - no support for mixer controls - no support for recording
1.7.4.6	27-Feb-2008	yamt	sync with head.
1.7.4.5	21-Jan-2008	yamt	sync with head
1.7.4.4	03-Sep-2007	yamt	sync with head.
1.7.4.3	26-Feb-2007	yamt	sync with head.
1.7.4.2	30-Dec-2006	yamt	sync with head.
1.7.4.1	21-Jun-2006	yamt	sync with head.
1.7.2.39	27-May-2007	bouyer	Pull up following revision(s) (requested by kent in ticket #1780): sys/dev/pci/azalia_codec.c: revisions 1.25, 1.28, 1.30 - 1.42 sys/dev/pci/azalia.c: revisions 1.43, 1.46 - 1.49 sys/dev/pci/azalia.h: revisions 1.15 - 1.16 Bug fixes, and add support for some more codecs. PR#34880, PR#35658, PR#36302
1.7.2.38	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.38 * azalia_init_corb() No need to confirm CORBRPRST is set even with ICH6/ICH7
1.7.2.37	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.37 For MCP55, don't check CORBRPRST flag a part of PR#33741
1.7.2.36	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.36 * azalia_pci_attach() print the vendor name of the device * azalia_pci_detach() add some debug messages * azalia_codec_delete() fix a possible NULL dereference
1.7.2.35	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.20 sys/dev/pci/azalia.h: revision 1.13 sys/dev/pci/azalia.c: revision 1.35 * disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.7.2.34	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.12 sys/dev/pci/azalia_codec.c: revision 1.19 sys/dev/pci/azalia.c: revision 1.34 ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.7.2.33	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.33 printf() -> DPRINTF(())
1.7.2.32	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.32 avoid infinite recursive call of azalia_rirb_kick_unsol_events()
1.7.2.31	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.11 sys/dev/pci/azalia.c: revision 1.31 handle unsolicited responses. A codec can receive unsolicited responses by codec_t::unsol_event()
1.7.2.30	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.30 If the encoding of an audio converter is 0, refers the audio function node. CMI9880 0.2 has such ADCs. PR#33787
1.7.2.29	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.29 check return values of azalia_init_corb() and azalia_init_rirb() a part of PR#33741
1.7.2.28	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.28 support for a converter group of which members have non-identical bits_rates.
1.7.2.27	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.27 - always prints the codec index in dmesg - add missing n
1.7.2.26	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.26 fix audio_format allocation for recording
1.7.2.25	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.25 sys/dev/pci/azalia.h: revision 1.9 sys/dev/pci/azalia_codec.c: revision 1.12 support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.7.2.24	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.24 If the codec supports for modem, print that we support for no modem explicitly.
1.7.2.23	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.23 sys/dev/pci/azalia.h: revision 1.8 sys/dev/pci/azalia_codec.c: revision 1.10 - print audio format only when the format is changed - print the name of STAC9220 codec
1.7.2.22	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.22 shorten audio format message
1.7.2.21	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.7 sys/dev/pci/azalia_codec.c: revision 1.9 sys/dev/pci/azalia.c: revision 1.21 * move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.7.2.20	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.20 Don't assume the device always has PCI_MAPREG_MEM_TYPE_64BIT. This change might help nForce 410/430 support.
1.7.2.19	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.h: revision 1.6 sys/dev/pci/azalia.c: revision 1.19 Don't crash when a connection list has invalid NIDs. PR#32485 by Samuel Tran
1.7.2.18	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.c: revision 1.18 add some trace messages in azalia_mixer_init().
1.7.2.17	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.h: revision 1.5 sys/dev/pci/azalia.c: revision 1.17 - make this compile with AZALIA_DEBUG Thanks to Tomokatsu HAYAKAWA. - introduce WIDGET_CHANNELS macro
1.7.2.16	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.3 sys/dev/pci/azalia_codec.c: revision 1.3 sys/dev/pci/azalia.c: revision 1.15 * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.7.2.15	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.2 sys/dev/pci/azalia_codec.c: revision 1.2 sys/dev/pci/azalia.c: revision 1.14 add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.7.2.14	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.1 sys/dev/pci/azalia_codec.c: revision 1.1 sys/dev/pci/azalia.c: revision 1.13 sys/dev/pci/files.pci: revision 1.239 split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.7.2.13	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.c: revision 1.12 * azalia_attach_intr() ignore errors of codec initializations if at least one codec is initialized successfully * azalia_alloc_dmamem() fail if the HDA controller does not support 64 bit addressing and buf_dmamap_* allocate a 64 bit address. Pointed out by Brett Lymn
1.7.2.12	14-Aug-2005	riz	branches: 1.7.2.12.2; Pull up revision 1.11 (requested by kent in ticket #647): * azalia_mixer_init() set AUDIO_MAX_GAIN/2 to all of volume items.
1.7.2.11	14-Aug-2005	riz	Pull up revision 1.10 (requested by kent in ticket #647): - handle widgets without `format override' or `amp param override' - support for volume knob
1.7.2.10	14-Aug-2005	riz	Pull up revision 1.9 (requested by kent in ticket #647): - make sure the power state of a node is D0 - ignore unsoliticited responses
1.7.2.9	14-Aug-2005	riz	Pull up revision 1.8 (requested by kent in ticket #647): make this compile with AZALIA_DEBUG. thanks to cube@
1.7.2.8	02-Jul-2005	tron	Pull up revision 1.7 (requested by kent in ticket #494): - free all of allocated resources when an error occurs - code cleanup
1.7.2.7	02-Jul-2005	tron	Pull up revision 1.6 (requested by kent in ticket #494): support for recording
1.7.2.6	02-Jul-2005	tron	Pull up revision 1.5 (requested by kent in ticket #494): - print the controller name if possible - make the max values of mixers AUDIO_MAX_GAIN - reset the stream in trigger_output()
1.7.2.5	02-Jul-2005	tron	Pull up revision 1.4 (requested by kent in ticket #494): - add capabilities to specify a DAC group or an ADC. - support for 4,6,8 channels playback
1.7.2.4	02-Jul-2005	tron	Pull up revision 1.3 (requested by kent in ticket #494): - add support for headphone-boost mixer controls - fix a bug of creation of output mute - unmute all controls by azalia_set_mixer() - if a bidirectional pin is green, orange, gray, or black, set it as "output" direction by default.
1.7.2.3	02-Jul-2005	tron	Pull up revision 1.2 (requested by kent in ticket #494): partial support for mixer. Still we have no ways to specify DACs or ADCs.
1.7.2.2	02-Jul-2005	tron	Pull up revision 1.1 (requested by kent in ticket #494): An audio driver for High Definition Audio. The driver is highly experimental at this moment: - limieted support for playback - support for no mixer controls - support for no recording
1.7.2.1	25-Jun-2005	tron	file azalia.c was added on branch netbsd-3 on 2005-07-02 16:39:31 +0000
1.7.2.12.2.1	02-Feb-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.c: revision 1.19 via patch Don't crash when a connection list has invalid NIDs. PR#32485 by Samuel Tran
1.15.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.15.6.1	29-Sep-2005	skrll	file azalia.c was added on branch ktrace-lwp on 2005-11-10 14:06:00 +0000
1.16.2.2	01-Feb-2006	yamt	sync with head.
1.16.2.1	15-Jan-2006	yamt	sync with head.
1.19.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.19.8.1	11-May-2006	elad	sync with head
1.19.6.4	03-Sep-2006	yamt	sync with head.
1.19.6.3	11-Aug-2006	yamt	sync with head
1.19.6.2	26-Jun-2006	yamt	sync with head.
1.19.6.1	24-May-2006	yamt	sync with head.
1.19.4.2	07-Jun-2006	kardel	Sync with head.
1.19.4.1	01-Jun-2006	kardel	Sync with head.
1.19.2.1	09-Sep-2006	rpaulo	sync with head
1.20.2.1	19-Jun-2006	chap	Sync with head.
1.27.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.39.4.2	10-Dec-2006	yamt	sync with head.
1.39.4.1	22-Oct-2006	yamt	sync with head
1.39.2.2	09-Feb-2007	ad	Sync with HEAD.
1.39.2.1	18-Nov-2006	ad	Sync with head.
1.42.4.1	03-Jun-2007	wrstuden	Catch up with now-somewhat-dated netbsd-4. These changes took longer than expected for me to actually get around to merging.
1.42.2.2	21-Jan-2010	snj	Apply patch (requested by tonnerre in ticket #1375): Signedness bug/crash in azalia. Fixes PR 42604.
1.42.2.1	20-May-2007	jdc	branches: 1.42.2.1.4; Pull up revisions 1.43, 1.46-1.49 (requested by kent in ticket #649). If both of AZALIA_DEBUG and AZALIA_DEBUG_DOT are defined, a DOT file for Graphviz representing the widget structure is printed. add diagnostic code in azalia_codec_construct_format() don't start dmesg lines with \t to do grep easily Some codecs need explicit initialization even for uni-directional pins. Pointed out by Deanna Phillips * azalia_code_connect_stream() set stream#0 for unused converters in a group PR#35658 by kikuma norimitsu
1.42.2.1.4.1	21-Jan-2010	snj	Apply patch (requested by tonnerre in ticket #1375): Signedness bug/crash in azalia. Fixes PR 42604.
1.43.2.3	17-May-2007	yamt	sync with head.
1.43.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.43.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.44.2.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.46.2.3	20-Aug-2007	ad	Sync with HEAD.
1.46.2.2	27-May-2007	ad	Sync with head.
1.46.2.1	13-Mar-2007	ad	Sync with head.
1.47.2.1	11-Jul-2007	mjf	Sync with head.
1.50.6.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.50.6.5	07-Nov-2007	joerg	Introduce device_has_power to fix a race between resuming a device and the device enabling interrupts as seen by jmcneill@ with uhci. Change ehci, ohci, uhci and azalia to use this function to protect the interrupt handler.
1.50.6.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.50.6.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.50.6.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.50.6.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.50.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.51.12.1	11-Dec-2007	yamt	sync with head.
1.51.10.1	26-Dec-2007	ad	Sync with head.
1.51.2.2	23-Mar-2008	matt	sync with HEAD
1.51.2.1	09-Jan-2008	matt	sync with HEAD
1.52.10.5	17-Jan-2009	mjf	Sync with HEAD.
1.52.10.4	29-Jun-2008	mjf	Sync with HEAD.
1.52.10.3	05-Jun-2008	mjf	Sync with HEAD. Also fix build.
1.52.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.52.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.52.6.1	24-Mar-2008	keiichi	sync with head.
1.54.4.4	11-Mar-2010	yamt	sync with head
1.54.4.3	16-May-2009	yamt	sync with head
1.54.4.2	04-May-2009	yamt	sync with head.
1.54.4.1	16-May-2008	yamt	sync with head.
1.54.2.3	17-Jun-2008	yamt	sync with head.
1.54.2.2	04-Jun-2008	yamt	sync with head
1.54.2.1	18-May-2008	yamt	sync with head.
1.55.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.55.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.63.2.1	27-Jun-2008	simonb	Sync with head.
1.64.8.2	12-Dec-2008	ad	Checkpoint work in progress.
1.64.8.1	11-Dec-2008	ad	Checkpoint work in progress.
1.64.6.2	21-Jan-2010	snj	Pull up following revision(s) (requested by tonnerre in ticket #1258): sys/dev/pci/azalia.c: revision 1.72 sys/dev/pci/hdaudio/hdaudio_afg.c: revision 1.19 Signedness bug/crash in azalia and hdaudio. Fixes PR 42604. By Pierre Pronchery.
1.64.6.1	02-Mar-2009	snj	branches: 1.64.6.1.2; 1.64.6.1.4; Pull up following revision(s) (requested by jmcneill in ticket #521): sys/dev/pci/azalia.c: revision 1.67 PR# port-i386/40143: Viewing an mpeg transport stream with mplayer causes crash Not directly related to the PR but this bug was discovered while testing. In azalia_stream_halt, invalidate the intr ptr to prevent a race condition where azalia_stream_intr can jump off into lala land during stream shutdown.
1.64.6.1.4.1	21-Apr-2010	matt	sync to netbsd-5
1.64.6.1.2.1	21-Jan-2010	snj	Pull up following revision(s) (requested by tonnerre in ticket #1258): sys/dev/pci/azalia.c: revision 1.72 Signedness bug/crash in azalia. Fixes PR 42604. By Pierre Pronchery.
1.64.4.3	28-Apr-2009	skrll	Sync with HEAD.
1.64.4.2	03-Mar-2009	skrll	Sync with HEAD.
1.64.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.66.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.72.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.77.4.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.77.4.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.77.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.77.2.1	17-Apr-2012	yamt	sync with head
1.79.12.1	18-May-2014	rmind	sync with head
1.79.8.2	03-Dec-2017	jdolecek	update from HEAD
1.79.8.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.83.2.1	28-Aug-2017	skrll	Sync with HEAD
1.84.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.84.10.1	10-Jun-2019	christos	Sync with HEAD
1.84.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.86.2.2	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.86.2.1	28-Apr-2019	isaki	Adapt to audio2.
1.88.4.1	29-Feb-2020	ad	Sync with head.
1.23	09-Feb-2020	jmcneill	Retire azalia(4).
1.22	08-May-2019	isaki	branches: 1.22.4; Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.21	23-Nov-2011	jmcneill	branches: 1.21.50; 1.21.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.20	14-Aug-2008	jmcneill	branches: 1.20.6; 1.20.28; 1.20.30; PR#38860: azalia record does not work on Realtek ALC662-GR o Don't add unconnected pins to generated mixer list, should help make auto-generated mixer lists slightly easier to work with. o Move COP_AWTYPE_AUDIO_MIXER type knobs to mix.* rather than polluting the inputs.* namespace and giving a few more characters for the auto-generated names. o Introduce an alternate to generic_mixer_init, generic_mixer_autoinit. generic_mixer_autoinit uses COP_AWTYPE_PIN_COMPLEX "default configuration" capability flags to determine proper pin widget control and eapd configuration, as well as creating virtual outputs.master, inputs.dac, and record.volume mixer controls that mirror the default dac and adc controls. o Introduce generic_mixer_init_widget which uses COP_AWTYPE_PIN_COMPLEX "default configuration" device type, gross location, and geometric locations to guess friendlier mixer control names. o On ALC662-GR, the primary ADC NID is 0x09, not 0x08, so correct the convgroupset in alc662_init_dacgroup. Audio capture now works on ALC662-GR (as found in the ASUS EeePC 701). Note that only ALC662-GR uses generic_mixer_autoinit and generic_mixer_init_widget; other codecs will need to be tested and converted to address PR#38578: azalia_codec.o is too large.
1.19	07-Jun-2008	freza	branches: 1.19.4; Split device_t/softc. Remove unused macros.
1.18	28-Apr-2008	martin	branches: 1.18.2; Remove clause 3 and 4 from TNF licenses
1.17	13-May-2007	kent	branches: 1.17.28; 1.17.30; 1.17.32; Add PCI_SUBCLASS_MULTIMEDIA_HDAUDIO to pcireg.h and azalia(4) uses it
1.16	08-May-2007	kent	fix a typo: PRESENSE -> PRESENCE From deanna@OpenBSD
1.15	05-Feb-2007	kent	branches: 1.15.2; 1.15.4; 1.15.6; 1.15.8; If both of AZALIA_DEBUG and AZALIA_DEBUG_DOT are defined, a DOT file for Graphviz representing the widget structure is printed.
1.14	21-Jul-2006	kent	branches: 1.14.4; 1.14.8; 1.14.10; support for EAPD, Balanced I/O, and LR-Swap
1.13	25-Jun-2006	kent	* disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.12	25-Jun-2006	kent	ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.11	25-Jun-2006	kent	handle unsolicited responses. A codec can receive unsolicited responses by codec_t::unsol_event()
1.10	23-Jun-2006	kent	add some constant macros for unsolicited response and jack sense
1.9	11-Jun-2006	kent	branches: 1.9.2; 1.9.4; support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.8	09-Jun-2006	kent	- print audio format only when the format is changed - print the name of STAC9220 codec
1.7	07-Jun-2006	kent	* move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.6	16-Jan-2006	kent	branches: 1.6.2; 1.6.4; 1.6.6; 1.6.12; Don't crash when a connection list has invalid NIDs. PR#32485 by Samuel Tran
1.5	03-Jan-2006	kent	- make this compile with AZALIA_DEBUG Thanks to Tomokatsu HAYAKAWA. - introduce WIDGET_CHANNELS macro
1.4	11-Dec-2005	christos	branches: 1.4.2; 1.4.4; merge ktrace-lwp.
1.3	29-Sep-2005	kent	branches: 1.3.6; * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.2	28-Sep-2005	kent	add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.1	28-Sep-2005	kent	split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.3.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.3.6.1	29-Sep-2005	skrll	file azalia.h was added on branch ktrace-lwp on 2005-11-10 14:06:00 +0000
1.4.4.2	01-Feb-2006	yamt	sync with head.
1.4.4.1	15-Jan-2006	yamt	sync with head.
1.4.2.15	27-May-2007	bouyer	Pull up following revision(s) (requested by kent in ticket #1780): sys/dev/pci/azalia_codec.c: revisions 1.25, 1.28, 1.30 - 1.42 sys/dev/pci/azalia.c: revisions 1.43, 1.46 - 1.49 sys/dev/pci/azalia.h: revisions 1.15 - 1.16 Bug fixes, and add support for some more codecs. PR#34880, PR#35658, PR#36302
1.4.2.14	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.23 sys/dev/pci/azalia.h: revision 1.14 support for EAPD, Balanced I/O, and LR-Swap
1.4.2.13	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.20 sys/dev/pci/azalia.h: revision 1.13 sys/dev/pci/azalia.c: revision 1.35 * disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.4.2.12	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.12 sys/dev/pci/azalia_codec.c: revision 1.19 sys/dev/pci/azalia.c: revision 1.34 ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.4.2.11	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.11 sys/dev/pci/azalia.c: revision 1.31 handle unsolicited responses. A codec can receive unsolicited responses by codec_t::unsol_event()
1.4.2.10	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.10 add some constant macros for unsolicited response and jack sense
1.4.2.9	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.25 sys/dev/pci/azalia.h: revision 1.9 sys/dev/pci/azalia_codec.c: revision 1.12 support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.4.2.8	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.23 sys/dev/pci/azalia.h: revision 1.8 sys/dev/pci/azalia_codec.c: revision 1.10 - print audio format only when the format is changed - print the name of STAC9220 codec
1.4.2.7	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.7 sys/dev/pci/azalia_codec.c: revision 1.9 sys/dev/pci/azalia.c: revision 1.21 * move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.4.2.6	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.h: revision 1.6 sys/dev/pci/azalia.c: revision 1.19 Don't crash when a connection list has invalid NIDs. PR#32485 by Samuel Tran
1.4.2.5	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia.h: revision 1.5 sys/dev/pci/azalia.c: revision 1.17 - make this compile with AZALIA_DEBUG Thanks to Tomokatsu HAYAKAWA. - introduce WIDGET_CHANNELS macro
1.4.2.4	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.3 sys/dev/pci/azalia_codec.c: revision 1.3 sys/dev/pci/azalia.c: revision 1.15 * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.4.2.3	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.2 sys/dev/pci/azalia_codec.c: revision 1.2 sys/dev/pci/azalia.c: revision 1.14 add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.4.2.2	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.1 sys/dev/pci/azalia_codec.c: revision 1.1 sys/dev/pci/azalia.c: revision 1.13 sys/dev/pci/files.pci: revision 1.239 split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.4.2.1	11-Dec-2005	riz	file azalia.h was added on branch netbsd-3 on 2005-12-29 19:33:40 +0000
1.6.12.1	19-Jun-2006	chap	Sync with head.
1.6.6.2	11-Aug-2006	yamt	sync with head
1.6.6.1	26-Jun-2006	yamt	sync with head.
1.6.4.1	07-Jun-2006	kardel	Sync with head.
1.6.2.1	09-Sep-2006	rpaulo	sync with head
1.9.4.5	03-Sep-2007	yamt	sync with head.
1.9.4.4	26-Feb-2007	yamt	sync with head.
1.9.4.3	30-Dec-2006	yamt	sync with head.
1.9.4.2	21-Jun-2006	yamt	sync with head.
1.9.4.1	11-Jun-2006	yamt	file azalia.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.9.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.14.10.1	03-Jun-2007	wrstuden	Catch up with now-somewhat-dated netbsd-4. These changes took longer than expected for me to actually get around to merging.
1.14.8.1	20-May-2007	jdc	Pull up revision 1.15-1.16 (requested by kent in ticket #649). If both of AZALIA_DEBUG and AZALIA_DEBUG_DOT are defined, a DOT file for Graphviz representing the widget structure is printed. fix a typo: PRESENSE -> PRESENCE >From deanna@OpenBSD
1.14.4.1	09-Feb-2007	ad	Sync with HEAD.
1.15.8.1	11-Jul-2007	mjf	Sync with head.
1.15.6.1	27-May-2007	ad	Sync with head.
1.15.4.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.15.2.1	17-May-2007	yamt	sync with head.
1.17.32.2	04-May-2009	yamt	sync with head.
1.17.32.1	16-May-2008	yamt	sync with head.
1.17.30.2	17-Jun-2008	yamt	sync with head.
1.17.30.1	18-May-2008	yamt	sync with head.
1.17.28.3	28-Sep-2008	mjf	Sync with HEAD.
1.17.28.2	29-Jun-2008	mjf	Sync with HEAD.
1.17.28.1	02-Jun-2008	mjf	Sync with HEAD.
1.18.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.18.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.19.4.1	19-Oct-2008	haad	Sync with HEAD.
1.20.30.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.20.28.1	17-Apr-2012	yamt	sync with head
1.20.6.2	12-Dec-2008	ad	Checkpoint work in progress.
1.20.6.1	11-Dec-2008	ad	Checkpoint work in progress.
1.21.54.1	28-Apr-2019	isaki	Adapt to audio2.
1.21.50.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.21.50.1	10-Jun-2019	christos	Sync with HEAD
1.22.4.1	29-Feb-2020	ad	Sync with head.
1.82	09-Feb-2020	jmcneill	Retire azalia(4).
1.81	28-Jul-2017	nat	branches: 1.81.4; 1.81.12; Mixer device bounds checking. Analysis by Ilja van Sprundel.
1.80	01-Jun-2017	chs	branches: 1.80.2; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.79	23-Nov-2011	jmcneill	branches: 1.79.8; 1.79.24; 1.79.26; 1.79.28; 1.79.36; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.78	03-May-2010	dholland	branches: 1.78.8; 1.78.10; Fix a crash during alc888 initialization caused by holding a pointer across realloc. Patch from Bj?rn Barkenfelt in PR kern/41957, rearranged somewhat by myself.
1.77	07-Apr-2009	stacktic	branches: 1.77.2; 1.77.4; Add Realtek ALC663 and treat like ALC662 (ok pooka@)
1.76	28-Feb-2009	jmcneill	Add proper Realtek ALC268 support. While we're here, fix a comment in the ALC262 quirk section. Tested by Sean Davis on an Acer Aspire One.
1.75	23-Feb-2009	jmcneill	* Realtek ALC885: Use generic_mixer_autoinit/init_widget * generic_mixer_create_virtual: if the front l/r DAC doesn't have a volume capability, enumerate all peer widgets and search for a volume mixer there.
1.74	27-Jan-2009	markd	branches: 1.74.2; Add Analog Devices AD1984A and treat like AD1984.
1.73	18-Jan-2009	cegger	Use aprint_error_dev. Eliminates XNAME macro.
1.72	12-Nov-2008	jmcneill	branches: 1.72.2; Add support for Realtek ALC269 as found in Eee PC 1000HA.
1.71	24-Aug-2008	kent	branches: 1.71.2; 1.71.4; * generic_mixer_create_virtual() Don't track items by pointers because the buffer can be moved by realloc(). PR kern/39382, reported and patch provided by Takahiro Hayashi
1.70	15-Aug-2008	jmcneill	AD1984: Take advantage of pin auto-config and auto-naming.
1.69	15-Aug-2008	jmcneill	Skip unconnected pins when creating a selector connlist.
1.68	14-Aug-2008	jmcneill	PR#38860: azalia record does not work on Realtek ALC662-GR o Don't add unconnected pins to generated mixer list, should help make auto-generated mixer lists slightly easier to work with. o Move COP_AWTYPE_AUDIO_MIXER type knobs to mix.* rather than polluting the inputs.* namespace and giving a few more characters for the auto-generated names. o Introduce an alternate to generic_mixer_init, generic_mixer_autoinit. generic_mixer_autoinit uses COP_AWTYPE_PIN_COMPLEX "default configuration" capability flags to determine proper pin widget control and eapd configuration, as well as creating virtual outputs.master, inputs.dac, and record.volume mixer controls that mirror the default dac and adc controls. o Introduce generic_mixer_init_widget which uses COP_AWTYPE_PIN_COMPLEX "default configuration" device type, gross location, and geometric locations to guess friendlier mixer control names. o On ALC662-GR, the primary ADC NID is 0x09, not 0x08, so correct the convgroupset in alc662_init_dacgroup. Audio capture now works on ALC662-GR (as found in the ASUS EeePC 701). Note that only ALC662-GR uses generic_mixer_autoinit and generic_mixer_init_widget; other codecs will need to be tested and converted to address PR#38578: azalia_codec.o is too large.
1.67	06-Jul-2008	abs	branches: 1.67.2; Commit patch from jmcneill@ to fix "No sound on Thinkpad azalia" Definitely fixes a T60p, and hopefully will fix PR#37141 for an R61 also
1.66	24-Jun-2008	gmcgarry	pcc doesn't like empty structures for initialisation. Fill with zeros.
1.65	07-Jun-2008	freza	branches: 1.65.2; Split device_t/softc. Remove unused macros.
1.64	28-Apr-2008	martin	branches: 1.64.2; Remove clause 3 and 4 from TNF licenses
1.63	10-Apr-2008	cegger	branches: 1.63.2; 1.63.4; use aprint_*_dev and device_xname
1.62	31-Mar-2008	jmcneill	Add outputs.master.mute and inputs.dac.mute knobs for ALC662. XXX these currently only control speaker mute, not headphone. Needs to be revisited.
1.61	31-Mar-2008	jmcneill	Initial support for the Realtek ALC662-GR as found in the ASUS Eee PC.
1.60	22-Feb-2008	oster	Make outputs.master and inputs.dac show up in mixerctl output for the ALC888. Patch from, and much thanks to, jmcneill @.
1.59	31-Jan-2008	markd	branches: 1.59.2; 1.59.6; s/AD1983_EVENT/AD198X_EVENT/
1.58	31-Jan-2008	markd	Use more meaningful names for AD1984 widgets. Select dac as default input for internal mono speaker. Handle jack pin detect for headphones and lineout on Dell Optiplex 755.
1.57	30-Jan-2008	jmcneill	AD1984: Create a dummy outputs.master mixer knob that mirrors inputs.dac, since AD1984 doesn't have a traditional master volume control.
1.56	30-Jan-2008	jmcneill	AD1984: Use the same DAC for speaker and headphones, so the headphone jack works as expected on Thinkpad T61 (and probably others).
1.55	06-Jan-2008	kent	recognized the names of Sigmatel STAC9227-9230
1.54	04-Jan-2008	kent	shorten mixer-input names by mixer-name automatic generation PR/37273
1.53	04-Jan-2008	kent	simplify conversion rule between device volume values and userland volume values. Some mixer values never be 255 by this change. PR#37274
1.52	03-Jan-2008	kent	code cleanup - use __arraycount - remove unnecessary M_ZERO flags
1.51	03-Jan-2008	kent	- add initialization for ATI HDMI audio - add custom mixer table for AD1986A by scw@ PR: kern/36140
1.50	09-Dec-2007	jmcneill	branches: 1.50.2; Merge jmcneill-pm branch.
1.49	26-Nov-2007	christos	branches: 1.49.2; 1.49.4; Add dell latitude 420 in the list of machines that need special treatment to silence speakers when headphones are plugged in. From Marcin Michal Jessa
1.48	07-Nov-2007	joerg	When converting mixer values to hardware values, round up. This makes stepping and restore actually work.
1.47	03-Nov-2007	markd	branches: 1.47.2; Add support for Analog Devices AD1983 codec as found in Dell Optiplex 745.
1.46	30-Oct-2007	joerg	Add another Dell machine where the head phone event needs to be handled.
1.45	22-Sep-2007	xtraeme	branches: 1.45.4; alc883_mixer_init: fix this->nmixers, __arraycount() wasn't correct.
1.44	22-Sep-2007	xtraeme	Apply revision 1.22 from OpenBSD, quoting deanna's commit: "Add support for Realtek ALC883. This is the "value" edition of the ALC882, and differs only by lacking an ADC and a mixer, so have it use ALC882's get_port(), set_port() and a slightly modified version of its mixer init."
1.43	13-Sep-2007	jmcneill	Add basic support for the Analog Devices AD1984 codec as found in the ThinkPad T61.
1.42	13-May-2007	kent	branches: 1.42.6; 1.42.8; * stac9221_gpio_unmute() The target node should be the audio function group, not 0. Pointed out by Deanna Phillips * stac9221_mixer_init() Enable GPIO unmute for Intel Macintosh
1.41	10-May-2007	sketch	Match recent addition as a STAC9200.
1.40	10-May-2007	sketch	Update previous to match what the author probably meant to do.
1.39	10-May-2007	kent	Support for headphone-sensing for DELL Inspiron 6400. PR#36302 by Aleksej Saushev
1.38	08-May-2007	kent	fix a typo: PRESENSE -> PRESENCE From deanna@OpenBSD
1.37	08-May-2007	kent	- fix incorrect codec name of STAC9200 - support for headphone-sensing in some DELL machines with STAC9200 From azalia_codec.c rev.1.19 by deanna@OpenBSD
1.36	19-Mar-2007	kent	readable mixer item names for AD1981HD
1.35	18-Mar-2007	kent	don't make a mixer item for a power widget
1.34	18-Mar-2007	kent	correct the DAC group definition for AD1988. reported by wiz
1.33	11-Mar-2007	kent	branches: 1.33.2; 1.33.4; * codec names for Realtek ALC262, ALC268, ALC861, ALC861-VD-GR, ALC883, ALC885, and ALC888 * DAC/ADC group definitions for Realtek ALC861, ALC861-VD-GR, ALC883, ALC885, and ALC888
1.32	11-Mar-2007	kent	remove DPRINTF() in generic_mixer_init() because they are useless in many cases.
1.31	11-Mar-2007	kent	- Add jack-sense code for STAC9220 in some DELL models (currently disabled) - Add DAC/ADC grouping for AD1988
1.30	11-Mar-2007	kent	reviced Intel Mac support code (currently disabled) PR#34880
1.29	21-Feb-2007	thorpej	branches: 1.29.4; Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.28	06-Jan-2007	kent	branches: 1.28.2; * generic_mixer_default() make the default values of bidirectional pins `output' because they were often the main reason of no sound problems. * add stac9221_mixer_init() It has a code for Intel Mac but it is disabled for now.
1.27	16-Nov-2006	christos	branches: 1.27.2; 1.27.4; __unused removal on arguments; approved by core.
1.26	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.25	03-Sep-2006	christos	branches: 1.25.2; 1.25.4; fix initializers
1.24	23-Jul-2006	kent	For ThinkPad T60, make the default value of outputs.linein.dir "output"
1.23	21-Jul-2006	kent	support for EAPD, Balanced I/O, and LR-Swap
1.22	19-Jul-2006	kent	fix crash by reading a mixer value with AD1981HD codec, which has invalid connection lists.
1.21	26-Jun-2006	kent	* custom mixer table for ALC880 * remove useless mixers of ALC882
1.20	25-Jun-2006	kent	* disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.19	25-Jun-2006	kent	ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.18	24-Jun-2006	kent	custom mixer table for CMI9880 PR#33787
1.17	24-Jun-2006	kent	DAC&ADC group for CMI9880 PR#33787
1.16	15-Jun-2006	kent	branches: 1.16.2; Fix a bug that AUDIO_MIXER_READ returns an incorrect mixer_ctrl_t::un.value.num_channels in the case of the combination of a mono pin and a stereo mixer.
1.15	13-Jun-2006	kent	branches: 1.15.2; * generic_mixer_default() set the default pin direction depending on the pin's device, not color
1.14	13-Jun-2006	kent	ALC882: - add usingdac and usingadc - remove alc882_init_widget() because it has no effect any more
1.13	11-Jun-2006	kent	codec-specific mixer initialization for Realtek ALC882
1.12	11-Jun-2006	kent	support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.11	11-Jun-2006	kent	* azalia_codec_init_vtbl() recognize the name of AD1983 * generic_codec_find_dac() check validity of a node in a connection list * introduce MIXER_DELTA, and simplify the code with it * codec-specific mixer initialization for STAC9220
1.10	09-Jun-2006	kent	- print audio format only when the format is changed - print the name of STAC9220 codec
1.9	07-Jun-2006	kent	* move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.8	10-May-2006	kent	branches: 1.8.2; fix incorrect ADC list for ALC880. PR kern/33326 by Uwe Klaus
1.7	16-Jan-2006	kent	branches: 1.7.2; 1.7.4; 1.7.6; 1.7.8; 1.7.10; readable names for AD1981HD codec
1.6	03-Jan-2006	kent	support for 8ch playback with STAC9221D codec. Thanks to Tomokatsu HAYAKAWA.
1.5	03-Jan-2006	kent	ALC260: improve readability of mixer item names.
1.4	11-Dec-2005	christos	branches: 1.4.2; 1.4.4; merge ktrace-lwp.
1.3	29-Sep-2005	kent	branches: 1.3.6; * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.2	28-Sep-2005	kent	add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.1	28-Sep-2005	kent	split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.3.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.3.6.1	29-Sep-2005	skrll	file azalia_codec.c was added on branch ktrace-lwp on 2005-11-10 14:06:00 +0000
1.4.4.2	01-Feb-2006	yamt	sync with head.
1.4.4.1	15-Jan-2006	yamt	sync with head.
1.4.2.26	27-May-2007	bouyer	Pull up following revision(s) (requested by kent in ticket #1780): sys/dev/pci/azalia_codec.c: revisions 1.25, 1.28, 1.30 - 1.42 sys/dev/pci/azalia.c: revisions 1.43, 1.46 - 1.49 sys/dev/pci/azalia.h: revisions 1.15 - 1.16 Bug fixes, and add support for some more codecs. PR#34880, PR#35658, PR#36302
1.4.2.25	31-Jul-2006	tron	Apply patch (requested by kent in ticket #1432): Fix build problem.
1.4.2.24	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.24 For ThinkPad T60, make the default value of outputs.linein.dir "output"
1.4.2.23	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.23 sys/dev/pci/azalia.h: revision 1.14 support for EAPD, Balanced I/O, and LR-Swap
1.4.2.22	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.22 fix crash by reading a mixer value with AD1981HD codec, which has invalid connection lists.
1.4.2.21	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.21 * custom mixer table for ALC880 * remove useless mixers of ALC882
1.4.2.20	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.20 sys/dev/pci/azalia.h: revision 1.13 sys/dev/pci/azalia.c: revision 1.35 * disconnect a stream and audio converters on a halt of the stream * enable S/PDIF for a digital audio converter when connecting a stream * rename "inputs.usingdac" to "playback.mode", and "record.usingadc" to "record.mode" * add S/PDIF output configuration to ALC882
1.4.2.19	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.12 sys/dev/pci/azalia_codec.c: revision 1.19 sys/dev/pci/azalia.c: revision 1.34 ALC260 & Fujitsu LOOX: mute the internal speaker when the headphone is inserted, and unmute the internal speaker when the headphone is pulled out.
1.4.2.18	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.18 custom mixer table for CMI9880 PR#33787
1.4.2.17	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.17 DAC&ADC group for CMI9880 PR#33787
1.4.2.16	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.16 Fix a bug that AUDIO_MIXER_READ returns an incorrect mixer_ctrl_t::un.value.num_channels in the case of the combination of a mono pin and a stereo mixer.
1.4.2.15	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.15 * generic_mixer_default() set the default pin direction depending on the pin's device, not color
1.4.2.14	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.14 ALC882: - add usingdac and usingadc - remove alc882_init_widget() because it has no effect any more
1.4.2.13	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.13 codec-specific mixer initialization for Realtek ALC882
1.4.2.12	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.25 sys/dev/pci/azalia.h: revision 1.9 sys/dev/pci/azalia_codec.c: revision 1.12 support for multi-channel (>=3) recording with ALC880, ALC882, or STAC9221
1.4.2.11	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia_codec.c: revision 1.11 * azalia_codec_init_vtbl() recognize the name of AD1983 * generic_codec_find_dac() check validity of a node in a connection list * introduce MIXER_DELTA, and simplify the code with it * codec-specific mixer initialization for STAC9220
1.4.2.10	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.c: revision 1.23 sys/dev/pci/azalia.h: revision 1.8 sys/dev/pci/azalia_codec.c: revision 1.10 - print audio format only when the format is changed - print the name of STAC9220 codec
1.4.2.9	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/azalia.h: revision 1.7 sys/dev/pci/azalia_codec.c: revision 1.9 sys/dev/pci/azalia.c: revision 1.21 * move mixer functions from azalia.c to azalia_codec.c, and rename them like azalia_mixer_foo() to generic_mixer_foo() * enable codec-specific code to hook any mixer operations such as creating mixer items, or special handling for a specific item. * provide a custom mixer table specific to ALC260 * provide a custom mixer table specific to Fujitsu LOOX
1.4.2.8	13-May-2006	tron	Pull up following revision(s) (requested by kent in ticket #1318): sys/dev/pci/azalia_codec.c: revision 1.8 fix incorrect ADC list for ALC880. PR kern/33326 by Uwe Klaus
1.4.2.7	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia_codec.c: revision 1.7 readable names for AD1981HD codec
1.4.2.6	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia_codec.c: revision 1.6 support for 8ch playback with STAC9221D codec. Thanks to Tomokatsu HAYAKAWA.
1.4.2.5	27-Jan-2006	tron	Pull up following revision(s) (requested by kent in ticket #1124): sys/dev/pci/azalia_codec.c: revision 1.5 ALC260: improve readability of mixer item names.
1.4.2.4	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.3 sys/dev/pci/azalia_codec.c: revision 1.3 sys/dev/pci/azalia.c: revision 1.15 * shorten _init_dacgroup() functions add a hook point for widget initialization and implement for ALC882
1.4.2.3	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.2 sys/dev/pci/azalia_codec.c: revision 1.2 sys/dev/pci/azalia.c: revision 1.14 add codec-specific code for ALC260, ALC880, ALC882, and STAC9221 to detect DAC/ADC cominations.
1.4.2.2	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.1 sys/dev/pci/azalia_codec.c: revision 1.1 sys/dev/pci/azalia.c: revision 1.13 sys/dev/pci/files.pci: revision 1.239 split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.4.2.1	11-Dec-2005	riz	file azalia_codec.c was added on branch netbsd-3 on 2005-12-29 19:33:40 +0000
1.7.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.7.8.1	11-May-2006	elad	sync with head
1.7.6.4	14-Sep-2006	yamt	sync with head.
1.7.6.3	11-Aug-2006	yamt	sync with head
1.7.6.2	26-Jun-2006	yamt	sync with head.
1.7.6.1	24-May-2006	yamt	sync with head.
1.7.4.2	07-Jun-2006	kardel	Sync with head.
1.7.4.1	01-Jun-2006	kardel	Sync with head.
1.7.2.1	09-Sep-2006	rpaulo	sync with head
1.8.2.1	19-Jun-2006	chap	Sync with head.
1.15.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.16.2.11	27-Feb-2008	yamt	sync with head.
1.16.2.10	04-Feb-2008	yamt	sync with head.
1.16.2.9	21-Jan-2008	yamt	sync with head
1.16.2.8	07-Dec-2007	yamt	sync with head
1.16.2.7	15-Nov-2007	yamt	sync with head.
1.16.2.6	27-Oct-2007	yamt	sync with head.
1.16.2.5	03-Sep-2007	yamt	sync with head.
1.16.2.4	26-Feb-2007	yamt	sync with head.
1.16.2.3	30-Dec-2006	yamt	sync with head.
1.16.2.2	21-Jun-2006	yamt	sync with head.
1.16.2.1	15-Jun-2006	yamt	file azalia_codec.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.25.4.2	10-Dec-2006	yamt	sync with head.
1.25.4.1	22-Oct-2006	yamt	sync with head
1.25.2.2	12-Jan-2007	ad	Sync with head.
1.25.2.1	18-Nov-2006	ad	Sync with head.
1.27.4.3	04-Sep-2008	skrll	Sync with netbsd-4.
1.27.4.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.27.4.1	03-Jun-2007	wrstuden	Catch up with now-somewhat-dated netbsd-4. These changes took longer than expected for me to actually get around to merging.
1.27.2.3	06-Jul-2008	bouyer	Pull up following revision(s) (requested by abs in ticket #1162): sys/dev/pci/azalia_codec.c: revision 1.67 Commit patch from jmcneill@ to fix "No sound on Thinkpad azalia" Definitely fixes a T60p, and hopefully will fix PR#37141 for an R61 also
1.27.2.2	01-Feb-2008	riz	Pull up following revision(s) (requested by joerg in ticket #982): sys/dev/pci/azalia_codec.c: revision 1.48 When converting mixer values to hardware values, round up. This makes stepping and restore actually work.
1.27.2.1	20-May-2007	jdc	Pull up revisions 1.28, 1.30-1.42 (requested by kent in ticket #649). * generic_mixer_default() make the default values of bidirectional pins `output' because they were often the main reason of no sound problems. * add stac9221_mixer_init() It has a code for Intel Mac but it is disabled for now. reviced Intel Mac support code (currently disabled) PR#34880 - Add jack-sense code for STAC9220 in some DELL models (currently disabled) - Add DAC/ADC grouping for AD1988 remove DPRINTF() in generic_mixer_init() because they are useless in many cases. * codec names for Realtek ALC262, ALC268, ALC861, ALC861-VD-GR, ALC883, ALC885, and ALC888 * DAC/ADC group definitions for Realtek ALC861, ALC861-VD-GR, ALC883, ALC885, and ALC888 correct the DAC group definition for AD1988. reported by wiz don't make a mixer item for a power widget readable mixer item names for AD1981HD - fix incorrect codec name of STAC9200 - support for headphone-sensing in some DELL machines with STAC9200 >From azalia_codec.c rev.1.19 by deanna@OpenBSD fix a typo: PRESENSE -> PRESENCE >From deanna@OpenBSD Support for headphone-sensing for DELL Inspiron 6400. PR#36302 by Aleksej Saushev Update previous to match what the author probably meant to do. Match recent addition as a STAC9200. * stac9221_gpio_unmute() The target node should be the audio function group, not 0. Pointed out by Deanna Phillips * stac9221_mixer_init() Enable GPIO unmute for Intel Macintosh
1.28.2.4	17-May-2007	yamt	sync with head.
1.28.2.3	24-Mar-2007	yamt	sync with head.
1.28.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.28.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.29.4.4	09-Oct-2007	ad	Sync with head.
1.29.4.3	27-May-2007	ad	Sync with head.
1.29.4.2	10-Apr-2007	ad	Sync with head.
1.29.4.1	13-Mar-2007	ad	Sync with head.
1.33.4.1	29-Mar-2007	reinoud	Pullup to -current
1.33.2.1	11-Jul-2007	mjf	Sync with head.
1.42.8.4	23-Mar-2008	matt	sync with HEAD
1.42.8.3	09-Jan-2008	matt	sync with HEAD
1.42.8.2	08-Nov-2007	matt	sync with -HEAD
1.42.8.1	06-Nov-2007	matt	sync with HEAD
1.42.6.7	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.42.6.6	11-Nov-2007	joerg	Sync with HEAD.
1.42.6.5	04-Nov-2007	jmcneill	Sync with HEAD.
1.42.6.4	31-Oct-2007	joerg	Sync with HEAD.
1.42.6.3	03-Oct-2007	jmcneill	Enable pnp messaging for the ALC262 codec
1.42.6.2	02-Oct-2007	joerg	Sync with HEAD.
1.42.6.1	13-Sep-2007	jmcneill	Sync with HEAD.
1.45.4.1	13-Nov-2007	bouyer	Sync with HEAD
1.47.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.47.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.47.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.47.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.49.4.1	11-Dec-2007	yamt	sync with head.
1.49.2.1	26-Dec-2007	ad	Sync with head.
1.50.2.1	08-Jan-2008	bouyer	Sync with HEAD
1.59.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.59.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.59.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.59.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.59.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.59.2.1	24-Mar-2008	keiichi	sync with head.
1.63.4.3	11-Aug-2010	yamt	sync with head.
1.63.4.2	04-May-2009	yamt	sync with head.
1.63.4.1	16-May-2008	yamt	sync with head.
1.63.2.2	17-Jun-2008	yamt	sync with head.
1.63.2.1	18-May-2008	yamt	sync with head.
1.64.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.64.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.65.2.2	18-Jul-2008	simonb	Sync with head.
1.65.2.1	27-Jun-2008	simonb	Sync with head.
1.67.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.67.2.1	19-Oct-2008	haad	Sync with HEAD.
1.71.4.3	02-Mar-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #533): sys/dev/pci/azalia_codec.c: revision 1.76 Add proper Realtek ALC268 support. While we're here, fix a comment in the ALC262 quirk section. Tested by Sean Davis on an Acer Aspire One.
1.71.4.2	24-Feb-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #491): sys/dev/pci/azalia_codec.c: revision 1.75 * Realtek ALC885: Use generic_mixer_autoinit/init_widget * generic_mixer_create_virtual: if the front l/r DAC doesn't have a volume capability, enumerate all peer widgets and search for a volume mixer there.
1.71.4.1	14-Nov-2008	snj	Pull up following revision(s) (requested by jmcneill in ticket #54): sys/dev/pci/azalia_codec.c: revision 1.72 Add support for Realtek ALC269 as found in Eee PC 1000HA.
1.71.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.71.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.71.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.72.2.2	13-Dec-2008	ad	Make it compile.
1.72.2.1	12-Dec-2008	ad	Checkpoint work in progress.
1.74.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.77.4.1	30-May-2010	rmind	sync with head
1.77.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.78.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.78.8.1	17-Apr-2012	yamt	sync with head
1.79.36.1	09-Aug-2017	snj	Pull up following revision(s) (requested by nat in ticket #1460): sys/dev/auconv.c: revision 1.30 via patch sys/dev/ic/ac97.c: revision 1.97 via patch sys/dev/pci/azalia_codec.c: revision 1.81 via patch Mixer device bounds checking. Analysis by Ilja van Sprundel.
1.79.28.1	09-Aug-2017	snj	Pull up following revision(s) (requested by nat in ticket #1460): sys/dev/auconv.c: revision 1.30 via patch sys/dev/ic/ac97.c: revision 1.97 via patch sys/dev/pci/azalia_codec.c: revision 1.81 via patch Mixer device bounds checking. Analysis by Ilja van Sprundel.
1.79.26.1	28-Aug-2017	skrll	Sync with HEAD
1.79.24.1	09-Aug-2017	snj	Pull up following revision(s) (requested by nat in ticket #1460): sys/dev/auconv.c: revision 1.30 via patch sys/dev/ic/ac97.c: revision 1.97 via patch sys/dev/pci/azalia_codec.c: revision 1.81 via patch Mixer device bounds checking. Analysis by Ilja van Sprundel.
1.79.8.1	03-Dec-2017	jdolecek	update from HEAD
1.80.2.1	01-Aug-2017	snj	Pull up following revision(s) (requested by nat in ticket #166): sys/dev/auconv.c: revision 1.30 sys/dev/audio.c: revision 1.372 sys/dev/ic/ac97.c: revision 1.97 sys/dev/pci/azalia_codec.c: revision 1.81 Mixer device bounds checking. Analysis by Ilja van Sprundel.
1.81.12.1	29-Feb-2020	ad	Sync with head.
1.81.4.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.41	13-Oct-2016	jdolecek	provide intr xname
1.40	18-Oct-2014	snj	branches: 1.40.2; 1.40.4; src is too big these days to tolerate superfluous apostrophes. It's "its", people!
1.39	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.38	27-Oct-2012	chs	branches: 1.38.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.37	26-Nov-2009	njoly	branches: 1.37.12; 1.37.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.36	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.35	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.34	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.33	28-Apr-2008	martin	branches: 1.33.14; Remove clause 3 and 4 from TNF licenses
1.32	10-Apr-2008	cegger	branches: 1.32.2; 1.32.4; use aprint_*_dev and device_xname
1.31	19-Oct-2007	ad	branches: 1.31.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.30	16-Nov-2006	christos	branches: 1.30.8; 1.30.22; 1.30.24; 1.30.28; __unused removal on arguments; approved by core.
1.29	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.28	11-Dec-2005	christos	branches: 1.28.20; 1.28.22; merge ktrace-lwp.
1.27	28-Jun-2005	thorpej	branches: 1.27.2; Use ANSI function decls and static.
1.26	04-Feb-2005	perry	de-__P
1.25	31-Jan-2003	thorpej	branches: 1.25.2; 1.25.10; 1.25.12; Use aprint_*().
1.24	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.23	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.22	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.21	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.20	13-Nov-2001	lukem	add RCSID
1.19	03-May-2001	ross	branches: 1.19.2; Split pci and eisa/isa attachment calls, plus misc cleanup.
1.18	25-Apr-2001	bouyer	Pull up the thorpej_scsipi branch to main branch. This is a completely rewritten scsipi_xfer execution engine, and the associated changes to HBA drivers. Overview of changes & features: - All xfers are queued in the mid-layer, rather than doing so in an ad-hoc fashion in individual adapter drivers. - Adapter/channel resource management in the mid-layer, avoids even trying to start running an xfer if the adapter/channel doesn't have the resources. - Better communication between the mid-layer and the adapters. - Asynchronous event notification mechanism from adapter to mid-layer and peripherals. - Better peripheral queue management: freeze/thaw, sorted requeueing during recovery, etc. - Clean separation of peripherals, adapters, and adapter channels (no more scsipi_link). - Kernel thread for each scsipi_channel makes error recovery much easier (no more dealing with interrupt context when recovering from an error). - Mid-layer support for tagged queueing: commands can have the tag type set explicitly, tag IDs are allocated in the mid-layer (thus eliminating the need to use buggy tag ID allocation schemes in many adapter drivers). - support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command will be requeued, or a REQUEST SENSE will be sent as appropriate. Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
1.17	28-Dec-2000	sommerfeld	branches: 1.17.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.16	15-Aug-1998	mycroft	branches: 1.16.12; Assign my copyrights to TNF.
1.15	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.14	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.13	07-Jun-1997	thorpej	branches: 1.13.2; 1.13.4; Pull the thorpej-bus-dma branch into the mainline.
1.12	13-Apr-1997	cgd	branches: 1.12.2; rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.11	13-Apr-1997	cgd	use pci_map_register().
1.10	28-Mar-1997	mycroft	Use a temporary data structure for holding probe information, rather than a full softc.
1.9	13-Mar-1997	cgd	clean up the __BROKEN_INDIRECT_CONFIG chunks of the last change
1.8	13-Mar-1997	cgd	fixes from Matt Jacob so that these can compile and run on the Alpha. (aic7xxx has been tested and works on the Alpha, bha has not yet been tested on thne Alpha.)
1.7	17-Feb-1997	jonathan	Change bha_pci product names to match latest pcidevs: PCI_PRODUCT_BUSLOGIC_OLD946C -> PCI_PRODUCT_BUSLOGIC_MULTIMASTER_NC PCI_PRODUCT_BUSLOGIC_946C -> PCI_PRODUCT_BUSLOGIC_MULTIMASTER
1.6	05-Nov-1996	jonathan	branches: 1.6.4; * Add command struct and modifier definitions for controlling the ISA-compatible port space of PCI buslogic cards. * Add call to bha_pci.c to disable the ISA-compatible ports of a PCI device. The ISA-compatible ports are enabled by default, which causes the card to be autoconfigured a second time as an ISA device, which appears to deadlock the card. * Change bha_cmd() to return the number of bytes it actually received in response to a command, or -1 on error. * Use heuristics (checking for bha-only registers, and checking the size of the response to BHA_INQURE_EXTENDED) to bha_find, to make sure the bha driver never matches an aha (Adaptec 1542 or compatible) device. A single kernel should now boot on either Adaptec or BusLogic controllers, provided we always probe for BusLogic devices before Adaptec devices, but this has not yet been verified.
1.5	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.4	13-Oct-1996	christos	backout kprintf changes
1.3	10-Oct-1996	christos	- printf -> kprintf, sprintf -> ksprintf - include systm.h
1.2	01-Sep-1996	mycroft	Minor changes.
1.1	31-Aug-1996	mycroft	Split the BusLogic driver apart, and rename it to `bha (like BSDi).
1.6.4.1	12-Mar-1997	is	Merge in changes from The Trunk
1.12.2.1	13-May-1997	thorpej	Stash the bus dma tag and bus-specific DMA flags in the softc.
1.13.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.13.2.1	01-Jul-1997	bouyer	Updates for new scsipi subsystem. Actally known to work on i386 and sparc.
1.16.12.2	27-Mar-2001	bouyer	Convert these drivers to thorpej_scsipi (untested).
1.16.12.1	05-Jan-2001	bouyer	Sync with HEAD
1.17.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.17.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.17.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.19.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.19.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.25.12.1	12-Feb-2005	yamt	sync with head.
1.25.10.1	29-Apr-2005	kent	sync with -current
1.25.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.25.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.27.2.2	27-Oct-2007	yamt	sync with head.
1.27.2.1	30-Dec-2006	yamt	sync with head.
1.28.22.2	10-Dec-2006	yamt	sync with head.
1.28.22.1	22-Oct-2006	yamt	sync with head
1.28.20.1	18-Nov-2006	ad	Sync with head.
1.30.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.30.24.1	06-Nov-2007	matt	sync with HEAD
1.30.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.30.8.1	23-Oct-2007	ad	Sync with head.
1.31.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.32.4.3	11-Mar-2010	yamt	sync with head
1.32.4.2	16-May-2009	yamt	sync with head
1.32.4.1	16-May-2008	yamt	sync with head.
1.32.2.1	18-May-2008	yamt	sync with head.
1.33.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.37.22.3	03-Dec-2017	jdolecek	update from HEAD
1.37.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.37.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.37.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.12.1	30-Oct-2012	yamt	sync with head
1.38.2.1	18-May-2014	rmind	sync with head
1.40.4.1	04-Nov-2016	pgoyette	Sync with HEAD
1.40.2.1	05-Dec-2016	skrll	Sync with HEAD
1.37	05-Dec-2023	thorpej	b3_617_map_vme(): use VM_NOSLEEP. EX_NOWAIT is 0, to lack of EX_WAITOK is the same as EX_NOWAIT.
1.36	05-Dec-2023	thorpej	Use vmem(9) rather than extent(9) to manage VME<->PCI address translations. While here, malloc(9) -> kmem(9), and fix a couple of things that the newer compiler gripes about.
1.35	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.34	24-Apr-2021	thorpej	branches: 1.34.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.33	24-Aug-2020	msaitoh	branches: 1.33.4; s/ressource/resource/. No functional change.
1.32	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.31	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.30	29-Mar-2014	christos	branches: 1.30.28; 1.30.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.29	27-Oct-2012	chs	branches: 1.29.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.28	30-Jan-2012	drochner	branches: 1.28.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.27	27-Jan-2012	para	converting extent(9) from malloc(9) to kmem(9) preceding kmem-vmem-pool-uvm patch releng@ acknowledged
1.26	26-Nov-2009	njoly	branches: 1.26.12; 1.26.16; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.25	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.24	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.23	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.22	16-Mar-2009	dsl	ANSIfy functions with function-pointer arguments
1.21	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.20	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.19	16-Dec-2008	christos	branches: 1.19.2; replace bitmask_snprintf(9) with snprintb(3)
1.18	10-Apr-2008	cegger	branches: 1.18.4; 1.18.12; use aprint_*_dev and device_xname
1.17	19-Oct-2007	ad	branches: 1.17.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	10-Jan-2007	cube	branches: 1.16.6; 1.16.18; 1.16.20; 1.16.24; Make it compile.
1.15	24-Nov-2006	wiz	s/independant/independent/, from Zafer.
1.14	11-Dec-2005	christos	branches: 1.14.20; 1.14.22; merge ktrace-lwp.
1.13	04-Feb-2005	perry	branches: 1.13.6; de-__P
1.12	31-Jan-2003	thorpej	branches: 1.12.2; 1.12.10; 1.12.12; Use aprint_*().
1.11	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.10	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.9	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.8	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.7	04-Mar-2002	simonb	Don't "extern int cold;" - this is in <sys/kernel.h>.
1.6	13-Nov-2001	lukem	add RCSID
1.5	12-Jun-2001	wiz	branches: 1.5.2; receive, not recieve
1.4	28-Dec-2000	sommerfeld	branches: 1.4.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.3	08-Aug-2000	tv	sizeof(fmtstring) + 8 wasn't enough for the string; that won't cope with the printed version. Make a little more room for the bitmask_snprintf() call.
1.2	08-Aug-2000	tv	%b -> bitmask_snprintf()
1.1	30-Jun-1999	drochner	branches: 1.1.2; 1.1.4; add driver for the Bit3 PCI-VME adapter, without the DMA parts for now
1.1.4.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.4.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	30-Jun-1999	thorpej	file btvmei.c was added on branch chs-ubc2 on 1999-07-01 23:34:02 +0000
1.4.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.4.2.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.4.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.5.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.5.2.2	16-Mar-2002	jdolecek	Catch up with -current.
1.5.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.12.12.1	12-Feb-2005	yamt	sync with head.
1.12.10.1	29-Apr-2005	kent	sync with -current
1.12.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.13.6.3	27-Oct-2007	yamt	sync with head.
1.13.6.2	26-Feb-2007	yamt	sync with head.
1.13.6.1	30-Dec-2006	yamt	sync with head.
1.14.22.1	10-Dec-2006	yamt	sync with head.
1.14.20.1	12-Jan-2007	ad	Sync with head.
1.16.24.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.20.1	06-Nov-2007	matt	sync with HEAD
1.16.18.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.6.1	23-Oct-2007	ad	Sync with head.
1.17.16.2	17-Jan-2009	mjf	Sync with HEAD.
1.17.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.18.12.2	28-Apr-2009	skrll	Sync with HEAD.
1.18.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.18.4.3	11-Mar-2010	yamt	sync with head
1.18.4.2	16-May-2009	yamt	sync with head
1.18.4.1	04-May-2009	yamt	sync with head.
1.19.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.26.16.1	18-Feb-2012	mrg	merge to -current.
1.26.12.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.26.12.2	30-Oct-2012	yamt	sync with head
1.26.12.1	17-Apr-2012	yamt	sync with head
1.28.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.28.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.29.2.1	18-May-2014	rmind	sync with head
1.30.30.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.30.30.1	10-Jun-2019	christos	Sync with HEAD
1.30.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.33.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.34.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.28	24-Apr-2024	thorpej	b3_2706_map_vme(): Use VM_BESTFIT.
1.27	05-Dec-2023	thorpej	Use vmem(9) rather than extent(9) to manage VME<->PCI address translations. While here, malloc(9) -> kmem(9), and fix a couple of things that the newer compiler gripes about.
1.26	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.25	24-Apr-2021	thorpej	branches: 1.25.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.24	10-Nov-2019	chs	branches: 1.24.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.23	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.22	27-Oct-2012	chs	branches: 1.22.38; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.21	30-Jun-2011	wiz	branches: 1.21.2; 1.21.12; dependant -> dependent
1.20	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.19	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.18	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.17	16-Mar-2009	dsl	ANSIfy functions with function-pointer arguments
1.16	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.15	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.14	10-Apr-2008	cegger	branches: 1.14.4; 1.14.12; 1.14.18; use aprint_*_dev and device_xname
1.13	19-Oct-2007	ad	branches: 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	24-Nov-2006	wiz	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; s/independant/independent/, from Zafer.
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	04-Feb-2005	perry	branches: 1.10.6; de-__P
1.9	31-Jan-2003	thorpej	branches: 1.9.2; 1.9.10; 1.9.12; Use aprint_*().
1.8	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.7	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.6	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.5	04-Mar-2002	simonb	Don't "extern int cold;" - this is in <sys/kernel.h>.
1.4	13-Nov-2001	lukem	add RCSID
1.3	07-Jul-2001	thorpej	branches: 1.3.2; bcopy -> memcpy
1.2	12-Mar-2000	drochner	branches: 1.2.6; 1.2.8; support bus probes a la "badaddr" and VME interrupts, fix error handling after mapping problems
1.1	25-Feb-2000	drochner	add (minimal, PIO only) support for the popular "Tundra Universe" PCI-VME interface chip and a frontend for the "Bit3 Mod. 2706" PCI-VME adapter
1.2.8.4	18-Oct-2002	nathanw	Catch up to -current.
1.2.8.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.8.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.8.1	24-Aug-2001	nathanw	Catch up with -current.
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	12-Mar-2000	bouyer	file btvmeii.c was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.3.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.2	16-Mar-2002	jdolecek	Catch up with -current.
1.3.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.12.1	12-Feb-2005	yamt	sync with head.
1.9.10.1	29-Apr-2005	kent	sync with -current
1.9.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.10.6.2	27-Oct-2007	yamt	sync with head.
1.10.6.1	30-Dec-2006	yamt	sync with head.
1.11.22.1	10-Dec-2006	yamt	sync with head.
1.11.20.1	12-Jan-2007	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.14.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.14.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.14.4.2	16-May-2009	yamt	sync with head
1.14.4.1	04-May-2009	yamt	sync with head.
1.21.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.21.2.1	30-Oct-2012	yamt	sync with head
1.22.38.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.22.38.1	10-Jun-2019	christos	Sync with HEAD
1.24.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.25.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	11-Dec-2005	christos	merge ktrace-lwp.
1.4	24-Feb-2004	wiz	Spell interrupt with two rs. From Peter Postma.
1.3	16-Sep-2001	wiz	branches: 1.3.18; Spell 'occurred' with two 'r's.
1.2	12-Jun-2001	wiz	branches: 1.2.2; 1.2.4; receive, not recieve
1.1	30-Jun-1999	drochner	branches: 1.1.2; 1.1.16; add driver for the Bit3 PCI-VME adapter, without the DMA parts for now
1.1.16.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.16.1	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	30-Jun-1999	thorpej	file btvmeireg.h was added on branch chs-ubc2 on 1999-07-01 23:34:02 +0000
1.2.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.2.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.18.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.18.2	18-Sep-2004	skrll	Sync with HEAD.
1.3.18.1	03-Aug-2004	skrll	Sync with HEAD
1.6	05-Dec-2023	thorpej	Use vmem(9) rather than extent(9) to manage VME<->PCI address translations. While here, malloc(9) -> kmem(9), and fix a couple of things that the newer compiler gripes about.
1.5	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.4	11-Dec-2005	christos	branches: 1.4.110; 1.4.120; merge ktrace-lwp.
1.3	04-Feb-2005	perry	de-__P
1.2	24-Mar-2004	drochner	branches: 1.2.8; 1.2.10; remove license clauses 3 and 4 from my cpoyright notices
1.1	30-Jun-1999	drochner	branches: 1.1.2; 1.1.38; add driver for the Bit3 PCI-VME adapter, without the DMA parts for now
1.1.38.4	04-Feb-2005	skrll	Sync with HEAD.
1.1.38.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.38.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.38.1	03-Aug-2004	skrll	Sync with HEAD
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	30-Jun-1999	thorpej	file btvmeivar.h was added on branch chs-ubc2 on 1999-07-01 23:34:03 +0000
1.2.10.1	12-Feb-2005	yamt	sync with head.
1.2.8.1	29-Apr-2005	kent	sync with -current
1.4.120.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.4.110.1	30-Oct-2012	yamt	sync with head
1.36	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.35	27-Sep-2016	pgoyette	branches: 1.35.14; 1.35.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.34	29-Mar-2014	christos	branches: 1.34.6; 1.34.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.33	27-Oct-2012	chs	branches: 1.33.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.32	26-Nov-2009	njoly	branches: 1.32.12; 1.32.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.31	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.30	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.29	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.28	28-Apr-2008	martin	branches: 1.28.14; Remove clause 3 and 4 from TNF licenses
1.27	10-Apr-2008	cegger	branches: 1.27.2; 1.27.4; use aprint_*_dev and device_xname
1.26	19-Oct-2007	ad	branches: 1.26.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.25	04-Mar-2007	christos	branches: 1.25.2; 1.25.14; 1.25.16; 1.25.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.24	28-Nov-2006	ad	branches: 1.24.4; Improve error handling. Related to PR/21900.
1.23	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.22	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.21	28-Aug-2006	christos	branches: 1.21.2; 1.21.4; static goes first.
1.20	11-Dec-2005	christos	branches: 1.20.4; 1.20.8; merge ktrace-lwp.
1.19	28-Jun-2005	thorpej	branches: 1.19.2; Use ANSI function decls and static.
1.18	27-Feb-2005	perry	nuke trailing whitespace
1.17	31-Jan-2003	thorpej	branches: 1.17.2; 1.17.10; 1.17.12; Use aprint_*().
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	15-May-2002	augustss	Adapt to new bridge name.
1.12	25-Jan-2002	ad	- Always validate the return value read from the outbound FIFO. - Copy access method info into the softc so we don't double dereference. - Remove static on functions.
1.11	13-Nov-2001	lukem	add RCSID
1.10	10-Jan-2001	ad	branches: 1.10.2; 1.10.4; Make this work with the SA 431 and 42xx series controllers.
1.9	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.8	09-Nov-2000	ad	branches: 1.8.2; Match controllers hiding behind a Symbios 53c1510; thanks to Jonathan Lemon <jlemon@flugsvamp.com> for providing clarification.
1.7	19-Oct-2000	ad	Nuke unused includes, sort decls.
1.6	26-Sep-2000	ad	Pasto; shouldn't have bitten anyone.
1.5	01-Sep-2000	ad	- Interface cleanup. static, const in places. - cac_pci_l0 has been moved to cac.c, since it's useful for EISA boards. - Model-specific linkage now is responsible for DMA synchronization. - If we don't recognise the board, print the subsystem ID. - Add support for RAID LC2 and Smart Array 431. - Start firmware background tasks on controllers that need it.
1.4	13-Jun-2000	ad	branches: 1.4.2; Use my proper name.
1.3	23-Mar-2000	ad	branches: 1.3.2; Identify individual models correctly.
1.2	21-Mar-2000	ad	Try to use memory mapped I/O on everything but the SMART-2/E.
1.1	16-Mar-2000	ad	Driver for Compaq array controllers and disks (cac(4)/ca(4)).
1.3.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.4.2.2	29-Jan-2002	he	Pull up revision 1.12 (requested by ad): Some fixes: o Always validate return value read from outbound FIFO o Copy access method into to softc to avoid double dereference o Remove static on functions Fixes PR#14453.
1.4.2.1	25-Oct-2001	he	Pull up revisions 1.5-1.10 (via patch, requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.8.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.8.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.8.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.8.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.2.1	09-Nov-2000	bouyer	file cac_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.10.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.10.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.10.4.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.10.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.10.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.10.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.10.2.2	28-Feb-2002	nathanw	Catch up to -current.
1.10.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.17.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.17.10.1	29-Apr-2005	kent	sync with -current
1.17.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.17.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.3	27-Oct-2007	yamt	sync with head.
1.19.2.2	03-Sep-2007	yamt	sync with head.
1.19.2.1	30-Dec-2006	yamt	sync with head.
1.20.8.1	03-Sep-2006	yamt	sync with head.
1.20.4.1	09-Sep-2006	rpaulo	sync with head
1.21.4.2	10-Dec-2006	yamt	sync with head.
1.21.4.1	22-Oct-2006	yamt	sync with head
1.21.2.2	12-Jan-2007	ad	Sync with head.
1.21.2.1	18-Nov-2006	ad	Sync with head.
1.24.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.25.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.25.16.1	06-Nov-2007	matt	sync with HEAD
1.25.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.25.2.1	23-Oct-2007	ad	Sync with head.
1.26.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.27.4.3	11-Mar-2010	yamt	sync with head
1.27.4.2	16-May-2009	yamt	sync with head
1.27.4.1	16-May-2008	yamt	sync with head.
1.27.2.1	18-May-2008	yamt	sync with head.
1.28.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.32.22.3	03-Dec-2017	jdolecek	update from HEAD
1.32.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.32.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.32.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.32.12.1	30-Oct-2012	yamt	sync with head
1.33.2.1	18-May-2014	rmind	sync with head
1.34.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.34.6.1	05-Oct-2016	skrll	Sync with HEAD
1.35.16.1	10-Jun-2019	christos	Sync with HEAD
1.35.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.31	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.30	22-Jul-2011	njoly	branches: 1.30.2; 1.30.6; Fix CHIPSFB_DEBUG build
1.29	11-May-2011	dyoung	Constify pci_attach_args.
1.28	05-Apr-2011	dyoung	Bitwise-OR PCI_COMMAND_IO_ENABLE \| PCI_COMMAND_MEM_ENABLE into the PCI Command Register instead of PCI_FLAGS_IO_ENABLED \| PCI_FLAGS_MEM_ENABLED. The latter amounts to the same value, but the names are in the pa_flags namespace instead of the PCI Command Register bits namespace.
1.27	23-Mar-2011	macallan	complete the driver/bus frontend split and make sure the driver can deal with both PCI and VLB variants of the chip
1.26	09-Feb-2011	macallan	use the bus independent stuff from ic/ct65550* and leave only the PCI specific bits in pci/chipsfb.c
1.25	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.24	23-Dec-2010	cegger	branches: 1.24.2; 1.24.4; build fix: add missing initializers
1.23	16-Dec-2010	cegger	ioctl: KNF switch-case
1.22	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.21	04-May-2010	macallan	use alternate font if available ( for autogenerated box drawing characters )
1.20	20-Aug-2009	macallan	branches: 1.20.2; 1.20.4; use vcons_replay_msgbuf() when appropriate
1.19	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.18	06-May-2009	elad	Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.17	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.16	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.15	08-May-2008	macallan	branches: 1.15.12; nuke clause 3
1.14	10-Apr-2008	cegger	branches: 1.14.2; 1.14.4; 1.14.6; use aprint_*_dev and device_xname
1.13	27-Feb-2008	macallan	get rid of the last #ifdef macppc and use PCI_MAGIC_IO_RANGE to select at which offset we mmap PCI IO space
1.12	19-Aug-2007	he	branches: 1.12.2; 1.12.18; 1.12.22; Use bool with prop_dictionary_get_bool.
1.11	02-Aug-2007	macallan	branches: 1.11.2; 1.11.4; don't bus_space_map more IO space than we really need
1.10	28-Jul-2007	mjf	Implement bounds checking in some places in display driver code to avoid the possibility of a local user panic.
1.9	04-Mar-2007	christos	branches: 1.9.2; 1.9.10; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.8	22-Jan-2007	macallan	branches: 1.8.2; get rid of OpenFirmware dependencies, use device properties instead
1.7	20-Jan-2007	he	Unify the declaration of OF_interpret() between sparc{,64} and macppc by adding the "nargs" argument to the macppc version, and fix the macppc ports uses of OF_interpret() accordingly. Also move the declaration of OF_interpt() from macppc's autoconf.h to ofw/openfirm.h. This fixes the build of the macppc port. Approved by macallan@.
1.6	10-Nov-2006	macallan	branches: 1.6.2; 1.6.6; remove some leftover #includes
1.5	27-Oct-2006	macallan	fix a typo in chipsfb_bitblt()
1.4	16-Oct-2006	macallan	branches: 1.4.2; detect video RAM size
1.3	27-Sep-2006	macallan	the chip wants all data we send with 64bit padding, no matter what padding individual scanlines have so we send an extra 0 when we have to.
1.2	27-Sep-2006	macallan	cleanup, implement putchar() using the blitter
1.1	23-Sep-2006	macallan	a half done wsdisplay driver for Chips & Technologies 65550 graphics chips So far it only uses the blitter for scrolling and rectangle filling, characters are still drawn in software and there's no support for video mode switching. Virtual consoles are supported via vcons. Works fine on a PowerBook 3400c.
1.4.2.3	10-Dec-2006	yamt	sync with head.
1.4.2.2	22-Oct-2006	yamt	sync with head
1.4.2.1	16-Oct-2006	yamt	file chipsfb.c was added on branch yamt-splraiseipl on 2006-10-22 06:06:16 +0000
1.6.6.5	17-Mar-2008	yamt	sync with head.
1.6.6.4	03-Sep-2007	yamt	sync with head.
1.6.6.3	26-Feb-2007	yamt	sync with head.
1.6.6.2	30-Dec-2006	yamt	sync with head.
1.6.6.1	10-Nov-2006	yamt	file chipsfb.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:43 +0000
1.6.2.3	01-Feb-2007	ad	Sync with head.
1.6.2.2	18-Nov-2006	ad	Sync with head.
1.6.2.1	10-Nov-2006	ad	file chipsfb.c was added on branch newlock2 on 2006-11-18 21:34:29 +0000
1.8.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.9.10.2	03-Sep-2007	skrll	Sync with HEAD.
1.9.10.1	15-Aug-2007	skrll	Sync with HEAD.
1.9.2.1	20-Aug-2007	ad	Sync with HEAD.
1.11.4.2	02-Aug-2007	macallan	don't bus_space_map more IO space than we really need
1.11.4.1	02-Aug-2007	macallan	file chipsfb.c was added on branch matt-mips64 on 2007-08-02 06:56:44 +0000
1.11.2.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.12.22.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.22.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.18.1	24-Mar-2008	keiichi	sync with head.
1.12.2.1	23-Mar-2008	matt	sync with HEAD
1.14.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.14.4.4	11-Aug-2010	yamt	sync with head.
1.14.4.3	16-Sep-2009	yamt	sync with head
1.14.4.2	16-May-2009	yamt	sync with head
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.20.4.4	31-May-2011	rmind	sync with head
1.20.4.3	21-Apr-2011	rmind	sync with head
1.20.4.2	05-Mar-2011	rmind	sync with head
1.20.4.1	30-May-2010	rmind	sync with head
1.20.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.24.4.2	17-Feb-2011	bouyer	Sync with HEAD
1.24.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.24.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.30.6.1	18-Feb-2012	mrg	merge to -current.
1.30.2.1	17-Apr-2012	yamt	sync with head
1.4	09-Feb-2011	macallan	this is no longer needed, register definitions for chipsfb now live in dev/ic/ct65550reg.h
1.3	08-May-2008	macallan	branches: 1.3.20; 1.3.26; 1.3.28; nuke clause 3
1.2	27-Sep-2006	macallan	branches: 1.2.2; 1.2.4; 1.2.8; 1.2.54; 1.2.56; 1.2.58; 1.2.60; cleanup, implement putchar() using the blitter
1.1	23-Sep-2006	macallan	a half done wsdisplay driver for Chips & Technologies 65550 graphics chips So far it only uses the blitter for scrolling and rectangle filling, characters are still drawn in software and there's no support for video mode switching. Virtual consoles are supported via vcons. Works fine on a PowerBook 3400c.
1.2.60.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.2.58.1	16-May-2008	yamt	sync with head.
1.2.56.1	18-May-2008	yamt	sync with head.
1.2.54.1	02-Jun-2008	mjf	Sync with HEAD.
1.2.8.2	30-Dec-2006	yamt	sync with head.
1.2.8.1	27-Sep-2006	yamt	file chipsfbreg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:43 +0000
1.2.4.2	18-Nov-2006	ad	Sync with head.
1.2.4.1	27-Sep-2006	ad	file chipsfbreg.h was added on branch newlock2 on 2006-11-18 21:34:29 +0000
1.2.2.2	22-Oct-2006	yamt	sync with head
1.2.2.1	27-Sep-2006	yamt	file chipsfbreg.h was added on branch yamt-splraiseipl on 2006-10-22 06:06:16 +0000
1.3.28.1	17-Feb-2011	bouyer	Sync with HEAD
1.3.26.1	06-Jun-2011	jruoho	Sync with HEAD.
1.3.20.1	05-Mar-2011	rmind	sync with head
1.23	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.22	14-Jul-2020	jdolecek	match newer HP Smart Array controllers, list from FreeBSD
1.21	14-Jul-2020	jdolecek	fix __arraycount() parameter
1.20	14-Jul-2020	jdolecek	make ciss_pci_devices[] more compact
1.19	14-Jul-2020	jdolecek	remove CISS_NO_INTERRUPT_HACK, discussed with joerg@ and spz@
1.18	14-Jul-2020	jdolecek	add support for PERFORMANT mode, and allow MSI/MSI-X (only) for adapters which support it code adapted from FreeBSD, but with fix for setting the performant bit and pull count on command submittion as seen in hpsa Linux driver tested with INTx and MSI-X on HP Smart Array 11 thanks to Andreas Gustafsson for initial testing, and providing access to test machine
1.17	05-Jul-2020	jdolecek	there is more to MSI/MSI-X support in ciss(4) than just allocating the right interrupt, it needs some explicit support; disable for now until the full support is there PR kern/55461
1.16	04-Jul-2020	jdolecek	switch to pci_intr_alloc() so this uses MSI/MSI-X if device supports it
1.15	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.14	12-Feb-2018	joerg	branches: 1.14.2; 1.14.4; Add a new option CISS_NO_INTERRUPT_HACK for driving ciss(4) by callouts. This is intended as workaround for certain Xen issues with dom0 kernels and will likely want to have a high HZ value as well for decent performance.
1.13	14-Jul-2016	msaitoh	branches: 1.13.10; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.12	17-Jun-2016	christos	catch up with new names.
1.11	29-Mar-2014	christos	branches: 1.11.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.10	27-Oct-2012	chs	branches: 1.10.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.9	26-Nov-2009	njoly	branches: 1.9.12; 1.9.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.8	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.7	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.6	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.5	10-Apr-2008	cegger	branches: 1.5.4; 1.5.18; use aprint_*_dev and device_xname
1.4	19-Oct-2007	ad	branches: 1.4.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.3	16-Nov-2006	christos	branches: 1.3.8; 1.3.22; 1.3.24; 1.3.28; __unused removal on arguments; approved by core.
1.2	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.1	21-Mar-2006	he	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.12; 1.1.18; 1.1.22; 1.1.24; 1.1.26; Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.1.26.2	10-Dec-2006	yamt	sync with head.
1.1.26.1	22-Oct-2006	yamt	sync with head
1.1.24.2	09-Sep-2006	rpaulo	sync with head
1.1.24.1	21-Mar-2006	rpaulo	file ciss_pci.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:16 +0000
1.1.22.1	18-Nov-2006	ad	Sync with head.
1.1.18.4	27-Oct-2007	yamt	sync with head.
1.1.18.3	30-Dec-2006	yamt	sync with head.
1.1.18.2	21-Jun-2006	yamt	sync with head.
1.1.18.1	21-Mar-2006	yamt	file ciss_pci.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:03 +0000
1.1.12.2	22-Apr-2006	simonb	Sync with head.
1.1.12.1	21-Mar-2006	simonb	file ciss_pci.c was added on branch simonb-timecounters on 2006-04-22 11:39:13 +0000
1.1.10.2	19-Apr-2006	elad	sync with head.
1.1.10.1	21-Mar-2006	elad	file ciss_pci.c was added on branch elad-kernelauth on 2006-04-19 03:25:33 +0000
1.1.8.2	11-Apr-2006	yamt	sync files somehow mis-tagged by yamt-pdpolicy-base2.
1.1.8.1	21-Mar-2006	yamt	file ciss_pci.c was added on branch yamt-pdpolicy on 2006-04-11 12:20:52 +0000
1.1.6.2	11-Apr-2006	riz	Pull up following revision(s) (requested by he in ticket #10409): sys/dev/ic/cissreg.h: revision 1.1 sys/dev/pci/ciss_pci.c: revision 1.1 sys/dev/ic/cissvar.h: revision 1.1 sys/dev/pci/files.pci: revision 1.245 sys/dev/ic/ciss.c: revision 1.1 via patch Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.1.6.1	21-Mar-2006	riz	file ciss_pci.c was added on branch netbsd-2 on 2006-04-11 01:15:22 +0000
1.1.4.2	08-Apr-2006	tron	Pull up following revision(s) (requested by he in ticket #1247): sys/dev/ic/cissreg.h: revision 1.1 sys/dev/pci/ciss_pci.c: revision 1.1 sys/dev/ic/cissvar.h: revision 1.1 sys/dev/pci/files.pci: revision 1.245 sys/dev/ic/ciss.c: revision 1.1 Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.1.4.1	21-Mar-2006	tron	file ciss_pci.c was added on branch netbsd-3 on 2006-04-08 23:31:38 +0000
1.1.2.2	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.1.2.1	21-Mar-2006	tron	file ciss_pci.c was added on branch peter-altq on 2006-03-28 09:42:13 +0000
1.3.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.3.24.1	06-Nov-2007	matt	sync with HEAD
1.3.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.8.1	23-Oct-2007	ad	Sync with head.
1.4.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.5.4.2	11-Mar-2010	yamt	sync with head
1.5.4.1	16-May-2009	yamt	sync with head
1.9.22.3	03-Dec-2017	jdolecek	update from HEAD
1.9.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.9.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.9.12.1	30-Oct-2012	yamt	sync with head
1.10.2.1	18-May-2014	rmind	sync with head
1.11.6.2	05-Oct-2016	skrll	Sync with HEAD
1.11.6.1	09-Jul-2016	skrll	Sync with HEAD
1.13.10.1	09-Apr-2018	bouyer	Pull up following revision(s) (requested by spz in ticket #708): sys/conf/files: revision 1.1194 sys/dev/ic/cissvar.h: revision 1.7 sys/dev/ic/ciss.c: revision 1.38 sys/dev/pci/ciss_pci.c: revision 1.14 sys/arch/amd64/conf/ALL: revision 1.81 Add a new option CISS_NO_INTERRUPT_HACK for driving ciss(4) by callouts. This is intended as workaround for certain Xen issues with dom0 kernels and will likely want to have a high HZ value as well for decent performance.
1.14.4.1	10-Jun-2019	christos	Sync with HEAD
1.14.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.47	26-Dec-2021	andvar	fix various typos, mainly in comments.
1.46	17-Jul-2020	jdolecek	remove explicit irqack() call for unexpected interrupt; this happens normally during boot when running identify for the drive as polled command, and fixes hard hang caused by the irqack() on a sun machine the unexpected interrupt happens because cmdide doesn't respect WDCTL_IDS bit and triggers interrupt also for polled commands also reclassify the "bogus intr" as aprint_verbose() to avoid noise on boot discussed and tested by Martin Husemann
1.45	01-Jul-2020	jdolecek	revert, not working
1.44	01-Jul-2020	jdolecek	ignore interrupt when not expecting one, particularly don't invoke the irqack in this case seems cmdide doesn't honour WDCTL_IDS when executing polled commands, and there is a race where the irqack call interferes with polled command and causes a lockup reported by Martin Husemann
1.43	22-Oct-2017	jdolecek	branches: 1.43.10; do not share queue between the non-indepedant channels; instead make sure only one of the channels is ever active on the same controller fixes PR kern/52606 by Martin Husemann, thanks for report and testing
1.42	20-Oct-2017	jdolecek	actually 0680 uses different chip map func, which doesn't call the function handing non-independant channels, so don't set IDE_SHARED_CHANNELS for it
1.41	20-Oct-2017	jdolecek	move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.40	19-Oct-2017	jdolecek	replace the chek for the shared channel of cmdide(4) a flag of the product array, rather than switch inside attach routine XXX judging from product name, Silicon Image 0680 might be newer than 0649 XXX and hence have actually independant channels, but I don't have the hw XXX so keeping as-is no functional change, just to improve visibility in course of fixing PR kern/52606
1.39	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.38	03-Sep-2012	kiyohara	branches: 1.38.2; 1.38.28; Fix no wd on cmd680. Also move setting wdc_maxdrives into cmd*_chip_map().
1.37	31-Jul-2012	bouyer	Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.36	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.35	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.34	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.33	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.32	04-Apr-2011	dyoung	branches: 1.32.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.31	06-Nov-2010	jakllsch	branches: 1.31.2; Enable detachment. Tested on SiI 0680 (native mapped).
1.30	05-Nov-2010	jakllsch	Pave the way for detachment of pciide(4)-family ATA controllers.
1.29	19-Oct-2009	bouyer	branches: 1.29.2; 1.29.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.28	18-Mar-2008	cube	branches: 1.28.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.27	09-Feb-2007	ad	branches: 1.27.20; 1.27.36; 1.27.40; Merge newlock2 to head.
1.26	24-Nov-2006	wiz	s/independant/independent/, from Zafer.
1.25	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.24	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.23	07-Feb-2006	bouyer	branches: 1.23.14; 1.23.16; Add another missing ch_ndrive initialization.
1.22	17-Jan-2006	bouyer	branches: 1.22.2; 1.22.4; Add missing ch_ndrive initialisations. Pointed out by Juan RP.
1.21	11-Dec-2005	christos	branches: 1.21.2; merge ktrace-lwp.
1.20	21-Sep-2005	christos	PR/31364: Douglas Fraser: cmd680 IDE controller not configured properly.
1.19	24-May-2005	lukem	branches: 1.19.2; add __KERNEL_RCSID
1.18	27-Feb-2005	perry	nuke trailing whitespace
1.17	21-Aug-2004	thorpej	branches: 1.17.4; 1.17.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.16	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.15	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.14	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.13	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.12	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.11	02-Aug-2004	bouyer	branches: 1.11.2; If acknowledge "bogus" interrupts (that is, interrupts that have happened according to the controller-dependant registers while wdc was not waiting for an interrupt). A reset can generate spurious interrupts (or even a flow of spurious interrupts on the promise) that needs to be properly acknowledged.
1.10	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.9	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.8	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.7	13-Dec-2003	thorpej	Split out the Silicon Image SATALink support into its own driver, "satalink".
1.6	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.5	10-Nov-2003	wiz	Spell address with two d's. Inspired by similar changes in OpenBSD, originating from Jonathon Gray and forwarded by jmc@openbsd.
1.4	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.11.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.11.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.11.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.11.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.11.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.11.2.2	03-Aug-2004	skrll	Sync with HEAD
1.11.2.1	02-Aug-2004	skrll	file cmdide.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.17.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.17.4.1	29-Apr-2005	kent	sync with -current
1.19.2.4	24-Mar-2008	yamt	sync with head.
1.19.2.3	26-Feb-2007	yamt	sync with head.
1.19.2.2	30-Dec-2006	yamt	sync with head.
1.19.2.1	21-Jun-2006	yamt	sync with head.
1.21.2.2	18-Feb-2006	yamt	sync with head.
1.21.2.1	01-Feb-2006	yamt	sync with head.
1.22.4.1	22-Apr-2006	simonb	Sync with head.
1.22.2.1	09-Sep-2006	rpaulo	sync with head
1.23.16.2	10-Dec-2006	yamt	sync with head.
1.23.16.1	22-Oct-2006	yamt	sync with head
1.23.14.3	06-Feb-2007	ad	Quiten noisy boot messages.
1.23.14.2	12-Jan-2007	ad	Sync with head.
1.23.14.1	18-Nov-2006	ad	Sync with head.
1.27.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.27.36.1	24-Mar-2008	keiichi	sync with head.
1.27.20.1	23-Mar-2008	matt	sync with HEAD
1.28.4.1	11-Mar-2010	yamt	sync with head
1.29.4.2	21-Apr-2011	rmind	sync with head
1.29.4.1	05-Mar-2011	rmind	sync with head
1.29.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.31.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.32.4.1	30-Oct-2012	yamt	sync with head
1.38.28.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.38.2.2	03-Dec-2017	jdolecek	update from HEAD
1.38.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.43.10.1	17-Aug-2020	martin	Pull up following revision(s) (requested by jdolecek in ticket #1054): sys/dev/pci/cmdide.c: revision 1.46 remove explicit irqack() call for unexpected interrupt; this happens normally during boot when running identify for the drive as polled command, and fixes hard hang caused by the irqack() on a sun machine the unexpected interrupt happens because cmdide doesn't respect WDCTL_IDS bit and triggers interrupt also for polled commands also reclassify the "bogus intr" as aprint_verbose() to avoid noise on boot discussed and tested by Martin Husemann
1.2	28-Dec-2001	itohy	Move this figure to more suitable place: http://www.netbsd.org/Documentation/Hardware/Chips/C-Media/cmi8738-mixer.fig
1.1	27-Dec-2001	itohy	Analyzed structure of CMI8738/PCI-6ch-* mixer for the benefit of those who edit the driver. This is a Fig file. Use Xfig to edit/view, use Transfig package to convert to other graphic formats.
1.61	08-Jan-2024	chs	mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.60	31-May-2022	andvar	branches: 1.60.4; fix various typos in comments, documentation and messages.
1.59	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.58	24-Apr-2021	thorpej	branches: 1.58.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.57	29-Feb-2020	isaki	branches: 1.57.8; Remove rounding by 4 bytes on round_blocksize(). For drivers which supports only 16bit * 2channels sampling, rounding by 4 bytes no longer meaningful.
1.56	08-Jun-2019	isaki	branches: 1.56.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.55	12-May-2019	maya	Remove unused function following isaki-audio2 merge. Shows up when doing clang builds. From a.rin.
1.54	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.53	16-Mar-2019	isaki	branches: 1.53.2; White space.
1.52	16-Mar-2019	isaki	Use C99 style struct initializer to audio_hw_if.
1.51	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.50	01-Jun-2017	chs	branches: 1.50.8; 1.50.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.49	15-Aug-2016	maxv	Use the exact same argument for kmem_alloc and kmem_free; from brainy
1.48	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.47	29-Mar-2014	christos	branches: 1.47.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.46	27-Oct-2012	chs	branches: 1.46.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.45	30-Jan-2012	drochner	branches: 1.45.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.44	24-Nov-2011	mrg	branches: 1.44.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.43	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.42	26-Nov-2009	njoly	branches: 1.42.12; 1.42.14; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.41	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.40	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.39	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.38	10-Apr-2008	cegger	branches: 1.38.4; 1.38.16; 1.38.18; use aprint_*_dev and device_xname
1.37	01-Apr-2008	xtraeme	Fix fallout from device_t/softc split to mpu(4): always pass the mpu_softc to mpu_intr(). Found by Takeshi Nakayama.
1.36	19-Oct-2007	ad	branches: 1.36.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.35	04-Mar-2007	christos	branches: 1.35.2; 1.35.14; 1.35.16; 1.35.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.34	16-Nov-2006	christos	branches: 1.34.4; 1.34.6; __unused removal on arguments; approved by core.
1.33	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.32	29-Aug-2006	christos	branches: 1.32.2; 1.32.4; fix incomplete initializer.
1.31	11-May-2006	mrg	quell GCC 4.1 uninitialised variable warnings.
1.30	24-Dec-2005	perry	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.12; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.29	11-Dec-2005	christos	merge ktrace-lwp.
1.28	15-Jan-2005	kent	branches: 1.28.10; ansify and KNF
1.27	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.26	29-Oct-2004	yamt	branches: 1.26.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.25	26-Oct-2004	xtraeme	Fix typo in a comment: alalog -> analog.
1.24	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.23	06-Dec-2003	itohy	Back out the following commit and apply the real fix. A variable was actually used in a wrong way. > revision 1.20 > date: 2003/10/25 21:31:43; author: christos; state: Exp; lines: +3 -3 > Fix uninitialized variable warnings.
1.22	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.21	22-Nov-2003	itohy	- Correctly handle bit 24 of miscellaneous control register, which has different meanings for read and write. This fixes 48kHz playback to SPDIF. - Fix the problem where the SPDIF output voltage setting (0.5V, 5V) was swapped. The default value is changed from 0.5V to 5V, which correctly selects 5V (so the default physical behavior is unchanged). Should fix PRs kern/16047 and kern/16817 (but not tested since I don't have other SPDIF hardware to test with). Approved by tshiozak.
1.20	25-Oct-2003	christos	Fix uninitialized variable warnings.
1.19	25-Oct-2003	christos	Fix uninitialized variable warnings
1.18	01-Feb-2003	thorpej	branches: 1.18.2; Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.17	31-Jan-2003	thorpej	Use aprint_*().
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	02-Feb-2002	augustss	Fix typo so that 48kHz can be set. From Stephen Ma in kern/15456.
1.12	27-Dec-2001	itohy	Fix comment.
1.11	13-Nov-2001	lukem	add RCSID
1.10	04-Nov-2001	itohy	Re-implement the mixer o Implement SPDIF selection/monitoring function. Now, say, playback to both analog output and SPDIF is possible. o Implement support of AUX input, MIC preamp and MIC recording gain. o L-R selection of record source doesn't seem to work on cmpci. Therefore, the .swap mixer functions are and deleted. o Following SB mixer registers don't exist on cmpci, and they are removed. input gain, output gain, AGC, equalization (bass, treble) o Other mixer changes, including inputs.XXX.mute -> (deleted) outputs.XXX.mute -> inputs.XXX.mute inputs.pc_speaker -> inputs.speaker spdif. -> reorganized to spdif.input., spdif.output. o Current status: I have tested these and confirmed to work fine. - Output and recording from Line-in, AUX, CD and MIC analog inputs, - Output and recording from FM synthesizer, - Output from PC speaker input, - Output wave playback. - SPDIF (44.1kHz) input selection (#1, #2 (6ch version only), wave to spdin), phase selection, monitoring and recording, - SPDIF (44.1kHz) playback, through (SPDIF in to SPDIF out) and monitoring. I haven't tested these but may work. - SPDIF 48kHz input and output, - Full-duplex operation, - Recording wave output. I don't think these are working. - Legacy (wave + FM synthesizer) to SPDIF output (and the monitoring), - Exchanging front and rear outputs, - Surround. These are not implemented. - 4ch / 6ch support, - Joystick port support.
1.9	03-Oct-2001	augustss	branches: 1.9.2; Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.8	04-Sep-2001	itohy	branches: 1.8.2; - Add opl at cmpci attachment. - Add mpu at cmpci attachment. - Fix panic during attach on alpha. - Change register of SPDIF input phase (but undocumented, either). - Change code for mixer initialization. - Return 1, not 0, when an interrupt is processed. More changes to the mixer are pending....
1.7	12-Feb-2001	tshiozak	branches: 1.7.2; 1.7.4; some improvement for cmpci; SPDIF supports, etc.
1.6	18-Jan-2001	jdolecek	constify
1.5	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.4	26-Jun-2000	simonb	branches: 1.4.2; Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.3	08-Jun-2000	gmcgarry	branches: 1.3.2; Make attach messages consistent with PCI drivers. Make function pointer usage consistent with other drivers.
1.2	30-Apr-2000	augustss	branches: 1.2.2; Remove unneeded #include "mpu.h"
1.1	30-Apr-2000	augustss	Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.2.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.3.2.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.4.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.4.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.4.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.4.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	26-Jun-2000	bouyer	file cmpci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.7.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.7.4.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.7.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.7.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.7.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.7.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.7.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.8.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.9.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.18.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.18.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.18.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.18.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.18.2.1	03-Aug-2004	skrll	Sync with HEAD
1.26.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.26.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.28.10.4	27-Oct-2007	yamt	sync with head.
1.28.10.3	03-Sep-2007	yamt	sync with head.
1.28.10.2	30-Dec-2006	yamt	sync with head.
1.28.10.1	21-Jun-2006	yamt	sync with head.
1.30.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.30.8.2	03-Sep-2006	yamt	sync with head.
1.30.8.1	24-May-2006	yamt	sync with head.
1.30.6.1	01-Jun-2006	kardel	Sync with head.
1.30.4.1	09-Sep-2006	rpaulo	sync with head
1.32.4.2	10-Dec-2006	yamt	sync with head.
1.32.4.1	22-Oct-2006	yamt	sync with head
1.32.2.1	18-Nov-2006	ad	Sync with head.
1.34.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.34.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.35.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.35.16.1	06-Nov-2007	matt	sync with HEAD
1.35.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.35.2.1	23-Oct-2007	ad	Sync with head.
1.36.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.36.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.38.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.38.16.1	11-Dec-2008	ad	Checkpoint work in progress.
1.38.4.2	11-Mar-2010	yamt	sync with head
1.38.4.1	16-May-2009	yamt	sync with head
1.42.14.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.42.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.42.12.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.42.12.2	30-Oct-2012	yamt	sync with head
1.42.12.1	17-Apr-2012	yamt	sync with head
1.44.2.1	18-Feb-2012	mrg	merge to -current.
1.45.6.3	03-Dec-2017	jdolecek	update from HEAD
1.45.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.45.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.46.2.1	18-May-2014	rmind	sync with head
1.47.6.3	28-Aug-2017	skrll	Sync with HEAD
1.47.6.2	05-Oct-2016	skrll	Sync with HEAD
1.47.6.1	09-Jul-2016	skrll	Sync with HEAD
1.50.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.50.10.1	10-Jun-2019	christos	Sync with HEAD
1.50.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.53.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.53.2.2	28-Apr-2019	isaki	Adapt to audio2.
1.53.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.56.4.1	29-Feb-2020	ad	Sync with head.
1.57.8.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.58.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.60.4.1	14-Jan-2024	martin	Pull up following revision(s) (requested by chs in ticket #542): sys/dev/pci/yds.c: revision 1.69 sys/dev/pci/cmpci.c: revision 1.61 sys/dev/pci/sv.c: revision 1.62 sys/dev/pci/fms.c: revision 1.50 mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.7	11-Dec-2005	christos	merge ktrace-lwp.
1.6	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	22-Nov-2003	itohy	- Correctly handle bit 24 of miscellaneous control register, which has different meanings for read and write. This fixes 48kHz playback to SPDIF. - Fix the problem where the SPDIF output voltage setting (0.5V, 5V) was swapped. The default value is changed from 0.5V to 5V, which correctly selects 5V (so the default physical behavior is unchanged). Should fix PRs kern/16047 and kern/16817 (but not tested since I don't have other SPDIF hardware to test with). Approved by tshiozak.
1.4	04-Nov-2001	itohy	branches: 1.4.16; Re-implement the mixer o Implement SPDIF selection/monitoring function. Now, say, playback to both analog output and SPDIF is possible. o Implement support of AUX input, MIC preamp and MIC recording gain. o L-R selection of record source doesn't seem to work on cmpci. Therefore, the .swap mixer functions are and deleted. o Following SB mixer registers don't exist on cmpci, and they are removed. input gain, output gain, AGC, equalization (bass, treble) o Other mixer changes, including inputs.XXX.mute -> (deleted) outputs.XXX.mute -> inputs.XXX.mute inputs.pc_speaker -> inputs.speaker spdif. -> reorganized to spdif.input., spdif.output. o Current status: I have tested these and confirmed to work fine. - Output and recording from Line-in, AUX, CD and MIC analog inputs, - Output and recording from FM synthesizer, - Output from PC speaker input, - Output wave playback. - SPDIF (44.1kHz) input selection (#1, #2 (6ch version only), wave to spdin), phase selection, monitoring and recording, - SPDIF (44.1kHz) playback, through (SPDIF in to SPDIF out) and monitoring. I haven't tested these but may work. - SPDIF 48kHz input and output, - Full-duplex operation, - Recording wave output. I don't think these are working. - Legacy (wave + FM synthesizer) to SPDIF output (and the monitoring), - Exchanging front and rear outputs, - Surround. These are not implemented. - 4ch / 6ch support, - Joystick port support.
1.3	04-Sep-2001	itohy	branches: 1.3.4; - Add opl at cmpci attachment. - Add mpu at cmpci attachment. - Fix panic during attach on alpha. - Change register of SPDIF input phase (but undocumented, either). - Change code for mixer initialization. - Return 1, not 0, when an interrupt is processed. More changes to the mixer are pending....
1.2	12-Feb-2001	tshiozak	branches: 1.2.2; 1.2.4; some improvement for cmpci; SPDIF supports, etc.
1.1	30-Apr-2000	augustss	branches: 1.1.6; Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.1.6.3	12-Mar-2001	bouyer	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	30-Apr-2000	bouyer	file cmpcireg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.2.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.2.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.3.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.4.16.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.16.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.16.1	03-Aug-2004	skrll	Sync with HEAD
1.12	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.11	23-Nov-2011	jmcneill	branches: 1.11.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.10	04-Mar-2007	christos	branches: 1.10.54; 1.10.76; 1.10.78; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.9	11-Dec-2005	christos	branches: 1.9.26; 1.9.28; merge ktrace-lwp.
1.8	15-Jan-2005	kent	branches: 1.8.10; ansify and KNF
1.7	10-Nov-2004	wiz	Fix typo in comment; from Sergey Svishchev.
1.6	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	22-Nov-2003	itohy	- Correctly handle bit 24 of miscellaneous control register, which has different meanings for read and write. This fixes 48kHz playback to SPDIF. - Fix the problem where the SPDIF output voltage setting (0.5V, 5V) was swapped. The default value is changed from 0.5V to 5V, which correctly selects 5V (so the default physical behavior is unchanged). Should fix PRs kern/16047 and kern/16817 (but not tested since I don't have other SPDIF hardware to test with). Approved by tshiozak.
1.4	04-Nov-2001	itohy	branches: 1.4.16; Re-implement the mixer o Implement SPDIF selection/monitoring function. Now, say, playback to both analog output and SPDIF is possible. o Implement support of AUX input, MIC preamp and MIC recording gain. o L-R selection of record source doesn't seem to work on cmpci. Therefore, the .swap mixer functions are and deleted. o Following SB mixer registers don't exist on cmpci, and they are removed. input gain, output gain, AGC, equalization (bass, treble) o Other mixer changes, including inputs.XXX.mute -> (deleted) outputs.XXX.mute -> inputs.XXX.mute inputs.pc_speaker -> inputs.speaker spdif. -> reorganized to spdif.input., spdif.output. o Current status: I have tested these and confirmed to work fine. - Output and recording from Line-in, AUX, CD and MIC analog inputs, - Output and recording from FM synthesizer, - Output from PC speaker input, - Output wave playback. - SPDIF (44.1kHz) input selection (#1, #2 (6ch version only), wave to spdin), phase selection, monitoring and recording, - SPDIF (44.1kHz) playback, through (SPDIF in to SPDIF out) and monitoring. I haven't tested these but may work. - SPDIF 48kHz input and output, - Full-duplex operation, - Recording wave output. I don't think these are working. - Legacy (wave + FM synthesizer) to SPDIF output (and the monitoring), - Exchanging front and rear outputs, - Surround. These are not implemented. - 4ch / 6ch support, - Joystick port support.
1.3	04-Sep-2001	itohy	branches: 1.3.4; - Add opl at cmpci attachment. - Add mpu at cmpci attachment. - Fix panic during attach on alpha. - Change register of SPDIF input phase (but undocumented, either). - Change code for mixer initialization. - Return 1, not 0, when an interrupt is processed. More changes to the mixer are pending....
1.2	12-Feb-2001	tshiozak	branches: 1.2.2; 1.2.4; some improvement for cmpci; SPDIF supports, etc.
1.1	30-Apr-2000	augustss	branches: 1.1.6; Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.1.6.3	12-Mar-2001	bouyer	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	30-Apr-2000	bouyer	file cmpcivar.h was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.2.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.2.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.3.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.4.16.5	17-Jan-2005	skrll	Sync with HEAD.
1.4.16.4	14-Nov-2004	skrll	Sync with HEAD.
1.4.16.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.16.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.16.1	03-Aug-2004	skrll	Sync with HEAD
1.8.10.1	03-Sep-2007	yamt	sync with head.
1.9.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.9.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.10.78.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.10.76.2	30-Oct-2012	yamt	sync with head
1.10.76.1	17-Apr-2012	yamt	sync with head
1.10.54.1	11-Dec-2008	ad	Checkpoint work in progress.
1.11.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.27	25-Mar-2021	rin	Introduce sc_poll_ticks and obsolete COM_HW_POLL bit in sc_hwflags. Polling is scheduled at every sc_poll_ticks ticks. This is useful to work around H/W bug, by which interrupts are lost sometimes; interrupt-based I/O mostly works and no need for polling every counter ticks.
1.26	08-Dec-2018	thorpej	branches: 1.26.12; 1.26.14; Clean up initialization of com_regs structure, in preparation for some additional changers.
1.25	30-Nov-2018	jmcneill	Add support for polling com devices when no interrupt is available.
1.24	27-Apr-2017	msaitoh	branches: 1.24.8; 1.24.10; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.23	23-May-2014	msaitoh	branches: 1.23.4; 1.23.18; Print the type of serial port from the interface field in the PCI class register.
1.22	29-Mar-2014	christos	branches: 1.22.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.21	31-Jul-2013	soren	On x86, print the I/O address of lpt_puc as for com_puc so that hacks using iopl/ioperm can be told where to find the parallel port. Closes PR kern/43683.
1.20	22-Jul-2013	soren	Since puc(4) serial ports are typically not identified in the BIOS COM[1234] table, the I/O address must be manually set using installboot(8) in order to enable a serial console. Print the address here so the user doesn't have to dig through PCI configuration space to find it. Enable suspend/resume hooks.
1.19	26-Nov-2009	njoly	branches: 1.19.12; 1.19.22; 1.19.26; 1.19.34; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.18	14-Mar-2008	cube	branches: 1.18.4; Split device_t and softc for all com(4) devices (well, everything that uses a com_softc backend). Use proper types and ansify where appropriate.
1.17	19-Oct-2007	ad	branches: 1.17.12; 1.17.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	16-Nov-2006	christos	branches: 1.16.8; 1.16.22; 1.16.24; 1.16.28; __unused removal on arguments; approved by core.
1.15	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.14	13-Jul-2006	gdamore	branches: 1.14.4; 1.14.6; Add an option COM_REGMAP to allow com(4) to use an array of register indices. This allows us to convert aucom to just another com attachment, and cleanup some code in the com_arbus.c. Additionally, we use a common com_cleanup routine rather than having a zillion copies of it in the attachment points. This has been tested on a number architectures, and it has been shown to get close to comparable performance when COM_REGMAP is defined, and comparable when it is not defined. Approved by core@. Fixes PR port-evbmips/32362.
1.13	11-Dec-2005	christos	branches: 1.13.4; 1.13.8; 1.13.16; merge ktrace-lwp.
1.12	28-Jun-2005	thorpej	branches: 1.12.2; Use ANSI function decls and static.
1.11	04-Feb-2005	perry	de-__P
1.10	03-Feb-2004	fredb	branches: 1.10.8; 1.10.10; Fix a warning about possibly uninitialized variable.
1.9	03-Feb-2004	fredb	Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.8	31-Jan-2003	thorpej	branches: 1.8.2; Use aprint_*().
1.7	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.6	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.5	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.4	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.3	13-Nov-2001	lukem	add RCSID
1.2	16-Mar-2001	bouyer	branches: 1.2.2; Ops, forgot to commit this file along with the "variable com freq" support. Thanks to Bill Sommerfeld for pointing this out.
1.1	26-Jun-1998	cgd	branches: 1.1.14; 1.1.24; 1.1.26; add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.1.26.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.26.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.26.2	14-Nov-2001	nathanw	Catch up to -current.
1.1.26.1	09-Apr-2001	nathanw	Catch up with -current.
1.1.24.1	22-Mar-2001	he	Pull up revision 1.2 (requested by bouyer): Allow attaching ``com'' devices at frequency other than the standard COM_FREQ.
1.1.14.1	27-Mar-2001	bouyer	Sync with HEAD.
1.2.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.8.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.8.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.8.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.8.2.1	03-Aug-2004	skrll	Sync with HEAD
1.10.10.1	12-Feb-2005	yamt	sync with head.
1.10.8.1	29-Apr-2005	kent	sync with -current
1.12.2.3	17-Mar-2008	yamt	sync with head.
1.12.2.2	27-Oct-2007	yamt	sync with head.
1.12.2.1	30-Dec-2006	yamt	sync with head.
1.13.16.3	17-Jun-2006	gdamore	Undo the undo. Restore COM_INIT_REGS.
1.13.16.2	17-Jun-2006	gdamore	Revert COM_INIT_REGS.
1.13.16.1	15-Jun-2006	gdamore	Adapt to new com framework.
1.13.8.1	11-Aug-2006	yamt	sync with head
1.13.4.1	09-Sep-2006	rpaulo	sync with head
1.14.6.2	10-Dec-2006	yamt	sync with head.
1.14.6.1	22-Oct-2006	yamt	sync with head
1.14.4.1	18-Nov-2006	ad	Sync with head.
1.16.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.24.2	23-Mar-2008	matt	sync with HEAD
1.16.24.1	06-Nov-2007	matt	sync with HEAD
1.16.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.8.1	23-Oct-2007	ad	Sync with head.
1.17.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.17.12.1	24-Mar-2008	keiichi	sync with head.
1.18.4.1	11-Mar-2010	yamt	sync with head
1.19.34.1	23-Jul-2013	riastradh	sync with HEAD
1.19.26.2	18-May-2014	rmind	sync with head
1.19.26.1	28-Aug-2013	rmind	sync with head
1.19.22.2	03-Dec-2017	jdolecek	update from HEAD
1.19.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.19.12.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.22.2.1	10-Aug-2014	tls	Rebase.
1.23.18.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.23.4.1	28-Aug-2017	skrll	Sync with HEAD
1.24.10.1	10-Jun-2019	christos	Sync with HEAD
1.24.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.26.14.1	03-Apr-2021	thorpej	Sync with HEAD.
1.26.12.1	03-Apr-2021	thorpej	Sync with HEAD.
1.21	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.20	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.19	24-Apr-2021	thorpej	branches: 1.19.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.18	23-Dec-2019	thorpej	branches: 1.18.10; No need to use I2C_F_POLL here.
1.17	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.16	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.15	06-Jun-2018	maya	branches: 1.15.2; Remove duplicate ;
1.14	01-Jun-2017	chs	branches: 1.14.8; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.13	29-Mar-2014	christos	branches: 1.13.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.12	16-Oct-2013	christos	remove unused variables.
1.11	29-Oct-2012	chs	branches: 1.11.2; remove an unnecessary cast. use the standard name for the third arg to attach. adjust some whitespace.
1.10	30-Jan-2012	drochner	branches: 1.10.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.9	29-Aug-2011	jmcneill	branches: 1.9.2; 1.9.6; have pci register itself as a module, and make PCI driver modules depend on it
1.8	09-Aug-2011	jmcneill	modify the dtv device api so hardware drivers can be loaded independently of the dtv module
1.7	06-Aug-2011	jmcneill	build fix: forgot to add sc_board to softc
1.6	06-Aug-2011	jmcneill	print devname not subdevname on attach, and report the board name to dtv subsystem instead of generic CX23885
1.5	06-Aug-2011	jmcneill	don't attach iic@coram unless CORAM_ATTACH_I2C is defined
1.4	05-Aug-2011	jmcneill	hide debug messages under CORAM_DEBUG
1.3	05-Aug-2011	jmcneill	return status from demod instead of 0, and don't ignore errors when setting demod parameters
1.2	04-Aug-2011	jmcneill	modularize coram and add detach + childdet methods
1.1	04-Aug-2011	jakllsch	Add coram(4). A driver for CX23885-based TV cards, such as the Hauppauge WinTV HVR-1250 "Coram" board. Currently only supports ATSC 8VSB reception.
1.9.6.1	18-Feb-2012	mrg	merge to -current.
1.9.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.9.2.2	30-Oct-2012	yamt	sync with head
1.9.2.1	17-Apr-2012	yamt	sync with head
1.10.6.3	03-Dec-2017	jdolecek	update from HEAD
1.10.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.10.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.11.2.1	18-May-2014	rmind	sync with head
1.13.6.1	28-Aug-2017	skrll	Sync with HEAD
1.14.8.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.14.8.1	25-Jun-2018	pgoyette	Sync with HEAD
1.15.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.15.2.1	10-Jun-2019	christos	Sync with HEAD
1.18.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.19.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.4	09-Aug-2011	jmcneill	modify the dtv device api so hardware drivers can be loaded independently of the dtv module
1.3	06-Aug-2011	jmcneill	build fix: forgot to add sc_board to softc
1.2	04-Aug-2011	jmcneill	modularize coram and add detach + childdet methods
1.1	04-Aug-2011	jakllsch	Add coram(4). A driver for CX23885-based TV cards, such as the Hauppauge WinTV HVR-1250 "Coram" board. Currently only supports ATSC 8VSB reception.
1.77	08-Feb-2024	andvar	fix misplaced or missing "e" in words with "ment" ending (argument, implement, increment, decrement, alignment), in comments, documentation, log messages.
1.76	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.75	24-Dec-2022	andvar	s/reqest/request/, s/requst/request/ and s/reuqest/request/ in comments.
1.74	09-Aug-2021	andvar	s/fist/first/
1.73	03-Feb-2021	isaki	Remove code no longer used. The code that used sc->{halt_input,halt_output} function pointer was removed in 2004. (see cs428x.c rev 1.7)
1.72	08-May-2019	isaki	branches: 1.72.10; Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.71	16-Mar-2019	isaki	branches: 1.71.2; Use C99 style struct initializer to audio_hw_if.
1.70	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.69	07-Jul-2016	msaitoh	branches: 1.69.16; 1.69.18; KNF. Remove extra spaces. No functional change.
1.68	21-Sep-2014	christos	branches: 1.68.2; remove more old pci_findvendor code
1.67	21-Sep-2014	christos	Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.66	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.65	27-Oct-2012	chs	branches: 1.65.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.64	30-Jan-2012	drochner	branches: 1.64.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.63	24-Nov-2011	mrg	branches: 1.63.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.62	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.61	04-Apr-2011	dyoung	branches: 1.61.4; 1.61.6; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.60	25-May-2010	pgoyette	branches: 1.60.2; Revert earlier changes to callers of pci_find{vendor,product} since the original routine entrypoint name has been restored. Requested by dyoung@ and mrg@
1.59	24-May-2010	pgoyette	Update all callers of the pci_find{vendor,product} routines to now call these routines through their global pointers.
1.58	22-May-2010	pgoyette	When PCIVERBOSE is not included, pci_findvendor() and pci_findproduct() can return NULL. Defend against this and print the values in hex if we can't decode them properly.
1.57	24-Feb-2010	dyoung	branches: 1.57.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.56	08-Jan-2010	dyoung	branches: 1.56.2; Expand PMF_FN_* macros.
1.55	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.54	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.53	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.52	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.51	21-Mar-2008	dyoung	branches: 1.51.4; 1.51.16; 1.51.18; Use aprint_*_dev() instead of accessing struct device member dv_xname directly.
1.50	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.49	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.48	09-Dec-2007	jmcneill	branches: 1.48.6; 1.48.10; Merge jmcneill-pm branch.
1.47	19-Oct-2007	ad	branches: 1.47.4; 1.47.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.46	15-Jun-2007	joerg	branches: 1.46.6; 1.46.8; 1.46.12; Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.45	16-Nov-2006	christos	branches: 1.45.2; 1.45.6; 1.45.8; 1.45.10; 1.45.16; __unused removal on arguments; approved by core.
1.44	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.43	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.42	29-Aug-2006	christos	branches: 1.42.2; 1.42.4; fix incomplete initializer
1.41	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.40	06-Aug-2006	jmcneill	PR/12843: Workaround for CS4280 on IBM ThinkPads. From Stephane Witzmann. Note: This workaround conflicts with piixpm(4); you will need to disable the piixpm driver if you need this fix.
1.39	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.38	15-Apr-2006	jmcneill	branches: 1.38.2; 1.38.4; Add support for the Turtle Beach Santa Cruz. Special thanks to Stephane Witzmann for spending so much time debugging this. Fixes: kern/18880, kern/24488, port-i386/14260, kern/12603, kern/12723, kern/24957, kern/23584 Tested on i386 (Stephane Witzmann), alpha (me).
1.37	29-Jan-2006	dsl	branches: 1.37.2; 1.37.4; 1.37.6; 1.37.8; 1.37.10; Make almost everything #include <sys/bswap.h> instead of <machine/bswap.h> The bswap.h and endian.h files are all rather incestuous, but I want to get the constant folding stuff into one place - sys/bswap.h
1.36	11-Dec-2005	christos	branches: 1.36.2; merge ktrace-lwp.
1.35	28-Jun-2005	thorpej	branches: 1.35.2; Use ANSI function decls and static.
1.34	15-Jan-2005	kent	ansify and KNF
1.33	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.32	09-Nov-2004	kent	branches: 1.32.2; remove mixer setting code in *_attach() because ac97_attach() does it.
1.31	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.30	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.29	05-Aug-2004	drochner	simplify some pointer arithmetics which uses typecasts on the LHS so that gcc-3.4 doesn't complain
1.28	09-Jul-2004	mycroft	Attempt to make these safer against stray interrupts.
1.27	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.26	03-May-2003	wiz	branches: 1.26.2; 1.26.4; DMA, not dma nor Dma.
1.25	31-Jan-2003	thorpej	Use aprint_*().
1.24	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.23	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.22	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.21	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.20	31-May-2002	augustss	Some casts in printf().
1.19	15-Nov-2001	lukem	branches: 1.19.8; don't need <sys/types.h> when including <sys/param.h>
1.18	13-Nov-2001	lukem	add RCSID
1.17	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.16	18-Apr-2001	tacha	branches: 1.16.2; 1.16.4; rewrite suspend/resume routine. If you suspend while playing audio, clcs and clct continue playing when resume.
1.15	13-Feb-2001	tacha	branches: 1.15.2; Check and set PCI power management register. clcs now works when warm boot from Windows.
1.14	07-Feb-2001	tacha	Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.13	27-Jan-2001	augustss	Sleep a little when closing the midi device to allow the UART to drain.
1.12	18-Jan-2001	jdolecek	constify
1.11	15-Jan-2001	perry	powerhook fixes, from thorpej, tested by me. Thinkpad T20 sound now perfect.
1.10	15-Jan-2001	perry	Fixes from OpenBSD via chopps. Audio now works on my Thinkpad T20 provided I don't suspend the thing...
1.9	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.8	26-Nov-2000	takemura	Add new powerhook argument values, PWR_SOFTSUSPEND, PWR_SOFTSTANDBY and PWR_SOFTRESUME. Apm calls powerhook with the values in normal interrupt priority level while others are protected with splhigh().
1.7	20-Sep-2000	augustss	branches: 1.7.2; Apply a patch from the driver author (originally from OpenBSD).
1.6	19-Jul-2000	augustss	Add MIDI support. From Tatoku Ogaito.
1.5	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.4	15-May-2000	thorpej	branches: 1.4.4; Use <dev/ic/ac97reg.h>
1.3	15-May-2000	thorpej	ac97.h -> ac97var.h
1.2	14-Jan-2000	augustss	Apply patches from Tatoku Ogaito (the author of the code).
1.1	13-Dec-1999	augustss	branches: 1.1.2; Add a device driver for the Cirrus Logic CS4280 sound chip. The code has been contributed by Tatoku Ogaito.
1.1.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.4.4.8	06-May-2001	he	Pull up revisions 1.7-1.8, 1.10-1.14, 1.16 (via patch, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.4.4.7	13-Mar-2001	he	Apply patch (requested by tacha): Correct checking of AC97 Input Slot Valid register bits.
1.4.4.6	26-Feb-2001	he	Pull up revision 1.15 (requested by tacha): Fix problem when warm booting from Windows.
1.4.4.5	26-Feb-2001	he	Back out previous two pullups, they have "issues". I must be getting rusty...
1.4.4.4	26-Feb-2001	he	Apply patch (requested by he): Typo correction.
1.4.4.3	26-Feb-2001	he	Pull up revisions 1.7,1.10-1.11 (via patch, requested by perry): Fixes so that this audio driver should work (hardware as in e.g. IBM ThinkPad T20).
1.4.4.2	19-Jul-2000	augustss	Pull up 1.5-1.6 change: Add MIDI support. Approved by thorpej.
1.4.4.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.7.2.8	21-Apr-2001	bouyer	Sync with HEAD
1.7.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.7.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.7.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.7.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.7.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.7.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.1	20-Sep-2000	bouyer	file cs4280.c was added on branch thorpej_scsipi on 2000-11-20 11:42:15 +0000
1.15.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.15.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.15.2.5	20-Jun-2002	nathanw	Catch up to -current.
1.15.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.15.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.15.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.15.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.16.4.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.16.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.16.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.16.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.8.1	20-Jun-2002	gehenna	catch up with -current.
1.26.4.1	22-Sep-2004	jmc	Pullup rev 1.30 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.26.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.26.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.26.2.7	14-Nov-2004	skrll	Sync with HEAD.
1.26.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.26.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.26.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.26.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.26.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.26.2.1	03-Aug-2004	skrll	Sync with HEAD
1.32.2.3	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.32.2.2	03-Jan-2005	kent	* remove stream_filter_list_append and stream_filter_list_prepend * add stream_filter_list_t::{append(), prepend(), set()}
1.32.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.35.2.7	24-Mar-2008	yamt	sync with head.
1.35.2.6	17-Mar-2008	yamt	sync with head.
1.35.2.5	21-Jan-2008	yamt	sync with head
1.35.2.4	27-Oct-2007	yamt	sync with head.
1.35.2.3	03-Sep-2007	yamt	sync with head.
1.35.2.2	30-Dec-2006	yamt	sync with head.
1.35.2.1	21-Jun-2006	yamt	sync with head.
1.36.2.1	01-Feb-2006	yamt	sync with head.
1.37.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.37.8.1	19-Apr-2006	elad	sync with head.
1.37.6.4	03-Sep-2006	yamt	sync with head.
1.37.6.3	11-Aug-2006	yamt	sync with head
1.37.6.2	26-Jun-2006	yamt	sync with head.
1.37.6.1	24-May-2006	yamt	sync with head.
1.37.4.1	22-Apr-2006	simonb	Sync with head.
1.37.2.1	09-Sep-2006	rpaulo	sync with head
1.38.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.38.2.1	19-Jun-2006	chap	Sync with head.
1.42.4.2	10-Dec-2006	yamt	sync with head.
1.42.4.1	22-Oct-2006	yamt	sync with head
1.42.2.1	18-Nov-2006	ad	Sync with head.
1.45.16.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.45.10.1	11-Jul-2007	mjf	Sync with head.
1.45.8.2	23-Oct-2007	ad	Sync with head.
1.45.8.1	15-Jul-2007	ad	Sync with head.
1.45.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.45.2.1	18-Jun-2007	liamjfoy	Pull up following revision(s) (requested by joerg in ticket #733): sys/dev/pci/cs4280.c: revision 1.46 sys/dev/pci/cs4281.c: revision 1.34 sys/dev/pci/cs428x.h: revision 1.13 Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.46.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.46.8.3	23-Mar-2008	matt	sync with HEAD
1.46.8.2	09-Jan-2008	matt	sync with HEAD
1.46.8.1	06-Nov-2007	matt	sync with HEAD
1.46.6.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.46.6.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.46.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.46.6.1	04-Oct-2007	joerg	Convert clcs(4) and clct(4) to PNP power management. This should also allow more then one instance of each to resume correctly.
1.47.6.1	11-Dec-2007	yamt	sync with head.
1.47.4.1	26-Dec-2007	ad	Sync with head.
1.48.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.48.6.1	24-Mar-2008	keiichi	sync with head.
1.51.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.51.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.51.16.1	11-Dec-2008	ad	Checkpoint work in progress.
1.51.4.3	11-Aug-2010	yamt	sync with head.
1.51.4.2	11-Mar-2010	yamt	sync with head
1.51.4.1	16-May-2009	yamt	sync with head
1.56.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.56.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.57.2.2	21-Apr-2011	rmind	sync with head
1.57.2.1	30-May-2010	rmind	sync with head
1.60.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.61.6.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.61.6.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.61.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.61.4.2	30-Oct-2012	yamt	sync with head
1.61.4.1	17-Apr-2012	yamt	sync with head
1.63.2.1	18-Feb-2012	mrg	merge to -current.
1.64.6.3	03-Dec-2017	jdolecek	update from HEAD
1.64.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.64.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.65.2.1	18-May-2014	rmind	sync with head
1.68.2.1	09-Jul-2016	skrll	Sync with HEAD
1.69.18.1	10-Jun-2019	christos	Sync with HEAD
1.69.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.71.2.3	05-May-2019	isaki	Remove encoding conversions on recording. These are handled in the upper layer now.
1.71.2.2	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.71.2.1	21-Apr-2019	isaki	Adapt cs428x families to audio2. - recording on cs4280 seems to have its own conversion. I will see it later but first make it compilable.
1.72.10.1	03-Apr-2021	thorpej	Sync with HEAD.
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	28-Jun-2005	thorpej	Use ANSI function decls and static.
1.2	18-Jan-2001	jdolecek	branches: 1.2.24; constify
1.1	13-Dec-1999	augustss	branches: 1.1.2; 1.1.8; Add a device driver for the Cirrus Logic CS4280 sound chip. The code has been contributed by Tatoku Ogaito.
1.1.8.3	11-Feb-2001	bouyer	Sync with HEAD.
1.1.8.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.8.1	13-Dec-1999	bouyer	file cs4280_image.h was added on branch thorpej_scsipi on 2000-11-20 11:42:16 +0000
1.1.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.2.24.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9	08-Feb-2024	andvar	fix misplaced or missing "e" in words with "ment" ending (argument, implement, increment, decrement, alignment), in comments, documentation, log messages.
1.8	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.7	15-Apr-2006	jmcneill	branches: 1.7.152; 1.7.156; Add support for the Turtle Beach Santa Cruz. Special thanks to Stephane Witzmann for spending so much time debugging this. Fixes: kern/18880, kern/24488, port-i386/14260, kern/12603, kern/12723, kern/24957, kern/23584 Tested on i386 (Stephane Witzmann), alpha (me).
1.6	11-Dec-2005	christos	branches: 1.6.4; 1.6.6; 1.6.8; 1.6.10; 1.6.12; merge ktrace-lwp.
1.5	15-Jan-2005	kent	branches: 1.5.10; ansify and KNF
1.4	07-Feb-2001	tacha	branches: 1.4.24; Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.3	15-May-2000	thorpej	branches: 1.3.6; Use <dev/ic/ac97reg.h>
1.2	14-Jan-2000	augustss	Apply patches from Tatoku Ogaito (the author of the code).
1.1	13-Dec-1999	augustss	branches: 1.1.2; Add a device driver for the Cirrus Logic CS4280 sound chip. The code has been contributed by Tatoku Ogaito.
1.1.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.3.6.3	11-Feb-2001	bouyer	Sync with HEAD.
1.3.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.6.1	15-May-2000	bouyer	file cs4280reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:16 +0000
1.4.24.1	17-Jan-2005	skrll	Sync with HEAD.
1.5.10.1	21-Jun-2006	yamt	sync with head.
1.6.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.6.10.1	19-Apr-2006	elad	sync with head.
1.6.8.1	24-May-2006	yamt	sync with head.
1.6.6.1	22-Apr-2006	simonb	Sync with head.
1.6.4.1	09-Sep-2006	rpaulo	sync with head
1.7.156.1	05-May-2019	isaki	Remove encoding conversions on recording. These are handled in the upper layer now.
1.7.152.1	10-Jun-2019	christos	Sync with HEAD
1.61	07-Jan-2025	andvar	Fix typos in the word 'Control' in comments.
1.60	14-Sep-2024	andvar	s/intterupt/interrupt/ in comments.
1.59	20-Dec-2023	thorpej	branches: 1.59.2; Remove unnecessary <sys/malloc.h>.
1.58	29-Mar-2022	riastradh	cs4281(4): Fix lock ordering in suspend. No idea if this code works -- obviously this path has never been tested in the >decade it's been here!
1.57	03-Feb-2021	isaki	Remove code no longer used. The code that used sc->{halt_input,halt_output} function pointer was removed in 2004. (see cs428x.c rev 1.7)
1.56	08-May-2020	simonb	branches: 1.56.2; Whitespace nit.
1.55	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.54	16-Mar-2019	isaki	branches: 1.54.2; Use C99 style struct initializer to audio_hw_if.
1.53	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.52	07-Jul-2016	msaitoh	branches: 1.52.16; 1.52.18; KNF. Remove extra spaces. No functional change.
1.51	29-Mar-2014	christos	branches: 1.51.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.50	16-Oct-2013	christos	use __USE()
1.49	16-Oct-2013	christos	use variables that are only used in debugging.
1.48	27-Oct-2012	chs	branches: 1.48.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.47	30-Jan-2012	drochner	branches: 1.47.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.46	24-Nov-2011	mrg	branches: 1.46.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.45	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.44	24-Feb-2010	dyoung	branches: 1.44.10; 1.44.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.43	08-Jan-2010	dyoung	branches: 1.43.2; Expand PMF_FN_* macros.
1.42	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.41	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.40	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.39	21-Mar-2008	dyoung	branches: 1.39.4; 1.39.16; 1.39.18; Use aprint_*_dev() instead of accessing struct device member dv_xname directly.
1.38	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.37	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.36	09-Dec-2007	jmcneill	branches: 1.36.6; 1.36.10; Merge jmcneill-pm branch.
1.35	19-Oct-2007	ad	branches: 1.35.4; 1.35.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.34	15-Jun-2007	joerg	branches: 1.34.6; 1.34.8; 1.34.12; Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.33	16-Nov-2006	christos	branches: 1.33.2; 1.33.6; 1.33.8; 1.33.10; 1.33.16; __unused removal on arguments; approved by core.
1.32	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.31	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.30	29-Aug-2006	christos	branches: 1.30.2; 1.30.4; fix incomplete initializer
1.29	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.28	15-Apr-2006	jmcneill	branches: 1.28.2; 1.28.4; Add support for the Turtle Beach Santa Cruz. Special thanks to Stephane Witzmann for spending so much time debugging this. Fixes: kern/18880, kern/24488, port-i386/14260, kern/12603, kern/12723, kern/24957, kern/23584 Tested on i386 (Stephane Witzmann), alpha (me).
1.27	11-Dec-2005	christos	branches: 1.27.4; 1.27.6; 1.27.8; 1.27.10; 1.27.12; merge ktrace-lwp.
1.26	28-Jun-2005	thorpej	branches: 1.26.2; Use ANSI function decls and static.
1.25	06-Jun-2005	yamt	cs4281_set_params: fix printf format.
1.24	27-Feb-2005	perry	nuke trailing whitespace
1.23	15-Jan-2005	kent	branches: 1.23.2; 1.23.4; ansify and KNF
1.22	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.21	02-Nov-2004	yamt	branches: 1.21.2; make recording of clct(4) work. namely, - cs4281_intr: advance a pointer after copying data. - cs4281_trigger_input: fix an inverted condition about dma mode.
1.20	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.19	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.18	09-Jul-2004	mycroft	Attempt to make these safer against stray interrupts.
1.17	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.16	03-May-2003	wiz	branches: 1.16.2; 1.16.4; DMA, not dma nor Dma.
1.15	31-Jan-2003	thorpej	Use aprint_*().
1.14	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.13	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	15-May-2002	simonb	Print some error messages when the codec fails to initialise (some errors were already reported, others weren't). Sprinkle some KNF around.
1.9	13-Dec-2001	tacha	remove bogus printf.
1.8	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.7	13-Nov-2001	lukem	add RCSID
1.6	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.5	16-Sep-2001	wiz	Spell 'occurred' with two 'r's.
1.4	18-Apr-2001	tacha	branches: 1.4.2; 1.4.4; 1.4.6; rewrite suspend/resume routine. If you suspend while playing audio, clcs and clct continue playing when resume.
1.3	07-Feb-2001	tacha	branches: 1.3.2; 1.3.4; Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.2	22-Jan-2001	augustss	Clean up a little.
1.1	22-Jan-2001	augustss	Add clct driver for Cirrus Logic CrystalClear PCI Audio CS4281. Driver written by Tatoku Ogaito.
1.3.4.8	29-Dec-2002	thorpej	Sync with HEAD.
1.3.4.7	18-Oct-2002	nathanw	Catch up to -current.
1.3.4.6	20-Jun-2002	nathanw	Catch up to -current.
1.3.4.5	08-Jan-2002	nathanw	Catch up to -current.
1.3.4.4	14-Nov-2001	nathanw	Catch up to -current.
1.3.4.3	08-Oct-2001	nathanw	Catch up to -current.
1.3.4.2	21-Sep-2001	nathanw	Catch up to -current.
1.3.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.3.2.3	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.3.2.1	07-Feb-2001	bouyer	file cs4281.c was added on branch thorpej_scsipi on 2001-02-11 19:15:50 +0000
1.4.6.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.4.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.4.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.4.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.4.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.2.4	27-Dec-2001	he	Pull up revision 1.9 (requested by tacha): Remove spurious warning which is no longer valid.
1.4.2.3	29-Jul-2001	he	Apply patch (requested by minoura): Correct argument to pci_intr_map(). Fixes PR#13508.
1.4.2.2	06-May-2001	he	Pull up revisions 1.1-1.4 (new, via patch, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.4.2.1	18-Apr-2001	he	file cs4281.c was added on branch netbsd-1-5 on 2001-05-06 15:12:08 +0000
1.16.4.1	22-Sep-2004	jmc	branches: 1.16.4.1.2; Pullup rev 1.19 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.16.4.1.2.1	30-Jan-2005	he	Pull up revision 1.21 (requested by yamt in ticket #967): Make recording of clct(4) work, namely: o cs_4281_intr: advance a pointer after copying data o cs_4281_trigger_input: fix an inverted condition about dma mode
1.16.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.16.2.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.16.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.16.2.6	14-Nov-2004	skrll	Sync with HEAD.
1.16.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.16.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.16.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.16.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.16.2.1	03-Aug-2004	skrll	Sync with HEAD
1.21.2.3	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.21.2.2	03-Jan-2005	kent	* remove stream_filter_list_append and stream_filter_list_prepend * add stream_filter_list_t::{append(), prepend(), set()}
1.21.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.23.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.23.2.1	29-Apr-2005	kent	sync with -current
1.26.2.7	24-Mar-2008	yamt	sync with head.
1.26.2.6	17-Mar-2008	yamt	sync with head.
1.26.2.5	21-Jan-2008	yamt	sync with head
1.26.2.4	27-Oct-2007	yamt	sync with head.
1.26.2.3	03-Sep-2007	yamt	sync with head.
1.26.2.2	30-Dec-2006	yamt	sync with head.
1.26.2.1	21-Jun-2006	yamt	sync with head.
1.27.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.27.10.1	19-Apr-2006	elad	sync with head.
1.27.8.3	03-Sep-2006	yamt	sync with head.
1.27.8.2	26-Jun-2006	yamt	sync with head.
1.27.8.1	24-May-2006	yamt	sync with head.
1.27.6.1	22-Apr-2006	simonb	Sync with head.
1.27.4.1	09-Sep-2006	rpaulo	sync with head
1.28.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.28.2.1	19-Jun-2006	chap	Sync with head.
1.30.4.2	10-Dec-2006	yamt	sync with head.
1.30.4.1	22-Oct-2006	yamt	sync with head
1.30.2.1	18-Nov-2006	ad	Sync with head.
1.33.16.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.33.10.1	11-Jul-2007	mjf	Sync with head.
1.33.8.2	23-Oct-2007	ad	Sync with head.
1.33.8.1	15-Jul-2007	ad	Sync with head.
1.33.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.33.2.1	18-Jun-2007	liamjfoy	Pull up following revision(s) (requested by joerg in ticket #733): sys/dev/pci/cs4280.c: revision 1.46 sys/dev/pci/cs4281.c: revision 1.34 sys/dev/pci/cs428x.h: revision 1.13 Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.34.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.34.8.3	23-Mar-2008	matt	sync with HEAD
1.34.8.2	09-Jan-2008	matt	sync with HEAD
1.34.8.1	06-Nov-2007	matt	sync with HEAD
1.34.6.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.34.6.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.34.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.34.6.1	04-Oct-2007	joerg	Convert clcs(4) and clct(4) to PNP power management. This should also allow more then one instance of each to resume correctly.
1.35.6.1	11-Dec-2007	yamt	sync with head.
1.35.4.1	26-Dec-2007	ad	Sync with head.
1.36.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.36.6.1	24-Mar-2008	keiichi	sync with head.
1.39.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.39.16.1	12-Dec-2008	ad	Checkpoint work in progress.
1.39.4.2	11-Mar-2010	yamt	sync with head
1.39.4.1	16-May-2009	yamt	sync with head
1.43.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.44.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.44.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.44.10.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.44.10.2	30-Oct-2012	yamt	sync with head
1.44.10.1	17-Apr-2012	yamt	sync with head
1.46.2.1	18-Feb-2012	mrg	merge to -current.
1.47.6.3	03-Dec-2017	jdolecek	update from HEAD
1.47.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.47.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.48.2.1	18-May-2014	rmind	sync with head
1.51.6.1	09-Jul-2016	skrll	Sync with HEAD
1.52.18.1	10-Jun-2019	christos	Sync with HEAD
1.52.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.54.2.2	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.54.2.1	21-Apr-2019	isaki	Adapt cs428x families to audio2. - recording on cs4280 seems to have its own conversion. I will see it later but first make it compilable.
1.56.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.59.2.1	02-Aug-2025	perseant	Sync with HEAD
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	15-Jan-2005	kent	branches: 1.4.2; 1.4.4; ansify and KNF
1.3	07-Feb-2001	tacha	branches: 1.3.2; 1.3.6; 1.3.28; Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.2	22-Jan-2001	augustss	Clean up a little.
1.1	22-Jan-2001	augustss	Add clct driver for Cirrus Logic CrystalClear PCI Audio CS4281. Driver written by Tatoku Ogaito.
1.3.28.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.3.28.1	17-Jan-2005	skrll	Sync with HEAD.
1.3.6.2	06-May-2001	he	Pull up revisions 1.1-1.3 (new, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.3.6.1	07-Feb-2001	he	file cs4281reg.h was added on branch netbsd-1-5 on 2001-05-06 15:12:25 +0000
1.3.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.3.2.1	07-Feb-2001	bouyer	file cs4281reg.h was added on branch thorpej_scsipi on 2001-02-11 19:15:50 +0000
1.4.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.2.1	29-Apr-2005	kent	sync with -current
1.20	08-Jun-2019	isaki	Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.19	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.18	01-Jun-2017	chs	branches: 1.18.10; 1.18.14; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.17	27-Oct-2012	chs	branches: 1.17.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.16	23-Nov-2011	jmcneill	branches: 1.16.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.15	10-Apr-2008	cegger	branches: 1.15.16; 1.15.38; 1.15.40; use aprint_*_dev and device_xname
1.14	19-Oct-2007	ad	branches: 1.14.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.6; 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	15-Apr-2006	jmcneill	branches: 1.11.8; 1.11.10; Add support for the Turtle Beach Santa Cruz. Special thanks to Stephane Witzmann for spending so much time debugging this. Fixes: kern/18880, kern/24488, port-i386/14260, kern/12603, kern/12723, kern/24957, kern/23584 Tested on i386 (Stephane Witzmann), alpha (me).
1.10	11-Dec-2005	christos	branches: 1.10.4; 1.10.6; 1.10.8; 1.10.10; 1.10.12; merge ktrace-lwp.
1.9	15-Jan-2005	kent	branches: 1.9.10; ansify and KNF
1.8	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.7	09-Jul-2004	mycroft	branches: 1.7.2; There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.6	03-May-2003	wiz	branches: 1.6.2; DMA, not dma nor Dma.
1.5	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.4	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.3	13-Nov-2001	lukem	add RCSID
1.2	18-Apr-2001	tacha	branches: 1.2.2; 1.2.4; rewrite suspend/resume routine. If you suspend while playing audio, clcs and clct continue playing when resume.
1.1	07-Feb-2001	tacha	branches: 1.1.2; 1.1.4; Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.1.4.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.4.2	14-Nov-2001	nathanw	Catch up to -current.
1.1.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.3	21-Apr-2001	bouyer	Sync with HEAD
1.1.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.2.1	07-Feb-2001	bouyer	file cs428x.c was added on branch thorpej_scsipi on 2001-02-11 19:15:52 +0000
1.2.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.2	06-May-2001	he	Pull up revisions 1.1-1.2 (new, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.2.2.1	18-Apr-2001	he	file cs428x.c was added on branch netbsd-1-5 on 2001-05-06 15:11:14 +0000
1.6.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.6.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.9.10.3	27-Oct-2007	yamt	sync with head.
1.9.10.2	30-Dec-2006	yamt	sync with head.
1.9.10.1	21-Jun-2006	yamt	sync with head.
1.10.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.10.10.1	19-Apr-2006	elad	sync with head.
1.10.8.1	24-May-2006	yamt	sync with head.
1.10.6.1	22-Apr-2006	simonb	Sync with head.
1.10.4.1	09-Sep-2006	rpaulo	sync with head
1.11.10.2	10-Dec-2006	yamt	sync with head.
1.11.10.1	22-Oct-2006	yamt	sync with head
1.11.8.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.13.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.14.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.15.40.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.15.38.2	30-Oct-2012	yamt	sync with head
1.15.38.1	17-Apr-2012	yamt	sync with head
1.15.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.15.16.1	11-Dec-2008	ad	Checkpoint work in progress.
1.16.8.2	03-Dec-2017	jdolecek	update from HEAD
1.16.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.17.14.1	28-Aug-2017	skrll	Sync with HEAD
1.18.14.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.18.14.2	03-May-2019	isaki	Remove dev/auconv.[ch]. Now it is handled in dev/audio/audio.c and dev/audio/linear.c. Remove dev/mulaw.h. It already have been just a link to dev/audio/mulaw.h in this branch.
1.18.14.1	21-Apr-2019	isaki	Adapt cs428x families to audio2. - recording on cs4280 seems to have its own conversion. I will see it later but first make it compilable.
1.18.10.1	10-Jun-2019	christos	Sync with HEAD
1.18	03-Feb-2021	isaki	Remove code no longer used. The code that used sc->{halt_input,halt_output} function pointer was removed in 2004. (see cs428x.c rev 1.7)
1.17	08-May-2019	isaki	branches: 1.17.10; Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	27-Oct-2012	chs	branches: 1.16.38; 1.16.42; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.15	23-Nov-2011	jmcneill	branches: 1.15.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.14	09-Dec-2007	jmcneill	branches: 1.14.26; 1.14.48; 1.14.50; Merge jmcneill-pm branch.
1.13	15-Jun-2007	joerg	branches: 1.13.6; 1.13.8; 1.13.16; 1.13.18; Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.12	04-Mar-2007	christos	branches: 1.12.2; 1.12.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.11	06-Aug-2006	jmcneill	branches: 1.11.8; 1.11.10; 1.11.12; 1.11.14; PR/12843: Workaround for CS4280 on IBM ThinkPads. From Stephane Witzmann. Note: This workaround conflicts with piixpm(4); you will need to disable the piixpm driver if you need this fix.
1.10	15-Apr-2006	jmcneill	Add support for the Turtle Beach Santa Cruz. Special thanks to Stephane Witzmann for spending so much time debugging this. Fixes: kern/18880, kern/24488, port-i386/14260, kern/12603, kern/12723, kern/24957, kern/23584 Tested on i386 (Stephane Witzmann), alpha (me).
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; 1.9.12; merge ktrace-lwp.
1.8	15-Jan-2005	kent	branches: 1.8.10; ansify and KNF
1.7	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.6	03-May-2003	wiz	branches: 1.6.2; 1.6.6; DMA, not dma nor Dma.
1.5	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.4	18-Apr-2001	tacha	branches: 1.4.2; rewrite suspend/resume routine. If you suspend while playing audio, clcs and clct continue playing when resume.
1.3	07-Feb-2001	tacha	branches: 1.3.2; 1.3.4; Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.2	22-Jan-2001	augustss	Clean up a little.
1.1	22-Jan-2001	augustss	Add clct driver for Cirrus Logic CrystalClear PCI Audio CS4281. Driver written by Tatoku Ogaito.
1.3.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.3.2.3	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.3.2.1	07-Feb-2001	bouyer	file cs428x.h was added on branch thorpej_scsipi on 2001-02-11 19:15:52 +0000
1.4.2.2	06-May-2001	he	Pull up revisions 1.1-1.4 (new, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.4.2.1	18-Apr-2001	he	file cs428x.h was added on branch netbsd-1-5 on 2001-05-06 15:12:48 +0000
1.6.6.1	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.6.2.1	17-Jan-2005	skrll	Sync with HEAD.
1.8.10.4	21-Jan-2008	yamt	sync with head
1.8.10.3	03-Sep-2007	yamt	sync with head.
1.8.10.2	30-Dec-2006	yamt	sync with head.
1.8.10.1	21-Jun-2006	yamt	sync with head.
1.9.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.9.10.1	19-Apr-2006	elad	sync with head.
1.9.8.2	11-Aug-2006	yamt	sync with head
1.9.8.1	24-May-2006	yamt	sync with head.
1.9.6.1	22-Apr-2006	simonb	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.11.14.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.11.12.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.11.10.1	12-Mar-2007	rmind	Sync with HEAD.
1.11.8.1	18-Jun-2007	liamjfoy	Pull up following revision(s) (requested by joerg in ticket #733): sys/dev/pci/cs4280.c: revision 1.46 sys/dev/pci/cs4281.c: revision 1.34 sys/dev/pci/cs428x.h: revision 1.13 Fix ACPI suspend/resume support of clcs and clct. Based on auich and hints from jmcneill@.
1.12.4.1	11-Jul-2007	mjf	Sync with head.
1.12.2.1	15-Jul-2007	ad	Sync with head.
1.13.18.1	11-Dec-2007	yamt	sync with head.
1.13.16.1	26-Dec-2007	ad	Sync with head.
1.13.8.1	09-Jan-2008	matt	sync with HEAD
1.13.6.1	04-Oct-2007	joerg	Convert clcs(4) and clct(4) to PNP power management. This should also allow more then one instance of each to resume correctly.
1.14.50.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.14.48.2	30-Oct-2012	yamt	sync with head
1.14.48.1	17-Apr-2012	yamt	sync with head
1.14.26.2	12-Dec-2008	ad	Checkpoint work in progress.
1.14.26.1	11-Dec-2008	ad	Checkpoint work in progress.
1.15.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.16.42.2	05-May-2019	isaki	Remove encoding conversions on recording. These are handled in the upper layer now.
1.16.42.1	21-Apr-2019	isaki	Adapt cs428x families to audio2. - recording on cs4280 seems to have its own conversion. I will see it later but first make it compilable.
1.16.38.1	10-Jun-2019	christos	Sync with HEAD
1.17.10.1	03-Apr-2021	thorpej	Sync with HEAD.
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	15-Jan-2005	kent	ansify and KNF
1.2	22-Jan-2001	augustss	branches: 1.2.2; 1.2.6; 1.2.28; Clean up a little.
1.1	22-Jan-2001	augustss	Add clct driver for Cirrus Logic CrystalClear PCI Audio CS4281. Driver written by Tatoku Ogaito.
1.2.28.1	17-Jan-2005	skrll	Sync with HEAD.
1.2.6.2	06-May-2001	he	Pull up revisions 1.1-1.2 (new, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.2.6.1	22-Jan-2001	he	file cs428xreg.h was added on branch netbsd-1-5 on 2001-05-06 15:13:08 +0000
1.2.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.2.2.1	22-Jan-2001	bouyer	file cs428xreg.h was added on branch thorpej_scsipi on 2001-02-11 19:15:52 +0000
1.2	09-Aug-2011	jmcneill	add CLK_DELAY
1.1	04-Aug-2011	jakllsch	Add coram(4). A driver for CX23885-based TV cards, such as the Hauppauge WinTV HVR-1250 "Coram" board. Currently only supports ATSC 8VSB reception.
1.23	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.22	09-Feb-2024	andvar	s/discriptor/descriptor/ in comments.
1.21	10-May-2023	riastradh	cxdtv(4): Use config_detach_children.
1.20	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.19	24-Apr-2021	thorpej	branches: 1.19.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.18	23-Dec-2019	thorpej	branches: 1.18.10; No need to use I2C_F_POLL here.
1.17	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.16	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.15	01-Jun-2017	chs	branches: 1.15.8; 1.15.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.14	29-Mar-2014	christos	branches: 1.14.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.13	16-Oct-2013	christos	remove unused variables.
1.12	27-Oct-2012	chs	branches: 1.12.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.11	30-Jan-2012	drochner	branches: 1.11.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.10	16-Jan-2012	jmcneill	Only match supported boards.
1.9	26-Sep-2011	jakllsch	branches: 1.9.2; 1.9.6; Define and use symbolic bits in registers.
1.8	29-Aug-2011	jmcneill	have pci register itself as a module, and make PCI driver modules depend on it
1.7	09-Aug-2011	jmcneill	modify the dtv device api so hardware drivers can be loaded independently of the dtv module
1.6	25-Jul-2011	jakllsch	Remove unconditional debugging code. Also, a few other cleanups.
1.5	15-Jul-2011	jmcneill	use lg3303_get_snr, lg3303_get_signal_strength
1.4	15-Jul-2011	jmcneill	add pcHDTV HD-5500 support: cxdtv0 at pci0 dev 14 function 2: pcHDTV HD-5500 HDTV Video Capture (rev. 0x05) cxdtv0: interrupting at irq 10 cxdtv0: lg3303: found @ 0x59 cxdtv0: tvpll: LG TDVS-H06xF dtv0 at cxdtv0: CX23880 [ATSC]
1.3	15-Jul-2011	jmcneill	add detach support
1.2	14-Jul-2011	jmcneill	hook in lg3303 demod support and modularize
1.1	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.9.6.1	18-Feb-2012	mrg	merge to -current.
1.9.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.9.2.2	30-Oct-2012	yamt	sync with head
1.9.2.1	17-Apr-2012	yamt	sync with head
1.11.6.3	03-Dec-2017	jdolecek	update from HEAD
1.11.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.11.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.12.2.1	18-May-2014	rmind	sync with head
1.14.6.1	28-Aug-2017	skrll	Sync with HEAD
1.15.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.15.10.1	10-Jun-2019	christos	Sync with HEAD
1.15.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.18.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.19.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	14-Jul-2011	jmcneill	hook in lg3303 demod support and modularize
1.1	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.1	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.3	11-Jul-2015	kamil	Improve spelling: regsiter -> register This change is non-functional. Approved by <riastradh>, <pgoyette>
1.2	26-Sep-2011	jakllsch	branches: 1.2.12; 1.2.30; Define and use symbolic bits in registers.
1.1	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.2.30.1	22-Sep-2015	skrll	Sync with HEAD
1.2.12.1	03-Dec-2017	jdolecek	update from HEAD
1.4	09-Aug-2011	jmcneill	modify the dtv device api so hardware drivers can be loaded independently of the dtv module
1.3	25-Jul-2011	jakllsch	Remove unconditional debugging code. Also, a few other cleanups.
1.2	15-Jul-2011	jmcneill	add detach support
1.1	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.8	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.7	17-Jan-2008	jklos	branches: 1.7.10; 1.7.30; 1.7.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.6	19-Oct-2007	ad	branches: 1.6.2; 1.6.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	10-Jul-2007	jklos	branches: 1.5.2; 1.5.4; 1.5.10; 1.5.12; 1.5.14; 1.5.18; More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.4	09-Jul-2007	jklos	More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.3	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.2	05-Jul-2007	jklos	Updates of working copies (there will be many).
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.5.18.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.14.4	21-Jan-2008	yamt	sync with head
1.5.14.3	27-Oct-2007	yamt	sync with head.
1.5.14.2	03-Sep-2007	yamt	sync with head.
1.5.14.1	10-Jul-2007	yamt	file cxgb_adapter.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:28 +0000
1.5.12.2	23-Mar-2008	matt	sync with HEAD
1.5.12.1	06-Nov-2007	matt	sync with HEAD
1.5.10.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.4.3	23-Oct-2007	ad	Sync with head.
1.5.4.2	15-Jul-2007	ad	Sync with head.
1.5.4.1	10-Jul-2007	ad	file cxgb_adapter.h was added on branch vmlocking on 2007-07-15 13:21:22 +0000
1.5.2.2	11-Jul-2007	mjf	Sync with head.
1.5.2.1	10-Jul-2007	mjf	file cxgb_adapter.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:05 +0000
1.6.8.1	19-Jan-2008	bouyer	Sync with HEAD
1.6.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.7.32.1	30-May-2010	rmind	sync with head
1.7.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.7.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_ael1002.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:29 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_ael1002.c was added on branch vmlocking on 2007-07-15 13:21:22 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_ael1002.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:05 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.2	07-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.20; 1.2.26; Updating Chelsio files. Not quite compiling yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.2.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.2.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.14.3	21-Jan-2008	yamt	sync with head
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	07-Jul-2007	yamt	file cxgb_common.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:29 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	07-Jul-2007	ad	file cxgb_common.h was added on branch vmlocking on 2007-07-15 13:21:22 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	07-Jul-2007	mjf	file cxgb_common.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:06 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_config.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:30 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_config.h was added on branch vmlocking on 2007-07-15 13:21:23 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_config.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:07 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_ctl_defs.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:30 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_ctl_defs.h was added on branch vmlocking on 2007-07-15 13:21:23 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_ctl_defs.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:07 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_firmware_exports.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:30 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_firmware_exports.h was added on branch vmlocking on 2007-07-15 13:21:23 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_firmware_exports.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:08 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.7	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.6	17-Jan-2008	jklos	branches: 1.6.10; 1.6.30; 1.6.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.5	09-Jul-2007	jklos	branches: 1.5.2; 1.5.4; 1.5.12; 1.5.14; 1.5.20; 1.5.26; More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.4	08-Jul-2007	jklos	Updating more Chelsio files.
1.3	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.2	05-Jul-2007	jklos	Updates of working copies (there will be many).
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.5.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.5.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.5.14.3	21-Jan-2008	yamt	sync with head
1.5.14.2	03-Sep-2007	yamt	sync with head.
1.5.14.1	09-Jul-2007	yamt	file cxgb_include.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:31 +0000
1.5.12.1	23-Mar-2008	matt	sync with HEAD
1.5.4.2	15-Jul-2007	ad	Sync with head.
1.5.4.1	09-Jul-2007	ad	file cxgb_include.h was added on branch vmlocking on 2007-07-15 13:21:24 +0000
1.5.2.2	11-Jul-2007	mjf	Sync with head.
1.5.2.1	09-Jul-2007	mjf	file cxgb_include.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:08 +0000
1.6.32.1	30-May-2010	rmind	sync with head
1.6.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.6.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_ioctl.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:31 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_ioctl.h was added on branch vmlocking on 2007-07-15 13:21:24 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_ioctl.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:09 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	05-Jul-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	05-Jul-2007	yamt	file cxgb_jhash.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:32 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	05-Jul-2007	ad	file cxgb_jhash.h was added on branch vmlocking on 2007-07-15 13:21:24 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	05-Jul-2007	mjf	file cxgb_jhash.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:10 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.8	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.7	17-Jan-2008	jklos	branches: 1.7.10; 1.7.30; 1.7.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.6	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.5	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.4	26-Aug-2007	dyoung	branches: 1.4.2; 1.4.4; 1.4.10; 1.4.12; 1.4.14; 1.4.16; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.3	09-Jul-2007	jklos	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.10; More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.2	08-Jul-2007	jklos	Updating more Chelsio files.
1.1	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.3.10.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.3.6.1	03-Sep-2007	skrll	Sync with HEAD.
1.3.4.3	09-Oct-2007	ad	Sync with head.
1.3.4.2	15-Jul-2007	ad	Sync with head.
1.3.4.1	09-Jul-2007	ad	file cxgb_l2t.c was added on branch vmlocking on 2007-07-15 13:21:25 +0000
1.3.2.2	11-Jul-2007	mjf	Sync with head.
1.3.2.1	09-Jul-2007	mjf	file cxgb_l2t.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:10 +0000
1.4.16.3	19-Jan-2008	bouyer	Sync with HEAD
1.4.16.2	08-Jan-2008	bouyer	Sync with HEAD
1.4.16.1	13-Dec-2007	bouyer	Sync with HEAD
1.4.14.1	11-Dec-2007	yamt	sync with head.
1.4.12.1	26-Dec-2007	ad	Sync with head.
1.4.10.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.4.3	21-Jan-2008	yamt	sync with head
1.4.4.2	03-Sep-2007	yamt	sync with head.
1.4.4.1	26-Aug-2007	yamt	file cxgb_l2t.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:32 +0000
1.4.2.2	23-Mar-2008	matt	sync with HEAD
1.4.2.1	09-Jan-2008	matt	sync with HEAD
1.7.32.1	30-May-2010	rmind	sync with head
1.7.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.7.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.2	07-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.20; 1.2.26; Updating Chelsio files. Not quite compiling yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.2.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.2.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.14.3	21-Jan-2008	yamt	sync with head
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	07-Jul-2007	yamt	file cxgb_l2t.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:33 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	07-Jul-2007	ad	file cxgb_l2t.h was added on branch vmlocking on 2007-07-15 13:21:25 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	07-Jul-2007	mjf	file cxgb_l2t.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:11 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.7	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.6	17-Jan-2008	jklos	branches: 1.6.10; 1.6.30; 1.6.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.5	15-Dec-2007	perry	__FUNCTION__ -> __func__
1.4	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.3	19-Oct-2007	ad	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.2	08-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.10; 1.2.12; 1.2.14; 1.2.18; Updating more Chelsio files.
1.1	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.2.18.1	25-Oct-2007	bouyer	Sync with HEAD.
1.2.14.4	21-Jan-2008	yamt	sync with head
1.2.14.3	27-Oct-2007	yamt	sync with head.
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	08-Jul-2007	yamt	file cxgb_lro.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:33 +0000
1.2.12.3	23-Mar-2008	matt	sync with HEAD
1.2.12.2	09-Jan-2008	matt	sync with HEAD
1.2.12.1	06-Nov-2007	matt	sync with HEAD
1.2.10.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.2.4.3	23-Oct-2007	ad	Sync with head.
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	08-Jul-2007	ad	file cxgb_lro.c was added on branch vmlocking on 2007-07-15 13:21:25 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	08-Jul-2007	mjf	file cxgb_lro.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:12 +0000
1.3.8.3	19-Jan-2008	bouyer	Sync with HEAD
1.3.8.2	02-Jan-2008	bouyer	Sync with HEAD
1.3.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.3.6.1	11-Dec-2007	yamt	sync with head.
1.3.4.1	26-Dec-2007	ad	Sync with head.
1.3.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.6.32.1	30-May-2010	rmind	sync with head
1.6.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.6.10.1	11-Aug-2010	yamt	sync with head.
1.17	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.16	12-May-2009	cegger	branches: 1.16.2; 1.16.4; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.15	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.14	18-Apr-2009	tsutsui	Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.13	18-Mar-2009	cegger	bcopy -> memcpy
1.12	07-Nov-2008	dyoung	branches: 1.12.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.11	07-Feb-2008	dyoung	branches: 1.11.6; 1.11.10; 1.11.16; 1.11.18; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.10	17-Jan-2008	jklos	10 gigabit Chelsio cards now compile and work, although not very well yet.
1.9	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.8	19-Oct-2007	ad	branches: 1.8.2; 1.8.4; 1.8.6; 1.8.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.7	12-Jul-2007	he	branches: 1.7.2; 1.7.8; 1.7.10; 1.7.12; 1.7.16; Adapt two more uses to the new signature of callout_init().
1.6	10-Jul-2007	jklos	branches: 1.6.2; More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.5	09-Jul-2007	jklos	More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.4	08-Jul-2007	jklos	Updating more Chelsio files.
1.3	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.2	05-Jul-2007	jklos	Updates of working copies (there will be many).
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.6.2.2	11-Jul-2007	mjf	Sync with head.
1.6.2.1	10-Jul-2007	mjf	file cxgb_main.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:13 +0000
1.7.16.1	25-Oct-2007	bouyer	Sync with HEAD.
1.7.12.5	11-Feb-2008	yamt	sync with head.
1.7.12.4	21-Jan-2008	yamt	sync with head
1.7.12.3	27-Oct-2007	yamt	sync with head.
1.7.12.2	03-Sep-2007	yamt	sync with head.
1.7.12.1	12-Jul-2007	yamt	file cxgb_main.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:34 +0000
1.7.10.3	23-Mar-2008	matt	sync with HEAD
1.7.10.2	09-Jan-2008	matt	sync with HEAD
1.7.10.1	06-Nov-2007	matt	sync with HEAD
1.7.8.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.7.2.3	23-Oct-2007	ad	Sync with head.
1.7.2.2	15-Jul-2007	ad	Sync with head.
1.7.2.1	12-Jul-2007	ad	file cxgb_main.c was added on branch vmlocking on 2007-07-15 13:21:26 +0000
1.8.8.2	19-Jan-2008	bouyer	Sync with HEAD
1.8.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.8.6.1	11-Dec-2007	yamt	sync with head.
1.8.4.1	26-Dec-2007	ad	Sync with head.
1.8.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.11.18.2	28-Apr-2009	skrll	Sync with HEAD.
1.11.18.1	19-Jan-2009	skrll	Sync with HEAD.
1.11.16.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.11.10.3	11-Aug-2010	yamt	sync with head.
1.11.10.2	16-May-2009	yamt	sync with head
1.11.10.1	04-May-2009	yamt	sync with head.
1.11.6.1	17-Jan-2009	mjf	Sync with HEAD.
1.12.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.4.1	30-May-2010	rmind	sync with head
1.16.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	09-Jul-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	09-Jul-2007	yamt	file cxgb_mbuf.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:35 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	09-Jul-2007	ad	file cxgb_mbuf.h was added on branch vmlocking on 2007-07-15 13:21:26 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	09-Jul-2007	mjf	file cxgb_mbuf.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:13 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_mc5.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:35 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_mc5.c was added on branch vmlocking on 2007-07-15 13:21:27 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_mc5.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:14 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_mv88e1xxx.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:36 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_mv88e1xxx.c was added on branch vmlocking on 2007-07-15 13:21:27 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_mv88e1xxx.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:15 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.2	09-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.20; 1.2.26; More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.1	08-Jul-2007	jklos	Updating more Chelsio files.
1.2.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.2.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.14.3	21-Jan-2008	yamt	sync with head
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	09-Jul-2007	yamt	file cxgb_mvec.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:36 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	09-Jul-2007	ad	file cxgb_mvec.h was added on branch vmlocking on 2007-07-15 13:21:28 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	09-Jul-2007	mjf	file cxgb_mvec.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:15 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.7	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.6	17-Jan-2008	jklos	branches: 1.6.10; 1.6.30; 1.6.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.5	15-Dec-2007	perry	__FUNCTION__ -> __func__
1.4	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.3	19-Oct-2007	ad	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.2	09-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.10; 1.2.12; 1.2.14; 1.2.18; More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.1	07-Jul-2007	jklos	Adding another file for Chesio 10 gig card support.
1.2.18.1	25-Oct-2007	bouyer	Sync with HEAD.
1.2.14.4	21-Jan-2008	yamt	sync with head
1.2.14.3	27-Oct-2007	yamt	sync with head.
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	09-Jul-2007	yamt	file cxgb_offload.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:37 +0000
1.2.12.3	23-Mar-2008	matt	sync with HEAD
1.2.12.2	09-Jan-2008	matt	sync with HEAD
1.2.12.1	06-Nov-2007	matt	sync with HEAD
1.2.10.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.2.4.3	23-Oct-2007	ad	Sync with head.
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	09-Jul-2007	ad	file cxgb_offload.c was added on branch vmlocking on 2007-07-15 13:21:28 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	09-Jul-2007	mjf	file cxgb_offload.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:16 +0000
1.3.8.3	19-Jan-2008	bouyer	Sync with HEAD
1.3.8.2	02-Jan-2008	bouyer	Sync with HEAD
1.3.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.3.6.1	11-Dec-2007	yamt	sync with head.
1.3.4.1	26-Dec-2007	ad	Sync with head.
1.3.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.6.32.1	30-May-2010	rmind	sync with head
1.6.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.6.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.2	07-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.20; 1.2.26; Updating Chelsio files. Not quite compiling yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.2.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.2.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.14.3	21-Jan-2008	yamt	sync with head
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	07-Jul-2007	yamt	file cxgb_offload.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:38 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	07-Jul-2007	ad	file cxgb_offload.h was added on branch vmlocking on 2007-07-15 13:21:29 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	07-Jul-2007	mjf	file cxgb_offload.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:17 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.2	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.1	17-Jan-2008	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.14; 1.1.18; 1.1.38; 1.1.40; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.1.40.1	30-May-2010	rmind	sync with head
1.1.38.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.18.1	11-Aug-2010	yamt	sync with head.
1.1.14.2	23-Mar-2008	matt	sync with HEAD
1.1.14.1	17-Jan-2008	matt	file cxgb_osdep.c was added on branch matt-armv6 on 2008-03-23 02:04:45 +0000
1.1.6.2	18-Feb-2008	mjf	Sync with HEAD.
1.1.6.1	17-Jan-2008	mjf	file cxgb_osdep.c was added on branch mjf-devfs on 2008-02-18 21:05:56 +0000
1.1.4.2	21-Jan-2008	yamt	sync with head
1.1.4.1	17-Jan-2008	yamt	file cxgb_osdep.c was added on branch yamt-lazymbuf on 2008-01-21 09:43:45 +0000
1.1.2.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.2.1	17-Jan-2008	bouyer	file cxgb_osdep.c was added on branch bouyer-xeni386 on 2008-01-19 12:15:08 +0000
1.11	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.10	08-Sep-2008	gmcgarry	branches: 1.10.14; 1.10.16; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.9	17-Jan-2008	jklos	branches: 1.9.6; 1.9.10; 1.9.12; 1.9.16; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.8	07-Jan-2008	ad	Remove vestiges of simplelock debugging.
1.7	15-Dec-2007	perry	__FUNCTION__ -> __func__
1.6	14-Dec-2007	dogcow	remove typedef for caddr_t
1.5	19-Oct-2007	ad	branches: 1.5.2; 1.5.4; 1.5.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.4	10-Jul-2007	jklos	branches: 1.4.2; 1.4.4; 1.4.10; 1.4.12; 1.4.14; 1.4.18; More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.3	09-Jul-2007	jklos	More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.2	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.4.18.1	25-Oct-2007	bouyer	Sync with HEAD.
1.4.14.4	21-Jan-2008	yamt	sync with head
1.4.14.3	27-Oct-2007	yamt	sync with head.
1.4.14.2	03-Sep-2007	yamt	sync with head.
1.4.14.1	10-Jul-2007	yamt	file cxgb_osdep.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:38 +0000
1.4.12.3	23-Mar-2008	matt	sync with HEAD
1.4.12.2	09-Jan-2008	matt	sync with HEAD
1.4.12.1	06-Nov-2007	matt	sync with HEAD
1.4.10.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.4.4.3	23-Oct-2007	ad	Sync with head.
1.4.4.2	15-Jul-2007	ad	Sync with head.
1.4.4.1	10-Jul-2007	ad	file cxgb_osdep.h was added on branch vmlocking on 2007-07-15 13:21:29 +0000
1.4.2.2	11-Jul-2007	mjf	Sync with head.
1.4.2.1	10-Jul-2007	mjf	file cxgb_osdep.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:17 +0000
1.5.8.3	19-Jan-2008	bouyer	Sync with HEAD
1.5.8.2	08-Jan-2008	bouyer	Sync with HEAD
1.5.8.1	02-Jan-2008	bouyer	Sync with HEAD
1.5.4.1	26-Dec-2007	ad	Sync with head.
1.5.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.16.1	19-Oct-2008	haad	Sync with HEAD.
1.9.12.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.9.10.2	11-Aug-2010	yamt	sync with head.
1.9.10.1	04-May-2009	yamt	sync with head.
1.9.6.1	28-Sep-2008	mjf	Sync with HEAD.
1.10.16.1	30-May-2010	rmind	sync with head
1.10.14.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.14; 1.1.38; 1.1.60; 1.1.62; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.62.1	30-May-2010	rmind	sync with head
1.1.60.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.38.1	11-Aug-2010	yamt	sync with head.
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_regs.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:39 +0000
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_regs.h was added on branch vmlocking on 2007-07-15 13:21:29 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_regs.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:18 +0000
1.13	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.12	18-Mar-2009	cegger	branches: 1.12.2; 1.12.4; bzero -> memset
1.11	03-Jan-2009	yamt	branches: 1.11.2; remove extra semicolons.
1.10	19-Dec-2008	cegger	use M_ZERO on malloc() and remove subsequent bzero().
1.9	17-Jan-2008	jklos	branches: 1.9.6; 1.9.10; 1.9.18; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.8	14-Dec-2007	dogcow	caddr_t -> void *
1.7	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.6	19-Oct-2007	ad	branches: 1.6.2; 1.6.4; 1.6.6; 1.6.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	12-Jul-2007	he	branches: 1.5.2; 1.5.8; 1.5.10; 1.5.12; 1.5.16; Adapt two more uses to the new signature of callout_init().
1.4	10-Jul-2007	jklos	branches: 1.4.2; More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.3	09-Jul-2007	jklos	More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.2	07-Jul-2007	jklos	Updating Chelsio files. Not quite compiling yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.4.2.2	11-Jul-2007	mjf	Sync with head.
1.4.2.1	10-Jul-2007	mjf	file cxgb_sge.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:20 +0000
1.5.16.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.12.4	21-Jan-2008	yamt	sync with head
1.5.12.3	27-Oct-2007	yamt	sync with head.
1.5.12.2	03-Sep-2007	yamt	sync with head.
1.5.12.1	12-Jul-2007	yamt	file cxgb_sge.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:40 +0000
1.5.10.3	23-Mar-2008	matt	sync with HEAD
1.5.10.2	09-Jan-2008	matt	sync with HEAD
1.5.10.1	06-Nov-2007	matt	sync with HEAD
1.5.8.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.2.3	23-Oct-2007	ad	Sync with head.
1.5.2.2	15-Jul-2007	ad	Sync with head.
1.5.2.1	12-Jul-2007	ad	file cxgb_sge.c was added on branch vmlocking on 2007-07-15 13:21:30 +0000
1.6.8.3	19-Jan-2008	bouyer	Sync with HEAD
1.6.8.2	02-Jan-2008	bouyer	Sync with HEAD
1.6.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.6.6.1	11-Dec-2007	yamt	sync with head.
1.6.4.1	26-Dec-2007	ad	Sync with head.
1.6.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.18.2	28-Apr-2009	skrll	Sync with HEAD.
1.9.18.1	19-Jan-2009	skrll	Sync with HEAD.
1.9.10.2	11-Aug-2010	yamt	sync with head.
1.9.10.1	04-May-2009	yamt	sync with head.
1.9.6.1	17-Jan-2009	mjf	Sync with HEAD.
1.11.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.12.4.1	30-May-2010	rmind	sync with head
1.12.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_sge_defs.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:41 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_sge_defs.h was added on branch vmlocking on 2007-07-15 13:21:31 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_sge_defs.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:20 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.3	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.2	17-Jan-2008	jklos	branches: 1.2.10; 1.2.30; 1.2.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_t3_cpl.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:41 +0000
1.1.12.1	23-Mar-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_t3_cpl.h was added on branch vmlocking on 2007-07-15 13:21:31 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_t3_cpl.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:21 +0000
1.2.32.1	30-May-2010	rmind	sync with head
1.2.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.10.1	11-Aug-2010	yamt	sync with head.
1.6	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.5	17-Jan-2008	jklos	branches: 1.5.10; 1.5.30; 1.5.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.4	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.3	10-Jul-2007	jklos	branches: 1.3.2; 1.3.4; 1.3.12; 1.3.14; 1.3.20; 1.3.22; 1.3.24; 1.3.26; More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.2	09-Jul-2007	jklos	More updates to Chelsio support files. It compiles, but almost certainly doesn't work yet.
1.1	22-Jun-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.3.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.3.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.3.24.1	11-Dec-2007	yamt	sync with head.
1.3.22.1	26-Dec-2007	ad	Sync with head.
1.3.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.3.14.3	21-Jan-2008	yamt	sync with head
1.3.14.2	03-Sep-2007	yamt	sync with head.
1.3.14.1	10-Jul-2007	yamt	file cxgb_t3_hw.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:42 +0000
1.3.12.2	23-Mar-2008	matt	sync with HEAD
1.3.12.1	09-Jan-2008	matt	sync with HEAD
1.3.4.2	15-Jul-2007	ad	Sync with head.
1.3.4.1	10-Jul-2007	ad	file cxgb_t3_hw.c was added on branch vmlocking on 2007-07-15 13:21:31 +0000
1.3.2.2	11-Jul-2007	mjf	Sync with head.
1.3.2.1	10-Jul-2007	mjf	file cxgb_t3_hw.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:22 +0000
1.5.32.1	30-May-2010	rmind	sync with head
1.5.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.5.10.1	11-Aug-2010	yamt	sync with head.
1.2	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.14; 1.1.38; 1.1.60; 1.1.62; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.62.1	30-May-2010	rmind	sync with head
1.1.60.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.38.1	11-Aug-2010	yamt	sync with head.
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_tcb.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:43 +0000
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_tcb.h was added on branch vmlocking on 2007-07-15 13:21:32 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_tcb.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:23 +0000
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; More Chelsio 10 gigabit driver files to complete the ones checked in yesterday.
1.2	07-Jul-2007	jklos	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.20; 1.2.26; Updating Chelsio files. Not quite compiling yet.
1.1	05-Jul-2007	jklos	These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.2.26.1	19-Jan-2008	bouyer	Sync with HEAD
1.2.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.14.3	21-Jan-2008	yamt	sync with head
1.2.14.2	03-Sep-2007	yamt	sync with head.
1.2.14.1	07-Jul-2007	yamt	file cxgb_toedev.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:43 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	15-Jul-2007	ad	Sync with head.
1.2.4.1	07-Jul-2007	ad	file cxgb_toedev.h was added on branch vmlocking on 2007-07-15 13:21:32 +0000
1.2.2.2	11-Jul-2007	mjf	Sync with head.
1.2.2.1	07-Jul-2007	mjf	file cxgb_toedev.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:24 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.2	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.14; 1.1.38; 1.1.60; 1.1.62; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.62.1	30-May-2010	rmind	sync with head
1.1.60.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.38.1	11-Aug-2010	yamt	sync with head.
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_version.h was added on branch yamt-lazymbuf on 2007-09-03 14:36:44 +0000
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_version.h was added on branch vmlocking on 2007-07-15 13:21:33 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_version.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:24 +0000
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_vsc7323.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:44 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_vsc7323.c was added on branch vmlocking on 2007-07-15 13:21:33 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_vsc7323.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:25 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_vsc8211.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:45 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_vsc8211.c was added on branch vmlocking on 2007-07-15 13:21:33 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_vsc8211.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:26 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.4	21-Mar-2010	jklos	Reorganizing Chelsio 10 gig files into separate directory.
1.3	17-Jan-2008	jklos	branches: 1.3.10; 1.3.30; 1.3.32; 10 gigabit Chelsio cards now compile and work, although not very well yet.
1.2	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.1	22-Jun-2007	jklos	branches: 1.1.2; 1.1.4; 1.1.12; 1.1.14; 1.1.20; 1.1.22; 1.1.24; 1.1.26; These files (cxgb*) are for the Chelsio T3 10 gigabit ethernet card from the FreeBSD driver by Kip Macy (kmacy@FreeBSD.org).
1.1.26.2	19-Jan-2008	bouyer	Sync with HEAD
1.1.26.1	13-Dec-2007	bouyer	Sync with HEAD
1.1.24.1	11-Dec-2007	yamt	sync with head.
1.1.22.1	26-Dec-2007	ad	Sync with head.
1.1.20.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.14.3	21-Jan-2008	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	22-Jun-2007	yamt	file cxgb_xgmac.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:45 +0000
1.1.12.2	23-Mar-2008	matt	sync with HEAD
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	22-Jun-2007	ad	file cxgb_xgmac.c was added on branch vmlocking on 2007-07-15 13:21:34 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	22-Jun-2007	mjf	file cxgb_xgmac.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:26 +0000
1.3.32.1	30-May-2010	rmind	sync with head
1.3.30.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.10.1	11-Aug-2010	yamt	sync with head.
1.11	17-Aug-2024	thorpej	Make the cyhc lock a regular mutex rather than a spin mutex. Simplify some interaction with the lock by converting some open-coded test/panic blocks to KASSERT()s that can be safely tested outside the lock perimeter. PR port-alpha/58602
1.10	06-Jun-2018	maya	branches: 1.10.34; 1.10.40; Remove duplicate ;
1.9	20-Mar-2014	skrll	branches: 1.9.28; Mechanically replace simplelock with kmutex_t.
1.8	19-Apr-2010	dyoung	branches: 1.8.8; 1.8.18; 1.8.22; Add default implementations for bus_space_is_equal(9), bus_space_tag_create(9), and bus_space_tag_destroy(9). Use bus_space_is_equal(9) throughout the kernel to compare bus_space_tag_t's. Tested on i386 and on sparc64.
1.7	22-Mar-2010	dyoung	#if 0 some diagnostic code that compares two bus_space_tag_t's. MI code should not compare an opaque type such as bus_space_tag_t!
1.6	28-Apr-2008	martin	branches: 1.6.20; 1.6.22; Remove clause 3 and 4 from TNF licenses
1.5	04-Jan-2008	ad	branches: 1.5.6; 1.5.8; 1.5.10; Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.4	19-Oct-2007	ad	branches: 1.4.2; 1.4.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.3	13-Nov-2001	lukem	branches: 1.3.32; 1.3.74; 1.3.88; 1.3.90; 1.3.94; cleanup RCSID
1.2	24-Nov-2000	chs	branches: 1.2.2; 1.2.4; adjust the spinlock macros in the non-MULTIPROCESSOR, non-LOCKDEBUG case so that gcc will think that static spinlock are used. this allows us to remove the ugly conditionalization of static spinlock declarations.
1.1	06-Jun-2000	thorpej	branches: 1.1.4; 1.1.6; Common routines for read/writing Cypress 82c693 control registers. Needed by `pciide' and the Alpha `sio' (PCI-ISA bridge) driver.
1.1.6.3	08-Dec-2000	bouyer	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	06-Jun-2000	bouyer	file cy82c693.c was added on branch thorpej_scsipi on 2000-11-20 11:42:16 +0000
1.1.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.1.4.1	06-Jun-2000	minoura	file cy82c693.c was added on branch minoura-xpg4dl on 2000-06-22 17:07:17 +0000
1.2.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.3.94.1	25-Oct-2007	bouyer	Sync with HEAD.
1.3.90.2	09-Jan-2008	matt	sync with HEAD
1.3.90.1	06-Nov-2007	matt	sync with HEAD
1.3.88.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.74.1	23-Oct-2007	ad	Sync with head.
1.3.32.2	21-Jan-2008	yamt	sync with head
1.3.32.1	27-Oct-2007	yamt	sync with head.
1.4.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.4.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.5.10.2	11-Aug-2010	yamt	sync with head.
1.5.10.1	16-May-2008	yamt	sync with head.
1.5.8.1	18-May-2008	yamt	sync with head.
1.5.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.6.22.1	30-May-2010	rmind	sync with head
1.6.20.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.8.22.1	18-May-2014	rmind	sync with head
1.8.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.8.8.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.9.28.1	25-Jun-2018	pgoyette	Sync with HEAD
1.10.40.1	02-Aug-2025	perseant	Sync with HEAD
1.10.34.1	24-Aug-2024	martin	Pull up following revision(s) (requested by thorpej in ticket #809): sys/dev/pci/cy82c693.c: revision 1.11 Make the cyhc lock a regular mutex rather than a spin mutex. Simplify some interaction with the lock by converting some open-coded test/panic blocks to KASSERT()s that can be safely tested outside the lock perimeter. PR port-alpha/58602
1.2	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.1	06-Jun-2000	thorpej	branches: 1.1.4; 1.1.6; 1.1.128; 1.1.130; 1.1.132; Common routines for read/writing Cypress 82c693 control registers. Needed by `pciide' and the Alpha `sio' (PCI-ISA bridge) driver.
1.1.132.1	16-May-2008	yamt	sync with head.
1.1.130.1	18-May-2008	yamt	sync with head.
1.1.128.1	02-Jun-2008	mjf	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	06-Jun-2000	bouyer	file cy82c693reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:17 +0000
1.1.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.1.4.1	06-Jun-2000	minoura	file cy82c693reg.h was added on branch minoura-xpg4dl on 2000-06-22 17:07:18 +0000
1.2	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.1	06-Jun-2000	thorpej	branches: 1.1.4; 1.1.6; 1.1.128; 1.1.130; 1.1.132; Common routines for read/writing Cypress 82c693 control registers. Needed by `pciide' and the Alpha `sio' (PCI-ISA bridge) driver.
1.1.132.1	16-May-2008	yamt	sync with head.
1.1.130.1	18-May-2008	yamt	sync with head.
1.1.128.1	02-Jun-2008	mjf	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	06-Jun-2000	bouyer	file cy82c693var.h was added on branch thorpej_scsipi on 2000-11-20 11:42:17 +0000
1.1.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.1.4.1	06-Jun-2000	minoura	file cy82c693var.h was added on branch minoura-xpg4dl on 2000-06-22 17:07:18 +0000
1.26	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.25	07-Jul-2016	msaitoh	branches: 1.25.16; 1.25.18; KNF. Remove extra spaces. No functional change.
1.24	29-Mar-2014	christos	branches: 1.24.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.23	26-Mar-2008	matt	branches: 1.23.38; 1.23.48; 1.23.54; split device_t/softc. Use aprint_*_dev and device_xname. switch to device_lookup_private.
1.22	19-Oct-2007	ad	branches: 1.22.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.21	16-Nov-2006	christos	branches: 1.21.8; 1.21.22; 1.21.24; 1.21.28; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	11-Dec-2005	christos	branches: 1.19.20; 1.19.22; merge ktrace-lwp.
1.18	28-Jun-2005	thorpej	branches: 1.18.2; Use ANSI function decls and static.
1.17	31-Jan-2003	thorpej	branches: 1.17.2; Use aprint_*().
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	13-Nov-2001	lukem	add RCSID
1.12	20-Jan-2001	thorpej	branches: 1.12.2; 1.12.4; Only need to pass a cy_softc * to cy_attach().
1.11	20-Jan-2001	thorpej	Clean up the PCI attachment a bit, and adjust the other pieces to suit.
1.10	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.9	29-May-2000	tsubai	* Support (not so) new cards. (only Cyclom-8YsP+ is tested) * Make compilable with CY_DEBUG.
1.8	08-Jun-1998	thorpej	branches: 1.8.8; 1.8.14; 1.8.22; Nuke __BROKEN_INDIRECT_CONFIG.
1.7	17-Jun-1997	cgd	Change expected mapping register type argument for the pci_mapreg_map() call for the board's memory space to be PCI_MAPREG_MEM_TYPE_32BIT_1M or PCI_MAPREG_MEM_TYPE_32BIT depending on the board ID. Also, remove a bogus extra argument to an interrupt-establishment-error printf. Problems pointed out by Jarkko Torppa <torppa@cute.fi> in PR 3753, but fixed slightly differently than he suggested.
1.6	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.5	13-Apr-1997	cgd	use pci_map_register(). Also, almost a complete rewrite/cleanup of this code. (It still needs more!)
1.4	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.3	13-Oct-1996	christos	backout kprintf changes
1.2	10-Oct-1996	christos	- printf -> kprintf, sprintf -> ksprintf - remove unused variables. - fix cleanup code in case of probe failure.
1.1	24-Sep-1996	christos	New cyclades driver from Timo Rossi.
1.8.22.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.8.14.3	11-Feb-2001	bouyer	Sync with HEAD.
1.8.14.2	05-Jan-2001	bouyer	Sync with HEAD
1.8.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.8.1	03-Jun-2000	he	Pull up revision 1.9 (requested by tsubai): Add support for PCI variants of Cyclom-4Y and -8Y cards. Also fixes PR#7175 (new serial controller).
1.12.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.12.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.12.2.2	18-Oct-2002	nathanw	Catch up to -current.
1.12.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.2.2	27-Oct-2007	yamt	sync with head.
1.18.2.1	30-Dec-2006	yamt	sync with head.
1.19.22.2	10-Dec-2006	yamt	sync with head.
1.19.22.1	22-Oct-2006	yamt	sync with head
1.19.20.1	18-Nov-2006	ad	Sync with head.
1.21.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.21.24.1	06-Nov-2007	matt	sync with HEAD
1.21.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.21.8.1	23-Oct-2007	ad	Sync with head.
1.22.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.23.54.1	18-May-2014	rmind	sync with head
1.23.48.2	03-Dec-2017	jdolecek	update from HEAD
1.23.48.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.23.38.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.24.6.1	09-Jul-2016	skrll	Sync with HEAD
1.25.18.1	10-Jun-2019	christos	Sync with HEAD
1.25.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.5	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.4	19-Oct-2007	ad	branches: 1.4.16; 1.4.18; 1.4.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.3	11-Dec-2005	christos	branches: 1.3.30; 1.3.44; 1.3.46; 1.3.50; merge ktrace-lwp.
1.2	04-Feb-2004	fredb	branches: 1.2.4; 1.2.18; The variables to hold the registers don't need to be declared volatile. Pointed out by Simon Burge on tech-kern.
1.1	03-Feb-2004	fredb	Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.2.18.1	27-Oct-2007	yamt	sync with head.
1.2.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.4.2	03-Aug-2004	skrll	Sync with HEAD
1.2.4.1	04-Feb-2004	skrll	file cyber.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.3.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.3.46.1	06-Nov-2007	matt	sync with HEAD
1.3.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.30.1	23-Oct-2007	ad	Sync with head.
1.4.20.1	16-May-2008	yamt	sync with head.
1.4.18.1	18-May-2008	yamt	sync with head.
1.4.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.3	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.2	11-Dec-2005	christos	branches: 1.2.70; 1.2.72; 1.2.74; merge ktrace-lwp.
1.1	03-Feb-2004	fredb	branches: 1.1.4; Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	03-Feb-2004	skrll	file cyberreg.h was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.2.74.1	16-May-2008	yamt	sync with head.
1.2.72.1	18-May-2008	yamt	sync with head.
1.2.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.3	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.2	11-Dec-2005	christos	branches: 1.2.70; 1.2.72; 1.2.74; merge ktrace-lwp.
1.1	03-Feb-2004	fredb	branches: 1.1.4; Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	03-Feb-2004	skrll	file cybervar.h was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.2.74.1	16-May-2008	yamt	sync with head.
1.2.72.1	18-May-2008	yamt	sync with head.
1.2.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.32	20-Oct-2017	jdolecek	move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.31	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.30	07-Oct-2013	jakllsch	branches: 1.30.18; Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.29	31-Jul-2012	bouyer	branches: 1.29.2; 1.29.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.28	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.27	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.26	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.25	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.24	04-Apr-2011	dyoung	branches: 1.24.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.23	05-Nov-2010	jakllsch	branches: 1.23.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.22	19-Oct-2009	bouyer	branches: 1.22.2; 1.22.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.21	18-Mar-2008	cube	branches: 1.21.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.20	09-Feb-2007	ad	branches: 1.20.20; 1.20.36; 1.20.40; Merge newlock2 to head.
1.19	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	16-Jan-2006	bouyer	branches: 1.17.18; 1.17.20; properly use ata_channel->ch_ndrive: - initialize it properly in the bus front-ends (all 2, exept in wdc_pcmcia.c for the "Sandisk CompactFlash Card" where it's set to 1) - remplace hardcoded '2' by ata_channel->ch_ndrive in MI IDE drivers. From Christos Zoulas in kern/32501.
1.16	11-Dec-2005	christos	branches: 1.16.2; merge ktrace-lwp.
1.15	24-May-2005	lukem	branches: 1.15.2; add __KERNEL_RCSID
1.14	27-Feb-2005	perry	nuke trailing whitespace
1.13	20-Aug-2004	thorpej	branches: 1.13.4; 1.13.6; Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.12	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.11	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.10	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.9	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.8	02-Aug-2004	bouyer	branches: 1.8.2; Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. Fix kern/26502 from Christian Biere.
1.7	03-Jan-2004	thorpej	branches: 1.7.2; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.6	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.5	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.4	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.7.2.1	11-Aug-2004	jmc	Pullup rev 1.8 (requested by bouyer in ticket #734) Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. PR#26502
1.8.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.8.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.8.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.8.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.8.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.8.2.2	03-Aug-2004	skrll	Sync with HEAD
1.8.2.1	02-Aug-2004	skrll	file cypide.c was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.13.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.4.1	29-Apr-2005	kent	sync with -current
1.15.2.4	24-Mar-2008	yamt	sync with head.
1.15.2.3	26-Feb-2007	yamt	sync with head.
1.15.2.2	30-Dec-2006	yamt	sync with head.
1.15.2.1	21-Jun-2006	yamt	sync with head.
1.16.2.1	01-Feb-2006	yamt	sync with head.
1.17.20.2	10-Dec-2006	yamt	sync with head.
1.17.20.1	22-Oct-2006	yamt	sync with head
1.17.18.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.17.18.1	18-Nov-2006	ad	Sync with head.
1.20.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.20.36.1	24-Mar-2008	keiichi	sync with head.
1.20.20.1	23-Mar-2008	matt	sync with HEAD
1.21.4.1	11-Mar-2010	yamt	sync with head
1.22.4.2	21-Apr-2011	rmind	sync with head
1.22.4.1	05-Mar-2011	rmind	sync with head
1.22.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.23.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.24.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.24.4.1	30-Oct-2012	yamt	sync with head
1.29.4.1	18-May-2014	rmind	sync with head
1.29.2.3	03-Dec-2017	jdolecek	update from HEAD
1.29.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.29.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.30.18.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.66	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.65	26-Oct-2022	riastradh	cz(4): Convert to ttylock/ttyunlock.
1.64	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.63	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.62	07-Jul-2016	msaitoh	branches: 1.62.16; 1.62.18; KNF. Remove extra spaces. No functional change.
1.61	15-Nov-2014	christos	branches: 1.61.2; centralize the dialout/call unit macros.
1.60	25-Jul-2014	dholland	Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.59	29-Mar-2014	christos	branches: 1.59.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.58	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.57	12-Sep-2013	martin	Eliminate an unused variable - someone with docs and test hardware please check the XXX introduced.
1.56	27-Oct-2012	chs	branches: 1.56.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.55	24-Apr-2011	rmind	branches: 1.55.4; 1.55.14; Rename ttymalloc() to tty_alloc(), and ttyfree() to tty_free() for consistency. Remove some unnecessary malloc.h inclusions as well.
1.54	26-Jul-2010	jym	branches: 1.54.2; Add PAE to ALL kernel, so that most paddr_t format string errors get caught during compilation. While here, fix the compilation for ALL.
1.53	26-Nov-2009	njoly	branches: 1.53.2; 1.53.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.52	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.51	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.50	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.49	08-Jun-2008	tsutsui	branches: 1.49.12; Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.48	10-Apr-2008	cegger	branches: 1.48.2; 1.48.4; 1.48.6; use aprint_*_dev and device_xname
1.47	19-Nov-2007	ad	branches: 1.47.14; - Factor out too many copies of the same bit of tty code. - Fix another tty signalling/wakeup problem.
1.46	12-Nov-2007	ad	Call ttwakeup() with tty_lock held.
1.45	09-Jul-2007	ad	branches: 1.45.6; 1.45.8; 1.45.12; 1.45.14; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.44	08-May-2007	garbled	Add a missing brace in cztty_receive inside a ifdef HOSTRAMCODE. This would randomly cause internal compiler errors while compiling pmppc.
1.43	04-Mar-2007	christos	branches: 1.43.2; 1.43.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.42	16-Nov-2006	christos	branches: 1.42.4; __unused removal on arguments; approved by core.
1.41	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.40	01-Oct-2006	elad	More from Matt Fleming: Adapt to KAUTH_DEVICE_TTY_PRIVSET and KAUTH_DEVICE_TTY_OPEN.
1.39	01-Oct-2006	elad	Adapt MD code to KAUTH_DEVICE_TTY_OPEN, batch #2 from Matt Fleming, thanks! Also, add forgotten splx() calls in some places.
1.38	21-Jul-2006	ad	branches: 1.38.4; 1.38.6; - Use the LWP cached credentials where sane. - Minor cosmetic changes.
1.37	14-May-2006	elad	integrate kauth.
1.36	28-Mar-2006	thorpej	Use device_unit().
1.35	05-Mar-2006	christos	branches: 1.35.2; 1.35.4; cleanup more SET/CLR/ISSET lossage
1.34	11-Dec-2005	christos	branches: 1.34.4; 1.34.6; 1.34.8; merge ktrace-lwp.
1.33	06-Sep-2005	kleink	Change the driver open function's conditional for overriding exclusive tty use from checking the proc's uid to suser(9), and account for the use of privileges. Noted by David Holland in PR kern/31126.
1.32	28-Jun-2005	thorpej	branches: 1.32.2; Use ANSI function decls and static.
1.31	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.30	27-Feb-2005	perry	nuke trailing whitespace
1.29	27-Oct-2003	chs	branches: 1.29.8; 1.29.10; uninitialized variables
1.28	29-Jun-2003	fvdl	branches: 1.28.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.27	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.26	05-Feb-2003	perry	"Utilize" has exactly the same meaning as "use," but it is more difficult to read and understand. Most manuals of English style therefore say that you should use "use".
1.25	31-Jan-2003	thorpej	Use aprint_*().
1.24	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.23	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.22	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.21	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.20	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.19	17-Mar-2002	atatat	branches: 1.19.4; Convert ioctl code to use EPASSTHROUGH instead of -1 or ENOTTY for indicating an unhandled "command". ERESTART is -1, which can lead to confusion. ERESTART has been moved to -3 and EPASSTHROUGH has been placed at -4. No ioctl code should now return -1 anywhere. The ioctl() system call is now properly restartable.
1.18	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.17	13-Nov-2001	lukem	add RCSID
1.16	02-May-2001	scw	branches: 1.16.2; 1.16.4; Add `l_poll' to `struct linesw' and provide an xxxpoll() entry point in each tty driver to indirect through it. This allows tty line-disciplines to handle poll(2) system calls.
1.15	20-Jan-2001	thorpej	branches: 1.15.2; Use device_lookup() in cztty_getttysoftc().
1.14	20-Jan-2001	thorpej	Garbage-collect some unused macros.
1.13	20-Jan-2001	thorpej	Don't deref NULL when searching for the correct board.
1.12	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.11	02-Nov-2000	eeh	branches: 1.11.2; Adapt to the new line discipline scheme.
1.10	28-Jul-2000	thorpej	Handle booting without a serial expander box connected.
1.9	14-Jun-2000	thorpej	Make sure we use firmware->host communication channel in cz_intr(), from Bill Studenmund.
1.8	09-Jun-2000	wrstuden	Adapt to going back to decemal device numbering.
1.7	09-Jun-2000	wrstuden	Rather than assinging 64 tty's per pci card, assign tty's as found. Cyclades says driver should work with up to 128 tty's per pci card even though they only support 64 now. Also matches FreeBSD behavior. Print located tty's using new tty naming scheme - ttyCZ?? where ?? are two hex letters. Make firmware load on macppc - one le32toh() was missing.
1.6	24-May-2000	thorpej	branches: 1.6.2; Update from Bill Studenmund <wrstuden@zembu.com>: - Note that the FreeBSD driver distributed by Cyclades served as documentation on how the board works. - Use the wait-until-pending-I/O-completes-before-loading-parameters command in czttyparam(). - Pass BREAKs up to the tty later properly. - Garbage-collect some debugging code.
1.5	23-May-2000	thorpej	Typo in previous.
1.4	23-May-2000	thorpej	Update from Bill Studenmund <wrstuden@zembu.com>: Correctly send and receive BREAKs -- still needs to do something intelligent with them. Some additional debugging goo.
1.3	19-May-2000	thorpej	In tiocm_to_cztty(), use SET() instead of CLR() in the TIOCMBIS case. "Oops!"
1.2	18-May-2000	thorpej	Fix an inverted test.
1.1	17-May-2000	thorpej	Add a device driver for the Cyclades-Z series of intelligent multi-port serial boards. Written by Jason R. Thorpe and Bill Studenmund.
1.6.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.11.2.5	11-Feb-2001	bouyer	Sync with HEAD.
1.11.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.11.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.11.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.11.2.1	02-Nov-2000	bouyer	file cz.c was added on branch thorpej_scsipi on 2000-11-20 11:42:17 +0000
1.15.2.7	11-Nov-2002	nathanw	Catch up to -current
1.15.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.15.2.5	17-Sep-2002	nathanw	Catch up to -current.
1.15.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.15.2.3	28-Feb-2002	nathanw	Catch up to -current.
1.15.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.15.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.16.4.3	13-Oct-2001	fvdl	Revert the t_dev -> t_devvp change in struct tty. The way that tty structs are currently used (especially by console ttys) aren't ready for it, and this will require quite a few changes.
1.16.4.2	26-Sep-2001	fvdl	* add a VCLONED vnode flag that indicates a vnode representing a cloned device. * rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass to VOP_REVOKE * the revoke system call will revoke all aliases, as before, but not the clones * vdevgone is called when detaching a device, so make it use REVOKECLONE to get rid of all clones as well * clean up all uses of VOP_OPEN wrt. locking. * add a few VOPS to spec_vnops that need to do something when it's a clone vnode (access and getattr) * add a copy of the vnode vattr structure of the original 'master' vnode to the specinfo of a cloned vnode. could possibly redirect getattr to the 'master' vnode, but this has issues with revoke * add a vdev_reassignvp function that disassociates a vnode from its original device, and reassociates it with the specified dev_t. to be used by cloning devices only, in case a new minor is allocated. * change all direct references in drivers to v_devcookie and v_rdev to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes when debugging race conditions that still exist wrt. locking and revoking vnodes. * make the locking state of a vnode consistent when passed to d_open and d_close (unlocked). locked would be better, but has some deadlock issues
1.16.4.1	07-Sep-2001	thorpej	Commit my "devvp" changes to the thorpej-devvp branch. This replaces the use of dev_t in most places with a struct vnode *. This will form the basic infrastructure for real cloning device support (besides being architecurally cleaner -- it'll be good to get away from using numbers to represent objects).
1.16.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.16.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.16.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.16.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.4.1	16-May-2002	gehenna	Add the character device switch. Replace the local major search function with devsw API.
1.28.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.28.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.28.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.28.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.28.2.2	03-Aug-2004	skrll	Sync with HEAD
1.28.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.29.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.29.8.1	29-Apr-2005	kent	sync with -current
1.32.2.5	07-Dec-2007	yamt	sync with head
1.32.2.4	15-Nov-2007	yamt	sync with head.
1.32.2.3	03-Sep-2007	yamt	sync with head.
1.32.2.2	30-Dec-2006	yamt	sync with head.
1.32.2.1	21-Jun-2006	yamt	sync with head.
1.34.8.4	11-Aug-2006	yamt	sync with head
1.34.8.3	24-May-2006	yamt	sync with head.
1.34.8.2	01-Apr-2006	yamt	sync with head.
1.34.8.1	13-Mar-2006	yamt	sync with head.
1.34.6.2	01-Jun-2006	kardel	Sync with head.
1.34.6.1	22-Apr-2006	simonb	Sync with head.
1.34.4.1	09-Sep-2006	rpaulo	sync with head
1.35.4.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.35.4.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.35.2.4	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.35.2.3	19-Apr-2006	elad	sync with head.
1.35.2.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.35.2.1	08-Mar-2006	elad	Adapt to kernel authorization KPI.
1.38.6.2	10-Dec-2006	yamt	sync with head.
1.38.6.1	22-Oct-2006	yamt	sync with head
1.38.4.1	18-Nov-2006	ad	Sync with head.
1.42.4.2	17-May-2007	yamt	sync with head.
1.42.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.43.4.1	11-Jul-2007	mjf	Sync with head.
1.43.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.43.2.1	27-May-2007	ad	Sync with head.
1.45.14.2	08-Dec-2007	mjf	Sync with HEAD.
1.45.14.1	19-Nov-2007	mjf	Sync with HEAD.
1.45.12.2	21-Nov-2007	bouyer	Sync with HEAD
1.45.12.1	13-Nov-2007	bouyer	Sync with HEAD
1.45.8.1	09-Jan-2008	matt	sync with HEAD
1.45.6.2	21-Nov-2007	joerg	Sync with HEAD.
1.45.6.1	14-Nov-2007	joerg	Sync with HEAD.
1.47.14.2	29-Jun-2008	mjf	Sync with HEAD.
1.47.14.1	02-Jun-2008	mjf	Sync with HEAD.
1.48.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.48.4.4	11-Aug-2010	yamt	sync with head.
1.48.4.3	11-Mar-2010	yamt	sync with head
1.48.4.2	16-May-2009	yamt	sync with head
1.48.4.1	04-May-2009	yamt	sync with head.
1.48.2.1	17-Jun-2008	yamt	sync with head.
1.49.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.53.4.2	31-May-2011	rmind	sync with head
1.53.4.1	05-Mar-2011	rmind	sync with head
1.53.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.54.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.55.14.3	03-Dec-2017	jdolecek	update from HEAD
1.55.14.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.55.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.55.4.1	30-Oct-2012	yamt	sync with head
1.56.2.1	18-May-2014	rmind	sync with head
1.59.2.1	10-Aug-2014	tls	Rebase.
1.61.2.1	09-Jul-2016	skrll	Sync with HEAD
1.62.18.1	10-Jun-2019	christos	Sync with HEAD
1.62.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.62.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.4	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.3	11-Jul-2015	kamil	Improve spelling: regsiter -> register This change is non-functional. Approved by <riastradh>, <pgoyette>
1.2	25-Dec-2007	perry	branches: 1.2.54; 1.2.74; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.1	17-May-2000	thorpej	branches: 1.1.6; 1.1.46; 1.1.104; 1.1.110; 1.1.114; 1.1.118; Add a device driver for the Cyclades-Z series of intelligent multi-port serial boards. Written by Jason R. Thorpe and Bill Studenmund.
1.1.118.1	02-Jan-2008	bouyer	Sync with HEAD
1.1.114.1	26-Dec-2007	ad	Sync with head.
1.1.110.1	18-Feb-2008	mjf	Sync with HEAD.
1.1.104.1	09-Jan-2008	matt	sync with HEAD
1.1.46.1	21-Jan-2008	yamt	sync with head
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	17-May-2000	bouyer	file czreg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:18 +0000
1.2.74.1	22-Sep-2015	skrll	Sync with HEAD
1.2.54.1	03-Dec-2017	jdolecek	update from HEAD
1.14	21-Sep-2014	christos	Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.13	17-Nov-2008	matt	branches: 1.13.26; Store pcidevs in a more compact manner. Instead of many string with duplicate words, store each word individual in a long string and then store offsets in the string to that word. This reduces the space needed by half.
1.12	11-Dec-2005	christos	branches: 1.12.70; 1.12.74; 1.12.80; 1.12.84; merge ktrace-lwp.
1.11	28-Jun-2005	thorpej	static'ify pci_vendors[] and pci_products[]
1.10	27-Feb-2005	perry	don't leave whitespace at the end of $ stripped RCS Ids don't put unneeded tabs at the end of a line
1.9	02-Aug-2004	mycroft	branches: 1.9.4; 1.9.6; For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.8	15-Dec-2003	jmc	Fixes from PR#23177. Various lint/logic fixes: Fix some non-initialized variables close the output files when done Redo the printing for RCS strings so they don't expand in the awk script too Do proper tests for variables existance before accessing Verified output from all scripts is identical to original versions
1.7	05-Dec-2003	grant	remove some extraneous \'s to shut up warning from gawk.
1.6	18-Jan-2001	jdolecek	branches: 1.6.24; make pci_knowndevs[] const
1.5	09-Jan-1998	thorpej	branches: 1.5.14; Put RCS IDs into generated files, in the correct place.
1.4	08-Jan-1998	thorpej	Back out RCD ID related changes.
1.3	05-Jan-1998	perry	make script insert RCS ids into generated files
1.2	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.1	18-Jun-1995	cgd	learn about some PCI vendor and product numbers. not yet used
1.5.14.1	11-Feb-2001	bouyer	Sync with HEAD.
1.6.24.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.24.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.24.1	03-Aug-2004	skrll	Sync with HEAD
1.9.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.4.1	29-Apr-2005	kent	sync with -current
1.12.84.1	19-Jan-2009	skrll	Sync with HEAD.
1.12.80.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.12.74.1	04-May-2009	yamt	sync with head.
1.12.70.1	17-Jan-2009	mjf	Sync with HEAD.
1.13.26.1	03-Dec-2017	jdolecek	update from HEAD
1.28	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.27	29-Mar-2014	christos	branches: 1.27.28; 1.27.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.26	27-Oct-2012	chs	branches: 1.26.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.25	26-Nov-2009	njoly	branches: 1.25.12; 1.25.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.24	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.23	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.22	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.21	10-Apr-2008	cegger	branches: 1.21.4; 1.21.18; use aprint_*_dev and device_xname
1.20	19-Oct-2007	ad	branches: 1.20.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.19	16-Nov-2006	christos	branches: 1.19.8; 1.19.22; 1.19.24; 1.19.28; __unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	11-Dec-2005	christos	branches: 1.17.20; 1.17.22; merge ktrace-lwp.
1.16	27-Feb-2005	perry	branches: 1.16.4; nuke trailing whitespace
1.15	04-Dec-2003	keihan	branches: 1.15.8; 1.15.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.14	31-Jan-2003	thorpej	branches: 1.14.2; Use aprint_*().
1.13	07-Dec-2002	ad	Implement the DPT EATA ioctl() interface.
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	13-Nov-2001	lukem	branches: 1.9.10; add RCSID
1.8	25-Apr-2001	bouyer	branches: 1.8.2; Pull up the thorpej_scsipi branch to main branch. This is a completely rewritten scsipi_xfer execution engine, and the associated changes to HBA drivers. Overview of changes & features: - All xfers are queued in the mid-layer, rather than doing so in an ad-hoc fashion in individual adapter drivers. - Adapter/channel resource management in the mid-layer, avoids even trying to start running an xfer if the adapter/channel doesn't have the resources. - Better communication between the mid-layer and the adapters. - Asynchronous event notification mechanism from adapter to mid-layer and peripherals. - Better peripheral queue management: freeze/thaw, sorted requeueing during recovery, etc. - Clean separation of peripherals, adapters, and adapter channels (no more scsipi_link). - Kernel thread for each scsipi_channel makes error recovery much easier (no more dealing with interrupt context when recovering from an error). - Mid-layer support for tagged queueing: commands can have the tag type set explicitly, tag IDs are allocated in the mid-layer (thus eliminating the need to use buggy tag ID allocation schemes in many adapter drivers). - support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command will be requeued, or a REQUEST SENSE will be sent as appropriate. Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
1.7	28-Dec-2000	sommerfeld	branches: 1.7.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.6	13-Jun-2000	ad	Use my proper name.
1.5	25-Feb-2000	ad	branches: 1.5.2; Cosmetic change.
1.4	18-Jan-2000	ad	Fix register offset problem. This should make EISA boards work correctly, and is also needed for the ISA frontend (some time this week).
1.3	05-Jan-2000	ad	branches: 1.3.2; Spelling.
1.2	29-Sep-1999	ad	branches: 1.2.2; - Move initial reset code into dpt_readcfg() - Bus specific front-end is now responsible for reading EATA configuration - EATA configuration data is now saved in the softc - Make synchronous writes work properly - Don't ignore HBA timeouts in dpt_cmd() - Prefix eata_cfg's members with `ec_', not `dc_'
1.1	27-Sep-1999	ad	A driver for DPT EATA SCSI adapters. dpt.c contains a brief TODO list.
1.2.2.3	01-Apr-2001	ad	- Make this work. - Restructure a bit. - KNF, const.
1.2.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.3	18-Jan-2000	he	Pull up revision 1.4 (requested by ad): Fix register offset problem to make EISA boards work correctly.
1.3.2.2	17-Jan-2000	he	Pull up revisions 1.1-1.3 (requested by ad): Add driver for DPT SmartCache and SmartRAID III or IV SCSI adapters.
1.3.2.1	05-Jan-2000	he	file dpt_pci.c was added on branch netbsd-1-4 on 2000-01-17 18:35:41 +0000
1.5.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.7.2.4	11-Dec-2002	thorpej	Sync with HEAD.
1.7.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.8.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.10.1	12-Dec-2002	he	Pull up revision 1.13 (requested by ad in ticket #1031): Implement DPT EATA ioctl interface.
1.14.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.1	03-Aug-2004	skrll	Sync with HEAD
1.15.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.15.8.1	29-Apr-2005	kent	sync with -current
1.16.4.2	27-Oct-2007	yamt	sync with head.
1.16.4.1	30-Dec-2006	yamt	sync with head.
1.17.22.2	10-Dec-2006	yamt	sync with head.
1.17.22.1	22-Oct-2006	yamt	sync with head
1.17.20.1	18-Nov-2006	ad	Sync with head.
1.19.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.19.24.1	06-Nov-2007	matt	sync with HEAD
1.19.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.19.8.1	23-Oct-2007	ad	Sync with head.
1.20.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.21.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.21.4.2	11-Mar-2010	yamt	sync with head
1.21.4.1	16-May-2009	yamt	sync with head
1.25.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.25.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.25.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.25.12.1	30-Oct-2012	yamt	sync with head
1.26.2.1	18-May-2014	rmind	sync with head
1.27.30.1	10-Jun-2019	christos	Sync with HEAD
1.27.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.103	17-May-2024	nia	eap(4): Restore the full set of supported hardware encodings from netbsd-7 Tested in QEMU.
1.102	25-Jul-2019	msaitoh	branches: 1.102.2; 1.102.28; Make register address unsigned to avoid undefined behavior. Found by KUBSan.
1.101	08-Jun-2019	isaki	Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.100	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.99	16-Mar-2019	isaki	branches: 1.99.2; Use C99 style struct initializer to audio_hw_if.
1.98	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.97	01-Jun-2017	chs	branches: 1.97.8; 1.97.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.96	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.95	29-Mar-2014	christos	branches: 1.95.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.94	24-Nov-2011	mrg	branches: 1.94.8; 1.94.12; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.93	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.92	28-Apr-2008	martin	branches: 1.92.12; 1.92.34; 1.92.36; Remove clause 3 and 4 from TNF licenses
1.91	04-Mar-2008	cube	branches: 1.91.2; 1.91.4; Split device_t/softc, and other related cosmetic changes.
1.90	19-Oct-2007	ad	branches: 1.90.12; 1.90.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.89	04-Mar-2007	christos	branches: 1.89.2; 1.89.14; 1.89.16; 1.89.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.88	16-Nov-2006	christos	branches: 1.88.4; 1.88.6; __unused removal on arguments; approved by core.
1.87	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.86	30-Aug-2006	christos	branches: 1.86.2; 1.86.4; fix incomplete initializers
1.85	01-Jul-2006	chap	Correct build failures when MIDI is not enabled (the idea!). Specifically eap but no midi, which makes sense if you want sound, and umidi but no midi, which makes little sense but shouldn't fail the build. Suggest changing the evbarm conf files containing umidi but #midi to comment out umidi as well (even though this fix should let them build).
1.84	30-Jun-2006	chap	Ankh-Morpork, we have a MIDI driver.... Merge from chap-midi branch, after ~month for review Comments by thorpej@ drochner@ and Alexandre Ratchov Incorporated: points by thorpej@ drochner@; preliminary support for a stats-collecting ioctl suggested by martin@ from comments by A.R. PR kern/32441 kern/32442 kern/32567 kern/32588 kern/32694 kern/33590 kern/33614 and one instance of kern/32651 ok martin@
1.83	13-Jun-2006	rpaulo	branches: 1.83.2; Initialize the mixer delta value to 8 for the 1370 version. No objections in tech-kern.
1.82	10-Jun-2006	rpaulo	KNF.
1.81	24-Dec-2005	perry	branches: 1.81.4; 1.81.8; 1.81.14; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.80	11-Dec-2005	christos	merge ktrace-lwp.
1.79	28-Jun-2005	thorpej	branches: 1.79.2; Use ANSI function decls and static.
1.78	27-Feb-2005	perry	nuke trailing whitespace
1.77	13-Feb-2005	fredb	"output.source" on the Ensoniq 1371 really should be "input.source". Making it so lets the audio(4) interface -- play.port and play.gain -- attach properly. Analysis and patch provided by Joachim Kuebart in PR kern/29178.
1.76	15-Jan-2005	kent	branches: 1.76.2; 1.76.4; ansify and KNF
1.75	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.74	09-Nov-2004	kent	branches: 1.74.2; remove mixer setting code in *_attach() because ac97_attach() does it.
1.73	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.72	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.71	03-Aug-2004	drochner	Check return values of config_detach() for child devices and abort our own detach() on errors.
1.70	09-Jul-2004	mycroft	Clean up. Attempt to make these safer against stray interrupts.
1.69	08-Jul-2004	drochner	support the game port on eap1371 This is only usable with some caution because these soundcards only allow to map IO port 0x20x for this, thus bypassing PCI address management. Very likely this will only work on primary PCI buses, and there is some potential for conflicts with ISA devices as well. (XXX cannot be detached because the "joy" driver doesn't support it yet)
1.68	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.67	04-Dec-2003	keihan	branches: 1.67.2; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.66	03-May-2003	wiz	branches: 1.66.2; DMA, not dma nor Dma.
1.65	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.64	31-Jan-2003	thorpej	Use aprint_*().
1.63	14-Jan-2003	mrg	add detach support
1.62	26-Dec-2002	pooka	For the ES1370 do "best effort" rate selection for the FM DAC. The card contains only one clock, which is already used by the other DAC. The FM DAC can handle a few fixed-frequency choices. thanks to Matthew Green for testing
1.61	14-Dec-2002	pooka	Silly me for assuming that the CT5880 cards could be identified by looking at which ones had the Product ID CT5880. Try to get the checks right now. Also add ES1373_8 to the list.
1.60	05-Dec-2002	pooka	Recognize a few more CT5880 chips. Only cosmetic, does not affect functionality.
1.59	30-Nov-2002	pooka	Don't use tsleep() in attach, use delay() This change makes the driver work properly for the following card, which used to function a bit weird: eap0 at pci0 dev 10 function 0: Ensoniq CT5880 CT5880C (rev. 0x02) eap0: interrupting at irq 9 eap0: TriTech TR28602 codec; no 3D stereo
1.58	24-Nov-2002	scw	Fix uninitialised variable warnings.
1.57	14-Nov-2002	gson	eap.c 1.54 changed ES1371 specific code but failed to make the corresponding cases to the ES1370 case, causing a panic when attaching an ES1370.
1.56	11-Nov-2002	pooka	rename EAP_EN() to the more descriptive EAP_DAC_EN(), and use it everywhere.
1.55	07-Nov-2002	thorpej	Don't use newlines in string literals.
1.54	04-Nov-2002	pooka	add support for the second DAC found on the chip. to attach the second audio device, define EAP_USE_BOTH_DACS
1.53	10-Oct-2002	pooka	make a few debug printf's match reality
1.52	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.51	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.50	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.49	04-Jul-2002	gson	Mask MCCB interrupts since they are not currently handled and it is not clear how they should be handled. Fixes kern/17071.
1.48	13-Nov-2001	lukem	branches: 1.48.8; 1.48.10; add RCSID
1.47	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.46	03-Sep-2001	reinoud	branches: 1.46.2; Fixed compiling error when AUDIO_DEBUG was switched on ...
1.45	24-Apr-2001	mycroft	branches: 1.45.2; It appears there's a race condition when disabling a channel that causes us to need to explicitly relatch the interrupt when firing it up again. So, in the trigger routines, explicitly disable and reenable the interrupt to relatch it, like we do in the interrupt routine. Also clean up some broken loop overrun checks. My ES1371 seems to be more reliable now, but I'm not going to pretend to fully understand this chip.
1.44	27-Jan-2001	augustss	branches: 1.44.2; ANSIfy.
1.43	27-Jan-2001	augustss	Sleep a little when closing the midi device to allow the UART to drain.
1.42	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.41	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.40	01-Jun-2000	augustss	branches: 1.40.2; Update URLs.
1.39	15-May-2000	thorpej	branches: 1.39.2; ac97.h -> ac97var.h
1.38	01-May-2000	augustss	Change the way the codec is access to mimic how Linux&FreeBSD does it.
1.37	30-Apr-2000	augustss	Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.36	24-Apr-2000	augustss	Move a test if the MIDI is open outside a loop in the interrupt routine. Some stylistic changes.
1.35	19-Mar-2000	soren	Make it a bit more clear which functions are specific to which models. Split off register definitions into eapreg.h. Remove duplicate codec reset.
1.34	11-Mar-2000	soren	The shortened 1371 reset delay in revision 1.32 doesn't seem to be quite enough for my 1373, so increase it some.
1.33	06-Mar-2000	perry	Make Creative Soundblaster PCI 128's work -- from Masahiro Ishiyama Note that on device close, there's a sharp pop that probably should be fixed. Note also that the device is hanging the machine at times for brief periods inside audio.c::audiogetinfo(), but I haven't yet figured out why.
1.32	02-Nov-1999	augustss	Shorten some delay()s. From WITEK Wnuk <witek@atari.org>
1.31	02-Nov-1999	soren	Fix crash when both 1370 and 1371/1373 devices are configured. Update documentation links. XXX The initial volume of the 1371 mixers is set to 255. Should be ~200.
1.30	01-Nov-1999	augustss	Make little endian and big endian version of all functions that deal with 16 bit samples. Let the drivers decide which ones to use.
1.29	29-Oct-1999	augustss	Add MIDI support. Closes PR 8578.
1.28	27-Oct-1999	augustss	Add support for ES1371. From OpenBSD and Ezra Story <ezy@panix.com>.
1.27	01-Sep-1999	kleink	branches: 1.27.2; 1.27.4; 1.27.6; Don't make a human-readable device description being printed at attach time depend on PCIVERBOSE.
1.26	10-Jul-1999	kleink	Fix a buglet in the driver's DMA buffer list ipmlementation, causing the next- in-list (wrong) descriptor structure of the buffer in question to be freed.
1.25	18-Feb-1999	mycroft	branches: 1.25.4; Pass the correct address (PCI bus address, not PA) to the card. Works on an Alpha now.
1.24	18-Feb-1999	mycroft	Minor code munging.
1.23	17-Feb-1999	mycroft	Fix a silly type.
1.22	17-Feb-1999	mycroft	Update comments.
1.21	17-Feb-1999	mycroft	Pass the direction to the allocm and round_buffersize methods. Some drivers need this to properly allocate DMAable memory.
1.20	08-Jan-1999	augustss	Fix more `void *' arithmetic.
1.19	25-Nov-1998	augustss	Make the copyright header conform to the NetBSD template.
1.18	19-Nov-1998	mrg	fix problems in many d_mmap routines: - returned EOPNOTSUPP rather than -1. - no check for negative offset. many of these fix potential security problems in these drivers. XXX XXX XXX the d_mmap cdev routine should be changed to have a prototype like: paddr_t (*d_mmap) __P((dev_t, off_t, int)); by someone!
1.17	25-Aug-1998	thorpej	Add some braces to make egcs happy.
1.16	17-Aug-1998	augustss	* Redo the way the way the MIDI driver attaches to audio devices. * Improve the midisyn layer a little. * Add a driver for the Yamaha OPL[23] FM synths. The opl driver is not finished yet; it sounds pretty awful. For some strange reason I cannot get any FM sound from my SB64 cards, but a regular SB16 works fine.
1.15	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.14	12-Aug-1998	mycroft	Associate the preamp with the mic volume in the input class.
1.13	12-Aug-1998	mycroft	Oops; fix typos.
1.12	12-Aug-1998	mycroft	Add a knob to enable/disable the mic preamp.
1.11	10-Aug-1998	mycroft	For consistency, only do the rate coupling in full-duplex mode.
1.10	09-Aug-1998	mycroft	Minor cleanup to debugging cruft.
1.9	09-Aug-1998	mycroft	Several things: * Use the new trigger_{in,out}put interface. * Allow the play and record channels to have different encodings (but the same sample rate). * Play u-law and a-law as 16-bit data. (This may not work yet...)
1.8	06-Jul-1998	augustss	Add mixer output selection. From gson@araneus.fi (Andreas Gustafsson)
1.7	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.6	26-May-1998	augustss	Fix a recording problem that occurs when the EAP chip sends its interrupt just BEFORE the dma has finished rather than after.
1.5	06-May-1998	augustss	Remove some lint.
1.4	02-May-1998	mycroft	Adjust PCLKDIV by 2 to correct the output sample rate. (Unfortunately, this isn't documented.) Fix typo in mappage routine.
1.3	02-May-1998	mycroft	Oops; fix typo.
1.2	02-May-1998	mycroft	Fix reading of the master volume, and a hack that seems to make stereo output work correctly. (Will check the documentation as soon as I can read it.)
1.1	01-May-1998	augustss	Add a driver for the Ensoniq AudioPCI sound card. The driver still needs some testing, but it seems to produce sound. The driver was written by me, but since I don't have the hardware the debugging and testing was done by Andreas Gustafsson <gson@araneus.fi>, Chuck Cranor <chuck@maria.wustl.edu>, and Phil Nelson <phil@cs.wwu.edu>. Thanks.
1.25.4.1	02-Aug-1999	thorpej	Update from trunk.
1.27.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.27.4.1	15-Nov-1999	fvdl	Sync with -current
1.27.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.27.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.27.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.39.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.40.2.2	01-May-2001	he	Pull up revision 1.45 (requested by hubertf): Work around race condition by relatching the interrupt. Makes the eap driver not hang on certain occasions.
1.40.2.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.44.2.11	15-Jan-2003	thorpej	Sync with HEAD.
1.44.2.10	29-Dec-2002	thorpej	Sync with HEAD.
1.44.2.9	19-Dec-2002	thorpej	Sync with HEAD.
1.44.2.8	11-Dec-2002	thorpej	Sync with HEAD.
1.44.2.7	11-Nov-2002	nathanw	Catch up to -current
1.44.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.44.2.5	01-Aug-2002	nathanw	Catch up to -current.
1.44.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.44.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.44.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.44.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.45.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.45.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.45.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.45.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.46.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.48.10.1	15-Dec-2002	he	Pull up revision 1.59 (requested by pooka in ticket #1035): Don't use tsleep() in attach, use delay(). Makes the driver work properly with ``Ensoniq CT5880''-based audio card.
1.48.8.1	15-Jul-2002	gehenna	catch up with -current.
1.66.2.11	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.66.2.10	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.66.2.9	15-Feb-2005	skrll	Sync with HEAD.
1.66.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.66.2.7	14-Nov-2004	skrll	Sync with HEAD.
1.66.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.66.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.66.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.66.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.66.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.66.2.1	03-Aug-2004	skrll	Sync with HEAD
1.67.2.1	22-Sep-2004	jmc	Pullup rev 1.72 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.74.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.74.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.76.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.76.2.1	29-Apr-2005	kent	sync with -current
1.79.2.5	17-Mar-2008	yamt	sync with head.
1.79.2.4	27-Oct-2007	yamt	sync with head.
1.79.2.3	03-Sep-2007	yamt	sync with head.
1.79.2.2	30-Dec-2006	yamt	sync with head.
1.79.2.1	21-Jun-2006	yamt	sync with head.
1.81.14.2	19-Jun-2006	chap	Sync with head.
1.81.14.1	18-Jun-2006	chap	Updates PR kern/32651 The eap driver now makes use of the MIDI UART tx interrupt available on the Ensoniq chips. A test with nominally 20 seconds of continuous transmit data went from 26 s real / 25 s system to 22 s real / 0.16 s system. Implementation is quite simple, and anyone who knows any of the other MIDI UART chips well enough to get a tx interrupt out of it would do a world of good by updating their drivers similarly. Anyway, scratch one off the kern/32651 todo list. Also added a documentation reference to the ES1373 datasheet, and added more 1371- and 1373-specific #defines in eapreg.h. Enough there now that anyone who wanted to play with the S/PDIF capability on the 1373 could begin to do that.
1.81.8.3	03-Sep-2006	yamt	sync with head.
1.81.8.2	11-Aug-2006	yamt	sync with head
1.81.8.1	26-Jun-2006	yamt	sync with head.
1.81.4.1	09-Sep-2006	rpaulo	sync with head
1.83.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.86.4.2	10-Dec-2006	yamt	sync with head.
1.86.4.1	22-Oct-2006	yamt	sync with head
1.86.2.1	18-Nov-2006	ad	Sync with head.
1.88.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.88.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.89.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.89.16.2	23-Mar-2008	matt	sync with HEAD
1.89.16.1	06-Nov-2007	matt	sync with HEAD
1.89.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.89.2.1	23-Oct-2007	ad	Sync with head.
1.90.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.90.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.90.12.1	24-Mar-2008	keiichi	sync with head.
1.91.4.1	16-May-2008	yamt	sync with head.
1.91.2.1	18-May-2008	yamt	sync with head.
1.92.36.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.92.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.92.34.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.92.34.1	17-Apr-2012	yamt	sync with head
1.92.12.5	12-Dec-2008	ad	Checkpoint work in progress.
1.92.12.4	11-Dec-2008	ad	Checkpoint work in progress.
1.92.12.3	09-Dec-2008	ad	Checkpoint work on MIDI.
1.92.12.2	08-Dec-2008	ad	Death to NOWAIT.
1.92.12.1	07-Dec-2008	ad	Ressurect the ad-audiomp branch at request of matt@. With this commit: - core audio subsystem made MPSAFE - eap, btsco drivers made MPSAFE TODO: - midi - convert remaining drivers
1.94.12.1	18-May-2014	rmind	sync with head
1.94.8.2	03-Dec-2017	jdolecek	update from HEAD
1.94.8.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.95.6.2	28-Aug-2017	skrll	Sync with HEAD
1.95.6.1	09-Jul-2016	skrll	Sync with HEAD
1.97.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.97.10.1	10-Jun-2019	christos	Sync with HEAD
1.97.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.99.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.99.2.3	30-Apr-2019	isaki	More adapt to audio2. - Use set_format. - Remove an empty method. - XXX DAC1 and secondary audio device support should be removed. Now mixing two (or more) sources is done by audio layer.
1.99.2.2	21-Apr-2019	isaki	Adapt to audio2. - Drop INDEPENDENT property from es1370 which has only one clock.
1.99.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.102.28.1	17-Jun-2024	martin	Pull up following revision(s) (requested by nia in ticket #697): sys/dev/pci/eap.c: revision 1.103 eap(4): Restore the full set of supported hardware encodings from netbsd-7 Tested in QEMU.
1.102.2.1	17-Jun-2024	martin	Pull up following revision(s) (requested by nia in ticket #1843): sys/dev/pci/eap.c: revision 1.103 eap(4): Restore the full set of supported hardware encodings from netbsd-7 Tested in QEMU.
1.14	21-May-2024	nia	eap(4): re-organize mixer controls to put classes first some software (understandably) doesn't like it if you declare a mixer node's child before the parent, notably ossxmix in pkgsrc
1.13	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.12	30-Jun-2006	chap	branches: 1.12.54; 1.12.56; 1.12.58; Ankh-Morpork, we have a MIDI driver.... Merge from chap-midi branch, after ~month for review Comments by thorpej@ drochner@ and Alexandre Ratchov Incorporated: points by thorpej@ drochner@; preliminary support for a stats-collecting ioctl suggested by martin@ from comments by A.R. PR kern/32441 kern/32442 kern/32567 kern/32588 kern/32694 kern/33590 kern/33614 and one instance of kern/32651 ok martin@
1.11	11-Dec-2005	christos	branches: 1.11.4; 1.11.8; 1.11.14; 1.11.16; merge ktrace-lwp.
1.10	13-Feb-2005	fredb	branches: 1.10.6; "output.source" on the Ensoniq 1371 really should be "input.source". Making it so lets the audio(4) interface -- play.port and play.gain -- attach properly. Analysis and patch provided by Joachim Kuebart in PR kern/29178.
1.9	15-Jan-2005	kent	branches: 1.9.2; 1.9.4; ansify and KNF
1.8	08-Jul-2004	drochner	support the game port on eap1371 This is only usable with some caution because these soundcards only allow to map IO port 0x20x for this, thus bypassing PCI address management. Very likely this will only work on primary PCI buses, and there is some potential for conflicts with ISA devices as well. (XXX cannot be detached because the "joy" driver doesn't support it yet)
1.7	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	14-Dec-2002	pooka	branches: 1.6.6; Silly me for assuming that the CT5880 cards could be identified by looking at which ones had the Product ID CT5880. Try to get the checks right now. Also add ES1373_8 to the list.
1.5	05-Dec-2002	pooka	Recognize a few more CT5880 chips. Only cosmetic, does not affect functionality.
1.4	11-Nov-2002	pooka	rename EAP_EN() to the more descriptive EAP_DAC_EN(), and use it everywhere.
1.3	04-Nov-2002	pooka	add support for the second DAC found on the chip. to attach the second audio device, define EAP_USE_BOTH_DACS
1.2	30-Apr-2000	augustss	branches: 1.2.6; 1.2.8; Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.1	19-Mar-2000	soren	Make it a bit more clear which functions are specific to which models. Split off register definitions into eapreg.h. Remove duplicate codec reset.
1.2.8.3	19-Dec-2002	thorpej	Sync with HEAD.
1.2.8.2	11-Dec-2002	thorpej	Sync with HEAD.
1.2.8.1	11-Nov-2002	nathanw	Catch up to -current
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	30-Apr-2000	bouyer	file eapreg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:19 +0000
1.6.6.5	15-Feb-2005	skrll	Sync with HEAD.
1.6.6.4	17-Jan-2005	skrll	Sync with HEAD.
1.6.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.6.1	03-Aug-2004	skrll	Sync with HEAD
1.9.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.2.1	29-Apr-2005	kent	sync with -current
1.10.6.1	30-Dec-2006	yamt	sync with head.
1.11.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.11.14.1	18-Jun-2006	chap	Updates PR kern/32651 The eap driver now makes use of the MIDI UART tx interrupt available on the Ensoniq chips. A test with nominally 20 seconds of continuous transmit data went from 26 s real / 25 s system to 22 s real / 0.16 s system. Implementation is quite simple, and anyone who knows any of the other MIDI UART chips well enough to get a tx interrupt out of it would do a world of good by updating their drivers similarly. Anyway, scratch one off the kern/32651 todo list. Also added a documentation reference to the ES1373 datasheet, and added more 1371- and 1373-specific #defines in eapreg.h. Enough there now that anyone who wanted to play with the S/PDIF capability on the 1373 could begin to do that.
1.11.8.1	11-Aug-2006	yamt	sync with head
1.11.4.1	09-Sep-2006	rpaulo	sync with head
1.12.58.1	16-May-2008	yamt	sync with head.
1.12.56.1	18-May-2008	yamt	sync with head.
1.12.54.1	02-Jun-2008	mjf	Sync with HEAD.
1.6	05-Mar-2020	riastradh	Revert "Include opt_diagnostic.h for DIAGNOSTIC." This did not do what I thought it did. opt_diagnostic.h is only for the unused _DIAGNOSTIC, which seems like an abortive attempt to incrementally convert DIAGNOSTIC to an opt_*.h option rather than a command-line option.
1.5	05-Mar-2020	riastradh	Include opt_diagnostic.h for DIAGNOSTIC. ...at least, in header files, which may not have already included libkern.h.
1.4	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.3	06-May-2009	cegger	branches: 1.3.12; 1.3.14; struct device * -> device_t, no functional changes intended.
1.2	11-Dec-2005	christos	branches: 1.2.74; 1.2.88; 1.2.90; merge ktrace-lwp.
1.1	08-Jul-2004	drochner	branches: 1.1.2; support the game port on eap1371 This is only usable with some caution because these soundcards only allow to map IO port 0x20x for this, thus bypassing PCI address management. Very likely this will only work on primary PCI buses, and there is some potential for conflicts with ISA devices as well. (XXX cannot be detached because the "joy" driver doesn't support it yet)
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	03-Aug-2004	skrll	Sync with HEAD
1.1.2.1	08-Jul-2004	skrll	file eapvar.h was added on branch ktrace-lwp on 2004-08-03 10:49:06 +0000
1.2.90.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.2.88.1	12-Dec-2008	ad	Checkpoint work in progress.
1.2.74.1	16-May-2009	yamt	sync with head
1.3.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.3.12.1	17-Apr-2012	yamt	sync with head
1.78	31-Mar-2025	riastradh	ehci(4): Sprinkle KERNEL_LOCKED_P and KNF on usb__add/rem. Mark with XXXSMP comments to be fixed later without the kernel lock. No functional change intended: callers generally do this from autoconf _attach/detach routines, which run kernel-locked anyway (for the moment). Cleanup in preparation for: PR port-amd64/59180: System reboots instead of shutting down
1.77	24-Mar-2024	mrg	branches: 1.77.2; ehci(4): properly handle failed attach thinkpad a475 fails to attach an ehci instance: ehci0: pre-2.0 USB rev, device ignored which ends up stopping suspend/resume working as the device has no pmf handlers installed. put most of the teardown code into a new common function that is called from failed attach and detach. if attach fails, register NULL pmf handlers. don't check the return value of pmf_device_register*(). re-order several parts of detach to match the reverse attach order. tested on rockpro64, thinkpad a475, and xps 1645, the a475 can now suspend/resume almost fully successfully.
1.76	24-Jan-2023	mlelstv	Consistently use cached chipset tag value.
1.75	28-Oct-2022	riastradh	branches: 1.75.2; ehci(4): Mark PCI interrupt handler MP-safe. ehci_intr has its own intr lock to coordinate with the MP-safe softint it defers all its work to, other than reading and writing a few registers to get and acknowledge the interrupt status.
1.74	13-Mar-2022	riastradh	ehci(4): Serialize access to portsc registers. Both ehci_roothub_ctrl and ehci_suspend/resume do r/m/w on them, so use a mutex to serialize access to avoid stomping on each other.
1.73	22-Dec-2021	skrll	Three fixes - pass the 64bit DMA tag if the HCCPARAMS says ehci supports it and the 64bit DMA tag is available/valid. This should help with the "cannot create xfer" error on 64bit systems. - restrict the control structure memory allocation to the low 4GB (CTRLDSSEGMENT was always set to zero anyway) - if ehci_init fails then tidyup before returning the error.
1.72	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.71	24-Apr-2021	thorpej	branches: 1.71.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.70	13-Jun-2019	maxv	branches: 1.70.12; Fix the error handling in ehci_pci_attach(): if we got a USB<2 device we won't call ehci_init(), so don't call ehci_detach() in ehci_pci_detach(). Fixes a panic seen on a recent Lenovo machine, which has an USB 1.1 controller; ehci_detach() was getting called while 'sc' had not been completely initialized.
1.69	13-Jun-2019	maxv	Random style in ehci, also KM_SLEEP does not fail.
1.68	25-Oct-2018	jdolecek	enable MSI support where available
1.67	10-May-2018	msaitoh	branches: 1.67.2; KNF. No binary change.
1.66	09-Apr-2018	jakllsch	Stop potential misuse of vendor names and USB vendor IDs in root hub device and string descriptors. Firstly: Few vendors have identical PCI-SIG vendor IDs and USB-IF vendor IDs. As such, using the PCI vendor ID as a USB vendor ID may trample on whomever is allocated that USB vendor ID. Secondly: The vendor of the host controller hardware implementation has little to nothing to do with our usbroothub implementation. Thus we should not potentially associate any problems therewith to such third party. This change will result in root hubs being identified by USB Vendor ID 0x0000. Root hub vendor string will now be "NetBSD" (or, specifically: ostype). Product ID (0x0000) and product strings remain unchanged.
1.65	12-Jun-2017	sborrill	branches: 1.65.4; Deal with broken BIOSes that leave PCI interrupts disabled. Thanks to nick@
1.64	13-Oct-2016	jdolecek	branches: 1.64.8; provide intr xname
1.63	23-Apr-2016	skrll	branches: 1.63.2; Merge nick-nhusb - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix
1.62	31-Aug-2015	skrll	Improve error handling in attach code. From t-hash.
1.61	19-Aug-2015	skrll	Interrupt handlers aren't MP-safe yet so use IPL_USB which is IPL_VM which takes the kernel lock.
1.60	15-Jul-2015	msaitoh	Initialize some members little earlier. It's not a bug.
1.59	21-Sep-2014	christos	branches: 1.59.2; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.58	29-Mar-2014	christos	branches: 1.58.4; 1.58.8; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.57	22-Sep-2012	tsutsui	branches: 1.57.2; Fix PR kern/46828 (6.0_BETA2 and 6.0_RC1 won't start on DL320/G5p): In ehci_get_ownership(), don't explicitly clear EHCI_LEG_HC_BIOS_OWNED semaphore bit in the driver before asking BIOS to give up ownership. The EHCI spec implies that the semaphore should not be changed by the other agent and actually the previous one (introduced in rev 1.53 after 5.x) caused hangup during probe on at least two HP machines as mentioned in the PR. Analyzed and patch provided by Ryo ONODERA. Should be pulled up to netbsd-6 (fatal hangup during boot).
1.56	20-Jul-2012	uwe	branches: 1.56.2; EHCI_USBINTR is 4 bytes wide, use proper access ops.
1.55	10-Jun-2012	mrg	merge the jmcneill-usbmp branch. many thanks to jared for the initial work, and every one else who has tested things for me. this is largely my fault at this point :-) the main changes are something like: - usbd_bus_methods{} gains a get_lock() to enable the host controller to provide a lock for the USB code. if the lock isn't provided, old-style protection is (partially) applied. - ehci/ohci/uhci have been converted to the new interfaces, including mutex/cv/etc conversion. - usbdivar.h contains a discussion about locking and what locks are held for which method calls. more to come for usbdi(9) here. - audio drivers (uaudio, umidi, auvitek) have been properly SMPified now that USB is ready. - scsi drivers have been modified to take the kernel lock explicitly before calling into scsi code. - usb pipes are associated with a lock, that is the same as the controller lock. (this could be split up further in the future.) - several usbfoo_locked() or usbfoo_unlocked() functions have been added to the usbdi(9) to enable functionality with or without the USB lock (per controller) already being held. the TODO.usbmp file has specific details on what is left to do, including what device-specific changes should be done now that the whole framework is ready.
1.54	30-Jan-2012	drochner	branches: 1.54.2; 1.54.4; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.53	30-Jul-2011	jmcneill	branches: 1.53.2; 1.53.6; ehci_get_ownership: clear BIOS semaphore when setting OS semaphore, slow down poll rate, and make sure to clear all SMI bits when we're done
1.52	04-Apr-2011	dyoung	Deregister PMF handlers during detachment. Complete the host controller shutdown.
1.51	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.50	11-Dec-2010	matt	branches: 1.50.2; Make ehci deal with pci devices that have multiple companion functions and ehci functions. We can't assume that there will only be one ehci per device. The existing code could not deal with: Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 0 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 1 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 2 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 3 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 4 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 5 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 6 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 7 not configured
1.49	25-May-2010	pgoyette	Revert earlier changes to callers of pci_find{vendor,product} since the original routine entrypoint name has been restored. Requested by dyoung@ and mrg@
1.48	24-May-2010	pgoyette	Update all callers of the pci_find{vendor,product} routines to now call these routines through their global pointers.
1.47	24-Feb-2010	dyoung	branches: 1.47.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.46	08-Jan-2010	dyoung	branches: 1.46.2; Expand PMF_FN_* macros.
1.45	15-Jun-2009	cegger	Apply hw workaround required for all SB600 revisions and SB700 revisions A12 and A13 to avoid USB subsystem hang symptom. The USB subsystem hang symptom is observed when the system has multiple USB devices connected to it or one USB device is often re-connected. In some cases a USB hub may be required to observe this symptom. This patch works around the problem by correcting the internal register setting that will help by changing the behavior of the internal logic to avoid the USB subsystem hang issue. The change in the behavior of the logic does not impact the normal operation of the USB subsystem. This fix has been discussed, developped, reviewed, polished up and tested on current-users by several people. Thread starts at: http://mail-index.netbsd.org/current-users/2009/05/17/msg009460.html
1.44	26-Apr-2009	cegger	u_intNN_t -> uintNN_t
1.43	26-Apr-2009	cegger	- Fix error handling in attachment - use aprint_*_dev
1.42	17-Apr-2009	christos	appease gcc
1.41	17-Apr-2009	dyoung	Use device_t instead of 'struct device ' or device_ptr_t. Use cfdata_t instead of 'struct cfdata '. Delete some gratuitous parentheses.
1.40	17-Apr-2009	dyoung	Disable interrupts on the h/w before detaching the interrupt handler. Protects against the problem described in <http://mail-index.netbsd.org/port-i386/2009/04/10/msg001317.html>, Xintr_ioapic_levelN jumps to 0 during shutdown.
1.39	07-Apr-2009	dyoung	Detach ehci(4) and uhci(4) at shutdown.
1.38	28-Apr-2008	martin	branches: 1.38.8; 1.38.10; 1.38.14; 1.38.16; Remove clause 3 and 4 from TNF licenses
1.37	28-Mar-2008	drochner	branches: 1.37.2; 1.37.4; split device/softc for USB host controllers and the usb (control) device, this is hairy stuff, and I've only tested with uhci/ehci at pci, please test the rest and report problems
1.36	07-Mar-2008	dyoung	Use device_t and accessors. Setup power management in the PCI front-end (XXX needs to look more alike the Cardbus front-end). Establish the shutdown hook using PMF.
1.35	22-Feb-2008	dyoung	Add a method to detach child. Cosmetic: use device_t and accessors. Use aprint_*_dev(). Use PMF_FN_{ARGS,CALL,PROTO}.
1.34	28-Jan-2008	jmcneill	branches: 1.34.2; 1.34.6; Release ownership of EHCI to the BIOS when we suspend or detach the device driver.
1.33	27-Jan-2008	jmcneill	Don't attempt to claim ownership from the BIOS if the OS already holds the HC OS Owned semaphore. Unconditionally clear SMIs as long as the USB legacy capability is present.
1.32	09-Dec-2007	jmcneill	Merge jmcneill-pm branch.
1.31	19-Oct-2007	ad	branches: 1.31.2; 1.31.4; 1.31.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.30	04-Aug-2007	tsutsui	branches: 1.30.2; 1.30.6; 1.30.8; ATI's ehci also requires EHCIF_DROPPED_INTR_WORKAROUND. From FreeBSD.
1.29	01-Aug-2007	christos	branches: 1.29.2; Establish powerhooks in the proper order. From Anon Ymous
1.28	08-Jul-2007	jmcneill	branches: 1.28.2; If the USB legacy capability flag is not set, we mistakenly attempt to clear it anyway by writing to PCI configuration register 0x00. Instead, if this flag is not set, simply bail out from ehci_get_ownership.
1.27	09-Feb-2007	ad	branches: 1.27.6; 1.27.8; 1.27.14; Merge newlock2 to head.
1.26	16-Nov-2006	christos	branches: 1.26.2; 1.26.4; __unused removal on arguments; approved by core.
1.25	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.24	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.23	10-Mar-2006	jmcneill	branches: 1.23.10; 1.23.12; On suspend, capture PCI configuration registers. On resume, restore them as well as explicitly reclaiming ownership of the UHCI/EHCI from the BIOS.
1.22	17-Jan-2006	xtraeme	branches: 1.22.2; 1.22.4; 1.22.6; 1.22.8; From OpenBSD: -- Add a workaround for VIA EHCI controllers which, under load, signal qTD completion before they have performed writeback from the overlay qTD. This condition would exhibit itself as a umass stall that never recovers. -- This fixes the problem reported by Thomas Klausner on current-users@: http://mail-index.netbsd.org/current-users/2006/01/17/0000.html
1.21	20-Nov-2005	augustss	branches: 1.21.2; Print an error message if we cannot located the legacy capability.
1.20	20-Nov-2005	augustss	Make debug printout a little nicer.
1.19	20-Nov-2005	augustss	Go through the prescribed process to get ownership of the controller from BIOS before we start using it. This seems to fix the problems some people have been experiencing with keyboards not working.
1.18	28-Jun-2005	thorpej	branches: 1.18.2; 1.18.8; Use ANSI function decls and static.
1.17	27-Feb-2005	perry	branches: 1.17.2; nuke trailing whitespace
1.16	02-Aug-2004	mycroft	branches: 1.16.4; 1.16.6; For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.15	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.14	22-Apr-2004	itojun	sprintf -> snprintf
1.13	31-Jan-2003	thorpej	branches: 1.13.2; Use aprint_*().
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	22-Aug-2002	augustss	Fix an error message.
1.8	14-Feb-2002	augustss	branches: 1.8.8; Use the correct years for copyright.
1.7	09-Jan-2002	augustss	Split out some code so you can have cardbus ehci without pci.
1.6	13-Nov-2001	lukem	add RCSID
1.5	10-Nov-2001	augustss	Fill the companion controller array.
1.4	06-Nov-2001	augustss	Add some data structures to keep enable finding companion controllers for EHCI.
1.3	28-Dec-2000	sommerfeld	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.10; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.2	24-Dec-2000	augustss	Even stubs need improvements.
1.1	24-Dec-2000	augustss	Add some placeholders for the EHCI (USB 2) driver. Don't get your hopes up. I've not even finished reading the (100+) page spec, and I have no hardware.
1.3.10.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.3.6.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.6.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.6.2	16-Mar-2002	jdolecek	Catch up with -current.
1.3.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.4.5	18-Oct-2002	nathanw	Catch up to -current.
1.3.4.4	27-Aug-2002	nathanw	Catch up to -current.
1.3.4.3	28-Feb-2002	nathanw	Catch up to -current.
1.3.4.2	11-Jan-2002	nathanw	More catchup.
1.3.4.1	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.3.2.1	28-Dec-2000	bouyer	file ehci_pci.c was added on branch thorpej_scsipi on 2001-01-05 17:36:03 +0000
1.8.8.1	29-Aug-2002	gehenna	catch up with -current.
1.13.2.6	11-Dec-2005	christos	Sync with head.
1.13.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.13.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.2.1	03-Aug-2004	skrll	Sync with HEAD
1.16.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.16.4.1	29-Apr-2005	kent	sync with -current
1.17.2.2	22-Jan-2006	tron	Pull up following revision(s) (requested by xtraeme in ticket #1123): sys/dev/usb/ehcivar.h: revision 1.23 sys/dev/pci/ehci_pci.c: revision 1.22 sys/dev/usb/ehci.c: revision 1.108 From OpenBSD: -- Add a workaround for VIA EHCI controllers which, under load, signal qTD completion before they have performed writeback from the overlay qTD. This condition would exhibit itself as a umass stall that never recovers. -- This fixes the problem reported by Thomas Klausner on current-users@: http://mail-index.netbsd.org/current-users/2006/01/17/0000.html
1.17.2.1	07-Dec-2005	riz	Pull up following revision(s) (requested by gendalia in ticket #1034): sys/dev/pci/ehci_pci.c: revisions 1.18-1.21 sys/dev/usb/ehcireg.h: revisions 1.22-1.23 Go through the prescribed process to get ownership of the controller from BIOS before we start using it. This seems to fix the problems some people have been experiencing with keyboards not working. Pull up additional code formatting, debugging, and a few #defines. Fixes PR#32249.
1.18.8.1	22-Nov-2005	yamt	sync with head.
1.18.2.9	17-Mar-2008	yamt	sync with head.
1.18.2.8	27-Feb-2008	yamt	sync with head.
1.18.2.7	04-Feb-2008	yamt	sync with head.
1.18.2.6	21-Jan-2008	yamt	sync with head
1.18.2.5	27-Oct-2007	yamt	sync with head.
1.18.2.4	03-Sep-2007	yamt	sync with head.
1.18.2.3	26-Feb-2007	yamt	sync with head.
1.18.2.2	30-Dec-2006	yamt	sync with head.
1.18.2.1	21-Jun-2006	yamt	sync with head.
1.21.2.1	01-Feb-2006	yamt	sync with head.
1.22.8.1	19-Apr-2006	elad	sync with head.
1.22.6.1	13-Mar-2006	yamt	sync with head.
1.22.4.1	22-Apr-2006	simonb	Sync with head.
1.22.2.1	09-Sep-2006	rpaulo	sync with head
1.23.12.2	10-Dec-2006	yamt	sync with head.
1.23.12.1	22-Oct-2006	yamt	sync with head
1.23.10.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.23.10.1	18-Nov-2006	ad	Sync with head.
1.26.4.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.26.2.1	09-Jul-2007	liamjfoy	Pull up following revision(s) (requested by jmcneill in ticket #760): sys/dev/pci/ehci_pci.c: revision 1.28 If the USB legacy capability flag is not set, we mistakenly attempt to clear it anyway by writing to PCI configuration register 0x00. Instead, if this flag is not set, simply bail out from ehci_get_ownership.
1.27.14.1	22-May-2007	itohy	Overhaul of USB stack, mostly DMA related This applies to NetBSD 4.99.13 (March 1, 2007) usbdi(9) interface is based on FreeBSD version, excluding - removal of portability code Patch most NetBSD changes, excluding - DMA memory "reserve", since we don't need contiguous buffers any longer - volatiles in DMA structure, since it should not be needed with proper bus_dmamap_sync(9)s DMA/non-DMA memory management overhaul - Move all DMA related code to usb_mem.[ch] (add usb_alloc_buffer_dma(), usb_free_buffer_dma(), etc.). XXX Should usb_mem.[ch] be renamed as usb_mem_dma.[ch] ? - Add corresponding non-DMA code to usb_mem_nodma.[ch] . Currently just use malloc(9). - Above files are conditionally used by config framework (added attributes to conf/files and dev/usb/files.usb). - Add diagnostic panics when resource allocation is requested on interrupt context. - Change memory allocations (that require context) from NOWAIT to WAITOK. Allocate DMA/non-DMA buffer per host interface, not globally. advantage: Buffers can be freed on detaching host interface. Activity of a host interface does not affect others. disadvantages: It possibly consumes more memory. API changes - usbd_alloc_xfer() is changed: old: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev); new: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev, usbd_pipe_handle pipe); - pipe argument of usbd_setup_xfer() are now unused XXX the pipe argument should be removed? - add mapping APIs - async request will be processed as a task (kernel thread context), and delayed to some extent - usbdivar.h: struct usbd_xfer: renamed a member "allocbuf" to "hcbuffer" (mapped/allocated/refered buffer for HCI driver) - usb_port.h: change usb_proc_ptr from struct ptoc to struct lwp * - usb_port.h: add usb_sigproc_ptr for psignal(9) (struct proc ) - usb.h: add UE_MAXPKTSZ(ep) and UE_MAXPKTSZ_MASK macros for USB 2.0 changes to USB device drivers - atu, aue, axe, cdce, cue, kue, rum, udav, upl, ural, url, uaudio, ubt, ucom, ugen, uhidev, uirda, ulpt, umidi, urio, uscanner, ustir, utoppy: catch up API change of usbd_alloc_xfer() - umass, usscanner: * catch up API change of usbd_alloc_xfer() * eliminate memory copy for large transfer ohci - free resources on detach - add lots of bus_dmamap_sync() operations - simplify the code of loading std chain - rewrite code of looking up TD/ITD from DMA addr by using allocation chunk - add workaround for CMD Tech 670 and 673 chipsets - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf slhci - allocate xfer and slhci_xfer at once, and simplify relevant code - add slhci_detach() - remove second arg of slhci_attach() since it is the same as the first arg. - add support for "mapping" (no, it doesn't map since it doesn't do DMA) buffer and mbuf - add pcmcia frontend - NOT TESTED, missing hardware ehci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf - done only simple test uhci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf To do - review, test, debug - rewrite network drivers to utilize usbd_map_buffer_mbuf() - rewrite uaudio(4) to eliminate memcpy - "pipe" argument of usbd_setup_*xfer() should eventually be removed
1.27.8.1	11-Jul-2007	mjf	Sync with head.
1.27.6.3	23-Oct-2007	ad	Sync with head.
1.27.6.2	20-Aug-2007	ad	Sync with HEAD.
1.27.6.1	15-Jul-2007	ad	Sync with head.
1.28.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.29.2.7	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.29.2.6	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.29.2.5	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.29.2.4	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.29.2.3	21-Aug-2007	joerg	Disable interrupts etc before saving the config space and reenable them afterwards. This fixes resume on my Thinkpad R52.
1.29.2.2	04-Aug-2007	jmcneill	Sync with HEAD.
1.29.2.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.30.8.2	04-Aug-2007	tsutsui	ATI's ehci also requires EHCIF_DROPPED_INTR_WORKAROUND. From FreeBSD.
1.30.8.1	04-Aug-2007	tsutsui	file ehci_pci.c was added on branch matt-mips64 on 2007-08-04 10:36:07 +0000
1.30.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.30.2.3	23-Mar-2008	matt	sync with HEAD
1.30.2.2	09-Jan-2008	matt	sync with HEAD
1.30.2.1	06-Nov-2007	matt	sync with HEAD
1.31.6.1	11-Dec-2007	yamt	sync with head.
1.31.4.1	26-Dec-2007	ad	Sync with head.
1.31.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.34.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.34.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.34.2.1	24-Mar-2008	keiichi	sync with head.
1.37.4.4	11-Mar-2010	yamt	sync with head
1.37.4.3	20-Jun-2009	yamt	sync with head
1.37.4.2	04-May-2009	yamt	sync with head.
1.37.4.1	16-May-2008	yamt	sync with head.
1.37.2.1	18-May-2008	yamt	sync with head.
1.38.16.1	17-Jun-2009	bouyer	branches: 1.38.16.1.2; Pull up following revision(s) (requested by cegger in ticket #814): sys/dev/pci/ehci_pci.c: revision 1.45 Apply hw workaround required for all SB600 revisions and SB700 revisions A12 and A13 to avoid USB subsystem hang symptom. The USB subsystem hang symptom is observed when the system has multiple USB devices connected to it or one USB device is often re-connected. In some cases a USB hub may be required to observe this symptom. This patch works around the problem by correcting the internal register setting that will help by changing the behavior of the internal logic to avoid the USB subsystem hang issue. The change in the behavior of the logic does not impact the normal operation of the USB subsystem. This fix has been discussed, developped, reviewed, polished up and tested on current-users by several people. Thread starts at: http://mail-index.netbsd.org/current-users/2009/05/17/msg009460.html
1.38.16.1.2.3	04-Jan-2012	matt	Take changes from matt-nb5-pq3 which fixes how multiple ehci functions on a pci device claim companion ohci controllers.
1.38.16.1.2.2	03-Jan-2012	matt	Only do the pci_conf_write if PCI_COMMAND_MASTER_ENABLE isn't already set.
1.38.16.1.2.1	21-Apr-2010	matt	sync to netbsd-5
1.38.14.2	23-Jul-2009	jym	Sync with HEAD.
1.38.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.38.10.1	17-Jun-2009	bouyer	branches: 1.38.10.1.4; Pull up following revision(s) (requested by cegger in ticket #814): sys/dev/pci/ehci_pci.c: revision 1.45 Apply hw workaround required for all SB600 revisions and SB700 revisions A12 and A13 to avoid USB subsystem hang symptom. The USB subsystem hang symptom is observed when the system has multiple USB devices connected to it or one USB device is often re-connected. In some cases a USB hub may be required to observe this symptom. This patch works around the problem by correcting the internal register setting that will help by changing the behavior of the internal logic to avoid the USB subsystem hang issue. The change in the behavior of the logic does not impact the normal operation of the USB subsystem. This fix has been discussed, developped, reviewed, polished up and tested on current-users by several people. Thread starts at: http://mail-index.netbsd.org/current-users/2009/05/17/msg009460.html
1.38.10.1.4.1	07-Jan-2011	matt	Deal with multifunction pci devices better in choosing companion devices.
1.38.8.1	28-Apr-2009	skrll	Sync with HEAD.
1.46.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.47.2.2	21-Apr-2011	rmind	sync with head
1.47.2.1	05-Mar-2011	rmind	sync with head
1.50.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.53.6.4	12-Mar-2012	mrg	fix detach bugs: - need to disestablish the ehci softint's. - need to destroy needs_explore_cv - note that ehci.c inits locks, but ehci_pci.c destroys them due to the way that the (pci) front end does softint handling, and can only be trusted to destroy them. XXX need to fix this, by looking at the many ehci frontends as well and checking ohci/uhci.
1.53.6.3	18-Feb-2012	mrg	merge to -current.
1.53.6.2	09-Dec-2011	mrg	- make pipe->close method take the thread lock - convert usb_taskq to use mutex/cv - convert needs_explore usage into a cv on the thread lock - remove KERNEL_*LOCK from uaudio and umidi, since we're supposedly MPSAFE here now - use IPL_SCHED instead of IPL_USB (aka biglocked) interrupts - drop the audio thread lock when calling into usb when it may sleep, avoiding a deadlock between audiowrite and audioioctl. this fixes mixerctl -a vs. playing hanging the system XXX probably need to check this in a bunch more places.
1.53.6.1	04-Dec-2011	jmcneill	Make ehci mpsafe.
1.53.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.2.2	30-Oct-2012	yamt	sync with head
1.53.2.1	17-Apr-2012	yamt	sync with head
1.54.4.1	01-Nov-2012	matt	sync with netbsd-6-0-RELEASE.
1.54.2.1	01-Oct-2012	riz	Pull up following revision(s) (requested by tstsui in ticket #569): sys/dev/pci/ehci_pci.c: revision 1.57 Fix PR kern/46828 (6.0_BETA2 and 6.0_RC1 won't start on DL320/G5p): In ehci_get_ownership(), don't explicitly clear EHCI_LEG_HC_BIOS_OWNED semaphore bit in the driver before asking BIOS to give up ownership. The EHCI spec implies that the semaphore should not be changed by the other agent and actually the previous one (introduced in rev 1.53 after 5.x) caused hangup during probe on at least two HP machines as mentioned in the PR. Analyzed and patch provided by Ryo ONODERA. Should be pulled up to netbsd-6 (fatal hangup during boot).
1.56.2.3	03-Dec-2017	jdolecek	update from HEAD
1.56.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.56.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.57.2.1	18-May-2014	rmind	sync with head
1.58.8.2	14-Sep-2016	skrll	Use IPL_USB for interrupt handler
1.58.8.1	06-Sep-2016	skrll	First pass at netbsd-7 updated with USB code from HEAD
1.58.4.2	08-Jul-2017	snj	Pull up following revision(s) (requested by sborrill in ticket #1431): sys/dev/pci/ehci_pci.c: revision 1.65 Deal with broken BIOSes that leave PCI interrupts disabled. Thanks to nick@
1.58.4.1	05-Apr-2017	snj	Pull up following revision(s) (requested by skrll in ticket #1395): share/man/man4/axe.4: netbsd-7-nhusb share/man/man4/axen.4: netbsd-7-nhusb share/man/man4/cdce.4: netbsd-7-nhusb share/man/man4/uaudio.4: netbsd-7-nhusb share/man/man4/ucom.4: netbsd-7-nhusb share/man/man4/uep.4: netbsd-7-nhusb share/man/man4/urtw.4: netbsd-7-nhusb share/man/man4/usb.4: netbsd-7-nhusb share/man/man4/uyap.4: netbsd-7-nhusb share/man/man4/xhci.4: netbsd-7-nhusb share/man/man9/usbdi.9: netbsd-7-nhusb sys/arch/amd64/conf/ALL: netbsd-7-nhusb sys/arch/amd64/conf/GENERIC: netbsd-7-nhusb sys/arch/amiga/dev/slhci_zbus.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_otg.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_usb.c: netbsd-7-nhusb sys/arch/arm/amlogic/amlogic_dwctwo.c: netbsd-7-nhusb sys/arch/arm/at91/at91ohci.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm2835_dwctwo.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm53xx_usb.c: netbsd-7-nhusb sys/arch/arm/ep93xx/epohci.c: netbsd-7-nhusb sys/arch/arm/gemini/obio_ehci.c: netbsd-7-nhusb sys/arch/arm/imx/files.imx23: netbsd-7-nhusb sys/arch/arm/imx/imxusb.c: netbsd-7-nhusb sys/arch/arm/imx/imxusbreg.h: netbsd-7-nhusb sys/arch/arm/omap/obio_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/omap3_ehci.c: netbsd-7-nhusb sys/arch/arm/omap/omapl1x_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/tiotg.c: netbsd-7-nhusb sys/arch/arm/s3c2xx0/ohci_s3c24x0.c: netbsd-7-nhusb sys/arch/arm/samsung/exynos_usb.c: netbsd-7-nhusb sys/arch/arm/xscale/pxa2x0_ohci.c: netbsd-7-nhusb sys/arch/arm/zynq/zynq_usb.c: netbsd-7-nhusb sys/arch/hpcarm/dev/nbp_slhci.c: netbsd-7-nhusb sys/arch/hpcmips/dev/plumohci.c: netbsd-7-nhusb sys/arch/i386/conf/ALL: netbsd-7-nhusb sys/arch/i386/conf/GENERIC: netbsd-7-nhusb sys/arch/i386/pci/gcscehci.c: netbsd-7-nhusb sys/arch/luna68k/conf/GENERIC: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahci.c: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahcivar.h: netbsd-7-nhusb sys/arch/mips/alchemy/dev/ohci_aubus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ehci_arbus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ohci_arbus.c: netbsd-7-nhusb sys/arch/mips/conf/files.adm5120: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ehci.c: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ohci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ehci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ohci.c: netbsd-7-nhusb sys/arch/playstation2/dev/ohci_sbus.c: netbsd-7-nhusb sys/arch/powerpc/booke/dev/pq3ehci.c: netbsd-7-nhusb sys/arch/powerpc/ibm4xx/dev/dwctwo_plb.c: netbsd-7-nhusb sys/arch/x68k/dev/slhci_intio.c: netbsd-7-nhusb sys/conf/files: netbsd-7-nhusb sys/dev/cardbus/ehci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/ohci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/uhci_cardbus.c: netbsd-7-nhusb sys/dev/ic/sl811hs.c: netbsd-7-nhusb sys/dev/ic/sl811hsvar.h: netbsd-7-nhusb sys/dev/isa/slhci_isa.c: netbsd-7-nhusb sys/dev/marvell/ehci_mv.c: netbsd-7-nhusb sys/dev/pci/ehci_pci.c: netbsd-7-nhusb sys/dev/pci/ohci_pci.c: netbsd-7-nhusb sys/dev/pci/uhci_pci.c: netbsd-7-nhusb sys/dev/pci/xhci_pci.c: netbsd-7-nhusb sys/dev/pcmcia/slhci_pcmcia.c: netbsd-7-nhusb sys/dev/usb/Makefile.usbdevs: netbsd-7-nhusb sys/dev/usb/TODO: netbsd-7-nhusb sys/dev/usb/TODO.usbmp: netbsd-7-nhusb sys/dev/usb/aubtfwl.c: netbsd-7-nhusb sys/dev/usb/auvitek.c: netbsd-7-nhusb sys/dev/usb/auvitek_audio.c: netbsd-7-nhusb sys/dev/usb/auvitek_dtv.c: netbsd-7-nhusb sys/dev/usb/auvitek_i2c.c: netbsd-7-nhusb sys/dev/usb/auvitek_video.c: netbsd-7-nhusb sys/dev/usb/auvitekvar.h: netbsd-7-nhusb sys/dev/usb/ehci.c: netbsd-7-nhusb sys/dev/usb/ehcireg.h: netbsd-7-nhusb sys/dev/usb/ehcivar.h: netbsd-7-nhusb sys/dev/usb/emdtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_dtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_ir.c: netbsd-7-nhusb sys/dev/usb/emdtvvar.h: netbsd-7-nhusb sys/dev/usb/ezload.c: netbsd-7-nhusb sys/dev/usb/ezload.h: netbsd-7-nhusb sys/dev/usb/files.usb: netbsd-7-nhusb sys/dev/usb/hid.c: netbsd-7-nhusb sys/dev/usb/hid.h: netbsd-7-nhusb sys/dev/usb/if_athn_usb.c: netbsd-7-nhusb sys/dev/usb/if_athn_usb.h: netbsd-7-nhusb sys/dev/usb/if_atu.c: netbsd-7-nhusb sys/dev/usb/if_atureg.h: netbsd-7-nhusb sys/dev/usb/if_aue.c: netbsd-7-nhusb sys/dev/usb/if_auereg.h: netbsd-7-nhusb sys/dev/usb/if_axe.c: netbsd-7-nhusb sys/dev/usb/if_axen.c: netbsd-7-nhusb sys/dev/usb/if_axenreg.h: netbsd-7-nhusb sys/dev/usb/if_axereg.h: netbsd-7-nhusb sys/dev/usb/if_cdce.c: netbsd-7-nhusb sys/dev/usb/if_cdcereg.h: netbsd-7-nhusb sys/dev/usb/if_cue.c: netbsd-7-nhusb sys/dev/usb/if_cuereg.h: netbsd-7-nhusb sys/dev/usb/if_kue.c: netbsd-7-nhusb sys/dev/usb/if_kuereg.h: netbsd-7-nhusb sys/dev/usb/if_otus.c: netbsd-7-nhusb sys/dev/usb/if_otusvar.h: netbsd-7-nhusb sys/dev/usb/if_rum.c: netbsd-7-nhusb sys/dev/usb/if_rumreg.h: netbsd-7-nhusb sys/dev/usb/if_rumvar.h: netbsd-7-nhusb sys/dev/usb/if_run.c: netbsd-7-nhusb sys/dev/usb/if_runvar.h: netbsd-7-nhusb sys/dev/usb/if_smsc.c: netbsd-7-nhusb sys/dev/usb/if_smscreg.h: netbsd-7-nhusb sys/dev/usb/if_smscvar.h: netbsd-7-nhusb sys/dev/usb/if_udav.c: netbsd-7-nhusb sys/dev/usb/if_udavreg.h: netbsd-7-nhusb sys/dev/usb/if_upgt.c: netbsd-7-nhusb sys/dev/usb/if_upgtvar.h: netbsd-7-nhusb sys/dev/usb/if_upl.c: netbsd-7-nhusb sys/dev/usb/if_ural.c: netbsd-7-nhusb sys/dev/usb/if_uralreg.h: netbsd-7-nhusb sys/dev/usb/if_uralvar.h: netbsd-7-nhusb sys/dev/usb/if_url.c: netbsd-7-nhusb sys/dev/usb/if_urlreg.h: netbsd-7-nhusb sys/dev/usb/if_urndis.c: netbsd-7-nhusb sys/dev/usb/if_urndisreg.h: netbsd-7-nhusb sys/dev/usb/if_urtw.c: netbsd-7-nhusb sys/dev/usb/if_urtwn.c: netbsd-7-nhusb sys/dev/usb/if_urtwn_data.h: netbsd-7-nhusb sys/dev/usb/if_urtwnreg.h: netbsd-7-nhusb sys/dev/usb/if_urtwnvar.h: netbsd-7-nhusb sys/dev/usb/if_urtwreg.h: netbsd-7-nhusb sys/dev/usb/if_zyd.c: netbsd-7-nhusb sys/dev/usb/if_zydreg.h: netbsd-7-nhusb sys/dev/usb/irmce.c: netbsd-7-nhusb sys/dev/usb/moscom.c: netbsd-7-nhusb sys/dev/usb/motg.c: netbsd-7-nhusb sys/dev/usb/motgvar.h: netbsd-7-nhusb sys/dev/usb/ohci.c: netbsd-7-nhusb sys/dev/usb/ohcireg.h: netbsd-7-nhusb sys/dev/usb/ohcivar.h: netbsd-7-nhusb sys/dev/usb/pseye.c: netbsd-7-nhusb sys/dev/usb/slurm.c: netbsd-7-nhusb sys/dev/usb/stuirda.c: netbsd-7-nhusb sys/dev/usb/u3g.c: netbsd-7-nhusb sys/dev/usb/uark.c: netbsd-7-nhusb sys/dev/usb/uatp.c: netbsd-7-nhusb sys/dev/usb/uaudio.c: netbsd-7-nhusb sys/dev/usb/uberry.c: netbsd-7-nhusb sys/dev/usb/ubsa.c: netbsd-7-nhusb sys/dev/usb/ubsa_common.c: netbsd-7-nhusb sys/dev/usb/ubsavar.h: netbsd-7-nhusb sys/dev/usb/ubt.c: netbsd-7-nhusb sys/dev/usb/uchcom.c: netbsd-7-nhusb sys/dev/usb/ucom.c: netbsd-7-nhusb sys/dev/usb/ucomvar.h: netbsd-7-nhusb sys/dev/usb/ucycom.c: netbsd-7-nhusb sys/dev/usb/udl.c: netbsd-7-nhusb sys/dev/usb/udl.h: netbsd-7-nhusb sys/dev/usb/udsbr.c: netbsd-7-nhusb sys/dev/usb/udsir.c: netbsd-7-nhusb sys/dev/usb/uep.c: netbsd-7-nhusb sys/dev/usb/uftdi.c: netbsd-7-nhusb sys/dev/usb/uftdireg.h: netbsd-7-nhusb sys/dev/usb/ugen.c: netbsd-7-nhusb sys/dev/usb/ugensa.c: netbsd-7-nhusb sys/dev/usb/uhci.c: netbsd-7-nhusb sys/dev/usb/uhcireg.h: netbsd-7-nhusb sys/dev/usb/uhcivar.h: netbsd-7-nhusb sys/dev/usb/uhid.c: netbsd-7-nhusb sys/dev/usb/uhidev.c: netbsd-7-nhusb sys/dev/usb/uhidev.h: netbsd-7-nhusb sys/dev/usb/uhmodem.c: netbsd-7-nhusb sys/dev/usb/uhso.c: netbsd-7-nhusb sys/dev/usb/uhub.c: netbsd-7-nhusb sys/dev/usb/uipad.c: netbsd-7-nhusb sys/dev/usb/uipaq.c: netbsd-7-nhusb sys/dev/usb/uirda.c: netbsd-7-nhusb sys/dev/usb/uirdavar.h: netbsd-7-nhusb sys/dev/usb/ukbd.c: netbsd-7-nhusb sys/dev/usb/ukbdmap.c: netbsd-7-nhusb sys/dev/usb/ukyopon.c: netbsd-7-nhusb sys/dev/usb/ukyopon.h: netbsd-7-nhusb sys/dev/usb/ulpt.c: netbsd-7-nhusb sys/dev/usb/umass.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.h: netbsd-7-nhusb sys/dev/usb/umass_quirks.c: netbsd-7-nhusb sys/dev/usb/umass_quirks.h: netbsd-7-nhusb sys/dev/usb/umass_scsipi.c: netbsd-7-nhusb sys/dev/usb/umass_scsipi.h: netbsd-7-nhusb sys/dev/usb/umassvar.h: netbsd-7-nhusb sys/dev/usb/umcs.c: netbsd-7-nhusb sys/dev/usb/umct.c: netbsd-7-nhusb sys/dev/usb/umidi.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.h: netbsd-7-nhusb sys/dev/usb/umodem.c: netbsd-7-nhusb sys/dev/usb/umodem_common.c: netbsd-7-nhusb sys/dev/usb/umodemvar.h: netbsd-7-nhusb sys/dev/usb/ums.c: netbsd-7-nhusb sys/dev/usb/uplcom.c: netbsd-7-nhusb sys/dev/usb/urio.c: netbsd-7-nhusb sys/dev/usb/urio.h: netbsd-7-nhusb sys/dev/usb/usb.c: netbsd-7-nhusb sys/dev/usb/usb.h: netbsd-7-nhusb sys/dev/usb/usb_mem.c: netbsd-7-nhusb sys/dev/usb/usb_mem.h: netbsd-7-nhusb sys/dev/usb/usb_quirks.c: netbsd-7-nhusb sys/dev/usb/usb_quirks.h: netbsd-7-nhusb sys/dev/usb/usb_subr.c: netbsd-7-nhusb sys/dev/usb/usbdevices.config: netbsd-7-nhusb sys/dev/usb/usbdevs: netbsd-7-nhusb sys/dev/usb/usbdevs.h: netbsd-7-nhusb sys/dev/usb/usbdevs_data.h: netbsd-7-nhusb sys/dev/usb/usbdi.c: netbsd-7-nhusb sys/dev/usb/usbdi.h: netbsd-7-nhusb sys/dev/usb/usbdi_util.c: netbsd-7-nhusb sys/dev/usb/usbdi_util.h: netbsd-7-nhusb sys/dev/usb/usbdivar.h: netbsd-7-nhusb sys/dev/usb/usbhid.h: netbsd-7-nhusb sys/dev/usb/usbhist.h: netbsd-7-nhusb sys/dev/usb/usbroothub.c: netbsd-7-nhusb sys/dev/usb/usbroothub.h: netbsd-7-nhusb sys/dev/usb/usbroothub_subr.c: delete sys/dev/usb/usbroothub_subr.h: delete sys/dev/usb/uscanner.c: netbsd-7-nhusb sys/dev/usb/uslsa.c: netbsd-7-nhusb sys/dev/usb/usscanner.c: netbsd-7-nhusb sys/dev/usb/ustir.c: netbsd-7-nhusb sys/dev/usb/uthum.c: netbsd-7-nhusb sys/dev/usb/utoppy.c: netbsd-7-nhusb sys/dev/usb/uts.c: netbsd-7-nhusb sys/dev/usb/uvideo.c: netbsd-7-nhusb sys/dev/usb/uvisor.c: netbsd-7-nhusb sys/dev/usb/uvscom.c: netbsd-7-nhusb sys/dev/usb/uyap.c: netbsd-7-nhusb sys/dev/usb/uyap_firmware.h: netbsd-7-nhusb sys/dev/usb/uyurex.c: netbsd-7-nhusb sys/dev/usb/x1input_rdesc.h: netbsd-7-nhusb sys/dev/usb/xhci.c: netbsd-7-nhusb sys/dev/usb/xhcireg.h: netbsd-7-nhusb sys/dev/usb/xhcivar.h: netbsd-7-nhusb sys/dev/usb/xinput_rdesc.h: netbsd-7-nhusb sys/external/bsd/common/conf/files.linux: netbsd-7-nhusb sys/external/bsd/common/include/linux/err.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/kernel.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/workqueue.h: netbsd-7-nhusb sys/external/bsd/common/linux/linux_work.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/atombios_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/drm/files.drmkms: netbsd-7-nhusb sys/external/bsd/drm2/i915drm/files.i915drmkms: netbsd-7-nhusb sys/external/bsd/drm2/include/linux/err.h: delete sys/external/bsd/drm2/include/linux/workqueue.h: delete sys/external/bsd/drm2/linux/files.drmkms_linux: netbsd-7-nhusb sys/external/bsd/drm2/linux/linux_work.c: delete sys/external/bsd/dwc2/dwc2.c: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2var.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwctwo2netbsd: netbsd-7-nhusb sys/external/bsd/dwc2/conf/files.dwc2: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_coreintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdddma.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdqueue.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hw.h: netbsd-7-nhusb sys/modules/drmkms_linux/Makefile: netbsd-7-nhusb sys/modules/i915drmkms/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libugenhc/ugenhc.c: netbsd-7-nhusb sys/rump/dev/lib/libusb/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libusb/USB.ioconf: netbsd-7-nhusb sys/rump/dev/lib/libusb/usb_at_ugenhc.c: delete sys/rump/dev/lib/libusb/opt/opt_usb.h: delete sys/rump/dev/lib/libusb/opt/opt_usbverbose.h: delete sys/sys/mbuf.h: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.8: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.c: netbsd-7-nhusb Merge netbsd-7-nhusb: - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix - Change the SOFTINT level from NET to SERIAL for the USB softint handler. This gives the callback a chance of running when another softint handler at SOFTINT_NET has blocked holding a lock, e.g. softnet_lock and most of the network stack. - kern/49065 - ifconfig tun0 ... sequence locks up system / lockup: softnet_lock held across usb xfr - kern/50491 - unkillable wait in usbd_transfer while using usmsc0 on raspberry pi 2 - kern/51395 - USB Ethernet makes xhci hang - Various improvements to slhci(4) - Various improvements to dwc2(4)
1.59.2.6	28-Aug-2017	skrll	Sync with HEAD
1.59.2.5	05-Dec-2016	skrll	Sync with HEAD
1.59.2.4	22-Sep-2015	skrll	Sync with HEAD
1.59.2.3	05-Dec-2014	skrll	Use int for return type for [eou]chi_init and motg_init.
1.59.2.2	03-Dec-2014	skrll	The grand renaming of structure members. No functional change.
1.59.2.1	03-Dec-2014	skrll	Trailing whitespace.
1.63.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.64.8.1	15-Jun-2017	snj	Pull up following revision(s) (requested by sborrill in ticket #33): sys/dev/pci/ehci_pci.c: revision 1.65 Deal with broken BIOSes that leave PCI interrupts disabled. Thanks to nick@
1.65.4.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.65.4.2	21-May-2018	pgoyette	Sync with HEAD
1.65.4.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.67.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.67.2.1	10-Jun-2019	christos	Sync with HEAD
1.70.12.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.71.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.75.2.2	11-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #823): sys/dev/pci/ehci_pci.c: revision 1.77 ehci(4): properly handle failed attach thinkpad a475 fails to attach an ehci instance: ehci0: pre-2.0 USB rev, device ignored which ends up stopping suspend/resume working as the device has no pmf handlers installed. put most of the teardown code into a new common function that is called from failed attach and detach. if attach fails, register NULL pmf handlers. don't check the return value of pmf_device_register*(). re-order several parts of detach to match the reverse attach order. tested on rockpro64, thinkpad a475, and xps 1645, the a475 can now suspend/resume almost fully successfully.
1.75.2.1	11-Mar-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #631): sys/dev/pci/ehci_pci.c: revision 1.76 Consistently use cached chipset tag value.
1.77.2.1	02-Aug-2025	perseant	Sync with HEAD
1.79	08-Jun-2024	andvar	s/amound/amount/ in comment.
1.78	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.77	10-May-2023	riastradh	emuxki(4): Use config_detach_children.
1.76	07-Sep-2022	khorben	emuxki(4): import a list of board definitions This is inspired by the cxdtv(4) driver. Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870, PCI) and with a Sound Blaster Audigy Rx 7.1 (SB1550, PCIe). Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.75	07-Sep-2022	khorben	emuxki(4): place struct emuxki_softc into its own file NFCI
1.74	07-Sep-2022	khorben	emuxki(4): restrict DMA memory within the first 2GB As implemented in the driver, the EMU10K1 chip can only address memory up to 31-bit addresses. Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870, PCI) and with a Sound Blaster Audigy Rx 7.1 (SB1550, PCIe). Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.73	07-Sep-2022	khorben	emuxki(4): allow building as a module Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870) Note that this required setting outputs.master to the maximum value allowed (255) to get sound out, and then cranking the volume pretty high. Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.72	29-Aug-2022	khorben	emuxki(4): add support for the Sound Blaster Audigy Rx This is as per kern/56980. Confirmed working on one out of two different computers, albeit with a strong hissing noise (same card, different mainboard). The mixer settings seem to be irrelevant or not supported though. Original support inspired by the Linux driver, and confirmed with the OpenBSD driver, itself originally from NetBSD. Tested on NetBSD/amd64.
1.71	06-Feb-2021	isaki	Remove debug messages which are no longer necessary.
1.70	08-Jun-2019	isaki	branches: 1.70.10; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.69	29-May-2019	isaki	Fix build error on i386. These were only needed during development. Reported by msaitoh@.
1.68	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.67	16-Mar-2019	isaki	branches: 1.67.2; Use C99 style struct initializer to audio_hw_if.
1.66	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.65	01-Jun-2017	chs	branches: 1.65.8; 1.65.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.64	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.63	29-Mar-2014	christos	branches: 1.63.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.62	30-Jan-2012	drochner	branches: 1.62.6; 1.62.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.61	24-Nov-2011	mrg	branches: 1.61.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.60	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.59	26-Nov-2009	njoly	branches: 1.59.12; 1.59.14; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.58	09-Jun-2009	tsutsui	Split device_t/softc. Tested on: emuxki0 at pci2 dev 7 function 0: Creative Labs SBLive! EMU 10000 (audio multimedia, revision 0x05)
1.57	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.56	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.55	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.54	06-Sep-2008	gmcgarry	branches: 1.54.6; 1.54.8; Backout unintended change.
1.53	06-Sep-2008	gmcgarry	Use One exit point to avoid unreachable code.
1.52	28-Apr-2008	martin	branches: 1.52.2; 1.52.6; Remove clause 3 and 4 from TNF licenses
1.51	10-Apr-2008	cegger	branches: 1.51.2; 1.51.4; use aprint_*_dev and device_xname
1.50	04-Mar-2007	christos	branches: 1.50.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.49	16-Nov-2006	christos	branches: 1.49.4; 1.49.6; __unused removal on arguments; approved by core.
1.48	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.47	23-Sep-2006	jmcneill	Print "interrupt at irq <foo>" message immediately after we establish the interrupt. Patch from PR# 33567.
1.46	30-Aug-2006	christos	branches: 1.46.2; 1.46.4; - fix incomplete initializer - remove unsigned comparison < 0
1.45	14-Apr-2006	christos	branches: 1.45.4; Coverity CID 1537: Make it clear to coverity that there is no overflow.
1.44	14-Apr-2006	christos	Coverity CID 1538: Avoid array overrun.
1.43	11-Dec-2005	christos	branches: 1.43.4; 1.43.6; 1.43.8; 1.43.10; 1.43.12; merge ktrace-lwp.
1.42	11-Oct-2005	jdolecek	put back joystick enabling code, lost in Audigy support merge
1.41	27-Feb-2005	perry	branches: 1.41.4; nuke trailing whitespace
1.40	15-Jan-2005	kent	branches: 1.40.2; 1.40.4; ansify and KNF
1.39	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.38	29-Oct-2004	yamt	branches: 1.38.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.37	07-Oct-2004	dsainty	Corrected a scattering of typos in comments
1.36	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.35	27-May-2004	christos	PR/25731: Adam K kirchhoff: Audigy support with the emuxki driver.
1.34	23-May-2004	tsutsui	Add some byteswap ops around DMA descriptors for big endian machines. Tested on macppc by matthew green. XXX This driver completely lacks bus_dmamap_sync(9) ops.
1.33	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.32	03-Apr-2004	martti	Added another SB Live! 5.1 soundcard and joystick version
1.31	10-Nov-2003	wiz	branches: 1.31.2; Spell address with two d's. Inspired by similar changes in OpenBSD, originating from Jonathon Gray and forwarded by jmc@openbsd.
1.30	30-Oct-2003	hannken	Complete last commit: Remove assigned-to but otherwise unused variable.
1.29	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.28	25-Oct-2003	christos	Fix uninitialized variable warnings.
1.27	25-Oct-2003	christos	Fix uninitialized variable warnings
1.26	03-May-2003	wiz	branches: 1.26.2; DMA, not dma nor Dma.
1.25	26-Apr-2003	wiz	Management, not managment. Mostly from jmc@openbsd.
1.24	17-Mar-2003	toshii	Fix a memory leak in emuxki_voice_new. Pointed out by Izumi Tsutsui in the current-users mailing list.
1.23	17-Mar-2003	toshii	Make emuxki_voice_set_audioparms return EINVAL if invalid sample rate is passed.
1.22	16-Mar-2003	toshii	Fix recording with formats other than 16 bit slinear_le, based on the patch submitted as kern/20620.
1.21	16-Mar-2003	toshii	Remove a couple of false comments, as pointed out in kern/20620.
1.20	03-Mar-2003	fvdl	The intrmasks array is only used in if 0-ed code, so if 0 it as well.
1.19	25-Feb-2003	toshii	Add recording support for emuxki(4). The patch is basically the same as I submitted in PR kern/18093. Some of the patch needs refinement. Also, only the ADC recording source is supported, because the other sources are probably much less useful and the implementation isn't trivial.
1.18	22-Feb-2003	tsutsui	hz -> Hz
1.17	18-Feb-2003	wiz	s/intial/initial/, from jmc in .pl in PR 20414.
1.16	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.15	31-Jan-2003	thorpej	Use aprint_*().
1.14	01-Jan-2003	tron	Use bus_space_write_2() to write the 16 bit register "EMU_TIMER" which should fix PR kern/15663. Patch supplied by Yorick Hardy via GNATS.
1.13	24-Dec-2002	jdolecek	emuxki_intr(): return 1 if interrupt was for us, 0 if not this appear to fix occassional sample repeats on my dual CPU i386
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	02-Feb-2002	jdolecek	branches: 1.9.10; enable the joystick port on attach, too
1.8	27-Jan-2002	jdolecek	emuxki_set_params(): don't try to set parameters if the appropriate 'voice' is not initialized This fixes kern/15394 by Onno van der Linden. Code fragment to trigger the bug: AUDIO_INITINFO(&info); info.mode = AUMODE_PLAY; info.play.channels = 2; fd = open("/dev/audioctl0", O_RDWR); (void)ioctl(fd, AUDIO_SETINFO, &info); emuxki_voice_set_audioparms(): g/c check for voice == NULL, this is never called with null voice some minor style changes in emuxki_open()/emuxki_close()
1.7	23-Dec-2001	jdolecek	branches: 1.7.2; 1.7.4; Merge in changes sent by Michael Shalayeff, from OpenBSD port of this driver: - use better default for DAC level (previous produced distorted output) - preserve DAC setting in between openings
1.6	23-Dec-2001	jdolecek	Add the AC97_HOST_SWAPPED_CHANNELS quirk for emuxki, and do some whitespace cleanup. Streight from Yannick Montulet.
1.5	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.4	13-Nov-2001	lukem	add RCSID
1.3	22-Oct-2001	jdolecek	branches: 1.3.2; De-__P()ify. Suggested by Lennart Augustsson.
1.2	18-Oct-2001	jdolecek	Fix obviously wrong printf() format in pci_intr_establish() failure case. Pointed out by Chuck Silvers. XXX This has compiled without a warning for me - what's up? Add sc->sc_dev.dv_xname in another failure printf()
1.1	17-Oct-2001	jdolecek	Add driver for Creative Labs SBLive! EMU10000, possibly also PCI512. Written by Yannick Montulet (thanks!), with only couple minor touches by me.
1.3.2.8	03-Jan-2003	thorpej	Sync with HEAD.
1.3.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.3.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.3.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.3.2.1	22-Oct-2001	nathanw	file emuxki.c was added on branch nathanw_sa on 2001-10-22 20:41:23 +0000
1.7.4.3	09-Feb-2002	he	Pull up revision 1.8 (requested by jdolecek): Do not try to set parameters if the appropriate ``voice'' is not initialized. Fixes PR#15394.
1.7.4.2	19-Jan-2002	he	Pull up revisions 1.1-1.7 (new, via patch, requested by he): Add driver for Creative Labs SBLive! EMU10000 and (probably) PCI512. Fixes PR#15260.
1.7.4.1	23-Dec-2001	he	file emuxki.c was added on branch netbsd-1-5 on 2002-01-19 21:55:09 +0000
1.7.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.7.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.7.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.2.1	23-Dec-2001	thorpej	file emuxki.c was added on branch kqueue on 2002-01-10 19:56:32 +0000
1.9.10.1	05-Jan-2003	jmc	Pull up revisions 1.13-1.14 (requested by tron in ticket #1056) Use bus_space_write_2() to write the 16 bit register EMU_TIMER which should fix PR kern/15663. Patch supplied by Yorick Hardy via GNATS.
1.26.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.26.2.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.26.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.26.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.26.2.5	19-Oct-2004	skrll	Sync with HEAD
1.26.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.26.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.26.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.26.2.1	03-Aug-2004	skrll	Sync with HEAD
1.31.2.2	22-Sep-2004	jmc	branches: 1.31.2.2.2; Pullup rev 1.36 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.31.2.1	05-Apr-2004	tron	Pull up revision 1.32 (requested by martti in ticket #54): Added another SB Live! 5.1 soundcard and joystick version
1.31.2.2.2.1	13-Apr-2005	tron	Pull up revision 1.34 (requested by tsutsui in ticket #1071): Add some byteswap ops around DMA descriptors for big endian machines. Tested on macppc by matthew green. XXX This driver completely lacks bus_dmamap_sync(9) ops.
1.38.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.38.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.40.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.40.2.1	29-Apr-2005	kent	sync with -current
1.41.4.3	03-Sep-2007	yamt	sync with head.
1.41.4.2	30-Dec-2006	yamt	sync with head.
1.41.4.1	21-Jun-2006	yamt	sync with head.
1.43.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.43.10.1	19-Apr-2006	elad	sync with head.
1.43.8.2	03-Sep-2006	yamt	sync with head.
1.43.8.1	24-May-2006	yamt	sync with head.
1.43.6.1	22-Apr-2006	simonb	Sync with head.
1.43.4.1	09-Sep-2006	rpaulo	sync with head
1.45.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.46.4.2	10-Dec-2006	yamt	sync with head.
1.46.4.1	22-Oct-2006	yamt	sync with head
1.46.2.1	18-Nov-2006	ad	Sync with head.
1.49.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.49.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.50.36.2	28-Sep-2008	mjf	Sync with HEAD.
1.50.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.51.4.5	11-Mar-2010	yamt	sync with head
1.51.4.4	20-Jun-2009	yamt	sync with head
1.51.4.3	16-May-2009	yamt	sync with head
1.51.4.2	04-May-2009	yamt	sync with head.
1.51.4.1	16-May-2008	yamt	sync with head.
1.51.2.1	18-May-2008	yamt	sync with head.
1.52.6.1	19-Oct-2008	haad	Sync with HEAD.
1.52.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.54.8.2	23-Jul-2009	jym	Sync with HEAD.
1.54.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.54.6.3	13-Dec-2008	ad	Fix locking for ac97.
1.54.6.2	12-Dec-2008	ad	Checkpoint work in progress.
1.54.6.1	11-Dec-2008	ad	Checkpoint work in progress.
1.59.14.4	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.59.14.3	21-Nov-2011	jakllsch	Make emuxki(4) work w/ audiomp changes.
1.59.14.2	19-Nov-2011	jmcneill	adapt to audiomp api changes
1.59.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.59.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.59.12.1	17-Apr-2012	yamt	sync with head
1.61.2.1	18-Feb-2012	mrg	merge to -current.
1.62.10.1	18-May-2014	rmind	sync with head
1.62.6.2	03-Dec-2017	jdolecek	update from HEAD
1.62.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.63.6.2	28-Aug-2017	skrll	Sync with HEAD
1.63.6.1	09-Jul-2016	skrll	Sync with HEAD
1.65.10.1	10-Jun-2019	christos	Sync with HEAD
1.65.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.67.2.7	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.67.2.6	03-May-2019	isaki	Remove unnecessary header file.
1.67.2.5	01-May-2019	isaki	Use __nothing macro.
1.67.2.4	01-May-2019	isaki	Reimplement emuxki driver. - Use single voice per playback and per recording. - Use fixed format, 2ch/48kHz, to simplify. - Fix several problems in previous driver. And now it works even on alpha! The driver is written by Y.Sugahara. Thanks!
1.67.2.3	01-May-2019	isaki	Adapt to audio2. - XXX New set_format is not implemented at this point. This implementation can not adapt to audio2 (and it also has not worked since netbsd-8). The driver supports multiple hardware stream which is named 'voice'. This voice was allocated at open() and was deallocated at close(). However, in order to support mixing, some interfaces such as set_params() (or set_format in audio2) has to be called before open().
1.67.2.2	21-Apr-2019	isaki	Make it compilable (on this branch). flags here is F(READ\|WRITE), not AUOPEN_(READ\|WRITE).
1.67.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.70.10.1	03-Apr-2021	thorpej	Sync with HEAD.
1.1	07-Sep-2022	khorben	emuxki(4): import a list of board definitions This is inspired by the cxdtv(4) driver. Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870, PCI) and with a Sound Blaster Audigy Rx 7.1 (SB1550, PCIe). Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.1	07-Sep-2022	khorben	emuxki(4): import a list of board definitions This is inspired by the cxdtv(4) driver. Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870, PCI) and with a Sound Blaster Audigy Rx 7.1 (SB1550, PCIe). Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.8	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.7	11-Dec-2005	christos	branches: 1.7.70; 1.7.72; 1.7.74; merge ktrace-lwp.
1.6	15-Jan-2005	kent	ansify and KNF
1.5	27-May-2004	christos	PR/25731: Adam K kirchhoff: Audigy support with the emuxki driver.
1.4	24-Feb-2004	wiz	occured -> occurred. From Peter Postma.
1.3	28-Jan-2003	wiz	branches: 1.3.2; unknown, not unkown. Noted by mjl.
1.2	23-Dec-2001	jdolecek	branches: 1.2.2; 1.2.4; Whitespace cleanup, streight from Yannick Montulet.
1.1	17-Oct-2001	jdolecek	branches: 1.1.2; Add driver for Creative Labs SBLive! EMU10000, possibly also PCI512. Written by Yannick Montulet (thanks!), with only couple minor touches by me.
1.1.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.1.2.1	17-Oct-2001	nathanw	file emuxkireg.h was added on branch nathanw_sa on 2001-10-22 20:41:23 +0000
1.2.4.2	19-Jan-2002	he	Pull up revision 1.1-1.2 (new, requested by he): Add driver for Creative Labs SBLive! EMU10000 and (probably) PCI512. Fixes PR#15260.
1.2.4.1	23-Dec-2001	he	file emuxkireg.h was added on branch netbsd-1-5 on 2002-01-19 21:54:30 +0000
1.2.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.1	23-Dec-2001	thorpej	file emuxkireg.h was added on branch kqueue on 2002-01-10 19:56:32 +0000
1.3.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.3.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.3.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.74.1	16-May-2008	yamt	sync with head.
1.7.72.1	18-May-2008	yamt	sync with head.
1.7.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.16	07-Sep-2022	khorben	emuxki(4): import a list of board definitions This is inspired by the cxdtv(4) driver. Tested on NetBSD/amd64 with a Sound Blaster Live! Value (CT4870, PCI) and with a Sound Blaster Audigy Rx 7.1 (SB1550, PCIe). Additional sound cards sponsored by the NetBSD Foundation; thanks!
1.15	07-Sep-2022	khorben	emuxki(4): place struct emuxki_softc into its own file NFCI
1.14	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.13	23-Nov-2011	jmcneill	branches: 1.13.50; 1.13.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.12	09-Jun-2009	tsutsui	branches: 1.12.12; 1.12.14; Split device_t/softc. Tested on: emuxki0 at pci2 dev 7 function 0: Creative Labs SBLive! EMU 10000 (audio multimedia, revision 0x05)
1.11	28-Apr-2008	martin	branches: 1.11.12; 1.11.14; Remove clause 3 and 4 from TNF licenses
1.10	04-Mar-2007	christos	branches: 1.10.36; 1.10.38; 1.10.40; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.9	11-Dec-2005	christos	branches: 1.9.26; 1.9.28; merge ktrace-lwp.
1.8	15-Jan-2005	kent	branches: 1.8.10; ansify and KNF
1.7	27-May-2004	christos	PR/25731: Adam K kirchhoff: Audigy support with the emuxki driver.
1.6	03-May-2003	wiz	branches: 1.6.2; DMA, not dma nor Dma.
1.5	26-Apr-2003	wiz	Management, not managment. Mostly from jmc@openbsd.
1.4	18-Feb-2003	wiz	s/intial/initial/, from jmc in .pl in PR 20414.
1.3	23-Dec-2001	jdolecek	branches: 1.3.2; 1.3.4; Merge in changes sent by Michael Shalayeff, from OpenBSD port of this driver: - use better default for DAC level (previous produced distorted output) - preserve DAC setting in between openings
1.2	23-Dec-2001	jdolecek	Whitespace cleanup, streight from Yannick Montulet.
1.1	17-Oct-2001	jdolecek	branches: 1.1.2; Add driver for Creative Labs SBLive! EMU10000, possibly also PCI512. Written by Yannick Montulet (thanks!), with only couple minor touches by me.
1.1.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.1.2.1	17-Oct-2001	nathanw	file emuxkivar.h was added on branch nathanw_sa on 2001-10-22 20:41:23 +0000
1.3.4.2	19-Jan-2002	he	Pull up revisions 1.1-1.3 (new, requested by he): Add driver for Creative Labs SBLive! EMU10000 and (probably) PCI512. Fixes PR#15260.
1.3.4.1	23-Dec-2001	he	file emuxkivar.h was added on branch netbsd-1-5 on 2002-01-19 21:54:48 +0000
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	23-Dec-2001	thorpej	file emuxkivar.h was added on branch kqueue on 2002-01-10 19:56:32 +0000
1.6.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.6.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.10.1	03-Sep-2007	yamt	sync with head.
1.9.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.9.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.10.40.2	20-Jun-2009	yamt	sync with head
1.10.40.1	16-May-2008	yamt	sync with head.
1.10.38.1	18-May-2008	yamt	sync with head.
1.10.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.11.14.1	23-Jul-2009	jym	Sync with HEAD.
1.11.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.12.14.2	21-Nov-2011	jakllsch	Make emuxki(4) work w/ audiomp changes.
1.12.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.12.12.1	17-Apr-2012	yamt	sync with head
1.13.54.1	01-May-2019	isaki	Reimplement emuxki driver. - Use single voice per playback and per recording. - Use fixed format, 2ch/48kHz, to simplify. - Fix several problems in previous driver. And now it works even on alpha! The driver is written by Y.Sugahara. Thanks!
1.13.50.1	10-Jun-2019	christos	Sync with HEAD
1.66	10-May-2023	riastradh	esa(4): Use config_detach_children.
1.65	08-Jun-2019	isaki	Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.64	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.63	16-Mar-2019	isaki	branches: 1.63.2; Use C99 style struct initializer to audio_hw_if.
1.62	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.61	01-Jun-2017	chs	branches: 1.61.8; 1.61.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.60	29-Mar-2014	christos	branches: 1.60.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.59	16-Oct-2013	christos	- use proper types (uint16_t instead of uint32_t) - don't index past the end of arrays - use __arraycount() consistently instead of manual cruft. - remove useless cast
1.58	30-Jan-2012	drochner	branches: 1.58.6; 1.58.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.57	03-Dec-2011	ryo	branches: 1.57.2; the argument of esa_get_locks() is struct esa_voice , not struct esa_softc . ok by mrg@ jmcneill@
1.56	24-Nov-2011	mrg	put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.55	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.54	24-Feb-2010	dyoung	branches: 1.54.10; 1.54.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.53	08-Jan-2010	dyoung	branches: 1.53.2; Expand PMF_FN_* macros.
1.52	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.51	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.50	10-Jun-2008	dyoung	branches: 1.50.8; 1.50.10; Fix a couple of mistakes in the device_t/softc split that cause machines such as my HP Pavilion N3270 to freeze at boot. This should fix PR kern/38815, too.
1.49	13-May-2008	jmcneill	branches: 1.49.2; Hook hardware volume buttons into pmf.
1.48	27-Mar-2008	jmcneill	branches: 1.48.2; 1.48.4; 1.48.6; Split device_t and softc
1.47	21-Mar-2008	dyoung	Use aprint_*_dev() instead of accessing struct device member dv_xname directly.
1.46	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.45	22-Feb-2008	dyoung	Add methods for detaching self and for detaching children. Use device_t and accessors. Use aprint_*_dev().
1.44	27-Jan-2008	jmcneill	branches: 1.44.2; 1.44.6; On resume, ensure that esa is powered up, and ensure we power the AC97 codec once the host controller is up and running.
1.43	09-Dec-2007	jmcneill	Merge jmcneill-pm branch.
1.42	19-Oct-2007	ad	branches: 1.42.2; 1.42.4; 1.42.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.41	04-Mar-2007	christos	branches: 1.41.2; 1.41.14; 1.41.16; 1.41.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.40	16-Nov-2006	christos	branches: 1.40.4; 1.40.6; __unused removal on arguments; approved by core.
1.39	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.38	25-Sep-2006	jmcneill	No need to call pci_set_powerstate from our powerhook, the bus handles this for us now.
1.37	24-Sep-2006	jmcneill	PR# 23620: channels are swapped
1.36	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.35	24-Sep-2006	jmcneill	Formatting fixups, remove commented out code that should not have been here, and remove a few unnecessary XXX comments.
1.34	30-Aug-2006	christos	branches: 1.34.2; 1.34.4; missing comma
1.33	30-Aug-2006	christos	fix incomplete initializers
1.32	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.31	11-Dec-2005	christos	branches: 1.31.4; 1.31.8; 1.31.14; 1.31.16; merge ktrace-lwp.
1.30	29-Oct-2005	jmcneill	Wow, this is clearly wrong (found in esa_set_params): if (p->channels < 1 && p->channels > 2) return EINVAL; Change the 'and' to an 'or' so the check actually does something.
1.29	28-Jun-2005	thorpej	branches: 1.29.2; 1.29.4; Use ANSI function decls and static.
1.28	15-Jan-2005	kent	ansify and KNF
1.27	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.26	29-Oct-2004	yamt	branches: 1.26.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.25	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.24	21-Jul-2004	scw	- Don't hardcode buffer size and block size in esa_round_buffersize() and esa_round_blocksize(). - Fix esa_trigger_output() and esa_trigger_input() to initialise the channel's buffer/block size using the supplied parameters. - While here, simplify esa_intr(). This gets kphone working on my esa(4)-equipped laptop.
1.23	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.22	25-Oct-2003	christos	branches: 1.22.2; Fix uninitialized variable warnings
1.21	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.20	01-Feb-2003	thorpej	branches: 1.20.2; Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.19	31-Jan-2003	thorpej	Use aprint_*().
1.18	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.17	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.16	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.15	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.14	08-Aug-2002	itojun	audit bitmask test mistake (s/&&/&/). from openbsd
1.13	02-Jun-2002	jmcneill	Fix the mixer controls to work a bit better.
1.12	24-Mar-2002	jmcneill	branches: 1.12.2; Fix suspend/resume. Previously, suspending/resuming while playback was active would either: 1. Cause the machine to stop responding, or 2. Cause the currently playing voices to stop output. With this change, voices are stopped on suspend, and continue from where they left off on resume.
1.11	16-Mar-2002	jmcneill	Add multiple voice support to the ESS Allegro-1 / Maestro-3 family of audio accelerators. Mixing is done in hardware by the ASSP, but is limited to 4 simultaneous channels due to the restricted "minisrc" image that we are currently working with. Due to limitations in the audio subsystem, I'm currently attaching multiple 'audio' devices to 'esa', one for each voice. Because of this hack, the default ESA_NUM_VOICES is 1.
1.10	10-Mar-2002	jmcneill	Correctly enable/disable the low-pass filter, depending on the sample rate. This fixes timing issues when playing back at 48kHz (noticable using pkgsrc/graphics/mplayer and DivX ;-) streams).
1.9	26-Feb-2002	joda	in esa_resume, call esa_codec_reset; makes it much more happy after a suspend on my O500 (snatched from FreeBSD)
1.8	23-Jan-2002	pooka	branches: 1.8.2; free memory allocated for suspend save area in detach
1.7	14-Jan-2002	pooka	Use standard PCI power management definitions from pcireg.h instead of homegrown ones. suggested by Soren
1.6	13-Jan-2002	jmcneill	Every card I've seen has had their channels swapped with respect to the mixer. Ie: $ mixerctl -w outputs.master=0,191 Would result in the _right_ speaker being turned off. So, we will swap the left and right mixer channels to compensate for this.
1.5	13-Jan-2002	jmcneill	Use malloc's M_ZERO flag instead of malloc/memset.
1.4	13-Jan-2002	pooka	Add support for the driver to work after suspend/resume. Does nothing fancy, works for the basic output case, but does not support programs which use playback while going into suspend. Sketched after code found in FreeBSD and Linux drivers.
1.3	07-Jan-2002	jmcneill	branches: 1.3.2; 1.3.4; Integrate recording support into the driver. Not very heavily tested yet.
1.2	06-Jan-2002	augustss	Note that Don Kim wrote the original Linux driver.
1.1	06-Jan-2002	jmcneill	Import driver for the ESS Allegro-1 / Maestro-3 based PCI Audio Accelerators.
1.3.4.8	29-Dec-2002	thorpej	Sync with HEAD.
1.3.4.7	18-Oct-2002	nathanw	Catch up to -current.
1.3.4.6	13-Aug-2002	nathanw	Catch up to -current.
1.3.4.5	20-Jun-2002	nathanw	Catch up to -current.
1.3.4.4	12-Apr-2002	nathanw	Missed some revisions in the merge.
1.3.4.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.3.4.2	11-Jan-2002	nathanw	More catchup.
1.3.4.1	07-Jan-2002	nathanw	file esa.c was added on branch nathanw_sa on 2002-01-11 23:39:17 +0000
1.3.2.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.6	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.2.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.3.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	07-Jan-2002	thorpej	file esa.c was added on branch kqueue on 2002-01-10 19:56:32 +0000
1.8.2.3	09-Mar-2002	he	Apply patch (requested by jmcneill): Fix timing of playback at 48kHz.
1.8.2.2	24-Jan-2002	he	Pull up revisions 1.1-1.8 (new, via patch, requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.8.2.1	23-Jan-2002	he	file esa.c was added on branch netbsd-1-5 on 2002-01-24 22:35:57 +0000
1.12.2.2	29-Aug-2002	gehenna	catch up with -current.
1.12.2.1	20-Jun-2002	gehenna	catch up with -current.
1.20.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.20.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.20.2.3	02-Nov-2004	skrll	Sync with HEAD.
1.20.2.2	24-Sep-2004	skrll	Sync with HEAD.
1.20.2.1	03-Aug-2004	skrll	Sync with HEAD
1.22.2.1	22-Sep-2004	jmc	Pullup rev 1.25 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.26.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.26.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.29.4.1	02-Nov-2005	yamt	sync with head.
1.29.2.8	24-Mar-2008	yamt	sync with head.
1.29.2.7	27-Feb-2008	yamt	sync with head.
1.29.2.6	04-Feb-2008	yamt	sync with head.
1.29.2.5	21-Jan-2008	yamt	sync with head
1.29.2.4	27-Oct-2007	yamt	sync with head.
1.29.2.3	03-Sep-2007	yamt	sync with head.
1.29.2.2	30-Dec-2006	yamt	sync with head.
1.29.2.1	21-Jun-2006	yamt	sync with head.
1.31.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.31.14.1	19-Jun-2006	chap	Sync with head.
1.31.8.2	03-Sep-2006	yamt	sync with head.
1.31.8.1	26-Jun-2006	yamt	sync with head.
1.31.4.1	09-Sep-2006	rpaulo	sync with head
1.34.4.2	10-Dec-2006	yamt	sync with head.
1.34.4.1	22-Oct-2006	yamt	sync with head
1.34.2.1	18-Nov-2006	ad	Sync with head.
1.40.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.40.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.41.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.41.16.3	23-Mar-2008	matt	sync with HEAD
1.41.16.2	09-Jan-2008	matt	sync with HEAD
1.41.16.1	06-Nov-2007	matt	sync with HEAD
1.41.14.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.41.14.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.41.14.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.41.14.1	05-Oct-2007	joerg	Convert esa(4) to PNP power management.
1.41.2.1	23-Oct-2007	ad	Sync with head.
1.42.6.1	11-Dec-2007	yamt	sync with head.
1.42.4.1	26-Dec-2007	ad	Sync with head.
1.42.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.44.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.44.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.44.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.44.2.1	24-Mar-2008	keiichi	sync with head.
1.48.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.48.4.4	11-Mar-2010	yamt	sync with head
1.48.4.3	16-May-2009	yamt	sync with head
1.48.4.2	04-May-2009	yamt	sync with head.
1.48.4.1	16-May-2008	yamt	sync with head.
1.48.2.2	17-Jun-2008	yamt	sync with head.
1.48.2.1	18-May-2008	yamt	sync with head.
1.49.2.1	18-Jun-2008	simonb	Sync with head.
1.50.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.8.2	12-Dec-2008	ad	Checkpoint work in progress.
1.50.8.1	08-Dec-2008	ad	Convert some more drivers.
1.53.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.54.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.54.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.54.10.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.54.10.1	17-Apr-2012	yamt	sync with head
1.57.2.1	18-Feb-2012	mrg	merge to -current.
1.58.10.1	18-May-2014	rmind	sync with head
1.58.6.2	03-Dec-2017	jdolecek	update from HEAD
1.58.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.60.6.1	28-Aug-2017	skrll	Sync with HEAD
1.61.10.1	10-Jun-2019	christos	Sync with HEAD
1.61.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.63.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.63.2.3	04-May-2019	isaki	Remove obsoleted methods in audio_hw_if. - drain: is handled in audio upper layer now. - mappage: is handled in audio upper layer now. - setfd: no one uses and it's meaningless now.
1.63.2.2	28-Apr-2019	isaki	Adapt to audio2. - Fix wrong round_blocksize() calculation. ~0x20 -> -0x20.
1.63.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.7	16-Oct-2013	christos	- use proper types (uint16_t instead of uint32_t) - don't index past the end of arrays - use __arraycount() consistently instead of manual cruft. - remove useless cast
1.6	11-Dec-2005	christos	branches: 1.6.110; 1.6.120; 1.6.126; merge ktrace-lwp.
1.5	15-Jan-2005	kent	ansify and KNF
1.4	07-Jan-2002	jmcneill	branches: 1.4.2; 1.4.4; 1.4.6; 1.4.22; Integrate recording support into the driver. Not very heavily tested yet.
1.3	06-Jan-2002	augustss	Note that Don Kim wrote the original Linux driver.
1.2	06-Jan-2002	augustss	Change copyright somewhat. (These are not TNF copyright templates so I removed the advertising clause.)
1.1	06-Jan-2002	augustss	Part of the ESS Allegro-1 & Maestro3 driver.
1.4.22.1	17-Jan-2005	skrll	Sync with HEAD.
1.4.6.2	24-Jan-2002	he	Pull up revisions 1.1-1.4 (new, requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.4.6.1	07-Jan-2002	he	file esadsp.h was added on branch netbsd-1-5 on 2002-01-24 22:36:48 +0000
1.4.4.2	11-Jan-2002	nathanw	More catchup.
1.4.4.1	07-Jan-2002	nathanw	file esadsp.h was added on branch nathanw_sa on 2002-01-11 23:39:17 +0000
1.4.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.2.1	07-Jan-2002	thorpej	file esadsp.h was added on branch kqueue on 2002-01-10 19:56:33 +0000
1.6.126.1	18-May-2014	rmind	sync with head
1.6.120.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.110.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.11	13-May-2023	andvar	fix typo in unused definition s/ESA_KDATA_OUEUE_LEFT/ESA_KDATA_QUEUE_LEFT/.
1.10	11-Dec-2005	christos	merge ktrace-lwp.
1.9	15-Jan-2005	kent	ansify and KNF
1.8	06-Mar-2002	jmcneill	branches: 1.8.14; Fix another typo that I missed last time (foo -> ESA_foo)
1.7	06-Mar-2002	jmcneill	Fix a few typos (foo -> ESA_foo).
1.6	14-Jan-2002	pooka	branches: 1.6.2; Use standard PCI power management definitions from pcireg.h instead of homegrown ones. suggested by Soren
1.5	13-Jan-2002	jmcneill	Remove unnecessary (also unused) AC'97 definitions -- they already exist in syssrc/sys/dev/ic/ac97reg.h.
1.4	13-Jan-2002	pooka	* add power definitions * fix a few (hidden) typos in the definitions
1.3	06-Jan-2002	augustss	branches: 1.3.2; 1.3.4; Note that Don Kim wrote the original Linux driver.
1.2	06-Jan-2002	augustss	Change copyright somewhat. (These are not TNF copyright templates so I removed the advertising clause.)
1.1	06-Jan-2002	augustss	Part of the ESS Allegro-1 & Maestro3 driver.
1.3.4.4	12-Apr-2002	nathanw	Missed some revisions in the merge.
1.3.4.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.3.4.2	11-Jan-2002	nathanw	More catchup.
1.3.4.1	06-Jan-2002	nathanw	file esareg.h was added on branch nathanw_sa on 2002-01-11 23:39:18 +0000
1.3.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.3.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	06-Jan-2002	thorpej	file esareg.h was added on branch kqueue on 2002-01-10 19:56:33 +0000
1.6.2.2	24-Jan-2002	he	Pull up revisions 1.1-1.6 (new, requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.6.2.1	14-Jan-2002	he	file esareg.h was added on branch netbsd-1-5 on 2002-01-24 22:36:33 +0000
1.8.14.1	17-Jan-2005	skrll	Sync with HEAD.
1.12	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.11	23-Nov-2011	jmcneill	branches: 1.11.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.10	27-Mar-2008	jmcneill	branches: 1.10.16; 1.10.38; 1.10.40; Missed in previous device_t/softc split commit for esa, pointed out by xtraeme@
1.9	09-Dec-2007	jmcneill	branches: 1.9.10; Merge jmcneill-pm branch.
1.8	04-Mar-2007	christos	branches: 1.8.14; 1.8.16; 1.8.24; 1.8.26; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.7	11-Dec-2005	christos	branches: 1.7.26; 1.7.28; merge ktrace-lwp.
1.6	24-Mar-2005	jmcneill	branches: 1.6.2; Update my email address (the one listed hasn't been active in years).
1.5	15-Jan-2005	kent	branches: 1.5.2; 1.5.4; ansify and KNF
1.4	16-Mar-2002	jmcneill	branches: 1.4.12; Add multiple voice support to the ESS Allegro-1 / Maestro-3 family of audio accelerators. Mixing is done in hardware by the ASSP, but is limited to 4 simultaneous channels due to the restricted "minisrc" image that we are currently working with. Due to limitations in the audio subsystem, I'm currently attaching multiple 'audio' devices to 'esa', one for each voice. Because of this hack, the default ESA_NUM_VOICES is 1.
1.3	13-Jan-2002	pooka	branches: 1.3.2; Add support for the driver to work after suspend/resume. Does nothing fancy, works for the basic output case, but does not support programs which use playback while going into suspend. Sketched after code found in FreeBSD and Linux drivers.
1.2	07-Jan-2002	jmcneill	branches: 1.2.2; 1.2.4; Integrate recording support into the driver. Not very heavily tested yet.
1.1	06-Jan-2002	jmcneill	Import driver for the ESS Allegro-1 / Maestro-3 based PCI Audio Accelerators.
1.2.4.4	12-Apr-2002	nathanw	Merge miss.
1.2.4.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.4.2	11-Jan-2002	nathanw	More catchup.
1.2.4.1	07-Jan-2002	nathanw	file esavar.h was added on branch nathanw_sa on 2002-01-11 23:39:18 +0000
1.2.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.2.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.1	07-Jan-2002	thorpej	file esavar.h was added on branch kqueue on 2002-01-10 19:56:33 +0000
1.3.2.2	24-Jan-2002	he	Pull up revisions 1.1-1.3 (new, requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.3.2.1	13-Jan-2002	he	file esavar.h was added on branch netbsd-1-5 on 2002-01-24 22:36:13 +0000
1.4.12.2	01-Apr-2005	skrll	Sync with HEAD.
1.4.12.1	17-Jan-2005	skrll	Sync with HEAD.
1.5.4.1	26-Mar-2005	yamt	sync with head.
1.5.2.1	29-Apr-2005	kent	sync with -current
1.6.2.2	21-Jan-2008	yamt	sync with head
1.6.2.1	03-Sep-2007	yamt	sync with head.
1.7.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.7.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.8.26.1	11-Dec-2007	yamt	sync with head.
1.8.24.1	26-Dec-2007	ad	Sync with head.
1.8.16.1	09-Jan-2008	matt	sync with HEAD
1.8.14.1	05-Oct-2007	joerg	Nuke some unused prototypes and softc members after PNP conversion.
1.9.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.40.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.10.38.2	30-Oct-2012	yamt	sync with head
1.10.38.1	17-Apr-2012	yamt	sync with head
1.10.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.10.16.1	08-Dec-2008	ad	Convert some more drivers.
1.11.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.18	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.17	19-Oct-2009	bouyer	branches: 1.17.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.16	15-May-2009	tsutsui	Split device_t/softc. No crash on: esiop0 at pci0 dev 11 function 0: Symbios Logic 53c875 (ultra-wide scsi) siop0 at pci0 dev 9 function 0: Symbios Logic 53c810a (fast scsi) siop0 at gsc0 hpa 0xf0830000 path 2/0/7 irq 3 ipl 6: NCR53C720 rev 2
1.15	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.14	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.13	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.12	16-Nov-2006	christos	branches: 1.12.52; 1.12.68; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	27-Feb-2005	perry	nuke trailing whitespace
1.7	04-Feb-2005	perry	de-__P
1.6	09-Apr-2003	thorpej	branches: 1.6.2; 1.6.10; 1.6.12; Need <uvm/uvm_extern.h>.
1.5	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	23-Apr-2002	bouyer	branches: 1.2.6; 1.2.8; More copyright fixes, pointed out by Thomas. Thanks !
1.1	21-Apr-2002	bouyer	First cut at a esiop driver (enhanced siop). Doesn't implement tagged queuing yet. If is restricted to SIOP which implement the load/store instruction, and has 10 scratch registers (basically, 825 and newer, possibly 770). It implements a different interface between host and script, using a real ring for command starts, and improved support for reconnect which will allow 256 tag per device. It uses interrupt on the fly to signal complete command, which allows several commands to be serviced per interrupt and doesn't require the script to stop to signal command completion.
1.2.8.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.8.1	23-Apr-2002	jdolecek	file esiop_pci.c was added on branch kqueue on 2002-06-23 17:47:37 +0000
1.2.6.3	18-Oct-2002	nathanw	Catch up to -current.
1.2.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.2.6.1	23-Apr-2002	nathanw	file esiop_pci.c was added on branch nathanw_sa on 2002-06-20 03:45:22 +0000
1.6.12.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.6.12.1	12-Feb-2005	yamt	sync with head.
1.6.10.1	29-Apr-2005	kent	sync with -current
1.6.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.9.2.1	30-Dec-2006	yamt	sync with head.
1.10.22.2	10-Dec-2006	yamt	sync with head.
1.10.22.1	22-Oct-2006	yamt	sync with head
1.10.20.1	18-Nov-2006	ad	Sync with head.
1.12.68.2	23-Jul-2009	jym	Sync with HEAD.
1.12.68.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.12.52.2	11-Mar-2010	yamt	sync with head
1.12.52.1	16-May-2009	yamt	sync with head
1.17.4.1	05-Mar-2011	rmind	sync with head
1.67	01-Sep-2025	mlelstv	With the device_t split, esm(4) got damaged since the wrong pointer is passed around. Fixes PR install/59621 Pullups -11, -10, -9.
1.66	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.65	19-Apr-2020	isaki	Fix round_blocksize not to return 0.
1.64	05-Oct-2019	mrg	branches: 1.64.6; add missing break.
1.63	08-Jun-2019	isaki	branches: 1.63.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.62	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.61	16-Mar-2019	isaki	branches: 1.61.2; Use C99 style struct initializer to audio_hw_if.
1.60	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.59	07-Jul-2016	msaitoh	branches: 1.59.16; 1.59.18; KNF. Remove extra spaces. No functional change.
1.58	29-Mar-2014	christos	branches: 1.58.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.57	14-Sep-2013	joerg	GC wc_rdreg and wc_rdchctl
1.56	30-Jan-2012	drochner	branches: 1.56.6; 1.56.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.55	24-Nov-2011	mrg	branches: 1.55.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.54	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.53	24-Feb-2010	dyoung	branches: 1.53.10; 1.53.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.52	08-Jan-2010	dyoung	branches: 1.52.2; Expand PMF_FN_* macros.
1.51	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.50	12-Jun-2009	cegger	use __arraycount(). Same object code generated.
1.49	07-May-2009	cegger	device_t/softc split patch presented on tech-kern@, no comments on it
1.48	05-May-2009	cegger	struct cfdata * -> cfdata_t
1.47	10-Apr-2008	cegger	branches: 1.47.4; 1.47.16; 1.47.18; use aprint_*_dev and device_xname
1.46	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.45	23-Feb-2008	dyoung	Detach self and children. Make some functions static. Use device_t and accessors. Use PMF_FN_*.
1.44	09-Dec-2007	jmcneill	branches: 1.44.6; 1.44.10; Merge jmcneill-pm branch.
1.43	19-Oct-2007	ad	branches: 1.43.4; 1.43.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.42	04-Mar-2007	christos	branches: 1.42.2; 1.42.14; 1.42.16; 1.42.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.41	16-Nov-2006	christos	branches: 1.41.4; 1.41.6; __unused removal on arguments; approved by core.
1.40	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.39	25-Sep-2006	jmcneill	No need to call pci_set_powerstate from our powerhook, the bus handles this for us now.
1.38	24-Sep-2006	jmcneill	PR# 29372: notebook with Maestro 2E sound chip panics (w/ fix)
1.37	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.36	30-Aug-2006	christos	branches: 1.36.2; 1.36.4; fix incomplete initializers
1.35	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.34	14-Apr-2006	christos	branches: 1.34.2; 1.34.4; Coverity CID 1102: Make the code more coverity friendly.
1.33	08-Mar-2006	lukem	branches: 1.33.2; Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.32	24-Dec-2005	perry	branches: 1.32.4; 1.32.6; 1.32.8; 1.32.10; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.31	11-Dec-2005	christos	merge ktrace-lwp.
1.30	15-Jan-2005	kent	branches: 1.30.10; ansify and KNF
1.29	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.28	29-Oct-2004	yamt	branches: 1.28.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.27	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.26	08-Jul-2004	kleink	Fix AUDIO_DEBUG printf formats.
1.25	30-Apr-2004	simonb	Set the WAVCACHE_CHCTL_U8 flag for both specific-endian unsigned linear 8 bit encodings as well as for native. Fixes problems with playing 8 bit .wav files on my Dell Inspiron 5000e.
1.24	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.23	04-Dec-2003	keihan	branches: 1.23.2; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.22	03-Feb-2003	kleink	branches: 1.22.2; Fix printf format warnings.
1.21	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.20	31-Jan-2003	thorpej	Use aprint_*().
1.19	31-Dec-2002	fredette	Removed esm_freemem(), which isn't used any more.
1.18	30-Dec-2002	fredette	First cut at recording support for the ESS Maestro-2 and 2e. 16-bit mono recording seems to work OK. 16-bit stereo recording is missing the left channel for reasons unknown, fixes welcome. 8-bit recording still unsupported.
1.17	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	14-Jan-2002	pooka	Use standard PCI power management definitions from pcireg.h instead of homegrown ones. suggested by Soren
1.12	13-Nov-2001	lukem	add RCSID
1.11	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.10	25-Apr-2001	simonb	branches: 1.10.2; 1.10.4; 1.10.6; Spell "portege" correctly - it's a computer, not a car.
1.9	31-Mar-2001	rh	Fix DPRINTF statements so this compiles with debug enabled.
1.8	08-Mar-2001	ichiro	Fix powerhook routine. support resume/suspend.
1.7	12-Feb-2001	ichiro	branches: 1.7.2; Power Management Hook esm_powerhook();
1.6	26-Jan-2001	rh	Add left/right reversal quirk entry for Toshiba Protege 7020
1.5	26-Jan-2001	rh	Add quirk for Compaq Armada M700 to fix left/right channel reversal. Quirk entry provided by Joachim Koenig-Baltes <koenig@cms.tecmath.com>
1.4	18-Jan-2001	tv	No-op commit to force update to a non-"-kk" revision.
1.3	09-Jan-2001	rh	branches: 1.3.2; Add a quirk table for specific implementations. Change the NEC-specific subvendor test for GPIO initialization into a more generic quirk table flag. Add a quirk flag for hardware that has left and right channels reversed (currently the COMPAL 38W2 a.k.a. Dell Inspiron 5000e) Clean up PCM channel initialization and add channel swapping option.
1.2	09-Jan-2001	lukem	minor knf
1.1	08-Jan-2001	rh	Initial import of the Maestro 2/2e 'esm' PCI audio driver. This driver is based on the FreeBSD 'maestro' driver by Taku YAMAMOTO <taku@cent.saitama-u.ac.jp>
1.3.2.5	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.3.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.3.2.2	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.3.2.1	09-Jan-2001	bouyer	file esm.c was added on branch thorpej_scsipi on 2001-01-18 09:23:25 +0000
1.7.2.8	03-Jan-2003	thorpej	Sync with HEAD.
1.7.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.7.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.7.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.7.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.7.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.10.6.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.10.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.10.4.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.10.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.10.2.2	03-May-2001	he	Pull up revisions 1.1-1.10 (new, via patch, requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.10.2.1	25-Apr-2001	he	file esm.c was added on branch netbsd-1-5 on 2001-05-03 20:57:08 +0000
1.22.2.6	17-Jan-2005	skrll	Sync with HEAD.
1.22.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.22.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.22.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.22.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.22.2.1	03-Aug-2004	skrll	Sync with HEAD
1.23.2.1	22-Sep-2004	jmc	Pullup rev 1.27 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.28.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.28.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.30.10.7	24-Mar-2008	yamt	sync with head.
1.30.10.6	27-Feb-2008	yamt	sync with head.
1.30.10.5	21-Jan-2008	yamt	sync with head
1.30.10.4	27-Oct-2007	yamt	sync with head.
1.30.10.3	03-Sep-2007	yamt	sync with head.
1.30.10.2	30-Dec-2006	yamt	sync with head.
1.30.10.1	21-Jun-2006	yamt	sync with head.
1.32.10.1	19-Apr-2006	elad	sync with head.
1.32.8.4	03-Sep-2006	yamt	sync with head.
1.32.8.3	26-Jun-2006	yamt	sync with head.
1.32.8.2	24-May-2006	yamt	sync with head.
1.32.8.1	13-Mar-2006	yamt	sync with head.
1.32.6.1	22-Apr-2006	simonb	Sync with head.
1.32.4.1	09-Sep-2006	rpaulo	sync with head
1.33.2.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.34.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.34.2.1	19-Jun-2006	chap	Sync with head.
1.36.4.2	10-Dec-2006	yamt	sync with head.
1.36.4.1	22-Oct-2006	yamt	sync with head
1.36.2.1	18-Nov-2006	ad	Sync with head.
1.41.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.41.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.42.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.42.16.3	23-Mar-2008	matt	sync with HEAD
1.42.16.2	09-Jan-2008	matt	sync with HEAD
1.42.16.1	06-Nov-2007	matt	sync with HEAD
1.42.14.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.42.14.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.42.14.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.42.14.1	05-Oct-2007	joerg	Convert esm(4) to PNP power management.
1.42.2.1	23-Oct-2007	ad	Sync with head.
1.43.6.1	11-Dec-2007	yamt	sync with head.
1.43.4.1	26-Dec-2007	ad	Sync with head.
1.44.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.44.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.44.6.1	24-Mar-2008	keiichi	sync with head.
1.47.18.2	23-Jul-2009	jym	Sync with HEAD.
1.47.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.47.16.3	13-Dec-2008	ad	Fix locking for ac97.
1.47.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.47.16.1	08-Dec-2008	ad	Convert some more drivers.
1.47.4.3	11-Mar-2010	yamt	sync with head
1.47.4.2	20-Jun-2009	yamt	sync with head
1.47.4.1	16-May-2009	yamt	sync with head
1.52.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.53.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.53.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.53.10.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.10.1	17-Apr-2012	yamt	sync with head
1.55.2.1	18-Feb-2012	mrg	merge to -current.
1.56.10.1	18-May-2014	rmind	sync with head
1.56.6.2	03-Dec-2017	jdolecek	update from HEAD
1.56.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.58.6.1	09-Jul-2016	skrll	Sync with HEAD
1.59.18.3	21-Apr-2020	martin	Sync with HEAD
1.59.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.59.18.1	10-Jun-2019	christos	Sync with HEAD
1.59.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.61.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.61.2.2	28-Apr-2019	isaki	Adapt to audio2. - Recording seems a bit tricky so I left it untouched. (It should work on audio2 as well if it works on -current/-8.)
1.61.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.63.2.1	27-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #493): sys/dev/pci/esm.c: revision 1.64 add missing break.
1.64.6.1	20-Apr-2020	bouyer	Sync with HEAD
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.4	30-Dec-2002	fredette	branches: 1.4.2; First cut at recording support for the ESS Maestro-2 and 2e. 16-bit mono recording seems to work OK. 16-bit stereo recording is missing the left channel for reasons unknown, fixes welcome. 8-bit recording still unsupported.
1.3	14-Jan-2002	pooka	Use standard PCI power management definitions from pcireg.h instead of homegrown ones. suggested by Soren
1.2	18-Jan-2001	tv	branches: 1.2.2; 1.2.4; 1.2.6; No-op commit to force update to a non-"-kk" revision.
1.1	08-Jan-2001	rh	branches: 1.1.2; Initial import of the Maestro 2/2e 'esm' PCI audio driver. This driver is based on the FreeBSD 'maestro' driver by Taku YAMAMOTO <taku@cent.saitama-u.ac.jp>
1.1.2.2	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.1.2.1	08-Jan-2001	bouyer	file esmreg.h was added on branch thorpej_scsipi on 2001-01-18 09:23:25 +0000
1.2.6.1	11-Feb-2002	jdolecek	Sync w/ -current.
1.2.4.2	03-May-2001	he	Pull up revisions 1.1-1.2 (new, requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.2.4.1	18-Jan-2001	he	file esmreg.h was added on branch netbsd-1-5 on 2001-05-03 20:57:28 +0000
1.2.2.2	03-Jan-2003	thorpej	Sync with HEAD.
1.2.2.1	28-Feb-2002	nathanw	Catch up to -current.
1.4.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.19	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.18	23-Nov-2011	jmcneill	branches: 1.18.50; 1.18.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.17	07-May-2009	cegger	branches: 1.17.12; 1.17.14; device_t/softc split patch presented on tech-kern@, no comments on it
1.16	23-Feb-2008	dyoung	branches: 1.16.4; 1.16.16; 1.16.18; Detach self and children. Make some functions static. Use device_t and accessors. Use PMF_FN_*.
1.15	09-Dec-2007	jmcneill	branches: 1.15.6; 1.15.10; Merge jmcneill-pm branch.
1.14	04-Mar-2007	christos	branches: 1.14.14; 1.14.16; 1.14.24; 1.14.26; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.13	11-Dec-2005	christos	branches: 1.13.26; 1.13.28; merge ktrace-lwp.
1.12	15-Jan-2005	kent	branches: 1.12.10; ansify and KNF
1.11	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.10	22-Sep-2004	kent	branches: 1.10.2; ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.9	04-Dec-2003	keihan	branches: 1.9.2; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.8	01-Feb-2003	thorpej	branches: 1.8.2; Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.7	31-Dec-2002	fredette	Removed esm_freemem(), which isn't used any more.
1.6	30-Dec-2002	fredette	First cut at recording support for the ESS Maestro-2 and 2e. 16-bit mono recording seems to work OK. 16-bit stereo recording is missing the left channel for reasons unknown, fixes welcome. 8-bit recording still unsupported.
1.5	31-Dec-2001	thorpej	Avoid an "unmamed fields" warning (gcc 3.1).
1.4	12-Feb-2001	ichiro	branches: 1.4.2; 1.4.4; 1.4.6; Power Management Hook esm_powerhook();
1.3	18-Jan-2001	tv	No-op commit to force update to a non-"-kk" revision.
1.2	09-Jan-2001	rh	branches: 1.2.2; Add a quirk table for specific implementations. Change the NEC-specific subvendor test for GPIO initialization into a more generic quirk table flag. Add a quirk flag for hardware that has left and right channels reversed (currently the COMPAL 38W2 a.k.a. Dell Inspiron 5000e) Clean up PCM channel initialization and add channel swapping option.
1.1	08-Jan-2001	rh	Initial import of the Maestro 2/2e 'esm' PCI audio driver. This driver is based on the FreeBSD 'maestro' driver by Taku YAMAMOTO <taku@cent.saitama-u.ac.jp>
1.2.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.2.2.2	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.2.2.1	09-Jan-2001	bouyer	file esmvar.h was added on branch thorpej_scsipi on 2001-01-18 09:23:25 +0000
1.4.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.4.2	03-May-2001	he	Pull up revisions 1.1-1.4 (new, requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.4.4.1	12-Feb-2001	he	file esmvar.h was added on branch netbsd-1-5 on 2001-05-03 20:57:46 +0000
1.4.2.2	03-Jan-2003	thorpej	Sync with HEAD.
1.4.2.1	08-Jan-2002	nathanw	Catch up to -current.
1.8.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.8.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.8.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.8.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.8.2.1	03-Aug-2004	skrll	Sync with HEAD
1.9.2.1	22-Sep-2004	jmc	Pullup rev 1.10 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.10.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.10.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.12.10.3	27-Feb-2008	yamt	sync with head.
1.12.10.2	21-Jan-2008	yamt	sync with head
1.12.10.1	03-Sep-2007	yamt	sync with head.
1.13.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.13.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.14.26.1	11-Dec-2007	yamt	sync with head.
1.14.24.1	26-Dec-2007	ad	Sync with head.
1.14.16.2	23-Mar-2008	matt	sync with HEAD
1.14.16.1	09-Jan-2008	matt	sync with HEAD
1.14.14.1	05-Oct-2007	joerg	Nuke some unused prototypes and softc members after PNP conversion.
1.15.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.6.1	24-Mar-2008	keiichi	sync with head.
1.16.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.16.16.1	08-Dec-2008	ad	Convert some more drivers.
1.16.4.1	16-May-2009	yamt	sync with head
1.17.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.17.12.1	17-Apr-2012	yamt	sync with head
1.18.54.1	28-Apr-2019	isaki	Adapt to audio2. - Recording seems a bit tricky so I left it untouched. (It should work on audio2 as well if it works on -current/-8.)
1.18.50.1	10-Jun-2019	christos	Sync with HEAD
1.75	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.74	26-Apr-2021	thorpej	branches: 1.74.6; - The "eso" device does not need to carry the "midibus" interface attribute. It attaches "opl" and "mpu" instances, which themselves attach "midi". - Be explicit about specifying the "eso" interface attribute when attaching "opl", "mpu", and "joy" instances.
1.73	24-Apr-2021	thorpej	branches: 1.73.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.72	28-Oct-2019	joerg	branches: 1.72.10; Fix operand priority of bit op vs ternary operator.
1.71	08-Jun-2019	isaki	branches: 1.71.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.70	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.69	16-Mar-2019	isaki	branches: 1.69.2; Use C99 style struct initializer to audio_hw_if.
1.68	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.67	01-Jun-2017	chs	branches: 1.67.8; 1.67.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.66	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.65	29-Mar-2014	christos	branches: 1.65.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.64	16-Oct-2013	christos	remove unused variables.
1.63	17-Jul-2013	soren	A few "isa" files are so tightly integrated into the x86 platform code as to not really be part of the optional isa bus autoconf machinery. Allows configuring a kernel like so: include "arch/amd64/conf/GENERIC" no isa
1.62	27-Oct-2012	chs	branches: 1.62.2; 1.62.8; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.61	12-Aug-2012	gson	branches: 1.61.2; Add missing locking calls to eso_attach() to avoid panic on boot. Partly based on a patch from mrg. Fixes PR kern/46791.
1.60	15-May-2012	mrg	don't kassert that the sc_intr_lock is held in eso_reset(). it's only called from attach and doesn't need to have it held, and PR 46451 shows that it currently asserts.
1.59	24-Nov-2011	mrg	branches: 1.59.2; 1.59.4; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.58	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.57	26-Nov-2009	njoly	branches: 1.57.12; 1.57.14; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.56	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.55	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.54	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.53	10-Apr-2008	cegger	branches: 1.53.4; 1.53.16; 1.53.18; use aprint_*_dev and device_xname
1.52	01-Apr-2008	xtraeme	Fix fallout from device_t/softc split to mpu(4): always pass the mpu_softc to mpu_intr(). Found by Takeshi Nakayama.
1.51	19-Oct-2007	ad	branches: 1.51.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.50	04-Mar-2007	christos	branches: 1.50.2; 1.50.14; 1.50.16; 1.50.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.49	11-Jan-2007	kleink	branches: 1.49.2; 1.49.4; Replace home-grown list management for DMA buffers with SLISTs, and refactor.
1.48	18-Dec-2006	kleink	Redo rev. 1.42, fixing the original array subscript pasto the right way.
1.47	14-Dec-2006	christos	explicitly set clk to 1 or 0.
1.46	14-Dec-2006	kleink	Un-break clock selection after -Wsign-compare cleanup in rev. 1.43; from Stephen Ma in PR kern/35220.
1.45	16-Nov-2006	christos	branches: 1.45.2; __unused removal on arguments; approved by core.
1.44	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.43	30-Aug-2006	christos	branches: 1.43.2; 1.43.4; - fix missing initializer - fix < 0 comparison with unsigned
1.42	14-Apr-2006	christos	Coverity CID 1539: Remove bogus test that caused array overrun.
1.41	11-Dec-2005	christos	branches: 1.41.4; 1.41.6; 1.41.8; 1.41.10; 1.41.12; merge ktrace-lwp.
1.40	26-Jan-2005	kleink	branches: 1.40.6; Clean up the array of audio_formats supported.
1.39	15-Jan-2005	kent	branches: 1.39.2; 1.39.4; ansify and KNF
1.38	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.37	29-Oct-2004	yamt	branches: 1.37.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.36	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.35	08-Jul-2004	kleink	In addition to alpha, grossly work around the 24-bit limitation of the Audio 1 DMA engine by allocating through the ISA DMA tag on amd64 and i386.
1.34	25-May-2004	kleink	General mixer initialization improvements around the MONO_IN/MONO_OUT/ PreAmp register, including a new MONO_IN mixer stage bypass mixer control; inspired by (and fixing) PR kern/18342 from Stephen Ma.
1.33	17-Feb-2004	kleink	branches: 1.33.2; Fix a (non-dangerous) malloc size botch in eso_allocm(); from Stephen Ma in PR kern/18342.
1.32	17-Feb-2004	kleink	Audio 2's DMA current address register implements only the least significant 20 bits, so make sure its memory allocation won't cross a 1M boundary. From Stephen Ma in PR kern/18342.
1.31	05-Feb-2003	kleink	branches: 1.31.2; unask -> unmask
1.30	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.29	31-Jan-2003	thorpej	Use aprint_*().
1.28	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.27	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.26	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.25	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.24	25-Apr-2002	kleink	branches: 1.24.4; Add a joystick attachment to the ESS Solo-1 driver.
1.23	13-Nov-2001	lukem	add RCSID
1.22	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.21	19-Jul-2001	kleink	branches: 1.21.2; Hint BUS_DMA_READ and BUS_DMA_WRITE.
1.20	28-Dec-2000	sommerfeld	branches: 1.20.2; 1.20.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.19	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.18	22-Mar-2000	kleink	branches: 1.18.4; Scale the drain timeout for allowing auto-initialize DMA to complete.
1.17	18-Jan-2000	cjs	Fix the size of the playback buffer. Patch from Dave Sainty <dave@dtsp.co.nz>.
1.16	28-Dec-1999	kleink	Off-by-one error in limiting the DMAC buffer size for Audio 2 to 64K (actually 64K - 1); from Dave Sainty in PR kern/9064.
1.15	28-Dec-1999	kleink	Fix a buffer KVA comparison botch in eso_mappage(); from Dave Sainty in PR kern/9062.
1.14	10-Dec-1999	kleink	Add support code to have the Hardware Volume Control interact with the software mixer Master Volume state in a defined way by lazily updating the latter if input from the former was processed.
1.13	03-Dec-1999	kleink	A PCI Revision ID of 0x02 is what ESS happens to be calling ES1946 Revision E.
1.12	02-Nov-1999	kleink	Add a mixer control for the Master Volume Mute knob.
1.11	01-Nov-1999	augustss	Make little endian and big endian version of all functions that deal with 16 bit samples. Let the drivers decide which ones to use.
1.10	28-Oct-1999	leo	Remove htopci(). bus_space_write_4() takes care of necessary endian conversions.
1.9	10-Oct-1999	cgd	branches: 1.9.2; 1.9.4; don't print using string pointer one off the end of the eso_rev2model array. pointed out by Kouichi Hashikawa in PR#8601.
1.8	28-Sep-1999	kleink	branches: 1.8.2; Since the Audio 1 DMA controller implements the 24 low address bits only, add the gross hack originally implemented by Charles Hannum in the SonicVibes driver to force allocation of DMA memory for that channel through the Scatter- Gather Map to get an address range that fits in. The eso driver now works on Alphas except for DWLPX-based systems, whose owners are not likely to desire audio or likely can afford slightly less cheap audio hardware. :-)
1.7	23-Sep-1999	kleink	Oops, we do have to check ourselves whether a given value is valid within an enumeration.
1.6	02-Sep-1999	kleink	Use the approximated sampling rate to compute the filter rate.
1.5	14-Aug-1999	kleink	Add missing MPU interrupt glue.
1.4	14-Aug-1999	kleink	Forgot to initialize a pointer; from Simon Burge.
1.3	02-Aug-1999	augustss	branches: 1.3.2; Move the mpu device declaration to conf/files. Let the mpu device attach at the sb device, and then midi at the mpu. Update the mpu at eso attachment.
1.2	01-Aug-1999	augustss	Move the MPU driver isa/ to ic/. (I didn't move it in the repository because the revision history is very short and uninteresting. :)
1.1	12-Jul-1999	kleink	Add a driver for the ESS Technology Solo-1 PCI AudioDrive line of chips.
1.3.2.2	02-Aug-1999	thorpej	Update from trunk.
1.3.2.1	02-Aug-1999	thorpej	file eso.c was added on branch chs-ubc2 on 1999-08-02 22:03:56 +0000
1.8.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.9.4.1	15-Nov-1999	fvdl	Sync with -current
1.9.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.9.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.18.4.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.20.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.20.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.20.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.20.4.1	03-Aug-2001	lukem	update to -current
1.20.2.6	03-Jan-2003	thorpej	Sync with HEAD.
1.20.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.20.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.20.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.20.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.20.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.21.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.24.4.2	15-Mar-2004	jmc	Pullup rev 1.33 (requested by kleink in ticket #1614) Fix a (non-dangerous) malloc size botch in eso_allocm(). PR#18342
1.24.4.1	15-Mar-2004	jmc	Pullup rev 1.32 (requested by kleink in ticket #1613) Audio 2's DMA current address register implements only the least significant 20 bits, so make sure its memory allocation won't cross a 1M boundary.
1.31.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.31.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.31.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.31.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.31.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.31.2.1	03-Aug-2004	skrll	Sync with HEAD
1.33.2.2	10-Jul-2004	tron	Pull up revision 1.35 (requested by kleink in ticket #633): In addition to alpha, grossly work around the 24-bit limitation of the Audio 1 DMA engine by allocating through the ISA DMA tag on amd64 and i386.
1.33.2.1	29-May-2004	tron	Pull up revision 1.34 (requested by kleink in ticket #400): General mixer initialization improvements around the MONO_IN/MONO_OUT/ PreAmp register, including a new MONO_IN mixer stage bypass mixer control; inspired by (and fixing) PR kern/18342 from Stephen Ma.
1.37.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.37.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.39.4.1	12-Feb-2005	yamt	sync with head.
1.39.2.1	29-Apr-2005	kent	sync with -current
1.40.6.5	27-Oct-2007	yamt	sync with head.
1.40.6.4	03-Sep-2007	yamt	sync with head.
1.40.6.3	26-Feb-2007	yamt	sync with head.
1.40.6.2	30-Dec-2006	yamt	sync with head.
1.40.6.1	21-Jun-2006	yamt	sync with head.
1.41.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.41.10.1	19-Apr-2006	elad	sync with head.
1.41.8.2	03-Sep-2006	yamt	sync with head.
1.41.8.1	24-May-2006	yamt	sync with head.
1.41.6.1	22-Apr-2006	simonb	Sync with head.
1.41.4.1	09-Sep-2006	rpaulo	sync with head
1.43.4.4	21-Dec-2006	yamt	sync with head.
1.43.4.3	18-Dec-2006	yamt	sync with head.
1.43.4.2	10-Dec-2006	yamt	sync with head.
1.43.4.1	22-Oct-2006	yamt	sync with head
1.43.2.3	01-Feb-2007	ad	Sync with head.
1.43.2.2	12-Jan-2007	ad	Sync with head.
1.43.2.1	18-Nov-2006	ad	Sync with head.
1.45.2.1	17-Dec-2006	riz	Pull up following revision(s) (requested by kleink in ticket #278): sys/dev/pci/eso.c: revision 1.46 sys/dev/pci/eso.c: revision 1.47 Un-break clock selection after -Wsign-compare cleanup in rev. 1.43; from Stephen Ma in PR kern/35220. explicitly set clk to 1 or 0.
1.49.4.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.49.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.50.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.50.16.1	06-Nov-2007	matt	sync with HEAD
1.50.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.50.2.1	23-Oct-2007	ad	Sync with head.
1.51.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.51.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.53.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.53.16.1	11-Dec-2008	ad	Checkpoint work in progress.
1.53.4.2	11-Mar-2010	yamt	sync with head
1.53.4.1	16-May-2009	yamt	sync with head
1.57.14.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.57.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.57.12.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.57.12.3	30-Oct-2012	yamt	sync with head
1.57.12.2	23-May-2012	yamt	sync with head.
1.57.12.1	17-Apr-2012	yamt	sync with head
1.59.4.2	13-Aug-2012	riz	Pull up following revision(s) (requested by gson in ticket #494): sys/dev/pci/eso.c: revision 1.61 Add missing locking calls to eso_attach() to avoid panic on boot. Partly based on a patch from mrg. Fixes PR kern/46791.
1.59.4.1	13-Jun-2012	riz	Pull up following revision(s) (requested by mrg in ticket #332): sys/dev/pci/eso.c: revision 1.60 don't kassert that the sc_intr_lock is held in eso_reset(). it's only called from attach and doesn't need to have it held, and PR 46451 shows that it currently asserts.
1.59.2.1	02-Jun-2012	mrg	sync to latest -current.
1.61.2.3	03-Dec-2017	jdolecek	update from HEAD
1.61.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.61.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.62.8.1	23-Jul-2013	riastradh	sync with HEAD
1.62.2.2	18-May-2014	rmind	sync with head
1.62.2.1	28-Aug-2013	rmind	sync with head
1.65.6.2	28-Aug-2017	skrll	Sync with HEAD
1.65.6.1	09-Jul-2016	skrll	Sync with HEAD
1.67.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.67.10.1	10-Jun-2019	christos	Sync with HEAD
1.67.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.69.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.69.2.3	28-Apr-2019	isaki	Don't release sc_lock on eso_halt_{input,output}. halt_* is called with sc_lock && sc_intr_lock held. This lock order is first sc_lock and then sc_intr_lock. So unlocking sc_lock with sc_intr_lock held is wrong operation. And cv_wait(sc_intr_lock) will work even with sc_lock held.
1.69.2.2	28-Apr-2019	isaki	Adapt to audio2. - Select a few typical frequencies which doesn't have rounding error instead of whole range.
1.69.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.71.2.1	27-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #492): sys/dev/pci/sv.c: revision 1.57 sys/dev/pci/eso.c: revision 1.72 sys/dev/ieee1394/fwohci.c: revision 1.144 Fix operand priority of bit op vs ternary operator.
1.72.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.73.2.1	13-May-2021	thorpej	Sync with HEAD.
1.74.6.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.9	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.8	11-Dec-2005	christos	branches: 1.8.164; 1.8.168; merge ktrace-lwp.
1.7	15-Jan-2005	kent	ansify and KNF
1.6	25-May-2004	kleink	Fix a shift-by-one-bit glitch in the MONO_OUT mixer register's definition; from Stephen Ma in PR kern/18342.
1.5	10-Dec-1999	kleink	branches: 1.5.28; 1.5.30; Add support code to have the Hardware Volume Control interact with the software mixer Master Volume state in a defined way by lazily updating the latter if input from the former was processed.
1.4	10-Dec-1999	kleink	Add a symbolic name for the ES1946 rev. E in this place, too.
1.3	23-Sep-1999	kleink	branches: 1.3.2; 1.3.8; Add Game Port definitions.
1.2	14-Aug-1999	kleink	Add missing MPU interrupt glue.
1.1	12-Jul-1999	kleink	branches: 1.1.2; Add a driver for the ESS Technology Solo-1 PCI AudioDrive line of chips.
1.1.2.2	02-Aug-1999	thorpej	Update from trunk.
1.1.2.1	12-Jul-1999	thorpej	file esoreg.h was added on branch chs-ubc2 on 1999-08-02 22:03:56 +0000
1.3.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.3.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.30.1	29-May-2004	tron	Pull up revision 1.6 (requested by kleink in ticket #400): Fix a shift-by-one-bit glitch in the MONO_OUT mixer register's definition; from Stephen Ma in PR kern/18342.
1.5.28.4	17-Jan-2005	skrll	Sync with HEAD.
1.5.28.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.28.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.28.1	03-Aug-2004	skrll	Sync with HEAD
1.8.168.1	28-Apr-2019	isaki	Adapt to audio2. - Select a few typical frequencies which doesn't have rounding error instead of whole range.
1.8.164.1	10-Jun-2019	christos	Sync with HEAD
1.11	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.10	23-Nov-2011	jmcneill	branches: 1.10.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.9	06-May-2009	cegger	branches: 1.9.12; 1.9.14; struct device * -> device_t, no functional changes intended.
1.8	12-Jan-2007	kleink	branches: 1.8.4; 1.8.44; 1.8.58; 1.8.60; Replace home-grown list management for DMA buffers with SLISTs, and refactor.
1.7	11-Dec-2005	christos	branches: 1.7.20; merge ktrace-lwp.
1.6	15-Jan-2005	kent	branches: 1.6.10; ansify and KNF
1.5	25-May-2004	kleink	General mixer initialization improvements around the MONO_IN/MONO_OUT/ PreAmp register, including a new MONO_IN mixer stage bypass mixer control; inspired by (and fixing) PR kern/18342 from Stephen Ma.
1.4	22-Mar-2000	kleink	branches: 1.4.28; 1.4.30; Scale the drain timeout for allowing auto-initialize DMA to complete.
1.3	02-Nov-1999	kleink	Add a mixer control for the Master Volume Mute knob.
1.2	02-Aug-1999	augustss	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; Move the mpu device declaration to conf/files. Let the mpu device attach at the sb device, and then midi at the mpu. Update the mpu at eso attachment.
1.1	12-Jul-1999	kleink	Add a driver for the ESS Technology Solo-1 PCI AudioDrive line of chips.
1.2.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.2.6.1	15-Nov-1999	fvdl	Sync with -current
1.2.4.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.2	02-Aug-1999	thorpej	Update from trunk.
1.2.2.1	02-Aug-1999	thorpej	file esovar.h was added on branch chs-ubc2 on 1999-08-02 22:03:57 +0000
1.4.30.1	29-May-2004	tron	Pull up revision 1.5 (requested by kleink in ticket #400): General mixer initialization improvements around the MONO_IN/MONO_OUT/ PreAmp register, including a new MONO_IN mixer stage bypass mixer control; inspired by (and fixing) PR kern/18342 from Stephen Ma.
1.4.28.4	17-Jan-2005	skrll	Sync with HEAD.
1.4.28.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.28.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.28.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	26-Feb-2007	yamt	sync with head.
1.7.20.1	01-Feb-2007	ad	Sync with head.
1.8.60.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.8.58.1	11-Dec-2008	ad	Checkpoint work in progress.
1.8.44.1	16-May-2009	yamt	sync with head
1.8.4.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.9.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.9.12.2	30-Oct-2012	yamt	sync with head
1.9.12.1	17-Apr-2012	yamt	sync with head
1.10.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.11	02-Nov-2014	christos	revert changes, this needed to be done differently.
1.10	02-Nov-2014	christos	make the apple code compile again.
1.9	01-Nov-2014	christos	fix linking of DRMKMS kernel. XXX: apple does not compile!
1.8	29-Oct-2014	uebayasi	Revert "normalization" for now. XXX agp(4)'s config/attachment should be rewritten
1.7	10-Oct-2014	uebayasi	Normalize.
1.6	04-Aug-2007	kiyohara	branches: 1.6.34; Add support agp_amd64. Imported from FreeBSD.
1.5	25-Mar-2007	macallan	branches: 1.5.4; 1.5.8; add very experimental Apple UniNorth AGP driver
1.4	11-Dec-2005	christos	branches: 1.4.26; 1.4.30; 1.4.32; 1.4.34; merge ktrace-lwp.
1.3	30-Aug-2004	drochner	branches: 1.3.12; add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.2	15-Sep-2001	thorpej	branches: 1.2.2; 1.2.4; 1.2.8; 1.2.24; Give each AGP controller its own attribute, and let the "agpbus" device (rather, the device that carries that attribute) also carry one or more attributes indicating which type of controller it might be. This will allow systems that might have AGP, but would never have e.g. an Intel PCI-Host bridge, to trim out code that won't be used.
1.1	15-Sep-2001	thorpej	Move the AGP device declaration stuff into files.agp.
1.2.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.24.1	03-Sep-2004	skrll	Sync with HEAD
1.2.8.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.8.1	15-Sep-2001	thorpej	file files.agp was added on branch kqueue on 2002-01-10 19:56:34 +0000
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	15-Sep-2001	fvdl	file files.agp was added on branch thorpej-devvp on 2001-10-01 12:45:53 +0000
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	15-Sep-2001	nathanw	file files.agp was added on branch nathanw_sa on 2001-09-21 22:35:55 +0000
1.3.12.1	03-Sep-2007	yamt	sync with head.
1.4.34.1	29-Mar-2007	reinoud	Pullup to -current
1.4.32.1	11-Jul-2007	mjf	Sync with head.
1.4.30.2	20-Aug-2007	ad	Sync with HEAD.
1.4.30.1	10-Apr-2007	ad	Sync with head.
1.4.26.1	15-Apr-2007	yamt	sync with head.
1.5.8.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.5.4.1	15-Aug-2007	skrll	Sync with HEAD.
1.6.34.2	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.6.34.1	04-Aug-2007	kiyohara	file files.agp was added on branch matt-mips64 on 2007-08-04 09:34:44 +0000
1.2	25-Jun-2005	dyoung	Move the definitions in dev/pci/files.ath into dev/pci/files.pci. Stop including dev/pci/files.ath in arch/i386/conf/files.i386, since we get the same definitions by including dev/pci/files.pci, now. Remove dev/pci/files.ath. Add arch/macppc/conf/Makefile.macppc with directives for linking the Atheros HAL for PowerPC. In athhal-powerpc-be-eabi.opt_ah.h, #define AH_REGOPS_FUNC 1, since otherwise the linker complains that the PowerPC HAL cannot link with register-read/write subroutines. Add ath(4) to the GENERIC macppc kernel configuration; comment it out.
1.1	13-Oct-2003	dyoung	branches: 1.1.4; Files belonging to the ath(4) PCI front-end. We can only attach an ath(4) on architectures where we have a HAL (which is only x86, for now) so I list the files belonging to the PCI front-end separately from the other PCI files.
1.1.4.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	13-Oct-2003	skrll	file files.ath was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.451	04-Aug-2025	skrll	Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots.
1.450	16-Jun-2025	macallan	branches: 1.450.2; defflag PM3FB_DEBUG wipe glyphcache when re-entering terminal mode
1.449	27-May-2025	macallan	pm3fb needs glyphcache now
1.448	20-Oct-2023	msaitoh	branches: 1.448.6; eqos(4): Add initial support for Intel Elkhart Lake internal Ethernet devices. - Only tested on PSE SGMII 1G Ethernet MAC with MaxLinear GPY115. - I don't know why dmat64 doesn't work. eqos_attach() have a special code if EQOS_HW_FEATURE_ADDR64_32BIT(sc) is true, but it seems it doesn't work. - TODO: Multiqueue support. Detach support.
1.447	04-Oct-2023	rin	igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.446	12-Apr-2023	riastradh	ichsmb(4), tco(4): Add support for TCO on newer Intel chipsets. TCO (`Total Cost of Ownership', Intel's bizarre name for a watchdog timer) used to hang off the Intel I/O platform controller hub's (ICH) low-pin-count interface bridge (LPC IB), or ichlpcib(4). On newer devices, it hangs off the ICH SMBus instead. Tested on INTEL 100SERIES_SMB (works) and INTEL 100SERIES_LP_SMB (doesn't work, still not sure why). XXX kernel revbump: This breaks the module ABI -- tco(4) modules older than the change to make ta_has_rcba into ta_version will incorrectly attach at buses they do not understand. (However, the tco(4) driver is statically built into GENERIC, so maybe it's safe for pullup since the module wouldn't have worked anyway.)
1.445	14-Oct-2022	jmcneill	branches: 1.445.2; Add a PCI resource manager and use it on Arm ACPI platforms. The Arm ACPI code relied on PCI_NETBSD_CONFIGURE to configure devices that were not enabled by system firmware. This is not safe to do unless the firmware explicitly permits it using a device specific method defined in the PCI firmware spec. Introduce a new PCI resource manager that discovers what has already been configured by firmware and allocates from the remaining space. This will ensure that devices setup by firmware are untouched and only will program BARs of devices that are not enabled at boot time. The current implementation assumes that the parent PCI-PCI bridge's are already configured. A worthwhile improvement in the future would be to support programming windows for bridges that are not fully configured.
1.444	16-Sep-2022	knakahara	Add ALWAYS_TXDEFER option to ixl(4), too.
1.443	16-Sep-2022	knakahara	Add ALWAYS_TXDEFER option to vmx(4), too.
1.442	16-Sep-2022	knakahara	ixg(4) add an option for Tx to use deferred softint regardless of whether can get txq lock or not. That imporve (7%) and stabilize throughput. But that can cause latency degradation, so off by default. ok'ed by msaitoh@n.o.
1.441	07-Sep-2022	martin	Fix the build, add new emuxki_boards.c
1.440	17-Dec-2021	knakahara	Add missing WM_TX_{,INTR_}PROCESS_LIMIT_DEFAULT options same as RX side.
1.439	12-Nov-2021	skrll	Trailing whitespace
1.438	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.437	26-Apr-2021	thorpej	- The "eso" device does not need to carry the "midibus" interface attribute. It attaches "opl" and "mpu" instances, which themselves attach "midi". - Be explicit about specifying the "eso" interface attribute when attaching "opl", "mpu", and "joy" instances.
1.436	09-Mar-2021	msaitoh	branches: 1.436.4; Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.435	01-Mar-2021	jakllsch	Update rge(4) from older OpenBSD, finish porting. Should consider merging this all into re(4) and rgephy(4) someday. Some cleanup tasks remain here.
1.434	17-Feb-2021	knakahara	In 64 bit architectures, WM_EVENT_COUNTER is enabled by default. No objection from tech-kern@n.o and tech-net@n.o. ok'ed by msaitoh@n.o.
1.433	30-Jan-2021	jmcneill	Sync with OpenBSD r1.98. Lots of improvements including checksum offload, hardware vlan tagging, and support for multiple receive queues.
1.432	14-Oct-2020	ryo	branches: 1.432.2; vmx(4) should be MI. moved to sys/dev/pci from sys/arch/x86/pci
1.431	08-Sep-2020	yamaguchi	Added iavf(4) that is based on OpenBSD's iavf(4) implementation reviewed by msaitoh@n.o and knakahara@n.o
1.430	26-Jul-2020	jdolecek	Add driver for Intel XMM7360 LTE modem, based upon Linux driver available at https://github.com/xmm7360/xmm7360-pci This version works on Linux, OpenBSD, and NetBSD. OpenBSD port written for genua GmbH Modem requires python script from the master site to initialize the network, it will be added to pkgsrc shortly
1.429	24-Jun-2020	thorpej	Add a PCI front-end for the "amdccp" (AMD Cryptographic Coprocessor) driver.
1.428	21-May-2020	macallan	- remove i2cbus etc. on drivers that don't use iic - for private iic buses pull in only what we need and skip the bus attachment
1.427	20-May-2020	macallan	radeonfb doesn't attach an iic*, so pull in iic, not i2cbus
1.426	09-Feb-2020	jmcneill	Retire azalia(4).
1.425	25-Jan-2020	thorpej	Retire the le@pci attachment. It has been superseded (and matched at a higher priority) by the pcn(4) driver since NetBSD 1.6.
1.424	20-Jan-2020	thorpej	Remove FDDI support.
1.423	19-Jan-2020	thorpej	Remove the de(4) driver, which has long since been supplanted by the tlp(4) driver, which supports more chips and more board variants.
1.422	19-Jan-2020	thorpej	Remove HIPPI support and the esh(4) driver that uses it. There have not been any users of HIPPI for some time, and it is unlikely to be resurrected.
1.421	17-Jan-2020	ryo	support internal PHY temperature sensor
1.420	16-Jan-2020	yamaguchi	Use def{flag,param} for parameters in ixl(4)
1.419	01-Jan-2020	ryo	branches: 1.419.2; add support Aquantia AQC seriese 10G network adapters. this driver is based on the FreeBSD version https://github.com/Aquantia/aqtion-freebsd , but drastically rewritten for NetBSD.
1.418	10-Dec-2019	yamaguchi	Ported driver for Intel Ethernet 700 series reviewed by msaitoh and knakahara
1.417	26-Nov-2019	nisimura	- use mii(4) layer to control KSZ8841 builtin PHY. - handle PAUSE flow control properly according to ifconfig(8) mediaopt selection. - some style knits; use aprint(9) and modernise callout(9).
1.416	20-Nov-2019	hikaru	Add opencrypto driver for Intel QuickAssist.
1.415	22-Sep-2019	jmcneill	Correct mcx comment, drop mii dependency
1.414	22-Sep-2019	mrg	add mcx attachment.
1.413	08-May-2019	isaki	branches: 1.413.2; Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.412	12-Dec-2018	maxv	branches: 1.412.2; Retire the LMC driver, and its associated lmcconfig tool. LMC has been mentioned repeatedly as a non-MP-safe driver that is hard to maintain, and no one is taking care of it. LMC was removed from OpenBSD three years ago, and from FreeBSD a few months ago.
1.411	07-Dec-2018	msaitoh	- defflag PPB_USEINTR - Print "interrupting at "
1.410	24-Nov-2018	bouyer	Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6.
1.409	20-Nov-2018	skrll	Add support for MSI/MSI-X to ahcisata at pci. The options AHCISATA_DISABLE_MSI and AHCISATA_DISABLE_MSIX are available if required.
1.408	14-Nov-2018	skrll	defflag XHCI_DISABLE_MSIX
1.407	22-Sep-2018	maxv	Remove isic(4). It is part of ISDN, which we are now retiring.
1.406	22-Sep-2018	maxv	Unreference iwic (now removed), forgot that.
1.405	22-Sep-2018	maxv	Remove the "ifritz" driver (no man page). It is part of ISDN, which we are retiring.
1.404	22-Sep-2018	maxv	Remove ifpci(4). It is part of ISDN, which we are retiring.
1.403	22-Sep-2018	nakayama	ixg and ixv depend on mii and mii_phy.
1.402	21-Sep-2018	maxv	Remove iavc(4).
1.401	06-Sep-2018	maxv	Retire the 'midway' driver. Discussed on tech-net@ recently and also three years ago, part of removing the network ATM code.
1.400	25-Aug-2018	maxv	branches: 1.400.2; Retire NDIS. It appears that it has never worked, after 13 years it was still marked as "experimental", and nowadays it may be one more obstacle to MPification of the network stack. Discussed on tech-net@.
1.399	08-Aug-2018	maya	Remove NetOctave NSP2000 support This code has bitrotted - it hasn't been adjusted from the old module framework. Discussed on tech-kern: https://mail-index.netbsd.org/tech-kern/2018/07/10/msg023638.html
1.398	15-Jul-2018	maxv	Retire ipkdb entirely. The option was removed from the config files yesterday. ok kamil christos
1.397	06-Jun-2018	jakllsch	branches: 1.397.2; shuffle mechanics of files.virtio config(5) include location, without actually moving the contents thereof from dev/pci yet
1.396	19-May-2018	jdolecek	add config glue for ena(4)
1.395	11-May-2018	maya	add bwfm pci support, from openbsd Tested on BCM43602.
1.394	01-Mar-2018	pgoyette	branches: 1.394.2; Move the imc and imcsmb stuff out of general files.pci and into the architecture-specific files.x86 Should unbreak the sgimips build.
1.393	25-Feb-2018	pgoyette	Import imcsmb driver from FreeBSD. This driver allows access to the SMBus controllers which are part of the integrated memory controllers on certain modern Intel CPUs. These SMBus are attached only to the memory DIMMs, so we provide only a minimum amount of functionality. Deliberately not included in GENERIC, as on some motherboards there can be conflicting access between the driver and the motherboard. The motherboards generally will provide a mechanism to synchronize access, but the methods are likely proprietary; the driver provides a place for inserting user-provided synchronization.
1.392	03-Dec-2017	jdolecek	port ips(4) driver from OpenBSD; needs a lot more work, right now just compilable
1.391	05-Sep-2017	skrll	Add an XHCI_DISABLE_MSI option
1.390	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.389	02-Aug-2017	cherry	Reorg the virtio(4) config(9) code to be explicitly pci specific. We'll later use this for a non-pci virtio(4) usecase. ok martin
1.388	13-Apr-2017	knakahara	branches: 1.388.4; wm(4) can disable msi/msix by build option and ddb command. suggested by nonaka@n.o. reviewed by msaitoh@n.o and nonaka@n.o.
1.387	22-Mar-2017	knakahara	make kernel config flag WM_EVENT_COUNTERS. suggested by msaitoh@n.o.
1.386	03-Mar-2017	knakahara	make kernel config option WM_RX_PROCESS_LIMIT_DEFAULT and WM_RX_INTR_PROCESS_LIMIT_DEFAULT e.g. if - WM_RX_PROCESS_LIMIT_DEFAULT is set 0 - WM_RX_INTR_PROCESS_LIMIT_DEFAULT is set -1 (means almost infinite) that means wm(4) does not use polling mode.
1.385	02-Dec-2016	msaitoh	branches: 1.385.2; Add ixgbe_osdep.c.
1.384	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.383	28-Nov-2016	msaitoh	FreeBSD r280182 made new file ix_txrx.c and moved ixgbe.c and ixv's common code into it. Before sync with whole of them, just move ixgbe.c and ixv.c's common code into ix_txrx.c from ixgbe.c. In this commit, only ixgbe.c is split into the device dependent part and the common part. ixv.c isn't change to make this commit no functional change. To use ixv.c with ix_txrx.c, it's required to modify the common part's API and functions themselves. This commit is done to make the next change easy to understand.
1.382	21-Nov-2016	macallan	missing bits for pm3fb, should have went with previous...
1.381	01-May-2016	nonaka	branches: 1.381.2; Added nvme(4) for Non-Volatile Memory Host Controller Interface devices. Ported from OpenBSD.
1.380	05-Jan-2016	msaitoh	Add ismt(4).
1.379	29-Oct-2015	christos	Add vioscsi, compile tested only (toxic)
1.378	20-Oct-2015	tnn	add ifnet attributes to vioif(4) so it can be config(8)ed as root device
1.377	27-Aug-2015	nonaka	Added rtwn(4) for Realtek RTL8188CE/RTL8192CE PCIe 802.11b/g/n wireless network devices. Ported from OpenBSD.
1.376	28-Mar-2015	jmcneill	Split hdaudio and PCI attach glue. Even though the driver was written with this separation in mind, all of the code lived in sys/dev/pci/hdaudio. Move common parts to sys/dev/hdaudio and PCI attach glue to sys/dev/pci.
1.375	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.374	10-Mar-2015	msaitoh	Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X.
1.373	07-Feb-2015	pooka	Add a driver for Intel Centrino 7260 and similar wireless cards. Supported devices should more or less match ones supported by the Linux iwlwifi mvm driver. Sponsored by genua mbh for OpenBSD. This is probably the world's first Canadian cross device driver: it was created for OpenBSD by writing and porting a NetBSD driver which was developed in a rump kernel in Linux userspace. Note: I don't have access to the hardware anymore, so this version is not tested and not enabled by default. While I tried to be careful in adding the NetBSD bits back, it's probable that there's a snafu or two. Feel free to send private email in case you have the hardware and there are issues.
1.372	26-Oct-2014	tls	branches: 1.372.2; Add viornd(4), a driver for the VirtIO entropy source available on QEMU, KVM, and Google Compute Engine. From OpenBSD.
1.371	19-Mar-2014	nonaka	branches: 1.371.4; Add a driver for Realtek RTS5209/RTS5229 Card Reader. Ported from OpenBSD.
1.370	18-Mar-2014	riastradh	Merge riastradh-drm2 to HEAD.
1.369	26-Jan-2014	msaitoh	PUCCN improvements: - Fix a bug that the puc cn mechanism doesn't use the UART's frequency in pucdata.c's table. - Add a new option PUC_CNAUTO. If this option is set, consinit() in x86/x86/consinit.c checks puc com device to use it as console. Without this option, the behavior is the same as before. - Add a new config parameter PUC_CNBUS. The old code scans bus #0 only. If PUC_CNBUS is set, the specified number's bus will be scanned. - Rename comcnprobe() to puc_cnprobe() to make it clear. - Rename comcninit() to puc_cninit() to make it clear. - Add code for a device that a device's com register is MMIO (#if0 ed).
1.368	21-Jan-2014	mlelstv	wscons driver for Intel Graphics Media Accelerator. Initial commit that already works for a couple of Notebooks based on G35, G45, Sandy Bridge and Ivy Bridge chips. Despite the word 'Accelerator' there is nothing acclerated yet.
1.367	18-Sep-2013	macallan	a preliminary driver for nvidia geforce graphics chips so far it only supports the GeForce 2MX, tested on macppc only no acceleration yet, just some DAC setup the main advantage over genfb is that this driver knows how to setup the palette registers for the 2nd output
1.366	21-Aug-2013	jakllsch	We already have a config rule for the eventual core of xhci(4); may as well have the PCI attachment rule as well.
1.365	22-Jul-2013	soren	PUCCN is now implicit.
1.364	30-Mar-2013	christos	branches: 1.364.4; 1.364.6; patches for new wifi devices.
1.363	30-Jan-2013	rkujawa	Include rasops that are actually needed by this driver.
1.362	01-Jan-2013	macallan	add a compile option to override the fallback to software drawing of bitmap fonts on R3xx ( options RADEONFB_ALWAYS_ACCEL_PUTCHAR ) since this works just fine on my RV350. Next step: don't map VRAM if we don't need it.
1.361	17-Dec-2012	mbalmer	Adding ibmcd(4), a device driver for the IBM 4810 BSP cash drawer port as found e.g. in SurePOS 300 series point of sale terminals. The driver provides a gpio(4) device with three pins: pin 0 to open drawer, pin 1 to read the status and pin 2 to read whether a cash drawer is connected or not.
1.360	02-Aug-2012	macallan	branches: 1.360.2; add support for anti-aliased fonts
1.359	30-Jul-2012	degroote	Add malo(4)@pci driver for Marvell Libertas wireless adaptor Ported from OpenBSD Known issues : - contrary to OpenBSD one, only support pci at the moment, because I don't have the necessary hardware to test PCMCIA / CARDUS Marvell Card - not connected to pmf(9) (unable to test it)
1.358	18-Jul-2012	rkujawa	Add 3Dfx Voodoo2 driver. Still needs some cleanup and prettyfying, but hey it works.
1.357	19-Apr-2012	bouyer	Add mpii(4), a driver for LSI Logic Fusion-MPT Message Passing Interface II SAS controllers. Ported from OpenBSD.
1.356	02-Mar-2012	nonaka	Added Lemote Yeeloong Notebook support.
1.355	16-Feb-2012	macallan	cache glyphs in video memory when using anti-aliased fonts on r128fb and radeonfb
1.354	24-Jan-2012	rkujawa	branches: 1.354.2; Add missing i2c dependencies to voodoofb.
1.353	21-Jan-2012	rkujawa	Fix double i2cbus dependency for voyager
1.352	21-Jan-2012	rkujawa	Change dependency iic to i2cbus for fb drivers. OK macallan.
1.351	03-Jan-2012	macallan	support anti-aliased fonts in 32bit colour enable with options RADEONFB_DEPTH_32
1.350	30-Oct-2011	hannken	branches: 1.350.2; 1.350.6; Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.349	08-Oct-2011	kiyohara	Fix panic when boot time. Require attribute drm to voodoofb.
1.348	31-Aug-2011	macallan	split up voyagerfb into the framebuffer portion and a pseudo bus that we can attach all the other sub-devices to while there attach an i2c bus
1.347	27-Aug-2011	bouyer	Add a basic driver for the SIS315 Pro frame buffer found on some loogson2-based systems, from OpenBSD. At this time it can only map the framebuffer and attaches wscons, using whatever resolution was configured by the firmware. It doens't support any accelaration or mode change.
1.346	24-Aug-2011	dyoung	Add to pci_bus_devorder() an argument that tells the number of slots available in the devs array. Change the type of the devs array from char to uint8_t. Treat the return value of pci_bus_devorder() as the number of slots that it filled. Don't use the __PCI_BUS_DEVORDER #definition to configure the kernel but let the linker do it. Make pci_bus_devorder() available on all architectures by adding a default implementation that will DTRT on all architectures but hpcmips, the only architecture to #define __PCI_BUS_DEVORDER. On hpcmips, adapt the implementation to the new calling convention. XXX I can compile an hpcmips GENERIC kernel, but I don't have a XXX hpcmips box to test it on.
1.345	12-Aug-2011	dyoung	Attach ixg(4) to the kernel build (it's not in any kernel configurations, yet).
1.344	11-Aug-2011	mbalmer	pwdog(4) is a device driver for QUANCOM Electronic PWDOG1 PCI attached watchdog timers.
1.343	04-Aug-2011	jakllsch	Add coram(4). A driver for CX23885-based TV cards, such as the Hauppauge WinTV HVR-1250 "Coram" board. Currently only supports ATSC 8VSB reception.
1.342	14-Jul-2011	jmcneill	hook in lg3303 demod support and modularize
1.341	11-Jul-2011	jakllsch	Add cxdtv(4), a dtv(4) driver for Conexant CX23880-series DTV interface chips. Initially supports digital reception on ATI HDTV Wonder card.
1.340	04-Apr-2011	bouyer	Add rdcide(4), a driver for the IDE controller found in RDC's vortex86/PMX-1000 system-on-chip.
1.339	23-Feb-2011	jmcneill	add network driver for Atheros AR813x/AR815x ethernet controllers, based on a patch from fire crow on tech-net with additional bpf & detach fixes, module support, and a match for 8152 v2.0 devices. alc0 at pci3 dev 0 function 0: Attansic/Atheros L1C/L2C Ethernet alc0: ioapic0 pin 17 alc0: Ethernet address 00:26:6c:9e:d4:c1 ukphy0 at alc0 phy 0: L2 10/100 PHY (OUI 0x00c82e, model 0x0002), rev. 5 ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT-FDX, auto
1.338	09-Feb-2011	macallan	use the bus independent stuff from ic/ct65550* and leave only the PCI specific bits in pci/chipsfb.c
1.337	26-Jan-2011	bouyer	Add a driver for the RDC Semiconductor RDC R6040 10/100 Ethernet controller, as found in the vortex86 SoCs (http://www.vortex86dx.com). Ported from freebsd. Not added to amd64's GENERIC because this CPU is 32bit only. thanks to DM&P Electronics, Inc for providing documentation and sample devices for this work.
1.336	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.335	11-Dec-2010	matt	branches: 1.335.2; 1.335.4; arcmsr needs sysmon_envsys
1.334	13-Nov-2010	jnemeth	PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) network driver The et(4) driver supports PCI Express Ethernet adapters based on the Agere/LSI ET1310/ET1301 integrated MAC/PHY. The et(4) driver was written by Sepherosa Ziehau for DragonFlyBSD, ported to OpenBSD by Jonathan Gray and subsequently ported to NetBSD by Kaspar Brand.
1.333	10-Nov-2010	skrll	Add nside(4) - a driver for the National Semiconductor PC87415 IDE controller as found in many HP PA-RISC machines. From OpenBSD. Reviewed by Manuel Bouyer.
1.332	09-Nov-2010	skrll	Update sti(4) from OpenBSD bringing across support for sti @ pci. Thanks to Adam Hoka for testing.
1.331	06-Nov-2010	jakllsch	Add Intel SCH IDE controller driver. From OpenBSD via Tomokazu HARADA in PR#42310.
1.330	03-Nov-2010	macallan	pull in rasops8 for radeonfb, while there defflag a bunch of options and make them all use the same prefix
1.329	24-May-2010	pgoyette	Extract the vendor/product tables and related access routines into a separate kernel module. Update pci bus attach routine to load the module (if available) when we're about to start scanning the bus, and unload the module after the scan is finished. On architectures which support loading of modules by the boot loader, the 'pciverbose' module can be loaded and executed without needing to rebuild the kernel. On all architectures, using 'options PCIVERBOSE' in the kernel configuration file will create a 'builtin' module which is functionally equivalent to previous behavior. XXX Although not nearly as large as the vendor and product tables, XXX the PCI class and subclass tables might also be offloaded into XXX the module at a future time. XXX Cardbus (and possibly other) drivers should also be modified to XXX load the module before scanning/attaching devices.
1.328	01-Apr-2010	jakllsch	Add support for Domex 536 PCI SCSI controller to nca(4). This truly remarkable chip is found on the Domex DMX-3191D SCSI adapter.
1.327	21-Mar-2010	jklos	Reorganizing all Chelsio 10 gig files into separate directory.
1.326	25-Feb-2010	macallan	branches: 1.326.2; a very preliminary driver for Wildcat 5110 / Sun XVR-500 graphics boards Since we don't have any documentationthis driver is unaccelerated and thanks to the insane state we get the hardware in it's also slow and scrolling looks weird. Some hardware info came from OpenBSD's ifb driver.
1.325	22-Feb-2010	ahoka	Restore splashscreen support with genfb. genfb patch from Pierre Pronchery, other small changes to make it compile by me. Closes PR kern/42605. XXX doesnt work in 8bit, probably a cmap issue
1.324	07-Jan-2010	jdc	branches: 1.324.2; Add cas (Sun Cassini/Cassini+ (GigaSwift) Ethernet).
1.323	28-Oct-2009	macallan	A driver for the Permedia 2 graphics processor, so far it's been tested only on sparc64 with a Sun PGX32 / TechSource Raptor GFX 8P So far it accelerates block copies, rectangle fills and the cursor, it supports the usual wsdisplay / vcons stuff. TODO: - character drawing by hardware so we don't need to map the framebuffer - DDC2 support - mode setting This driver still relies on the firmware to set up graphics modes etc. so it won't work on x86 ( or non-OpenFirmware for that matter ) unless you use the VESA BIOS to set up graphics and pass the right properties ( width, height etc. ) to the driver. For the same reason it will work only if it's the OpenFirmware console.
1.322	26-Sep-2009	snj	"Intel High Definition Audio" -> "High Definition Audio"
1.321	26-Sep-2009	jakllsch	As, siisata attaches not just at pci, but for a while now, also at cardbus, move siisata core driver config to the propper config file.
1.320	20-Sep-2009	christos	PR/42100: Dave J. Barnes: Support for old toshiba PICCOLO IDE controllers.
1.319	06-Sep-2009	sborrill	hdaudio(4) is a standards-compliant driver for Intel High Definition Audio. It will replace azalia(4) after testing. To use, comment out azalia in your kernel configuration and uncomment the hdaudio and hdafg lines so it reads: # Intel High Definition Audio hdaudio* at pci? dev ? function ? hdafg* at hdaudiobus? You should also: cd /dev sh MAKEDEV audio
1.318	12-Aug-2009	macallan	add voyagerfb, a driver for the Gdium Liberty 1000's onboard graphics chip
1.317	27-Jul-2009	kiyohara	Support Marvell Hercules-I/II SATA Controllers.
1.316	21-Jun-2009	mrg	make external drm the default.
1.315	17-Jun-2009	jakllsch	A few changes for siisata(4): - Support detachment. From KIYOHARA Takashi. - Add PCI detachment functionality (albeit not very interesting when the bus can not yet be rescanned). - Rework interrupt handlers to reduce near-duplicate code. Borrowed from ahcisata(4). - Attempt to make polled I/O work. Untested. - Fix formatting of some messages. - For always-polled commands, disable interrupt at slot level rather than port level. - Instead of busy-waiting indefinitely for completion of some commands move on after 31 seconds. Use cv_timedwait(9) instead of DELAY(9). - Use abstracted SATA FIS code. - Enable use of disks that don't respond with the standard signature.
1.314	21-Apr-2009	nonaka	Added SD/MMC support from OpenBSD. tested on i386, amd64 at current-users ML by pgoyette@. tested on zaurus by myself.
1.313	20-Apr-2009	cegger	Add ale(4), a driver for Attansic/Atheros L1E gigabit ethernet device. Ported by me from OpenBSD, improvements from Kevin Lahey, successfully tested by Kevin Lahey and Stephen Borrill.
1.312	16-Jan-2009	cegger	branches: 1.312.2; age driver needs mii_phy
1.311	16-Jan-2009	cegger	Driver for Attansic L1 gigabit ethernet driver. Written by Pyun YongHyeon for FreeBSD, ported to DragonFlyBSD by Sepherosa Ziehau, ported to OpenBSD by Kevin Lo and then ported to NetBSD by me. XXX Driver needs testing.
1.310	09-Jan-2009	macallan	PCI attachment for the bwi driver From Urban Boquist
1.309	28-Dec-2008	jmcneill	PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.308	30-Oct-2008	darran	branches: 1.308.2; 1.308.8; NetOctave NSP2000 driver, ported from FreeBSD and integrated with opencrypto by CoyotePoint Systems. The FreeBSD driver source was recently made available by NBMK Encryption Technologies. The port includes some currently unused code which implements kernel and user space interfaces for the driver in FreeBSD. These are left in at this time to facilitate the port of these interface to NetBSD if they are of interest.
1.307	29-Oct-2008	jkunz	Ported alipm(4) and admtemp(4) from OpenBSD.
1.306	11-Oct-2008	bouyer	branches: 1.306.2; jme(4), a driver for JMicron Technologies JME250 Gigabit Ethernet and JME260 Fast Ethernet PCI Express controllers. Written with a lot of cut-n-paste from the FreeBSD jme(4) driver. No support for jumbo ethernet frames yet (but should come soon). Thanks to JMicron Technologies for providing me sample boards and documentation for this work.
1.305	21-Jul-2008	joerg	Express explicitly that VGA_POST needs X86EMU.
1.304	23-May-2008	jnemeth	branches: 1.304.2; 1.304.4; Import siisata(4) by Jonathan A. Kollasch. The siisata driver supports the Silicon Image SteelVine family of SATA-II controllers, interfacing the hardware with the ata(4) and atapi(4) sub- systems. The following controllers are supported by the siisata driver: Silicon Image SiI3124 4-port PCI/PCI-X Silicon Image SiI3132 2-port PCI-Express x1 Silicon Image SiI3531 1-port PCI-Express x1 SATA Native Command Queueing is not yet supported. Device hot swapping is not yet supported. Silicon Image's Software RAID is not yet supported by the ataraid(4) driver. Approved by: core (christos), releng (bouyer)
1.303	29-Mar-2008	cube	branches: 1.303.2; 1.303.4; 1.303.6; Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC. The linux driver was used as documentation. Many thanks to all the people who helped or at least supported me while I was fighting that bug that made the EeePC freeze. At the top of that list, Matt J. Fleming (mjf@) and David Gwynne (from the OpenBSD fame).
1.302	09-Feb-2008	ober	branches: 1.302.6; Intel PRO/Wireless LAN 4965AGN Mini-PCI Adapter Driver ported from OpenBSD by deroote@. OK nick@ moof@
1.301	17-Jan-2008	jklos	10 gigabit Chelsio cards now compile and work, although not very well yet.
1.300	25-Dec-2007	joerg	Add initial version of calling VGA POST from vga_resume. This is the equivalent to "vbetool post" using x86emu in the kernel.
1.299	15-Dec-2007	dyoung	Finish making the sip(4) and gsip(4) drivers share compiled code.
1.298	05-Dec-2007	xtraeme	branches: 1.298.4; Rename the Areca RAID driver (known as arc(4) to arcmsr(4) to avoid namespace conflict with NetBSD/arc. Found by tsutsui@.
1.297	04-Dec-2007	xtraeme	Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD. arc0 at pci2 dev 14 function 0: interrupting at ioapic0 pin 18 (irq 5) arc0: Areca ARC-1210 Host Adapter RAID controller arc0: 4 ports, 256MB SDRAM, firmware <V1.43 2007-4-17> scsibus0 at arc0: 16 targets, 8 luns per target [...] scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <Areca, ARC-1210-VOL#00, R001> disk fixed sd0: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors sd1 at scsibus0 target 0 lun 1: <Areca, ARC-1210-VOL#01, R001> disk fixed sd1: 465 GB, 56514 cyl, 36 head, 480 sec, 512 bytes/sect x 976562176 sectors bioctl(4) output with two RAID0 volumes: Volume Status Size Device arc0 0 Online 466G ARC-1210-VOL#00 RAID0 0 Online 234G 0:0.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:1.0 noencl <WDC WD2500YS-01SHB1 20.06C06> arc0 1 Online 466G ARC-1210-VOL#01 RAID0 0 Online 234G 0:2.0 noencl <WDC WD2500YS-01SHB1 20.06C06> 1 Online 234G 0:3.0 noencl <WDC WD2500YS-01SHB1 20.06C06> The driver still needs changes related to locking and talking to the firmware, which sometimes is unable to answer... Raid card donated by Areca Technology Corporation via Trent George. Disks used for testing were contributed by TNF. Thank you very much.
1.296	07-Nov-2007	macallan	branches: 1.296.2; add preliminary driver for ATI Rage 128 graphics controllers So far it supports some acceleration ( copies and rectangle fills ) but no video mode programming at all. Known problems: - tested only on macppc - matches only one r128 chip out of ca. 30 - character drawing is unaccelerated - no attempt is made to restore the console when X exits
1.295	28-Jul-2007	kiyohara	branches: 1.295.4; 1.295.6; 1.295.10; 1.295.12; 1.295.14; Add support for Intel ICH SMBus controller.
1.294	11-Jul-2007	kiyohara	branches: 1.294.2; Add support for NVIDIA nForce 2/3/4 SMBus controller and SMBus driver.
1.293	10-Jul-2007	jklos	More updates to the cxgb stuff. Now it actually starts configuring and talking to the hardware.
1.292	09-Jul-2007	jklos	Boo boo - .h files don't go in here...
1.291	09-Jul-2007	jklos	Added one last file for Chelsio support.
1.290	07-Jul-2007	jklos	Added two files for Chelsio 10 gig cards.
1.289	21-Jun-2007	jklos	Added cxgb entry for Chelsio T3 10 gigabit card.
1.288	15-May-2007	bouyer	add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work.
1.287	12-May-2007	bouyer	Split the ahcisata driver in pci front-end and bus-independant back-end.
1.286	26-Apr-2007	garbled	Remove the PCIIDE_SLIDE_SETIRQA hack. Recent revalations have uncovered the fact that it's useless.
1.285	11-Apr-2007	macallan	fix a typo
1.284	10-Apr-2007	macallan	move genfb's PCI frontend to dev/pci
1.283	24-Mar-2007	jmcneill	Include dev/pci/drm/files.pcidrm
1.282	23-Mar-2007	jmcneill	Allow viadrm to attach to unichromefb
1.281	23-Mar-2007	jmcneill	Add viadrm
1.280	21-Mar-2007	macallan	defflag RADEONFB_DEBUG and MACHFB_DEBUG, pull in drm for both, pull in edid for machfb
1.279	20-Mar-2007	drochner	Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11. Minor modifications by me: -use an mi device major number -(coarsly) divided into pci card specific and less specific parts, moved the latter to dev/drm -renamed autoconf attributes to reflect this Todo: -adapt all card frontends but i915 to drm include file location -review the mtrr change -make the change to agp_i810.c coexist with the fix for buggy VESA BIOSes which is commented out temporarily -RCS IDs etc style stuff -LKM support (rescan support for vga) -test
1.278	20-Mar-2007	garbled	Put a small hack in slide.c disabled by default. PCIIDE_SLIDE_SETIRQA. This sets IDE_IRQA in the IDE Control/Status Register. This is needed on prep machines with Motorola RAVEN host bridges, as they wire up the ide device oddly.
1.277	25-Jan-2007	macallan	branches: 1.277.2; 1.277.6; 1.277.8; 1.277.10; add some flags for chipsfb
1.276	20-Jan-2007	xtraeme	Updated viaenv(4) driver: * Support for the VIA VT8231 Hardware monitor. * Power Management Timer available for timecounters in both VT86C686A and VT8231 (code simplified thanks to dev/ic/acpipmtimer). * Remove viapm(4) code and manpage (which was a link to viaenv.4 anyway). From OpenBSD, tested by some users.
1.275	17-Dec-2006	bouyer	Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.274	17-Dec-2006	bouyer	Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64.
1.273	30-Nov-2006	bouyer	branches: 1.273.2; Add ahcisata(4), a driver for AHCI 1.0 and 1.1 controllers. Tested on the sata ports of a Intel 63xxESB chipset. Does not support NCQ yet.
1.272	31-Oct-2006	thorpej	In the pcn driver: - Remove the PCN_NO_PROM option. Instead, query the am79c970-no-eeprom property, and read the MAC address from the CSRs if that property is TRUE. In the ibmnws port: - Implement device_register(). - In device_register(), set the am79c970-no-eeprom property for the built-in Ethernet.
1.271	27-Oct-2006	bouyer	Add SATA native registers support for VIA SATA controllers, so that drives are probed using the SATA way (from FreeBSD). While here add the VT8237A SATA Controller to the tables, should fix PR kern/34927. Thanks to the C�sar Catri�n Carre�o and paul (at) whooppee.com for tests.
1.270	25-Oct-2006	bouyer	Map the SATA registers for SATA channels, and probe drives using them instead of using the old PATA way. Tested with a PDC20375 (2xSATA + 1xPATA). While there add the PDC20618-621 products (Ultra/133 controllers); untested. Yes, it's strange to support PATA-only devices in a driver called pdcsata, but that's how it is ...
1.269	23-Sep-2006	macallan	a half done wsdisplay driver for Chips & Technologies 65550 graphics chips So far it only uses the blitter for scrolling and rectangle filling, characters are still drawn in software and there's no support for video mode switching. Virtual consoles are supported via vcons. Works fine on a PowerBook 3400c.
1.268	09-Sep-2006	riz	branches: 1.268.2; Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.267	07-Sep-2006	itohy	branches: 1.267.2; Add PIOBM (busmastering transfer using ATA PIO mode) support. The PIOBM is used by only one driver (will be added later, stay tuned) and intruduce an attribute "ata_piobm" so that it will be conditionally compiled in. The "ata_dma" (busmastering transfer using ATA DMA mode) and "ata_udma" (busmastering transfer using ATA Ultra DMA mode) attributes are also added for consistency, but unused for now.
1.266	16-Aug-2006	gdamore	Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.265	13-Aug-2006	jmcneill	Add rasops16 attribute to unichromefb.
1.264	13-Aug-2006	simonb	Add a port of the OpenBSD Intel 3945ABG wpi(4) wireless driver, by Jean-Baptiste Campesato. From PR kern/33778.
1.263	10-Aug-2006	cube	iwi now needs firmload [hi nick!]
1.262	02-Aug-2006	jmcneill	branches: 1.262.2; Forgot to commit this last night -- add the unichromefb(4) driver.
1.261	11-Jul-2006	drochner	use timecounter code in dev/ic/acpipmtimer.*, kill interface attribute in device declaration
1.260	30-Jun-2006	nisimura	Add kse(4) driver for Micrel KSZ8842/8841 Ethernet controller.
1.259	26-Jun-2006	drochner	-make sure it is a PIIX4 before accessing counter stuff (there might be usable registers on 440mx and that ATI thing too, but that needs to be checked/tested) -make sure the I/O access to the power management stuff was enabled by the BIOS before trying to map/access stuff -most PIIX4s have a bug in the timer latch which causes jitter or worse if it is read naively - check the revision -use common code in dev/ic/acpipmtimer.c
1.258	07-Jun-2006	dogcow	branches: 1.258.2; revert my earlier commit - the /ic/ stuff should go in src/sys/conf/files, not here.
1.257	07-Jun-2006	dogcow	ral_pci was missing the /ic/ files.
1.256	06-Jun-2006	rpaulo	Attachment framework for the rt2561.c and rt2661.c drivers. From OpenBSD.
1.255	28-May-2006	martin	Move definition of the 'fb' attribute to global scope and make machfb@pci require it. On most archs this does not change anything, but on sparc{,64} it allows linking of kernels that have machfb as the only framebuffer. Solution suggested by Quentin.
1.254	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.253	07-May-2006	jmcneill	branches: 1.253.2; Add support for the Intel PIIX4 power management controller, from OpenBSD.
1.252	30-Apr-2006	xtraeme	Add missing 'arp' keyword for the nfe driver, reported by Murray Armfield in PR kern/33384.
1.251	26-Apr-2006	rpaulo	Comment out glxt.
1.250	18-Apr-2006	rittera	A number of minor changes for NDIS. (OK'd by phil) 1. Removed the makeoptions line from arch/i386/conf/GENERIC. Now ndis_driver_data.h is simply copied into the kernel build directory instead of editing the config file to specify its location. 2. Uncommented lines in files.i386 and files.pci related to NDIS. NDIS will not be compiled into the kernel unless the two lines are uncommented from GENERIC (I checked using nm), so there is no reason for this to be commented out. 3. Added ndiscvt to usr.sbin/Makefile.
1.249	17-Apr-2006	rpaulo	Pull firmload on ipw.
1.248	12-Apr-2006	macallan	convert to vcons while there, attach an fb device when running on sparc
1.247	11-Apr-2006	macallan	A driver for 3Dfx Voodoo3 graphics boards, may or may not work with Voodoo4, Voodoo5 or Banshee boards. So far it supports: - full acceleration in 8bit graphics mode - video mode switching - virtual consoles via vcons So far it hasn't been tested on anything else than macppc and even there it needs a hack to overload ofb. TODO: - test on i386 - don't hardcode video mode
1.246	31-Mar-2006	rittera	Updated to the latest NDIS code. I added commented out lines in the files arch/i386/conf/GENERIC, arch/i386/conf/files.i386 and dev/pci/files.pci which can simply be uncommented to compile NDIS into the kernel. I'll write some documentation on this soon. Note that NDIS is still somewhat experimental. It is currently tested and functions relatively well on on two cards: 1. Dell (Broadcom) TrueMobile 1400 Dual Band WLAN Mini-PCI 2. Intel EtherExpress Pro/100
1.245	21-Mar-2006	he	Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.244	12-Mar-2006	chs	branches: 1.244.2; add nfe.
1.243	06-Mar-2006	bouyer	branches: 1.243.2; Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.242	19-Feb-2006	tls	branches: 1.242.2; Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.241	06-Dec-2005	christos	branches: 1.241.2; 1.241.4; 1.241.6; New lmc driver from David Boggs
1.240	18-Oct-2005	joerg	Both iwi and ipw need arp.
1.239	28-Sep-2005	kent	split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.238	09-Sep-2005	ragge	Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.237	29-Aug-2005	drochner	remove the interface attribute (defining the "channel" locator) from all the XXXide mutants -- this is handled by the "ata" attribute already
1.236	11-Jul-2005	kiyohara	ieee1394 import from FreeBSD.
1.235	01-Jul-2005	drochner	branches: 1.235.2; add autoconf glue for the ralink wireless drivers, basically from PR kern/30449, but changed so that pci/cardbus and usb devices are all called "ral" to the user, so that code can be shared eventually
1.234	25-Jun-2005	dyoung	Move the definitions in dev/pci/files.ath into dev/pci/files.pci. Stop including dev/pci/files.ath in arch/i386/conf/files.i386, since we get the same definitions by including dev/pci/files.pci, now. Remove dev/pci/files.ath. Add arch/macppc/conf/Makefile.macppc with directives for linking the Atheros HAL for PowerPC. In athhal-powerpc-be-eabi.opt_ah.h, #define AH_REGOPS_FUNC 1, since otherwise the linker complains that the PowerPC HAL cannot link with register-read/write subroutines. Add ath(4) to the GENERIC macppc kernel configuration; comment it out.
1.233	18-Jun-2005	kent	An audio driver for High Definition Audio. The driver is highly experimental at this moment: - limieted support for playback - support for no mixer controls - support for no recording
1.232	27-Feb-2005	perry	branches: 1.232.2; nuke trailing whitespace
1.231	20-Feb-2005	jdolecek	Add device driver support for the VIA Networking Technologies VT6122 gigabit ethernet chip and integrated 10/100/1000 copper PHY. Obtained from: FreeBSD
1.230	12-Feb-2005	rearnsha	Change PCIIDE_I31244_ENABLEDMA to PCIIDE_I31244_DISABLEDMA as suggested by briggs. Put the flag in opt_pciide.h.
1.229	21-Jan-2005	yamt	branches: 1.229.2; - switch to dev/rtl8169.c from pci/if_re.c. - enable re@cardbus. both of re@pci and re@cardbus seem to work for me.
1.228	12-Jan-2005	reinoud	branches: 1.228.2; Initial commit of auixp(4) driver. It's a driver for the audio part of ATI IXP-200 and family motherboard chips. Its functional for both recording and playback. Note that quadraphonic and Dolby 5.1 audio are not tested by me but ought to work fine. I can set the number of channels etc. OK but i can't listen to the result. Any feedback on this performance would be greatly apreciated.
1.227	11-Jan-2005	skrll	Add iwi(4).
1.226	06-Dec-2004	cube	Add config wires for ixpide(4).
1.225	01-Dec-2004	grant	add iteide(4) driver for ITE8212-based IDE controllers. from OpenBSD, ported to NetBSD by me. ok'd by bouyer@, thorpej@.
1.224	24-Nov-2004	bouyer	pdcsata(4), a driver for the Promise SATA150 (aka PDC203xx) serie of controllers. Tested with a PDC20375 (2 SATA, one PATA) controller on sparc64. Added to all kernel config file which had pdcide(4).
1.223	13-Nov-2004	kent	auacer uses aurateconv
1.222	22-Oct-2004	scw	Make artsata depend on sata.
1.221	10-Oct-2004	augustss	Add a driver for the Acer Labs M5455 audio controller. This controller is integrated in some ALi Southbridges. XXX This driver is incomplete and slightly buggy, but it works enough to enable me to listen to music on my Sharp MM20.
1.220	26-Sep-2004	dyoung	Add rtw(4).
1.219	30-Aug-2004	drochner	add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.218	30-Aug-2004	drochner	there is (appearently) no reason for the "bktr" driver to carry a "pcibus" attribute, so remove it
1.217	26-Aug-2004	itohy	Workbit NinjaSCSI-32 PCI/CardBus SCSI driver (njs)
1.216	23-Aug-2004	lukem	Add ipw(4)
1.215	12-Aug-2004	thorpej	Don't assume wdc-specific ATA / ATAPI code should be brought in if atabus or atapibus are configured; use a separate wdc_common attribute to indicate that the shared wdc code is also present.
1.214	09-Jul-2004	bouyer	Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.213	08-Jul-2004	drochner	support the game port on eap1371 This is only usable with some caution because these soundcards only allow to map IO port 0x20x for this, thus bypassing PCI address management. Very likely this will only work on primary PCI buses, and there is some potential for conflicts with ISA devices as well. (XXX cannot be detached because the "joy" driver doesn't support it yet)
1.212	28-May-2004	thorpej	satalink now needs the "sata" attribute.
1.211	28-May-2004	toshii	Add re(4) RealTek 8139C+/8169/8169S/8110S PCI ethernet adapter driver, from FreeBSD. I haven't tested this code very well, but it seems to work fairly well for me.
1.210	12-Mar-2004	ragge	branches: 1.210.2; Add driver for Intel PRO/10GbE ethernet adapter. Now NetBSD has no problem with intercontinental TCP connections of over 4Gbit/s, which is where my test hardware runs out of bus bandwidth. Stuff that is on the TODO list: * HW VLAN support. * Large jumbo buffers (16k). * TCP Segmentation Offload * RAIDC (receive interrupt delay adaptation) * Understand how to use memory above 4GB.
1.209	14-Feb-2004	junyoung	device<tab> attach<tab>
1.208	03-Feb-2004	fredb	Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.207	25-Jan-2004	jdolecek	add puc(4) atppc(4) attachment; probes correctly, but needs further work to support DMA, and some actual transfer testing
1.206	20-Jan-2004	jdolecek	cleanup old lpt(4) attachment, and glue ppbus in so that they can coexist: * lpt device is defined in MI place (dev/ppbus/files.ppbus), dev/ic/lpt.c is included there too; dev/ic/lpt.c is not included if ppbus is configured or if there is alternative platform lpt (like for pc532) * g/c MD lpt definitions and custom puc/upc attachments, glue moved to conf/files and dev/pci/files.pci respectively; remove device lpt definition from dev/isa/files.isa * add ppbus parport attribute, atppc device attachments, adjust plip and lpt glue
1.205	31-Dec-2003	jonathan	Split opencrypto configuration into an attribute, usable by inkernel clients, and a pseudo-device for userspace access. The attribute is named `opencrypto'. The pseudo-device is renamed to "crypto", which has a dependency on "opencrypto". The sys/conf/majors entry and pseudo-device attach entrypoint are updated to match the new pseudo-device name. Fast IPsec (sys/netipsec/files.ipsec) now lists a dependency on the "opencrypto" attribute. Drivers for crypto accelerators (ubsec, hifn775x) also pull in opencrypto, as providers of opencrypto transforms.
1.204	14-Dec-2003	thorpej	Split the Intel i31244 SATA controller ("Artisea") driver out of piixide into its own driver (artsata).
1.203	13-Dec-2003	thorpej	Split out the Silicon Image SATALink support into its own driver, "satalink".
1.202	05-Dec-2003	cube	Remove bitbang code that was taken from FreeBSD to support recent SiS chipsets and use mii_bitbang interface instead. Reflect sip dependency in the config file. Support for SiS96x needs broader testing.
1.201	04-Nov-2003	mycroft	Re-add ServerWorks IDE support, as "rccide". Note: this is untested, as I no longer have a ServerWorks-based motherboard.
1.200	31-Oct-2003	nisimura	Added stpcide(4) driver for STMicroelectronics STPC IDE controllers.
1.199	19-Oct-2003	matt	Add PCN_NO_PROM flag. (read the macaddr from the chip assuming the "BIOS" has properly written it). From John Gordon.
1.198	08-Oct-2003	bouyer	Following Matt Thomas's request, rename ata attribute to ata_hl, and wdc_base to ata. We can now have atabus* at ata? in kernel config files.
1.197	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.196	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.195	30-Sep-2003	thorpej	New generic I2C framework. Supports bit-bang and "intelligent" I2C interface controllers (of varying intelligence levels). Contributed by Wasabi Systems, Inc. Primarily written by Steve Woodford, with some modification by me.
1.194	27-Sep-2003	mrg	add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.193	25-Sep-2003	pooka	autoconf goop for iavc
1.192	08-Sep-2003	thorpej	Remove unused "skge" defns.
1.191	26-Aug-2003	jdolecek	Add driver for SysKonnect SK-9821 and 3COM 3C940 gigabit ethernet boards From FreeBSD via OpenBSD, port to NetBSD done by Stephen Degler Changes relative to submitted version: * yukonreg.h and xmaciireg.h merged into if_skreg.h * bhack[] constified + other small editing changes * use 'Ethernet address' rather than 'address' in attach message XXX completely untested by me, needs further cleanup Driver provided in PR kern/22511 by Stephen Degler
1.190	08-Aug-2003	jonathan	Config hooks for ubsec (Bluesteelnet, Broadcom 582x) driver.
1.189	06-Jul-2003	dyoung	Add driver atw for PCI/Mini-PCI/Cardbus 802.11b NICs based on the ADMtek ADM8211. Read the man page for bugs and other outstanding issues.
1.188	04-Jul-2003	wiz	1.179 had the commit message "Use CFATTACH_DECL()." but actually added a paragraph for the non-existent (or just not in-tree) als driver -- remove that paragraph again to avoid confusion.
1.187	01-Jul-2003	drochner	branches: 1.187.2; VERY quick and VERY dirty port of the OpenBSD "txp" driver for 3Com 3C(R)990 cards. No TCP/UDP nor IPSEC offloading for now. The ifmedia part needs some work, but I only have a 100-FX card; hopefully someone will clean this up.
1.186	19-Apr-2003	fvdl	Add ahd at pci (to be added later).
1.185	16-Apr-2003	thorpej	Device driver for the LSI Logic Fusion-MPT based SCSI and Fibre Channel adapters. Currently supports: * LSI 53c1030 Ultra320 SCSI * LSI FC909, FC909A, FC919, and FC929 Fibre Channel Ported from the FreeBSD "mpt" driver, written by Greg Ansley. Thanks to Frank van der Linden for testing and some bug finding. This work was sponsored by Wasabi Systems, Inc.
1.184	07-Nov-2002	martin	Add a driver for the Myson Technology MTD803 3-in-1 Fast Ethernet Controller, provided by Peter Bex in PR 18675.
1.183	25-Oct-2002	leo	Fritz!PCI v2.0 ISDN driver from FreeBSD.
1.182	24-Oct-2002	junyoung	Add machfb, ATI Mach64/Rage framebuffer display driver. XXX this version is not fully functional yet. More to come shortly.
1.181	11-Oct-2002	thorpej	Fix up IPKDB_NE_PCI option dependencies: * IPKDB_NE_PCISLOT does not need any dependencies (it is merely a required parameter for IPKDB_NE_PCI). * IPKDB_NE_PCI does should not have an option-dependency on IPKDB_NE_PCISLOT. While IPKDB_NE_PCISLOT is a required parameter, that is not how option- dependencies work, and we don't want IPKDB_NE_PCI to imply IPKDB_NE_PCISLOT, as that would cause a bogus value for IPKDB_NE_PCISLOT to be used. Also, the IPKDB_NE_PCI selector should be lower-case; make it so.
1.180	08-Oct-2002	kent	auvia: - Add aurateconv - Update comment
1.179	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.178	24-Sep-2002	pooka	Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.177	03-Sep-2002	augustss	Replace the mr driver with the gtp driver. From OpenBSD.
1.176	06-Aug-2002	augustss	Pull in tea5757.c for the mr driver. Fixes kern/17864.
1.175	23-Jun-2002	fvdl	Add files for bge at pci attachment.
1.174	02-Jun-2002	enami	Collect random number from AMD 768MPX power management controller.
1.173	18-May-2002	matt	branches: 1.173.2; Add lfmiop (LSILogic Fusion-MPT I/O Processor) placeholders for driver to added latter.
1.172	26-Apr-2002	ad	branches: 1.172.2; Add a driver for Adaptec FSA RAID controllers, as often found in Dell servers. Based on the FreeBSD/OpenBSD versions.
1.171	25-Apr-2002	kleink	Add a joystick attachment to the ESS Solo-1 driver.
1.170	22-Apr-2002	ad	Add a driver for ICP-Vortex GDT and Intel Storage RAID controllers. Parts taken from OpenBSD. Test hardware kindly provided by Intel. This still needs management bits, and doesn't support older controllers, but that shouldn't be hard to fix.
1.169	22-Apr-2002	bouyer	Add esiop at pci.
1.168	22-Apr-2002	augustss	Rename the audio "bus" attribute audiobus to avoid confusion with audio device.
1.167	17-Apr-2002	mycroft	Remove TLP_MATCH_* options.
1.166	30-Mar-2002	uwe	Add (unfinished) igsfb driver. While here g/c stale cprofb entry, it's not in the tree and I plan to support CyberPro cards with igsfb driver eventually (where is my netwinder?)
1.165	26-Mar-2002	thorpej	Only build pciconf.o if pci is configured.
1.164	25-Mar-2002	martin	Now that we have all pieces in place (and enough granularity to specify B-channel and D-channel drivers separately) split the Fritz!PCI card driver out of the isic driver. The new device is called "ifpci" and uses the same D-channel driver as the isic devices, but has it's own B-channel driver.
1.163	09-Mar-2002	kent	Move sampling rate conversion functions to aurateconv.c. Introduce "aurateconv" attribute for audio devices. Add aurateconv to uaudio and auich. (due to kern/15845 and kern/15848)
1.162	02-Feb-2002	jdolecek	Reduce the massive code duplication regarding joy(4). Split it into MI and MD parts, and make ISA/ISAPNP/PCI joy(4) attachments MI.
1.161	30-Jan-2002	ad	Bare-bones driver for AMI RAID. Parts taken from FreeBSD. This was tried a good while ago and it had problems under load. Changes were made to address that, but I don't have the ability to test them. So, I'm committing it before it rots.
1.160	09-Jan-2002	augustss	Split out some code so you can have cardbus ehci without pci.
1.159	07-Jan-2002	jmcneill	Re-enable radio at bktr, only if NRADIO > 0 and an FM tuner is present.
1.158	07-Jan-2002	drochner	"attach radio at radio": using one word as both attribute and device doesn't work well, radio.c is pulled in even with no such device in the configuration, and the kernel doesn't link due to missing "radio_cd". So call the attribute "radiodev" to avoid confusion.
1.157	07-Jan-2002	tron	Backout revision 1.155 which caused build failures if bktr(4) is used without radio(4).
1.156	06-Jan-2002	jmcneill	Import driver for the ESS Allegro-1 / Maestro-3 based PCI Audio Accelerators.
1.155	06-Jan-2002	jmcneill	Add radio support to bktr.
1.154	03-Jan-2002	thorpej	Need ibm561.c for TGA.
1.153	01-Jan-2002	augustss	Add support for radio cards. Written by Maxim Tsyplakov and Vladimir Popov for OpenBSD, from where it was imported.
1.152	01-Jan-2002	hpeyerl	Add stanza for PC Weasel/PCI.
1.151	12-Dec-2001	elric	Added support for the PowerStorm 4d20, initial support that is. I plan on cleaning a few things up over the next few days, but this appears to be working for me.
1.150	05-Dec-2001	augustss	Do SIR framing on incoming frames.
1.149	05-Dec-2001	wiz	Use defparam instead of defflags for some bktr(4) related options that need a value. Add BKTR_REVERSE_MUTE to defflags.
1.148	05-Dec-2001	simonb	s/defopt/defparam/ in previous; from lukem.
1.147	05-Dec-2001	simonb	Change PCI_NETBSD_ENABLE_IDE back to a defopt - it's a mask.
1.146	02-Dec-2001	augustss	Add driver for the Toshiba's Oboe IrDA chip. From Jan Sparud.
1.145	28-Nov-2001	lukem	- convert usage of "defopt" to "defflag" where the relevant option does not support a value (e.g., it's to be used as "options FOO" instead of "options FOO=xxx"). options that take a value were converted to defparam recently. - minor whitespace & formatting cleanups
1.144	18-Nov-2001	someya	add Trident 4DWAVE based (4DWAVE DX/NX,SiS 7018,ALi M5451) PCI audio driver
1.143	03-Nov-2001	tsutsui	Add a driver for the Tekram DC-395U/UW/F and DC-315/U SCSI host adapters, which have the Tekram TRM-S1040 ASIC. This driver is written by Rui-Xiang Guo <rxg@ms25.url.com.tw>, and a number of cosmetic changes by me. Tested on i386 by the author, and on macppc and sparc64 by me. XXX On arc, kernel got panic in ltsleep() called from scsipi_execute_xs(), XXX but I'm not sure what is wrong...
1.142	17-Oct-2001	jdolecek	branches: 1.142.2; Add driver for Creative Labs SBLive! EMU10000, possibly also PCI512. Written by Yannick Montulet (thanks!), with only couple minor touches by me.
1.141	13-Oct-2001	ichiro	Add Intersil Prism2.5 Mini-PCI wavelan.
1.140	16-Sep-2001	eeh	Driver for Sun GEM gigabit ethernet, Sun ERI 10/100, and Apple GMAC.
1.139	13-Sep-2001	thorpej	Add an ioctl interface to the PCI bus. Add ioctls to read/write PCI configuration space registers, and to fetch bus info.
1.138	04-Sep-2001	itohy	branches: 1.138.2; Add "opl at cmpci" and "mpu at cmpci" attachment.
1.137	27-Aug-2001	thorpej	New driver for the AMD PCnet-PCI family of Ethernet chips. This driver uses direct DMA to mbufs (like other PCI network drivers, and unlike the old "le at pci" driver), and also supports communication with the MII-connected PHYs on the 10/100 boards.
1.136	22-Aug-2001	thorpej	Add a PCI_NETBSD_ENABLE_IDE option, that goes along with PCI_NETBSD_CONFIGURE, which tells machine dependent code to enable the specified (by a bitmask) PCI IDE channels in the southbridge, in the event that system firmware does not do so.
1.135	01-Aug-2001	ad	Hook in the mly control interface on i386.
1.134	30-Jul-2001	ad	Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.133	25-Jul-2001	thorpej	Driver for the Sundance Tech./Tamarack TC9021 Gigabit Ethernet controller.
1.132	07-Jul-2001	thorpej	branches: 1.132.2; Remove config glue for the old `ncr' driver.
1.131	19-Jun-2001	thorpej	Device driver for the Sundance Tech. ST-201 10/100 Ethernet. This chip is found on the D-Link DFE-550TX.
1.130	03-Jun-2001	tsutsui	Add a driver for the Initio INIC-940/950 PCI SCSI controllers. Based on OpenBSD's iha driver, and modified some structures by me. Tested on arc, i386 and macppc.
1.129	22-May-2001	thorpej	Add mii_bitbang attribute to gsip, pointed out by Andrew Gillham.
1.128	18-May-2001	thorpej	Add a driver for the National Semiconductor DP83820 Gigabit Ethernet chip. This is found on the NetGear GA-622 and Asante FriendlyNet GigaNIX.
1.127	30-Mar-2001	minoura	Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.126	25-Mar-2001	jmc	Actually fix fwlynx correctly. Got the right filename this time.
1.125	13-Mar-2001	jmc	Fix typo on fwlynx_pci driver. Brett McCoy reported this in.
1.124	08-Mar-2001	thorpej	advlib.c and advmcode.c should be built for all adv, not just adv_pci. From takashi.yamamoto@bigfoot.com, kern/12341.
1.123	24-Feb-2001	lukem	branches: 1.123.2; whitespace police
1.122	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.121	09-Feb-2001	briggs	Add an option (defopt) PCI_NETBSD_CONFIGURE that provides PCI bus configuration (assignment of bus numbers, BARs, timer values, interrupt lines, etc.). The interface must be called from m.d. code prior to probing the bus. It is meant to be called once for each primary (bus == 0) PCI bus in the system. It will configure any busses behind PCI-PCI bridges. Section 9 man page for pci_configure_bus() will come soon. In the meantime, sample usage is in arch/sandpoint/sandpoint/mainbus.c. [ Reviewed by thorpej ]
1.120	07-Feb-2001	tacha	Split out common code to cs428x.c and cosmetic change to introduce higher symmetry between cs4280.c and cs4281.c. Also fix the problem rebooting from Windows. Relevant patch is contributed from Shingo WATANABE <nabe@nabechan.org>.
1.119	04-Feb-2001	ad	Add a driver for the Mylex DAC960 family (including DEC SWXCR).
1.118	22-Jan-2001	ad	opt_twe.h isn't needed any more.
1.117	22-Jan-2001	augustss	Add clct driver for Cirrus Logic CrystalClear PCI Audio CS4281. Driver written by Tatoku Ogaito.
1.116	08-Jan-2001	thorpej	Only build "esm" if it's included in the kernel config file.
1.115	08-Jan-2001	rh	Add esm audio driver.
1.114	05-Jan-2001	martin	Enable the imported ISDN4BSD based ISDN subsystem. This is the kernel part (userland to follow soon) of the latest (and very probably last) release (version 0.96) of ISDN4BSD. ISDN4BSD has a homepage at http://www.freebsd-support.de/i4b/. It gives the user various ways to use the isdn connection: raw data (via the i4brbch "raw b-channel" device), ppp (via the isp "isdn PPP" device), voice/answering machine (the i4btel "telephone" device) and ip over isdn (the ipr device, "IP over raw ISDN"). Supported are a bunch of common and older cards, more to be added soon after some cleanup. Currently only the european E-DSS1 variant of the ISDN D channel protocol is supported.
1.113	24-Dec-2000	augustss	Add some placeholders for the EHCI (USB 2) driver. Don't get your hopes up. I've not even finished reading the (100+) page spec, and I have no hardware.
1.112	14-Dec-2000	onoe	Add PCI version of Aironet, not tested at all. ISA/ISApnp version won't come since I don't know about isa...
1.111	28-Nov-2000	thorpej	Device driver for the built-in audio on Intel ICH, ICH0, ICH2, and i440MX chipsets. Based on a driver by Michael Shalayeff, modified somewhat by me to use bus_dma properly, and fix some audio encoding emulation bugs. Thanks to Manuel Bouyer for testing and feedback.
1.110	26-Nov-2000	ad	lsu -> ld, by popular request.
1.109	14-Nov-2000	matt	Add fwlynx PCI attachment
1.108	08-Nov-2000	ad	Add glue for iop.
1.107	08-Nov-2000	ad	Move lsu stuff to more logical locations, and push queueing into the lsu driver.
1.106	05-Nov-2000	thorpej	Add NeoMagic 256 audio.
1.105	05-Nov-2000	thorpej	viaenv carries the sysmon_envsys attribute.
1.104	20-Oct-2000	ad	- tsleep()/wakeup_one() will hurt when under load, so use a queue instead. - Don't copy to or from the aligned buffer unless there's a need to.
1.103	19-Oct-2000	ad	Hook in twe.
1.102	12-Oct-2000	itojun	hifn7751 crypto card driver. from openbsd. does nothing useful at this moment - initailize the chip, that's all. TODO: crypto logic framework in kernel (see openbsd sys/crypto) TODO: ipsec frontend (need major rework in ipsec tree - should start with busywait) TODO: character device frontend
1.101	02-Aug-2000	bouyer	PCIIDE_CMD0646U_UDMA->PCIIDE_CMD0646U_ENABLEUDMA for consistency with PCIIDE_AMD756_ENABLEDMA defopt PCIIDE_CMD0646U_ENABLEUDMA, PCIIDE_AMD756_ENABLEDMA, PCIIDE_CMD064x_DISABLE Fix a typo pointed out by John Hawkinson
1.100	25-Jul-2000	jeffs	Add code to allow the PCI com serial ports to be used as the system console. This is not enabled by default, and is turned on with options PUCCN. Done by castor@netbsd.org.
1.99	25-Jun-2000	gmcgarry	cmpci requires mulaw dependency.
1.98	25-Jun-2000	sommerfeld	Defopt TLP_MATCH_*
1.97	24-Jun-2000	thorpej	Add a clearing-house pseudo-device for system monitoring devices such as the LM78 and VT82C686A (and eventually ACPI). Multiple sensor devices can be hooked registered with `sysmon', and eventually sysmon will also handle hardware (and software) watchdog timers. Convert the `lm' and `viaenv' drivers to the new interface.
1.96	06-Jun-2000	thorpej	branches: 1.96.2; Common routines for read/writing Cypress 82c693 control registers. Needed by `pciide' and the Alpha `sio' (PCI-ISA bridge) driver.
1.95	30-May-2000	matt	Add the PCI attachment for IEEE 1394 OHCI controllers. Files to follow later.
1.94	17-May-2000	thorpej	branches: 1.94.2; Add a device driver for the Cyclades-Z series of intelligent multi-port serial boards. Written by Jason R. Thorpe and Bill Studenmund.
1.93	15-May-2000	bouyer	Update for siop split.
1.92	10-May-2000	haya	Move if_rl_pci.c:1.7 to if_rtk_pci.c.
1.91	08-May-2000	joda	VIA VT82C686A hardware monitor
1.90	07-May-2000	wiz	some glue for bktr-driver
1.89	30-Apr-2000	augustss	Add driver for C-Media CMI8x38 Audio Chip. From Takuya SHIOZAKI <AoiMoe@imou.to>
1.88	21-Apr-2000	bouyer	Snapshot of work in progress: new driver for the NCR 53c8xx SCSI controller (the name 'siop' is still being discussed, may change). Only basic disconnect/reselect for now, no sync/wide negotiation. Tested with 810A, 875 and 895 on i386 only. The bus-independant part should also be able to handle the 53c720 and 53c770. A new driver with enhanced script should appear for the 825/875/895 'soon'.
1.87	20-Apr-2000	nathanw	Replace rcons console code with rasops console code. Console text should now be supported on all TGA cards; 8- and 32-bit TGA and 8-bit TGA2 have been tested. Implement accelerated character drawing, scrolling, and clearing. Stop clearing "odd" in VHCR; it's unnecessary and hurts performance. Use bus_space_vaddr() instead of a local hack.
1.86	19-Apr-2000	haya	Changes the name of RealTek driver. The new name is `rtk'. This used to be called `rl' and it conflict with RL vax disks, canonical and historical unix driver name. This changes are minimal: it only changes the name of RealTek driver. The filename of source code and a lot of the letter `rl' in source files should be changed shortly.
1.85	10-Apr-2000	haya	Incorporate the changes of RL81x9 driver provided by M. Kanaoka <kanaoka@ann.hi-ho.ne.jp>. He separated the driver into IC specific portion and bus attachment portion and added cardbus attachent.
1.84	05-Apr-2000	mrg	add "hme at pci" support.
1.83	02-Apr-2000	nathanw	Move bt463 code to dev/ic and adapt to the new ramdac interface. Set default window type to 24-bit truecolor. Correctly set pseudocolor mode.
1.82	31-Mar-2000	tsarna	Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.81	22-Mar-2000	ws	Make IPKDB working again. Add support for i386 debugging and pci-based ne2000 boards.
1.80	22-Mar-2000	cgd	add PCI_CONFIG_DUMP as a defopted option
1.79	22-Mar-2000	cgd	opt_pciverbose.h -> opt_pci.h (it's gonna be used by more things, soon)
1.78	16-Mar-2000	ad	Driver for Compaq array controllers and disks (cac(4)/ca(4)).
1.77	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.76	25-Feb-2000	drochner	add (minimal, PIO only) support for the popular "Tundra Universe" PCI-VME interface chip and a frontend for the "Bit3 Mod. 2706" PCI-VME adapter
1.75	26-Jan-2000	thorpej	Fix an oversight in the AIC-7xxx SEEPROM split.
1.74	04-Jan-2000	chopps	Add support for setting the clock and adapt the code to the new more generic ic/hd56470.c core which now supports the isa version of the same card.
1.73	13-Dec-1999	augustss	Add a device driver for the Cirrus Logic CS4280 sound chip. The code has been contributed by Tatoku Ogaito.
1.72	19-Nov-1999	thorpej	Make the ThunderLAN and VIA Rhine drivers use the common MII bit-bang module.
1.71	02-Nov-1999	augustss	Add the midibus attribute to the fms driver.
1.70	02-Nov-1999	thorpej	Get fms MPU and OPL module inclusion right.
1.69	01-Nov-1999	augustss	Add FM801 audio device driver. From Witold J. Wnuk ww181302@zodiac.mimuw.edu.pl Closes PR kern/8729.
1.68	30-Oct-1999	augustss	Use the right attribute for attaching USB controllers. From MAEKAWA Masahide <bishop@rr.iij4u.or.jp>
1.67	30-Oct-1999	augustss	Add midibus attribute for eap.
1.66	27-Oct-1999	augustss	Add support for ES1371. From OpenBSD and Ezra Story <ezy@panix.com>.
1.65	15-Oct-1999	haya	branches: 1.65.2; 1.65.4; This is the first check-in of CardBus driver. CardBus driver contains CardBus bus stub, YENTA PCI-CardBus bridge (cbb), 3Com 3C575TX driver (ex) and Intel fxp driver. TODO: o Conform to the KNF more strictly. o Be unified with pcmcia code as much as possible. o Add more drivers for CardBus card, such as APA-1480 or USB card. The affected files are listed below. sys/arch/i386/conf/files.i386 sys/arch/macppc/conf/files.macppc sys/conf/files sys/dev/ic/elinkxl.c sys/dev/ic/elinkxlvar.h sys/dev/ic/i82365.c sys/dev/ic/i82365var.h sys/dev/isa/i82365_isasubr.c sys/dev/pci/files.pci sys/dev/pcmcia/pcmcia.c sys/dev/pcmcia/pcmciachip.h The added files are listed below. sys/arch/i386/conf/CARDBUS sys/arch/i386/include/rbus_machdep.h sys/arch/i386/i386/rbus_machdep.c sys/arch/macppc/include/rbus_machdep.h sys/arch/macppc/macppc/rbus_machdep.c sys/dev/cardbus/if_ex_cardbus.c sys/dev/cardbus/Makefile.cardbusdevs sys/dev/cardbus/cardbus.c sys/dev/cardbus/cardbus_map.c sys/dev/cardbus/cardbusdevs sys/dev/cardbus/cardbusdevs.h sys/dev/cardbus/cardbusdevs_data.h sys/dev/cardbus/cardbusvar.h sys/dev/cardbus/cardslot.c sys/dev/cardbus/cardslotvar.h sys/dev/cardbus/devlist2h.awk sys/dev/cardbus/files.cardbus sys/dev/cardbus/if_fxp_cardbus.c sys/dev/cardbus/pccardcis.h sys/dev/cardbus/rbus.c sys/dev/cardbus/rbus.h sys/dev/pci/pccbb.c sys/dev/pci/pccbbreg.h sys/dev/pci/pccbbvar.h
1.64	13-Oct-1999	thorpej	Add declaration for the Adaptec AIC-6915 64-bit 10/100 Ethernet interface. Note: this is a placeholder right now; the driver should be ready in the next few days.
1.63	27-Sep-1999	ad	branches: 1.63.2; Add glue for 'dpt' driver.
1.62	01-Sep-1999	drochner	add "ti" - Alteon Gigabit Ethernet
1.61	01-Sep-1999	thorpej	This is the long-awaited "new Tulip driver", a complete, from-scratch rewrite of the driver for the DECchip 21x4x Ethernet chips, and a variety of clones. Currently, the driver supports the Winbond 89C840F (this works pretty well), and the Lite-On PNIC (e.g. NetGear PCI boards), however Lite-On support may be broken [I may simply have a busted test board]. Eventually, support for the Macronix and ASIX chips will filter into this driver, and then, slowly, support for the genuine DEC chips, and maybe even the DE-425 EISA model.
1.60	02-Aug-1999	augustss	Move the mpu device declaration to conf/files. Let the mpu device attach at the sb device, and then midi at the mpu. Update the mpu at eso attachment.
1.59	12-Jul-1999	kleink	Add a driver for the ESS Technology Solo-1 PCI AudioDrive line of chips.
1.58	30-Jun-1999	drochner	add driver for the Bit3 PCI-VME adapter, without the DMA parts for now
1.57	27-Jun-1999	drochner	add driver for realtek 8129/8139, from freebsd, with mii stuff kicked out
1.56	20-Jun-1999	thorpej	Intel i82557 driver is now split into bus and chip bits.
1.55	01-Jun-1999	thorpej	Glue in the SiS900 Ethernet controller driver.
1.54	25-Mar-1999	explorer	branches: 1.54.2; 1.54.4; 1.54.6; Start commit of Lan Media Corp T1/HSSI/DS3 driver
1.53	09-Feb-1999	sakamoto	add "mii" to vr driver.
1.52	21-Jan-1999	sakamoto	Add VIA Rhine/Rhine II Fast Ethernet files.
1.51	06-Jan-1999	thorpej	Device declaration for the PCscsi-PCI SCSI chip driver.
1.50	10-Dec-1998	augustss	Add a driver for the S3 SonicVibes chip. From OpenBSD and Constantine Paul Sapuntzakis (csapuntz@cvs.openbsd.org) with some changes by me. XXX The driver still needs work.
1.49	04-Nov-1998	fvdl	Add the 'ex' driver, a DMA driver for 3Com 90x and 90xB cards. Rename constants from EP_ to ELINK_ since they're now used in the ex driver as well.
1.48	31-Oct-1998	thorpej	Use the rtl80x9 common code for media selection.
1.47	12-Oct-1998	bouyer	Merge bouyer-ide
1.46	26-Sep-1998	dante	Add AdvanSys ULTRA WIDE SCSI controllers
1.45	26-Aug-1998	dante	Add AdvanSys support
1.44	11-Aug-1998	thorpej	Adapt "tl" driver to the new MII code, and add the "mii" attribute to the "fxp" driver.
1.43	26-Jul-1998	explorer	add ntwo* at pci? function ?, a T1 network driver.
1.42	21-Jul-1998	drochner	adapt to LANCE driver split; let PCNET/PCI use the 32-bit mode
1.41	12-Jul-1998	augustss	Add USB support. Supported so far: * UHCI and OHCI host controllers on PCI * Hubs * HID devices withe special drivers for mouse and keyboard * Printers
1.40	26-Jun-1998	cgd	add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.39	02-Jun-1998	thorpej	branches: 1.39.2; Device driver for the SMC 83c170 Ethernet PCI Integrated Controller (EPIC/100) used in the SMC EtherPower II. Media control isn't yet supported, due to some MII infrastructure problems which I hope to address soon. This isn't a huge deal, since the PHY defaults to auto-negotiate mode. Also, the device just programs the multicast hash table to accept all multicast, to avoid a hardware bug that causes the multicast address filter to lose in 10Mb/s mode. This bug will be fixed in a more sane way once the media control issues are dealt with.
1.38	31-May-1998	cgd	dd a PCI 'quirks' mechanism, meant to be used as the mechanism of last resort when trying to communicate information about bogus behaviour of PCI devices to the MI autoconfiguration code. In general, bogus behaviour should be handled by drivers, but there are some types of bogons which can't be addressed that way. The only quirk currently defined is one which indicates that the device is multi-function even though the device's header says otherwise. (Mmm, Intel 82371FB PCI-to-ISA Bridge (PIIX); you'd think that at least Intel would have gotten it right...)
1.37	14-May-1998	kml	Driver for Essential Communications' RoadRunner HIPPI (800 Mb/sec network) card. With some modification, this could probably also work for their Gigabit Ethernet card based on the same chipset...
1.36	01-May-1998	augustss	Add a driver for the Ensoniq AudioPCI sound card. The driver still needs some testing, but it seems to produce sound. The driver was written by me, but since I don't have the hardware the debugging and testing was done by Andreas Gustafsson <gson@araneus.fi>, Chuck Cranor <chuck@maria.wustl.edu>, and Phil Nelson <phil@cs.wwu.edu>. Thanks.
1.35	15-Apr-1998	drochner	add mi PCI VGA and TGA
1.34	04-Mar-1998	cgd	add (commented out) device/file declarations for the MI PCI IDE controller driver. These are commented out here until the wdc declaration mess is resolved, and until then need to go into MD files files in places where they play nice with the wdc declaration.
1.33	16-Feb-1998	thorpej	Remove "class" declarations, and add "devclass" declarations where appropriate. Fix several inconsistencies between device class and attributes. Mostly from Chris Demetriou.
1.32	12-Jan-1998	thorpej	Update for config changes.
1.31	21-Oct-1997	bouyer	Correct 'ThunderLAN' spelling
1.30	17-Oct-1997	bouyer	branches: 1.30.2; Driver for Texas Instruments' thunderland network controller (present in some Compaq products).
1.29	14-Oct-1997	thorpej	Pulldown from marc-pcmcia branch: PCI front-end for the "ne" driver.
1.28	13-Sep-1997	enami	Declare PCIVERBOSE by defopt in files.pci. Include opt_pciverbose.h in pci_subr.c.
1.27	12-Sep-1997	mycroft	Fix bogus comment.
1.26	30-Aug-1997	mycroft	Pull in new file.
1.25	20-Jul-1997	pk	branches: 1.25.2; config.new => config
1.24	05-Jun-1997	thorpej	Add device declaration for Intel EtherExpress PRO 10/100B driver.
1.23	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.22	12-Mar-1997	cgd	specs for ISP 10x0 (isp) driver PCI attachment. From Matt Jacob
1.21	12-Nov-1996	thorpej	branches: 1.21.4; Centralize the declaration of the "en" driver (Efficient Networks, Inc. 155Mb/sec ATM interface).
1.20	24-Sep-1996	christos	New cyclades driver from Timo Rossi.
1.19	01-Sep-1996	mycroft	Update.
1.18	14-Jul-1996	cgd	alphabetize by driver name
1.17	22-Jun-1996	chuck	pci specific code for the eni155p card
1.16	16-May-1996	mycroft	New version, with changes from Justin Gibbs and Noriyuki Soda.
1.15	07-May-1996	thorpej	Make the MI LANCE driver standalone, and use cfattach to resolve naming conflicts between bus attachments on ports that can have multiple instances of the LANCE. Add a real PCI front-end for PCnet-PCI Ethernet cards. Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.14	25-Apr-1996	thorpej	Break up the ep driver into isa, eisa, and pci front-ends and a bus-independent core driver. Tested on all three bus types, including an isa 3c509 masquerading as an eisa device (use ep* at eisa? slot ? in your kernel config file to catch this one). XXX Driver still needs to be converted to <machine/bus.h>
1.13	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.12	04-Mar-1996	cgd	the i386 port no longer attaches isa, eisa, and pci at root.
1.11	28-Feb-1996	cgd	make PCI bus match/attach and sub-device attachment machine-independent, add definition of 'ppb' (PCI-PCI bridge) device.
1.10	27-Feb-1996	cgd	alphabetize device list
1.9	13-Jan-1996	thorpej	Add the "ahe" driver, autoconfiguration support for the aic7xxx-based Adaptec 2[78]4x SCSI controllers, from Michael Graff <explorer@flame.org>. Fixes PR #1594 from Noriyuki Soda <soda@sra.co.jp> in a different way.
1.8	10-Oct-1995	mycroft	Add specs for ahc driver.
1.7	19-Aug-1995	cgd	add 'fpa' for Matt Thomas's driver for DEC PCI FDDI boards. alphabetize.
1.6	27-Jul-1995	mycroft	Add needs-flags for pci_subr.c.
1.5	18-Jun-1995	cgd	make a comment reflect the way things should be done
1.4	17-Jun-1995	cgd	more appropriate name, as this only has PCI support subroutines.
1.3	05-Jun-1995	cgd	enable the de driver
1.2	17-Apr-1995	cgd	try to actually (gasp!) describe the drivers.
1.1	17-Apr-1995	cgd	config.new files description for PCI mi files.
1.21.4.1	09-Mar-1997	is	netinet/if_ether.h -> netinet/if_inarp.h
1.25.2.4	14-Oct-1997	thorpej	Split off NE2000 support from the ISA "ed" driver, and create a new "ne" driver that deals with NE2000s and compatibles on ISA, PCI, and PCMCIA busses. This driver uses the generic DP8390 core chipset back-end.
1.25.2.3	22-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.25.2.2	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.25.2.1	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.30.2.1	21-Oct-1997	mrg	pull up from trunk: Correct ThunderLAN spelling
1.39.2.6	02-Oct-1998	bouyer	Sync with HEAD.
1.39.2.5	11-Sep-1998	bouyer	Sync with HEAD.
1.39.2.4	13-Aug-1998	bouyer	- sync with -current - full support of CMD PCI0640 chips. This needs a special probe routine, as this chip doesn't honnor the "PCI_COMMAND_IO_ENABLE" bit. sigh. - do not map the native pci interrupts if we don't have a native channel. some chips (e.g. PCI0640) have a valid interrupt value which gets established, which may prevent the compat channel or some other ISA device from working properly later. - b* -> mem*
1.39.2.3	27-Jul-1998	bouyer	Sync with -current
1.39.2.2	02-Jul-1998	bouyer	Sync with HEAD.
1.39.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.54.6.2	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.54.6.1	28-Jun-1999	itojun	KAME/NetBSD 1.4 SNAP kit, dated 19990628. NOTE: this branch (kame) is used just for refernce. this may not compile due to multiple reasons.
1.54.4.3	02-Aug-1999	thorpej	Update from trunk.
1.54.4.2	01-Jul-1999	thorpej	Sync w/ -current.
1.54.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.54.2.4	14-Aug-2000	he	Pull up revision 1.101 (via patch, requested by bouyer): Change PCIIDE_CMD0646U_UDMA to PCIIDE_CMD0646U_ENABLEUDMA for consistency with PCIIDE_AMD756_ENABLEDMA.
1.54.2.3	17-Jan-2000	he	Pull up revision 1.63 (requested by ad): Add driver for DPT SmartCache and SmartRAID III or IV SCSI adapters.
1.54.2.2	20-Oct-1999	he	Pull up revision 1.57 (requested by bouyer): Back-port of the Realtek ethernet driver.
1.54.2.1	05-Oct-1999	he	Pull up revision 1.62 (via patch, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.63.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.65.4.1	15-Nov-1999	fvdl	Sync with -current
1.65.2.9	21-Apr-2001	bouyer	Sync with HEAD
1.65.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.65.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.65.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.65.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.65.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.65.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.65.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.65.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.94.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.96.2.12	24-Jan-2002	he	Pull up revision 1.156 (requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.96.2.11	19-Jan-2002	he	Pull up revision 1.142 (requested by he): Add driver for Creative Labs SBLive! EMU10000 and (probably) PCI512. Fixes PR#15260.
1.96.2.10	25-Oct-2001	he	Apply patch (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.96.2.9	06-May-2001	he	Pull up revisions 1.117,1.120 (requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.96.2.8	06-May-2001	he	Pull up revision 1.106 (requested by sommerfeld): Add a driver for the NeoMagic 256 AC'97 chip.
1.96.2.7	03-May-2001	he	Pull up revisions 1.115-1.116 (requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.96.2.6	01-May-2001	he	Pull up revision 1.127 (requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.96.2.5	03-Aug-2000	bouyer	Pull up (approved by thorpej): sys/dev/pci/pciide.c 1.82 -> 1.84 sys/dev/pci/pciide_cmd_reg.h 1.8 -> 1.9 sys/dev/pci/files.pci 1.100 -> 1.101 share/man/man4/pciide.4 1.23 -> 1.24 Add support for the CMD PCI646U. Linux claims that this driver is brocken with UDMA, so enable Ultra-DMA only if "options PCIIDE_CMD0646U_ENABLEUDMA" is set. defopt PCIIDE_CMD0646U_ENABLEUDMA, PCIIDE_AMD756_ENABLEDMA, PCIIDE_CMD064x_DISABLE Fix a typo pointed out by John Hawkinson Update documentation.
1.96.2.4	30-Jul-2000	bouyer	Pull up (requested by thorpej): lm_pnpbios.c 1.2->1.3 files 1.374->1.375 nslm7x.c 1.3->1.4 nslm7xvar.h 1.2->1.3 lm_isa.c 1.2->1.3 files.pci 1.96->1.97 viaenv.c 1.2->1.3 sysmon.c 1.1 sysmonvar.h 1.1 Add a clearing-house pseudo-device for system monitoring devices such as the LM78 and VT82C686A (and eventually ACPI). Multiple sensor devices can be hooked registered with `sysmon', and eventually sysmon will also handle hardware (and software) watchdog timers. Convert the `lm' and `viaenv' drivers to the new interface.
1.96.2.3	03-Jul-2000	thorpej	Pull up rev. 1.99: cmpci requires mulaw dependency.
1.96.2.2	26-Jun-2000	soren	Remove part of sysmon change mistakenly included in previous pull-up.
1.96.2.1	25-Jun-2000	sommerfeld	defopt TLP_MATCH_*. pullup approved by thorpej
1.123.2.18	11-Nov-2002	nathanw	Catch up to -current
1.123.2.17	18-Oct-2002	nathanw	Catch up to -current.
1.123.2.16	17-Sep-2002	nathanw	Catch up to -current.
1.123.2.15	13-Aug-2002	nathanw	Catch up to -current.
1.123.2.14	01-Aug-2002	nathanw	Catch up to -current.
1.123.2.13	20-Jun-2002	nathanw	Catch up to -current.
1.123.2.12	17-Apr-2002	nathanw	Catch up to -current.
1.123.2.11	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.123.2.10	28-Feb-2002	nathanw	Catch up to -current.
1.123.2.9	11-Jan-2002	nathanw	More catchup.
1.123.2.8	11-Jan-2002	nathanw	Catch up to newer -current.
1.123.2.7	08-Jan-2002	nathanw	Catch up to -current.
1.123.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.123.2.5	22-Oct-2001	nathanw	Catch up to -current.
1.123.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.123.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.123.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.123.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.132.2.9	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.132.2.8	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.132.2.7	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.132.2.6	16-Mar-2002	jdolecek	Catch up with -current.
1.132.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.132.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.132.2.3	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.132.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.132.2.1	03-Aug-2001	lukem	update to -current
1.138.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.142.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.172.2.4	29-Aug-2002	gehenna	catch up with -current.
1.172.2.3	15-Jul-2002	gehenna	catch up with -current.
1.172.2.2	20-Jun-2002	gehenna	catch up with -current.
1.172.2.1	30-May-2002	gehenna	Catch up with -current.
1.173.2.4	11-Nov-2004	he	Pull up revision 1.217 (via patch, requested by itohy in ticket #1741): Add support for the Workbit NinjaSCSI-32 PCI/Cardbus SCSI driver, njs(4).
1.173.2.3	05-Oct-2003	tron	Pull up revision 1.194 (requested by mrg in ticket #1486): add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.173.2.2	28-Jan-2003	jmc	Pullup revisions 1.174-1.175 (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.173.2.1	07-Aug-2002	lukem	Pull up revision 1.176 (requested by augustss in ticket #643): Pull in tea5757.c for the mr driver. Fixes kern/17864.
1.187.2.16	11-Dec-2005	christos	Sync with head.
1.187.2.15	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.187.2.14	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.187.2.13	15-Feb-2005	skrll	Sync with HEAD.
1.187.2.12	24-Jan-2005	skrll	Sync with HEAD.
1.187.2.11	17-Jan-2005	skrll	Sync with HEAD.
1.187.2.10	18-Dec-2004	skrll	Sync with HEAD.
1.187.2.9	29-Nov-2004	skrll	Sync with HEAD.
1.187.2.8	02-Nov-2004	skrll	Sync with HEAD.
1.187.2.7	19-Oct-2004	skrll	Sync with HEAD
1.187.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.187.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.187.2.4	03-Sep-2004	skrll	Sync with HEAD
1.187.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.187.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.187.2.1	03-Aug-2004	skrll	Sync with HEAD
1.210.2.3	30-Aug-2004	tron	branches: 1.210.2.3.2; Pull up revision 1.217 via patch (requested by itohy in ticket #800): Workbit NinjaSCSI-32 PCI/CardBus SCSI driver (njs)
1.210.2.2	14-Jul-2004	tron	Pull up revision 1.214 (requested by bouyer in ticket #644): Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.210.2.1	21-Jun-2004	tron	Pull up revision 1.211 (requested by toshii in ticket #517): Add re(4) RealTek 8139C+/8169/8169S/8110S PCI ethernet adapter driver, from FreeBSD. I haven't tested this code very well, but it seems to work fairly well for me.
1.210.2.3.2.2	11-Apr-2006	riz	Pull up following revision(s) (requested by he in ticket #10409): sys/dev/ic/cissreg.h: revision 1.1 sys/dev/pci/ciss_pci.c: revision 1.1 sys/dev/ic/cissvar.h: revision 1.1 sys/dev/pci/files.pci: revision 1.245 sys/dev/ic/ciss.c: revision 1.1 via patch Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.210.2.3.2.1	02-Apr-2005	he	Pull up revision 1.224 (via patch, requested by bouyer in ticket #1019): Add pdcsata(4), a driver for the Promise SATA150 (aka PDC203xx) series of controllers.
1.228.2.1	29-Apr-2005	kent	sync with -current
1.229.2.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.229.2.1	12-Feb-2005	yamt	sync with head.
1.232.2.10	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1609): sys/arch/i386/conf/INSTALL: revision 1.298 sys/arch/i386/conf/GENERIC: revision 1.802 sys/dev/pci/files.pci: revision 1.274 sys/arch/amd64/conf/GENERIC: revision 1.122 sys/dev/pci/mfi_pci.c: revision 1.1 via patch distrib/sets/lists/man/mi: revision 1.958 sys/arch/amd64/conf/INSTALL: revision 1.61 sys/dev/ic/mfireg.h: revision 1.1 share/man/man4/Makefile: revision 1.416 via patch sys/dev/ic/mfi.c: revision 1.1 sys/arch/i386/conf/XEN2_DOM0: revision 1.19 sys/dev/ic/mfivar.h: revision 1.1 sys/conf/files: revision 1.821 share/man/man4/mfi.4: revision 1.1 share/man/man4/mfi.4: revision 1.2 sys/dev/pci/pcidevs via patch Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64. Punctuation issues.
1.232.2.9	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1608): sys/arch/i386/conf/INSTALL: revision 1.299 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.20 via patch sys/arch/i386/conf/GENERIC: revision 1.803 via patch sys/dev/pci/files.pci: revision 1.275 via patch sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 via patch sys/arch/amd64/conf/GENERIC: revision 1.123 via patch sys/dev/pci/if_bnx.c: revision 1.1 via patch share/man/man4/bnx.4: revision 1.1-1.2 via patch distrib/sets/lists/man/mi: revision 1.959 via patch sys/arch/amd64/conf/INSTALL: revision 1.62 via patch sys/dev/microcode/bnx/bnxfw.h: revision 1.1-1.2 via patch sys/dev/pci/if_bnxreg.h: revision 1.1 via patch share/man/man4/Makefile: revision 1.417 via patch sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 via patch Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64. Add a NetBSD RCS tag. Nx -> Ox; remove trailing whitespace.
1.232.2.8	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.232.2.7	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): share/man/man4/pciide.4: revision 1.60 share/man/man4/Makefile: revision 1.376 via patch sys/dev/pci/svwsata.c: revision 1.1 sys/dev/pci/files.pci: revision 1.243 sys/arch/i386/conf/INSTALL: revision 1.281 sys/dev/pci/pcidevs: revision 1.771 via patch sys/dev/pci/pciidevar.h: revision 1.33 sys/arch/i386/conf/XEN0: revision 1.34 sys/arch/i386/conf/GENERIC: revision 1.732 sys/dev/pci/pciide_svwsata_reg.h: revision 1.1 Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.232.2.6	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): sys/dev/pci/files.pci: revision 1.244 via patch add nfe.
1.232.2.5	08-Apr-2006	tron	Pull up following revision(s) (requested by he in ticket #1247): sys/dev/ic/cissreg.h: revision 1.1 sys/dev/pci/ciss_pci.c: revision 1.1 sys/dev/ic/cissvar.h: revision 1.1 sys/dev/pci/files.pci: revision 1.245 sys/dev/ic/ciss.c: revision 1.1 Add a driver, ciss(4), for the HP/Compaq drivers using the newer "Command Interface to SCSI-3 Support" command interface. Driver ported from OpenBSD by Tonnerre Lombard -- thanks!
1.232.2.4	20-Feb-2006	tron	Pull up following revision(s) (requested by tls in ticket #1179): sys/dev/pci/amdpm.c: revision 1.11 sys/dev/i2c/files.i2c: revision 1.7 sys/dev/i2c/adt7463.c: revision 1.1 sys/dev/pci/files.pci: revision 1.242 via patch sys/dev/pci/amdpmvar.h: revision 1.1 sys/dev/i2c/adt7463reg.h: revision 1.1 sys/dev/DEVNAMES: revision 1.192 via patch sys/dev/pci/amdpm_smbus.c: revision 1.1 sys/dev/pci/amdpm_smbusreg.h: revision 1.1 Add support for i2c bus on AMD-8111 used on many Opteron server motherboards, and ADT 7463c thermal and voltage monitor found on the Tyan S2881 and S2882-D (and probably other boards as well). We do not currently support any kind of detection of the i2c address of the thermal monitor; it appears to be at 0x2D on the S2881 and 0x2E on the S2882-D (kernel config examples forthcoming). From PR kern/32463 submitted by Anil Gopinath, anil_public@yahoo.com.
1.232.2.3	29-Dec-2005	riz	Pull up following revision(s) (requested by kent in ticket #916): sys/dev/pci/azalia.h: revision 1.1 sys/dev/pci/azalia_codec.c: revision 1.1 sys/dev/pci/azalia.c: revision 1.13 sys/dev/pci/files.pci: revision 1.239 split azalia.c into three files. * azalia_codec.c: code for specific codecs * azalia.c: other code * azalia.h: common definitions
1.232.2.2	13-Sep-2005	tron	Pull up following revision(s) (requested by ragge in ticket #768): sys/dev/pci/if_xge.c: revision 1.1 via patch sys/dev/pci/if_xgereg.h: revision 1.1 sys/dev/pci/files.pci: revision 1.238 Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.232.2.1	02-Jul-2005	tron	Pull up revision 1.233 (requested by kent in ticket #494): An audio driver for High Definition Audio. The driver is highly experimental at this moment: - limieted support for playback - support for no mixer controls - support for no recording
1.235.2.8	11-Feb-2008	yamt	sync with head.
1.235.2.7	21-Jan-2008	yamt	sync with head
1.235.2.6	07-Dec-2007	yamt	sync with head
1.235.2.5	15-Nov-2007	yamt	sync with head.
1.235.2.4	03-Sep-2007	yamt	sync with head.
1.235.2.3	26-Feb-2007	yamt	sync with head.
1.235.2.2	30-Dec-2006	yamt	sync with head.
1.235.2.1	21-Jun-2006	yamt	sync with head.
1.241.6.3	07-Jun-2006	kardel	Sync with head.
1.241.6.2	01-Jun-2006	kardel	Sync with head.
1.241.6.1	22-Apr-2006	simonb	Sync with head.
1.241.4.1	09-Sep-2006	rpaulo	sync with head
1.241.2.1	01-Mar-2006	yamt	sync with head.
1.242.2.7	14-Sep-2006	yamt	sync with head.
1.242.2.6	03-Sep-2006	yamt	sync with head.
1.242.2.5	11-Aug-2006	yamt	sync with head
1.242.2.4	26-Jun-2006	yamt	sync with head.
1.242.2.3	24-May-2006	yamt	sync with head.
1.242.2.2	01-Apr-2006	yamt	sync with head.
1.242.2.1	13-Mar-2006	yamt	sync with head.
1.243.2.2	11-May-2006	elad	sync with head
1.243.2.1	19-Apr-2006	elad	sync with head.
1.244.2.3	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.244.2.2	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.244.2.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.253.2.1	19-Jun-2006	chap	Sync with head.
1.258.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.262.2.2	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/if_skreg.h: revision 1.8 sys/dev/pci/if_msk.c: revision 1.1 sys/dev/pci/files.pci: revision 1.268 sys/dev/pci/if_mskvar.h: revision 1.1 Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.262.2.1	15-Aug-2006	ghen	Pull up following revision(s) (requested by jmcneill in ticket #18): sys/dev/pci/files.pci: revision 1.265 Add rasops16 attribute to unichromefb.
1.267.2.3	01-Feb-2007	ad	Sync with head.
1.267.2.2	12-Jan-2007	ad	Sync with head.
1.267.2.1	18-Nov-2006	ad	Sync with head.
1.268.2.3	18-Dec-2006	yamt	sync with head.
1.268.2.2	10-Dec-2006	yamt	sync with head.
1.268.2.1	22-Oct-2006	yamt	sync with head
1.273.2.5	19-Apr-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1128): sys/dev/pci/if_lii.c: revisions 1.1, 1.3 via patch sys/arch/i386/conf/ALL: revision 1.161 via patch share/man/man4/lii.4: revision 1.1 via patch sys/dev/pci/files.pci: revision 1.303 via patch sys/dev/DEVNAMES: revision 1.238 via patch sys/arch/i386/conf/GENERIC: revision 1.885 via patch distrib/sets/lists/man/mi: revisions 1.1062, 1.1063 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.43 via patch share/man/man4/Makefile: revision 1.460 via patch sys/arch/i386/conf/INSTALL: revision 1.323 via patch sys/dev/pci/if_liireg.h: revision 1.1 via patch Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC.
1.273.2.4	11-Jan-2008	xtraeme	Pull up following revision(s) (requested by christos in ticket #1037): sys/dev/pci/arcmsr.c: patch sys/dev/pci/arcmsrvar.h: patch sys/dev/pci/pcidevs: patch sys/dev/pci/files.pci: patch sys/arch/i386/conf/GENERIC: patch sys/arch/i386/conf/XEN2_DOM0: patch sys/arch/i386/conf/INSTALL_LARGE: patch sys/arch/amd64/conf/GENERIC: patch sys/arch/amd64/conf/INSTALL: patch share/man/man4/Makefile: patch share/man/man4/arcmsr.4: patch distrib/sets/lists/man/mi: patch Add the Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD.
1.273.2.3	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.273.2.2	21-Dec-2006	tron	branches: 1.273.2.2.2; Pull up following revision(s) (requested by bouyer in ticket #287): sys/arch/i386/conf/INSTALL: revision 1.299 sys/arch/i386/conf/XEN2_DOM0: revision 1.20 sys/arch/i386/conf/GENERIC: revision 1.803 sys/dev/pci/files.pci: revision 1.275 sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 sys/arch/amd64/conf/GENERIC: revision 1.123 sys/dev/pci/if_bnx.c: revision 1.1 share/man/man4/bnx.4: revision 1.1 distrib/sets/lists/man/mi: revision 1.959 sys/arch/amd64/conf/INSTALL: revision 1.62 sys/dev/microcode/bnx/bnxfw.h: revision 1.1 sys/arch/i386/conf/ALL: revision 1.76 sys/dev/pci/if_bnxreg.h: revision 1.1 share/man/man4/Makefile: revision 1.417 sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.273.2.1	21-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #286): sys/arch/i386/conf/INSTALL: revision 1.298 sys/arch/i386/conf/GENERIC: revision 1.802 sys/dev/pci/files.pci: revision 1.274 sys/arch/amd64/conf/GENERIC: revision 1.122 sys/dev/pci/mfi_pci.c: revision 1.1 distrib/sets/lists/man/mi: revision 1.958 sys/arch/amd64/conf/INSTALL: revision 1.61 sys/arch/i386/conf/ALL: revision 1.75 sys/dev/ic/mfireg.h: revision 1.1 share/man/man4/Makefile: revision 1.416 sys/dev/ic/mfi.c: revision 1.1 sys/arch/i386/conf/XEN2_DOM0: revision 1.19 sys/dev/ic/mfivar.h: revision 1.1 sys/conf/files: revision 1.821 share/man/man4/mfi.4: revision 1.1 Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64.
1.273.2.2.2.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.273.2.2.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.277.10.1	29-Mar-2007	reinoud	Pullup to -current
1.277.8.1	11-Jul-2007	mjf	Sync with head.
1.277.6.4	20-Aug-2007	ad	Sync with HEAD.
1.277.6.3	15-Jul-2007	ad	Sync with head.
1.277.6.2	27-May-2007	ad	Sync with head.
1.277.6.1	10-Apr-2007	ad	Sync with head.
1.277.2.4	17-May-2007	yamt	sync with head.
1.277.2.3	07-May-2007	yamt	sync with head.
1.277.2.2	15-Apr-2007	yamt	sync with head.
1.277.2.1	24-Mar-2007	yamt	sync with head.
1.294.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.295.14.2	28-Jul-2007	kiyohara	Add support for Intel ICH SMBus controller.
1.295.14.1	28-Jul-2007	kiyohara	file files.pci was added on branch matt-mips64 on 2007-07-28 10:51:58 +0000
1.295.12.4	18-Feb-2008	mjf	Sync with HEAD.
1.295.12.3	27-Dec-2007	mjf	Sync with HEAD.
1.295.12.2	08-Dec-2007	mjf	Sync with HEAD.
1.295.12.1	19-Nov-2007	mjf	Sync with HEAD.
1.295.10.1	13-Nov-2007	bouyer	Sync with HEAD
1.295.6.3	23-Mar-2008	matt	sync with HEAD
1.295.6.2	09-Jan-2008	matt	sync with HEAD
1.295.6.1	08-Nov-2007	matt	sync with -HEAD
1.295.4.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.295.4.1	11-Nov-2007	joerg	Sync with HEAD.
1.296.2.2	26-Dec-2007	ad	Sync with head.
1.296.2.1	08-Dec-2007	ad	Sync with head.
1.298.4.2	19-Jan-2008	bouyer	Sync with HEAD
1.298.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.302.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.302.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.302.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.302.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.303.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.303.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.303.4.7	11-Aug-2010	yamt	sync with head.
1.303.4.6	11-Mar-2010	yamt	sync with head
1.303.4.5	16-Sep-2009	yamt	sync with head
1.303.4.4	19-Aug-2009	yamt	sync with head.
1.303.4.3	18-Jul-2009	yamt	sync with head.
1.303.4.2	20-Jun-2009	yamt	sync with head
1.303.4.1	04-May-2009	yamt	sync with head.
1.303.2.1	04-Jun-2008	yamt	sync with head
1.304.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.304.4.1	19-Oct-2008	haad	Sync with HEAD.
1.304.2.1	28-Jul-2008	simonb	Sync with head.
1.306.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.306.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.308.8.2	05-Nov-2013	matt	Pull down xhci support from HEAD
1.308.8.1	21-Apr-2010	matt	sync to netbsd-5
1.308.2.7	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.308.2.6	28-Oct-2009	bouyer	branches: 1.308.2.6.4; Pull up the following revisions, requested by sborrill in ticket #1114: share/man/man4/sdhc.4 1.1-1.2 sys/dev/pci/sdhc_pci.c 1.1-1.3 distrib/sets/lists/man/mi patch share/man/man4/Makefile patch sys/arch/amd64/conf/GENERIC patch sys/arch/i386/conf/ALL patch sys/arch/i386/conf/GENERIC patch sys/dev/pci/files.pci patch sys/dev/pci/pcidevs patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add sdhc(4), a driver for SD controllers following the SD Host Controller Standard Simplified Specification.
1.308.2.5	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/conf/files: revision 1.956 via patch sys/dev/pci/files.pci: revision 1.321 As, siisata attaches not just at pci, but for a while now, also at cardbus, move siisata core driver config to the propper config file.
1.308.2.4	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ic/siisata.c: revision 1.3 sys/dev/ic/siisatareg.h: revision 1.3 sys/dev/ic/siisatavar.h: revision 1.3 sys/dev/pci/files.pci: revision 1.315 sys/dev/pci/siisata_pci.c: revision 1.3 A few changes for siisata(4): - Support detachment. From KIYOHARA Takashi. - Add PCI detachment functionality (albeit not very interesting when the bus can not yet be rescanned). - Rework interrupt handlers to reduce near-duplicate code. Borrowed from ahcisata(4). - Attempt to make polled I/O work. Untested. - Fix formatting of some messages. - For always-polled commands, disable interrupt at slot level rather than port level. - Instead of busy-waiting indefinitely for completion of some commands move on after 31 seconds. Use cv_timedwait(9) instead of DELAY(9). - Use abstracted SATA FIS code. - Enable use of disks that don't respond with the standard signature.
1.308.2.3	26-Sep-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #946): distrib/sets/lists/man/mi: revision 1.1155 doc/CHANGES: revision 1.1285 etc/MAKEDEV.tmpl: revision 1.128 share/man/man4/Makefile: revision 1.498 via patch share/man/man4/hdaudio.4: revision 1.1-1.3 share/man/man8/MAKEDEV.8: revision 1.36 sys/arch/amd64/conf/GENERIC: revision 1.250-1.251 sys/arch/amd64/conf/XEN3_DOM0: revision 1.44-1.45 sys/arch/i386/conf/ALL: revision 1.206-1.207 sys/arch/i386/conf/GENERIC: revision 1.942-1.943 sys/arch/i386/conf/XEN3_DOM0: patch sys/conf/majors: revision 1.47 via patch sys/dev/pci/hdaudio/files.hdaudio: revision 1.1-1.2 sys/dev/pci/hdaudio/hdaudio.c: revision 1.1-1.4 sys/dev/pci/hdaudio/hdaudio_afg.c: revisions 1.1-1.14 sys/dev/pci/hdaudio/hdaudio_mixer.h: revisions 1.1-1.3 sys/dev/pci/hdaudio/hdaudio_pci.c: revisions 1.1-1.2 sys/dev/pci/hdaudio/hdaudioio.h: revisions 1.1-1.2 sys/dev/pci/hdaudio/hdaudioreg.h: revisions 1.1-1.3 sys/dev/pci/hdaudio/hdaudiovar.h: revisions 1.1-1.4 sys/dev/pci/files.pci: revisions 1.319 and 1.322 via patch hdaudio(4) is a standards-compliant driver for High Definition Audio. It will replace azalia(4) after testing. To use, comment out azalia in your kernel configuration and uncomment the hdaudio and hdafg lines so it reads: hdaudio* at pci? dev ? function ? hdafg* at hdaudiobus? You should also: cd /dev sh MAKEDEV audio
1.308.2.2	03-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #735): sys/dev/pci/if_age.c: revisions 1.1-1.28 sys/dev/pci/if_agereg.h: revisions 1.1-1.2 sys/dev/pci/if_ale.c: revisions 1.1-1.3 sys/dev/pci/if_alereg.h: revisions 1.1-1.2 sys/dev/pci/files.pci: revisions 1.310-1.313 sys/dev/mii/atphy.c: revisions 1.1-1.5 sys/dev/mii/files.mii: revision 1.42 share/man/man4/age.4: revisions 1.1-1.3 share/man/man4/ale.4: revisions 1.1-1.2 share/man/man4/atphy.4: revisions 1.1-1.2 share/man/man4/Makefile: revisions 1.486, 1.489 distrib/sets/lists/man/mi: revisions 1.1118, 1.1132 sys/arch/i386/conf/GENERIC: revisions 1.928, 1.933 sys/arch/i386/conf/XEN2_DOM0: revisions 1.57, 1.60 sys/arch/amd64/conf/GENERIC: revisions 1.239, 1.241 sys/arch/amd64/conf/XEN3_DOM0: revisions 1.38, 1.42 Add the age(4) and ale(4) drivers for Attansic L1 and Atheros AR8121/AR8113/AR8114 devices respectively.
1.308.2.1	01-May-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #260): sys/dev/pci/files.pci: revision 1.309 sys/dev/pci/gcscaudio.c: revision 1.1 sys/dev/pci/gcscaudioreg.h: revision 1.1 PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.308.2.6.4.1	07-Jan-2011	matt	arcmsr needs sysmon_envsys
1.312.2.2	23-Jul-2009	jym	Sync with HEAD.
1.312.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.324.2.3	06-Nov-2010	uebayasi	Sync with HEAD.
1.324.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.324.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.326.2.3	21-Apr-2011	rmind	sync with head
1.326.2.2	05-Mar-2011	rmind	sync with head
1.326.2.1	30-May-2010	rmind	sync with head
1.335.4.3	05-Mar-2011	bouyer	Sync with HEAD
1.335.4.2	17-Feb-2011	bouyer	Sync with HEAD
1.335.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.335.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.350.6.5	29-Apr-2012	mrg	sync to latest -current.
1.350.6.4	06-Mar-2012	mrg	sync to -current
1.350.6.3	06-Mar-2012	mrg	sync to -current
1.350.6.2	04-Mar-2012	mrg	sync to latest -current.
1.350.6.1	18-Feb-2012	mrg	merge to -current.
1.350.2.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.350.2.4	23-Jan-2013	yamt	sync with head
1.350.2.3	30-Oct-2012	yamt	sync with head
1.350.2.2	23-May-2012	yamt	sync with head.
1.350.2.1	17-Apr-2012	yamt	sync with head
1.354.2.1	23-Apr-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #193): sys/arch/i386/conf/GENERIC: revision 1.1072 sys/dev/pci/mpii.c: revision 1.1 sys/arch/i386/conf/XEN3_DOM0: revision 1.66 sys/dev/pci/files.pci: revision 1.357 share/man/man4/Makefile: revision 1.584 distrib/sets/lists/man/mi: revision 1.1387 share/man/man4/mpii.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.354 sys/arch/amd64/conf/XEN3_DOM0: revision 1.83 sys/arch/i386/conf/ALL: revision 1.337 Add mpii(4), a driver for LSI Logic Fusion-MPT Message Passing Interface II SAS controllers. Ported from OpenBSD.
1.360.2.5	03-Dec-2017	jdolecek	fix back some more incorrect sync
1.360.2.4	03-Dec-2017	jdolecek	update from HEAD
1.360.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.360.2.2	23-Jun-2013	tls	resync from head
1.360.2.1	25-Feb-2013	tls	resync with head
1.364.6.3	04-Mar-2014	riastradh	Rework drm2 configuration and modularization. - Split drm2.kmod into drmkms_linux, drmkms, drmkms_pci. - Split up drmkms-related files.* similarly. - Make drm_agp_* routines hooks that the drmkms_pci module installs. - Reuse drm device major 180 for (old) drm and drmkms. Now old drm and new drmkms can coexist happily in the same tree, with or without PCI, and require no changes to userland.
1.364.6.2	29-Jan-2014	riastradh	First draft of in-kernel drm2 attachment. The drm2 `drm' device is now called `drmkms' (likewise `i915drmkms', &c.) so that it can coexist with the old drm code to reduce diffs from HEAD. (Can't call a device `drm2'.)
1.364.6.1	23-Jul-2013	riastradh	sync with HEAD
1.364.4.2	18-May-2014	rmind	sync with head
1.364.4.1	28-Aug-2013	rmind	sync with head
1.371.4.5	25-Jul-2017	snj	Apply patch (requested by nonaka in ticket #1413): iwm(4): Add a driver for Intel Wireless 726x, 316x, 826x and 416x series. Ported from OpenBSD.
1.371.4.4	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1338): sys/arch/amd64/conf/GENERIC: revision 1.418 sys/dev/pci/files.pci: revision 1.379 sys/dev/pci/vioscsi.c: revisions 1.1-1.8 sys/dev/pci/vioscsireg.h: revision 1.1 sys/dev/pci/virtioreg.h: revisions 1.5, 1.6 add the common flag bits -- Add vioscsi, compile tested only (toxic) -- this is working now, remove debugging. -- add vioscsi -- kill some more debugging. -- Fake mode sense data for illegal targets. qemu reports 256 targets... -- fix the status return. -- Use SCSIPI_CHAN_NOSETTLE. -- vioscsi_req_get()/virtio_enqueue_prep() failing is actually perfectly normal - observed failures included 10, 27, 61 in-flight commands, so probably depends on particular command mix; return with XS_RESOURCE_SHORTAGE rather then panic do vioscsi_req_put() when initial bus_dmamap_load() fails, as suggested by the XXX; the vq_done hook is called by virtio, but in that case we never get to commit the request to it -- react on ADAPTER_REQ_SET_XFER_MODE so that we set tagged queuing pass tag type and set id
1.371.4.3	19-Apr-2015	riz	branches: 1.371.4.3.4; Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.371.4.2	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.371.4.1	02-Nov-2014	martin	Pull up following revision(s) (requested by tls in ticket #174): doc/CHANGES: revision 1.2005 sys/dev/pci/viornd.c: revision 1.1 sys/dev/pci/files.pci: revision 1.372 sys/arch/amd64/conf/GENERIC: revision 1.401 share/man/man4/viornd.4: revision 1.1 Add viornd(4), a driver for the VirtIO entropy source available on QEMU, KVM, and Google Compute Engine. From OpenBSD.
1.371.4.3.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.372.2.7	28-Aug-2017	skrll	Sync with HEAD
1.372.2.6	05-Dec-2016	skrll	Sync with HEAD
1.372.2.5	29-May-2016	skrll	Sync with HEAD
1.372.2.4	19-Mar-2016	skrll	Sync with HEAD
1.372.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.372.2.2	22-Sep-2015	skrll	Sync with HEAD
1.372.2.1	06-Apr-2015	skrll	Sync with HEAD
1.381.2.3	26-Apr-2017	pgoyette	Sync with HEAD
1.381.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.381.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.385.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.388.4.8	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.388.4.7	23-Oct-2021	martin	Pull up the following, requested by msaitoh in ticket #1701: share/man/man4/wm.4 1.42 sys/dev/pci/files.pci 1.434 sys/dev/pci/if_wm.c 1.686-1.687,1.692-1.693, 1.697,1.699-1.703, 1.706-1.715 via patch sys/dev/pci/if_wmvar.h 1.46 - Add missing drain for pcq in wm_stop_locked(). - Add support for I219V 15-19 and I219LM 16-19. - Fix Tx stall. - Use wm_flush_desc_rings() workaround more on I219. - Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. - Add some sysctl values for debugging TX/RX queues. - WM_EVENT_COUNTER is enabled by default on 64 bit architectures. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxq->rxq_ptr updating. - Stop legacy interrupts before calling softint. - Disable printf()s in wm_flush_desc_rings() because the code is verified. - Print I219's version number. - Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. - Fix return value of interrupt handler. - Only print an error about missing I/O BARs for chips that need it. - Do not return a void value from a void function. - Add new flag named WM_F_CRC_STRIP and use it. No functional change. - Whitespace fixes. Fix comments. No functional change.
1.388.4.6	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.388.4.5	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.388.4.4	07-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1129): sys/arch/amd64/conf/XEN3_DOM0: revision 1.159 sys/arch/amd64/conf/GENERIC: revision 1.508 sys/arch/i386/conf/ALL: revision 1.458 sys/dev/ic/mfi.c: revision 1.60 sys/dev/pci/mpiireg.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.1628 sys/dev/pci/mfii.c: revision 1.1,1.2 (adapted) sys/arch/i386/conf/GENERIC: revision 1.1194 sys/dev/pci/mpii.c: revision 1.13 sys/dev/ic/mfireg.h: revision 1.9 share/man/man4/mfii.4: revision 1.1 share/man/man4/Makefile: revision 1.673 (patch) sys/dev/pci/files.pci: revision 1.410 share/man/man4/mfii.4: revision 1.2 sys/arch/amd64/conf/ALL: revision 1.108 sys/arch/i386/conf/XEN3PAE_DOM0: revision 1.8 (patch, in XEN3_DOM0) Add some definitions from OpenBSD, needed by the upcoming mfii driver. No functionnal change. - Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change. - Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6. - Add my name in copyright list - add a man page for the new mpii, mostly from OpenBSD. - Sort SEE ALSO. Fix date. Whitespace fixes.
1.388.4.3	31-Oct-2018	martin	Add bwfm config glue, requested by maya in ticket #1072.
1.388.4.2	23-Sep-2018	martin	Pull up following revision(s) (requested by nakayama in ticket #1035): sys/dev/pci/files.pci: revision 1.403 ixg and ixv depend on mii and mii_phy.
1.388.4.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.394.2.7	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.394.2.6	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.394.2.5	30-Sep-2018	pgoyette	Ssync with HEAD
1.394.2.4	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.394.2.3	28-Jul-2018	pgoyette	Sync with HEAD
1.394.2.2	25-Jun-2018	pgoyette	Sync with HEAD
1.394.2.1	21-May-2018	pgoyette	Sync with HEAD
1.397.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.397.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.397.2.1	10-Jun-2019	christos	Sync with HEAD
1.400.2.1	24-Sep-2018	jdolecek	partial HEAD sync - ISDN device removal
1.412.2.1	05-May-2019	isaki	Remove obsoleted au{,rate,vol}conv and mulaw attributes. audio provides the equivalent of them inseparably.
1.413.2.6	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.413.2.5	21-Oct-2021	martin	Pull up the following, requested by msaitoh in ticket #1363: share/man/man4/wm.4 1.42 sys/dev/pci/files.pci 1.434 sys/dev/pci/if_wm.c 1.690,1.692-1.693,1.697-1.704, 1.706-1.715 via patch - Add missing drain for pcq in wm_stop_locked(). - Add support for I219V 15-19 and I219LM 16-19. - Fix Tx stall. - Use wm_flush_desc_rings() workaround more on I219. - Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. - Add some sysctl values for debugging TX/RX queues. - Use atomic_{load,store}_relaxed for evcnt 64 bit counter. - WM_EVENT_COUNTER is enabled by default on 64 bit architectures. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxq->rxq_ptr updating. - Stop legacy interrupts before calling softint. - Disable printf()s in wm_flush_desc_rings() because the code is verified. - Print I219's version number. - Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. - Fix return value of interrupt handler. - Only print an error about missing I/O BARs for chips that need it. - Do not return a void value from a void function. - Fix comment. - Whitespace fixes. No functional change.
1.413.2.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.413.2.3	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.413.2.2	07-Jul-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #980): sys/dev/pci/if_aq.c: revision 1.4 sys/dev/pci/if_aq.c: revision 1.5 sys/arch/amd64/conf/GENERIC: revision 1.553 sys/dev/pci/files.pci: revision 1.419 sys/arch/amd64/conf/XEN3_DOM0: revision 1.170 sys/dev/pci/if_aq.c: revision 1.9 share/man/man4/Makefile: revision 1.693 sys/dev/pci/pcidevs: revision 1.1411 share/man/man4/aq.4: revision 1.1 share/man/man4/aq.4: revision 1.3 sys/arch/i386/conf/ALL: revision 1.479 share/man/man4/aq.4: revision 1.4 sys/dev/pci/if_aq.c: revision 1.10 sys/dev/pci/files.pci: revision 1.421 sys/dev/pci/if_aq.c: revision 1.11 sys/dev/pci/if_aq.c: revision 1.12 sys/dev/pci/if_aq.c: revision 1.13 sys/dev/pci/if_aq.c: revision 1.14 sys/dev/pci/if_aq.c: revision 1.15 sys/dev/pci/if_aq.c: revision 1.16 sys/dev/pci/pcidevs: revision 1.1408 sys/arch/amd64/conf/ALL: revision 1.135 sys/net/ethertypes.h: revision 1.19 sys/arch/i386/conf/GENERIC: revision 1.1218 distrib/sets/lists/man/mi: revision 1.1668 sys/dev/pci/if_aq.c: revision 1.1 sys/dev/pci/if_aq.c: revision 1.2 sys/dev/pci/pcidevs: revision 1.1395 sys/dev/pci/if_aq.c: revision 1.3 sys/arch/evbarm/conf/GENERIC64: revision 1.125 Add the ETHERTYPE_QINQ for 802.1ad VLAN stacking add Aquantia AQC 10G network adapters add support Aquantia AQC seriese 10G network adapters. this driver is based on the FreeBSD version https://github.com/Aquantia/aqtion-freebsd , but drastically rewritten for NetBSD. add aq(4) Add Aquantia AQC100, AQC100S and D100. add support VLAN HW filter set/clear IFF_OACTIVE flag only on txring 0 make counters per queue support internal PHY temperature sensor Found by kUBSan: - Use unsigned to avoid undefined behavior in aq_hw_init(). - Cast to unsigned to avoid undefined behavior in aq_set_mac_addr(). fix descriptions of register map in comment return the ifmedia active status correctly even while the link is not up after attach. pointed out by msaitoh@. thanks. On FIBRE devices, there are times when linkstat interrupt doesn't occur? reported from Andrius V. thanks. - use polling instead of linkstat interrupt when FIBRE - add AQ_FORCE_POLL_LINKSTAT options (not by default) sort product table, and tabify add support AQC100S and D100. not tested, but they are probably the same as the AQC100.
1.413.2.1	23-Sep-2019	martin	Pull up following revision(s) (requested by jmcneill in ticket #230): sys/dev/pci/files.pci: revision 1.415 distrib/sets/lists/man/mi: revision 1.1655 sys/dev/pci/if_mcx.c: revision 1.1 share/man/man4/Makefile: revision 1.683 share/man/man4/mcx.4: revision 1.1 sys/arch/evbarm/conf/GENERIC64: revision 1.107 sys/dev/pci/files.pci: revision 1.414 Add support for Mellanox 5th generation ethernet devices. This driver supports the ConnectX-4 Lx EN, ConnectX-4 EN, ConnectX-5 EN, and ConnectX-6 EN adapters. Driver ported from OpenBSD. add mcx attachment. Correct mcx comment, drop mii dependency
1.419.2.3	29-Feb-2020	ad	Sync with head.
1.419.2.2	25-Jan-2020	ad	Sync with head.
1.419.2.1	17-Jan-2020	ad	Sync with head.
1.432.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.436.4.1	13-May-2021	thorpej	Sync with HEAD.
1.445.2.4	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #1148): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.445.2.3	03-Nov-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #446): sys/dev/pci/if_eqos_pci.c: revision 1.3 sys/arch/i386/conf/GENERIC: revision 1.1251 sys/arch/i386/conf/GENERIC: revision 1.1252 sys/arch/amd64/conf/GENERIC: revision 1.607 sys/arch/amd64/conf/GENERIC: revision 1.608 sys/dev/ic/dwc_eqos.c: revision 1.20 sys/dev/ic/dwc_eqos.c: revision 1.21 share/man/man4/eqos.4: revision 1.2 sys/dev/ic/dwc_eqos.c: revision 1.22 sys/dev/ic/dwc_eqos_reg.h: revision 1.7 sys/dev/ic/dwc_eqos.c: revision 1.23 sys/dev/ic/dwc_eqos_reg.h: revision 1.8 sys/dev/ic/dwc_eqos.c: revision 1.24 sys/dev/ic/dwc_eqos.c: revision 1.25 sys/dev/ic/dwc_eqos.c: revision 1.26 sys/dev/ic/dwc_eqos.c: revision 1.27 sys/dev/ic/dwc_eqos_var.h: revision 1.5 sys/dev/ic/dwc_eqos.c: revision 1.28 sys/dev/ic/dwc_eqos_var.h: revision 1.6 sys/dev/ic/dwc_eqos.c: revision 1.29 sys/dev/ic/dwc_eqos.c: revision 1.18 sys/dev/ic/dwc_eqos.c: revision 1.19 sys/dev/pci/files.pci: revision 1.448 sys/dev/pci/if_eqos_pci.c: revision 1.1 sys/dev/pci/if_eqos_pci.c: revision 1.2 eqos(4): Fix definition of GMAC_MAC_HW_FEATURE1_RXFIFOSIZE. eqos(4): Fix a bug that the MAC address is swapped. Don't swap the MAC address in eqos_get_eaddr(). Other OSes except FreeBSD (which was based on NetBSD's) don't swap it. With this change, my own OnLogic Helix 330's MAC address becomes correct. The OUI is 84:8b:cd:4d. It's owned by Logic Supply and they were acquired by OnLogic. On Quartz64 with UEFI, the MAC address is wrongly set and the multicast bit might be set. To do workaround, clear the bit if it's set. eqos(4): Add missing clock range. eqos(4): Accept if snpsver == 0x52. Tested with Intel Elkhart Lake. TODO: Multiqueue support. Add watchdog timer. Add detach function. eqos(4): Add initial support for Intel Elkhart Lake internal Ethernet devices. - Only tested on PSE SGMII 1G Ethernet MAC with MaxLinear GPY115. - I don't know why dmat64 doesn't work. eqos_attach() have a special code if EQOS_HW_FEATURE_ADDR64_32BIT(sc) is true, but it seems it doesn't work. - TODO: Multiqueue support. Detach support. eqos(4): Fix compile error for arch that sizeof(bus_size_t) == 4 (i386). Trailing whitespace eqos(4): Disable eqos(4) by default because it's not stable on x86. eqos(4): KNF. No functional change. eqos(4): Add and modify some DPRINTF()s. eqos(4): Add sysctls for debugging. eqos(4): Use EQOS_TXLOCK() more to be stable. Fix a bug that sc_tx.{cur,next,queued} become inconsitent. Use txlock when accessing TX data. eqos(4): Set TX/RX DMA burst length to improve performance. eqos(4): Set flow control correctly. eqos_pci: Limit to 32bit DMA only for PSE devices. eqos(4): Fix typo in comment.
1.445.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.445.2.1	01-Aug-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #282): sys/dev/pci/ichsmb.c: revision 1.82 sys/arch/amd64/conf/GENERIC: revision 1.602 sys/arch/x86/pci/tco.c: revision 1.10 sys/arch/x86/pci/tco.h: revision 1.5 sys/arch/x86/pci/ichlpcib.c: revision 1.59 sys/dev/ic/i82801lpcreg.h: revision 1.17 sys/arch/x86/pci/files.pci: revision 1.27 sys/dev/pci/files.pci: revision 1.446 ichsmb(4), tco(4): Add support for TCO on newer Intel chipsets. TCO (`Total Cost of Ownership', Intel's bizarre name for a watchdog timer) used to hang off the Intel I/O platform controller hub's (ICH) low-pin-count interface bridge (LPC IB), or ichlpcib(4). On newer devices, it hangs off the ICH SMBus instead. Tested on INTEL 100SERIES_SMB (works) and INTEL 100SERIES_LP_SMB (doesn't work, still not sure why). XXX kernel revbump: This breaks the module ABI -- tco(4) modules older than the change to make ta_has_rcba into ta_version will incorrectly attach at buses they do not understand. (However, the tco(4) driver is statically built into GENERIC, so maybe it's safe for pullup since the module wouldn't have worked anyway.)
1.448.6.1	02-Aug-2025	perseant	Sync with HEAD
1.450.2.1	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #6): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.3	26-Jul-2025	martin	PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.2	28-Oct-2019	ozaki-r	branches: 1.2.32; Implement a front-end driver of virtio-9p called vio9p In conjunction with mount_9p, it enables a NetBSD system running as a VM guest to mount an exported filesystem by the host via virtio-9p. It exports a 9p end-point of virtio-9p via a character device file for mount_9p. Reviewed by yamaguchi@
1.1	02-Aug-2017	cherry	branches: 1.1.2; 1.1.4; 1.1.8; Reorg the virtio(4) config(9) code to be explicitly pci specific. We'll later use this for a non-pci virtio(4) usecase. ok martin
1.1.8.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.4.2	03-Dec-2017	jdolecek	update from HEAD
1.1.4.1	02-Aug-2017	jdolecek	file files.virtio was added on branch tls-maxphys on 2017-12-03 11:37:07 +0000
1.1.2.2	28-Aug-2017	skrll	Sync with HEAD
1.1.2.1	02-Aug-2017	skrll	file files.virtio was added on branch nick-nhusb on 2017-08-28 17:52:05 +0000
1.2.32.1	02-Aug-2025	perseant	Sync with HEAD
1.50	08-Jan-2024	chs	mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.49	07-Aug-2021	thorpej	branches: 1.49.6; Merge thorpej-cfargs2.
1.48	24-Apr-2021	thorpej	branches: 1.48.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.47	08-Jun-2019	isaki	branches: 1.47.12; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.46	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.45	16-Mar-2019	isaki	branches: 1.45.2; Use C99 style struct initializer to audio_hw_if.
1.44	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.43	01-Jun-2017	chs	branches: 1.43.8; 1.43.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.42	29-Mar-2014	christos	branches: 1.42.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.41	27-Oct-2012	chs	branches: 1.41.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.40	24-Nov-2011	mrg	branches: 1.40.8; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.39	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.38	13-Nov-2010	uebayasi	branches: 1.38.8; 1.38.10; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.37	26-Nov-2009	njoly	branches: 1.37.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.36	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.35	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.34	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.33	28-Apr-2008	martin	branches: 1.33.12; 1.33.14; Remove clause 3 and 4 from TNF licenses
1.32	10-Apr-2008	cegger	branches: 1.32.2; 1.32.4; use aprint_*_dev and device_xname
1.31	01-Apr-2008	xtraeme	Fix fallout from device_t/softc split to mpu(4): always pass the mpu_softc to mpu_intr(). Found by Takeshi Nakayama.
1.30	19-Oct-2007	ad	branches: 1.30.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.29	04-Mar-2007	christos	branches: 1.29.2; 1.29.14; 1.29.16; 1.29.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.28	16-Nov-2006	christos	branches: 1.28.4; 1.28.6; __unused removal on arguments; approved by core.
1.27	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.26	30-Aug-2006	christos	branches: 1.26.2; 1.26.4; static goes first
1.25	11-Dec-2005	christos	branches: 1.25.4; 1.25.8; merge ktrace-lwp.
1.24	28-Jun-2005	thorpej	branches: 1.24.2; Use ANSI function decls and static.
1.23	15-Jan-2005	kent	ansify and KNF
1.22	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.21	09-Nov-2004	kent	branches: 1.21.2; remove mixer setting code in *_attach() because ac97_attach() does it.
1.20	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.19	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.18	03-May-2003	wiz	branches: 1.18.2; 1.18.4; DMA, not dma nor Dma.
1.17	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.16	31-Jan-2003	thorpej	Use aprint_*().
1.15	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.14	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.13	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.12	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.11	13-Nov-2001	lukem	add RCSID
1.10	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.9	28-Dec-2000	sommerfeld	branches: 1.9.2; 1.9.4; 1.9.6; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.8	07-Dec-2000	abs	Wrap call to mpu_intr in NMPU - nowbuildswithout 'mpu at fms' in config
1.7	14-Nov-2000	thorpej	branches: 1.7.2; NBPG -> PAGE_SIZE
1.6	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.5	15-May-2000	thorpej	branches: 1.5.4; ac97.h -> ac97var.h
1.4	08-Apr-2000	tsarna	Fix fms_free like auvia_free
1.3	02-Nov-1999	augustss	branches: 1.3.4; 1.3.6; Shorten some delay()s. From WITEK Wnuk <witek@atari.org>
1.2	01-Nov-1999	augustss	Fold some long lines and make it compile with the new auconv functions.
1.1	01-Nov-1999	augustss	Add FM801 audio device driver. From Witold J. Wnuk ww181302@zodiac.mimuw.edu.pl Closes PR kern/8729.
1.3.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.3.4.2	15-Nov-1999	fvdl	Sync with -current
1.3.4.1	02-Nov-1999	fvdl	file fms.c was added on branch fvdl-softdep on 1999-11-15 00:41:00 +0000
1.5.4.2	13-Dec-2000	he	Pull up revision 1.8 (requested by abs): Make this build without ``mpu at fms''.
1.5.4.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.7.2.5	05-Jan-2001	bouyer	Sync with HEAD
1.7.2.4	08-Dec-2000	bouyer	Sync with HEAD.
1.7.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.7.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.1	14-Nov-2000	bouyer	file fms.c was added on branch thorpej_scsipi on 2000-11-20 11:42:19 +0000
1.9.6.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.9.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.9.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.9.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.9.2.1	08-Oct-2001	nathanw	Catch up to -current.
1.18.4.1	22-Sep-2004	jmc	Pullup rev 1.19 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.18.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.18.2.3	14-Nov-2004	skrll	Sync with HEAD.
1.18.2.2	02-Nov-2004	skrll	Sync with HEAD.
1.18.2.1	24-Sep-2004	skrll	Sync with HEAD.
1.21.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.21.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.24.2.3	27-Oct-2007	yamt	sync with head.
1.24.2.2	03-Sep-2007	yamt	sync with head.
1.24.2.1	30-Dec-2006	yamt	sync with head.
1.25.8.1	03-Sep-2006	yamt	sync with head.
1.25.4.1	09-Sep-2006	rpaulo	sync with head
1.26.4.2	10-Dec-2006	yamt	sync with head.
1.26.4.1	22-Oct-2006	yamt	sync with head
1.26.2.1	18-Nov-2006	ad	Sync with head.
1.28.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.28.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.29.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.29.16.1	06-Nov-2007	matt	sync with HEAD
1.29.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.29.2.1	23-Oct-2007	ad	Sync with head.
1.30.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.30.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.32.4.3	11-Mar-2010	yamt	sync with head
1.32.4.2	16-May-2009	yamt	sync with head
1.32.4.1	16-May-2008	yamt	sync with head.
1.32.2.1	18-May-2008	yamt	sync with head.
1.33.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.33.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.37.4.1	05-Mar-2011	rmind	sync with head
1.38.10.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.38.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.38.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.38.8.2	30-Oct-2012	yamt	sync with head
1.38.8.1	17-Apr-2012	yamt	sync with head
1.40.8.3	03-Dec-2017	jdolecek	update from HEAD
1.40.8.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.40.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.41.2.1	18-May-2014	rmind	sync with head
1.42.6.1	28-Aug-2017	skrll	Sync with HEAD
1.43.10.1	10-Jun-2019	christos	Sync with HEAD
1.43.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.45.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.45.2.2	28-Apr-2019	isaki	Adapt to audio2. - Searching nearest frequency is unnecessary in audio2.
1.45.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.47.12.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.48.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.49.6.1	14-Jan-2024	martin	Pull up following revision(s) (requested by chs in ticket #542): sys/dev/pci/yds.c: revision 1.69 sys/dev/pci/cmpci.c: revision 1.61 sys/dev/pci/sv.c: revision 1.62 sys/dev/pci/fms.c: revision 1.50 mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.7	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.6	23-Nov-2011	jmcneill	branches: 1.6.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.5	06-May-2009	cegger	branches: 1.5.12; 1.5.14; struct device * -> device_t, no functional changes intended.
1.4	28-Apr-2008	martin	branches: 1.4.12; 1.4.14; Remove clause 3 and 4 from TNF licenses
1.3	11-Dec-2005	christos	branches: 1.3.28; 1.3.70; 1.3.72; 1.3.74; merge ktrace-lwp.
1.2	15-Jan-2005	kent	ansify and KNF
1.1	01-Nov-1999	augustss	branches: 1.1.4; 1.1.6; 1.1.12; 1.1.36; Add a missing file for the fms driver.
1.1.36.1	17-Jan-2005	skrll	Sync with HEAD.
1.1.12.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.12.1	01-Nov-1999	bouyer	file fmsvar.h was added on branch thorpej_scsipi on 2000-11-20 11:42:19 +0000
1.1.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.4.2	15-Nov-1999	fvdl	Sync with -current
1.1.4.1	01-Nov-1999	fvdl	file fmsvar.h was added on branch fvdl-softdep on 1999-11-15 00:41:00 +0000
1.3.74.2	16-May-2009	yamt	sync with head
1.3.74.1	16-May-2008	yamt	sync with head.
1.3.72.1	18-May-2008	yamt	sync with head.
1.3.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.3.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.4.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.4.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.5.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.5.12.2	30-Oct-2012	yamt	sync with head
1.5.12.1	17-Apr-2012	yamt	sync with head
1.6.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.19	29-Mar-2010	kiyohara	Remove obsolete file.
1.18	26-Nov-2009	njoly	branches: 1.18.2; 1.18.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.17	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.16	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.15	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.14	28-Apr-2008	martin	branches: 1.14.14; Remove clause 3 and 4 from TNF licenses
1.13	10-Apr-2008	cegger	branches: 1.13.2; 1.13.4; use aprint_*_dev and device_xname
1.12	19-Oct-2007	ad	branches: 1.12.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	11-Dec-2005	christos	branches: 1.11.30; 1.11.44; 1.11.46; 1.11.50; merge ktrace-lwp.
1.10	27-Feb-2005	perry	branches: 1.10.4; nuke trailing whitespace
1.9	04-Feb-2005	perry	de-__P
1.8	23-Apr-2004	itojun	branches: 1.8.4; 1.8.6; pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.7	31-Jan-2003	thorpej	branches: 1.7.2; Use aprint_*().
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.2	13-Nov-2001	lukem	add RCSID
1.1	06-Mar-2001	matt	branches: 1.1.2; 1.1.4; 1.1.6; Add the PCI frontend for the TI LYNX firewire chip. Note that the backend still needs to be worked on.
1.1.6.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.5	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.4	08-Jan-2002	nathanw	Catch up to -current.
1.1.4.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.4.2	09-Apr-2001	nathanw	Catch up with -current.
1.1.4.1	06-Mar-2001	nathanw	file fwlynx_pci.c was added on branch nathanw_sa on 2001-04-09 01:56:59 +0000
1.1.2.2	12-Mar-2001	bouyer	Sync with HEAD.
1.1.2.1	06-Mar-2001	bouyer	file fwlynx_pci.c was added on branch thorpej_scsipi on 2001-03-12 13:31:06 +0000
1.7.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.7.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.6.1	12-Feb-2005	yamt	sync with head.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.4.1	27-Oct-2007	yamt	sync with head.
1.11.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.46.1	06-Nov-2007	matt	sync with HEAD
1.11.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.30.1	23-Oct-2007	ad	Sync with head.
1.12.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.13.4.4	11-Aug-2010	yamt	sync with head.
1.13.4.3	11-Mar-2010	yamt	sync with head
1.13.4.2	16-May-2009	yamt	sync with head
1.13.4.1	16-May-2008	yamt	sync with head.
1.13.2.1	18-May-2008	yamt	sync with head.
1.14.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.4.1	30-May-2010	rmind	sync with head
1.18.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.46	31-Mar-2018	sevan	There was only one G3 PowerBook with Firewire support, the Pismo. Issue is not exclusive to the Pismo, confirmed on a G4 Titanium PowerBook. Chip is a Texas Instruments TSB41AB1.
1.45	31-Mar-2018	sevan	PG3_FW device has been renamed UNINORTH_FW in pcidevs.
1.44	10-May-2017	msaitoh	branches: 1.44.8; Use pci_intr_establish_xname().
1.43	07-Jul-2016	msaitoh	branches: 1.43.8; KNF. Remove extra spaces. No functional change.
1.42	29-Mar-2014	christos	branches: 1.42.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.41	04-Aug-2012	riastradh	branches: 1.41.2; 1.41.4; Fix error branches and config pending races in firewire init. This way, if anything fails, it just fails; you don't panic. This can happen if suspending and resuming of firewire is broken (e.g., as I encountered in PR kern/44581).
1.40	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.39	29-Apr-2010	kiyohara	branches: 1.39.8; 1.39.12; Adds hw quirk for SUN firewire controllers. From FreeBSD. Thanks for cegger@.
1.38	24-Apr-2010	kiyohara	Support detach by drvctl(8).
1.37	29-Mar-2010	kiyohara	Bye-bye fw_port.h.
1.36	24-Feb-2010	dyoung	branches: 1.36.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.35	03-Feb-2010	macallan	branches: 1.35.2; do not attach to Powrbook G3 onboard Firewire controllers until someone with access to the hardware can figure out why it deadlocks during device discovery
1.34	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	28-Apr-2008	martin	branches: 1.32.10; 1.32.14; Remove clause 3 and 4 from TNF licenses
1.31	29-Mar-2008	kiyohara	branches: 1.31.2; 1.31.4; Sorry forgot. Split device_t/softc, and cosmetic change.
1.30	29-Mar-2008	ad	Make it compile again.
1.29	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.28	09-Dec-2007	jmcneill	branches: 1.28.6; 1.28.10; Merge jmcneill-pm branch.
1.27	06-Nov-2007	kiyohara	branches: 1.27.2; 1.27.4; The argument of *_intr_establish() is changed to fwohci_flit. Our ieee1394 coordinates from FreeBSD implementation. Add fwohci_pci_shutdown() and fwohci_pci_power(). Move __KERNEL_RCSID to ahead.
1.26	19-Oct-2007	ad	branches: 1.26.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.25	16-Nov-2006	christos	branches: 1.25.8; 1.25.22; 1.25.24; 1.25.28; __unused removal on arguments; approved by core.
1.24	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.23	11-Dec-2005	christos	branches: 1.23.20; 1.23.22; merge ktrace-lwp.
1.22	11-Jul-2005	kiyohara	ieee1394 import from FreeBSD.
1.21	27-Feb-2005	perry	branches: 1.21.4; nuke trailing whitespace
1.20	04-Feb-2005	perry	de-__P
1.19	23-Apr-2004	itojun	branches: 1.19.4; 1.19.6; pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.18	31-Jan-2003	thorpej	branches: 1.18.2; Use aprint_*().
1.17	04-Dec-2002	haya	New Feature: add pseudo device for IEEE 1394 isochronous stream and isochronous reception routine for IEEE 1394 OHCI (fwohci). The transmission part is under construction. The minimum configuration options for this feature are: # IEEE 1394 (i.LINK) fwohci* at pci? dev ? function ? pseudo-device fwiso 1
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	26-Jan-2002	ichiro	OCHI -> OHCI
1.12	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.11	13-Nov-2001	lukem	add RCSID
1.10	01-May-2001	jmc	branches: 1.10.2; Change to IPL_BIO
1.9	12-Apr-2001	thorpej	Use IPL_NET for now -- the only use of 1394 right now is for network interfaces (IPv{4,6}-over-1394), and this allows us to get rid of some evil interrupt mixing issues in the 1394 network code right now.
1.8	15-Mar-2001	enami	Make fwohci works on big endian cpu (e.g., powermac g4).
1.7	15-Mar-2001	enami	Fix indentation and white space usage.
1.6	28-Dec-2000	sommerfeld	branches: 1.6.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.5	13-Dec-2000	enami	Add RCS Id.
1.4	05-Nov-2000	matt	branches: 1.4.2; Merge in my some of my stuff.
1.3	05-Nov-2000	onoe	First Prototype implementation of network interface part for IEEE1394 (if_fw). Current status: Only OHCI chip is supported (fwohci). ping (IPv4) works with Sony's implementation (SmartConnect) on Win98. sometimes works but not stable. Not implemented yet: IRM (Isochronous Resource Manager) functionality. Link layer fragmentation. Topology map. More to do: clean ups MCAP charactor device part dhcp There is no entry in GENERIC config file yet. Follow sys/dev/ieee1394/IMPLEMENTATION to enable if_fw.
1.2	30-May-2000	augustss	branches: 1.2.4; Make this compile too.
1.1	30-May-2000	matt	Beginning of a IEEE 1394 framework. An attachment for PCI OHCI controllers and bus-independent module that just begins to print things out. No real code behind it. THIS IS A WORK IN PROGRESS. The *reg.h are woefully incomplete.
1.2.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.2.4.1	30-May-2000	minoura	file fwohci_pci.c was added on branch minoura-xpg4dl on 2000-06-22 17:07:23 +0000
1.4.2.7	21-Apr-2001	bouyer	Sync with HEAD
1.4.2.6	27-Mar-2001	bouyer	Sync with HEAD.
1.4.2.5	05-Jan-2001	bouyer	Sync with HEAD
1.4.2.4	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.4.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.4.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	05-Nov-2000	bouyer	file fwohci_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:20 +0000
1.6.2.7	11-Dec-2002	thorpej	Sync with HEAD.
1.6.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.6.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.6.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.6.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.6.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.6.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.10.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.10.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.10.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.18.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.18.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.18.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.18.2.1	03-Aug-2004	skrll	Sync with HEAD
1.19.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.19.6.1	12-Feb-2005	yamt	sync with head.
1.19.4.1	29-Apr-2005	kent	sync with -current
1.21.4.6	17-Mar-2008	yamt	sync with head.
1.21.4.5	21-Jan-2008	yamt	sync with head
1.21.4.4	15-Nov-2007	yamt	sync with head.
1.21.4.3	27-Oct-2007	yamt	sync with head.
1.21.4.2	30-Dec-2006	yamt	sync with head.
1.21.4.1	21-Jun-2006	yamt	sync with head.
1.23.22.2	10-Dec-2006	yamt	sync with head.
1.23.22.1	22-Oct-2006	yamt	sync with head
1.23.20.1	18-Nov-2006	ad	Sync with head.
1.25.28.2	13-Nov-2007	bouyer	Sync with HEAD
1.25.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.25.24.4	23-Mar-2008	matt	sync with HEAD
1.25.24.3	09-Jan-2008	matt	sync with HEAD
1.25.24.2	08-Nov-2007	matt	sync with -HEAD
1.25.24.1	06-Nov-2007	matt	sync with HEAD
1.25.22.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.25.22.4	01-Dec-2007	jmcneill	If fwohci happens to fail to attach, register a NULL power handler.
1.25.22.3	14-Nov-2007	joerg	Convert to suspend/resume framework. Not fully functional, but a start.
1.25.22.2	06-Nov-2007	joerg	Sync with HEAD.
1.25.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.25.8.1	23-Oct-2007	ad	Sync with head.
1.26.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.26.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.27.4.1	11-Dec-2007	yamt	sync with head.
1.27.2.1	26-Dec-2007	ad	Sync with head.
1.28.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.28.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.28.6.1	24-Mar-2008	keiichi	sync with head.
1.31.4.4	11-Aug-2010	yamt	sync with head.
1.31.4.3	11-Mar-2010	yamt	sync with head
1.31.4.2	16-May-2009	yamt	sync with head
1.31.4.1	16-May-2008	yamt	sync with head.
1.31.2.1	18-May-2008	yamt	sync with head.
1.32.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.32.10.1	21-Nov-2010	riz	Pull up following revision(s) (requested by macallan in ticket #1470): sys/dev/pci/fwohci_pci.c: revision 1.35 do not attach to Powrbook G3 onboard Firewire controllers until =20 someone with access to the hardware can figure out why it deadlocks during device discovery
1.35.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.36.2.1	30-May-2010	rmind	sync with head
1.39.12.1	18-Feb-2012	mrg	merge to -current.
1.39.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.39.8.2	30-Oct-2012	yamt	sync with head
1.39.8.1	17-Apr-2012	yamt	sync with head
1.41.4.1	18-May-2014	rmind	sync with head
1.41.2.2	03-Dec-2017	jdolecek	update from HEAD
1.41.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.42.6.2	28-Aug-2017	skrll	Sync with HEAD
1.42.6.1	09-Jul-2016	skrll	Sync with HEAD
1.43.8.1	11-May-2017	pgoyette	Sync with HEAD
1.44.8.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.20	07-Feb-2024	msaitoh	Remove ryo@'s mail addresses.
1.19	29-Feb-2020	isaki	round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4.
1.18	08-Jun-2019	isaki	branches: 1.18.2; 1.18.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.17	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	09-Dec-2018	jdolecek	branches: 1.16.2; use pci_intr_establish_xname() everywhere
1.15	01-Jun-2017	chs	branches: 1.15.8; 1.15.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.14	29-Mar-2014	christos	branches: 1.14.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.13	31-Jan-2014	ryo	fix a problem to set sample rate of AC97 codec.
1.12	02-Mar-2012	nonaka	branches: 1.12.2; 1.12.4; convert to device_t.
1.11	25-Nov-2011	jmcneill	branches: 1.11.2; remove an unnecessary call to vtbl->unlock
1.10	25-Nov-2011	jmcneill	need to hold thread lock when using AC97_IS_* macros
1.9	24-Nov-2011	mrg	put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.8	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.7	09-Jun-2011	joerg	branches: 1.7.2; 1.7.4; Use a format string when printing strings
1.6	13-Nov-2010	uebayasi	branches: 1.6.6; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.5	24-Feb-2010	dyoung	branches: 1.5.2; 1.5.4; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.4	08-Jan-2010	dyoung	branches: 1.4.2; Expand PMF_FN_* macros.
1.3	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.2	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.1	28-Dec-2008	jmcneill	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.1.10.4	11-Mar-2010	yamt	sync with head
1.1.10.3	16-May-2009	yamt	sync with head
1.1.10.2	04-May-2009	yamt	sync with head.
1.1.10.1	28-Dec-2008	yamt	file gcscaudio.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:55 +0000
1.1.8.2	01-May-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #260): sys/dev/pci/files.pci: revision 1.309 sys/dev/pci/gcscaudio.c: revision 1.1 sys/dev/pci/gcscaudioreg.h: revision 1.1 PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.1.8.1	28-Dec-2008	snj	file gcscaudio.c was added on branch netbsd-5 on 2009-05-01 01:19:09 +0000
1.1.6.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.4.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.4.1	28-Dec-2008	skrll	file gcscaudio.c was added on branch nick-hppapmap on 2009-01-19 13:18:25 +0000
1.1.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.2.1	28-Dec-2008	mjf	file gcscaudio.c was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.4.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.5.4.2	21-Apr-2010	matt	sync to netbsd-5
1.5.4.1	24-Feb-2010	matt	file gcscaudio.c was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.5.2.2	12-Jun-2011	rmind	sync with head
1.5.2.1	05-Mar-2011	rmind	sync with head
1.6.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.7.4.4	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.7.4.3	20-Nov-2011	jmcneill	need to hold intr_lock when calling audio_pint/audio_rint
1.7.4.2	19-Nov-2011	jmcneill	add gcscaudio_get_locks()
1.7.4.1	19-Nov-2011	jmcneill	adapt to audiomp api changes
1.7.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.2.1	17-Apr-2012	yamt	sync with head
1.11.2.3	06-Mar-2012	mrg	sync to -current
1.11.2.2	06-Mar-2012	mrg	sync to -current
1.11.2.1	04-Mar-2012	mrg	sync to latest -current.
1.12.4.1	18-May-2014	rmind	sync with head
1.12.2.2	03-Dec-2017	jdolecek	update from HEAD
1.12.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.14.6.1	28-Aug-2017	skrll	Sync with HEAD
1.15.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.15.10.1	10-Jun-2019	christos	Sync with HEAD
1.15.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.16.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.16.2.3	04-May-2019	isaki	Remove obsoleted methods in audio_hw_if. - drain: is handled in audio upper layer now. - mappage: is handled in audio upper layer now. - setfd: no one uses and it's meaningless now.
1.16.2.2	28-Apr-2019	isaki	Adapt to audio2.
1.16.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.18.4.1	29-Feb-2020	ad	Sync with head.
1.18.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by isaki in ticket #792): sys/arch/amiga/dev/aucc.c: revision 1.48 sys/dev/pci/auixp.c: revision 1.49 sys/dev/pci/gcscaudio.c: revision 1.19 sys/dev/pci/auich.c: revision 1.159 sys/dev/sbus/dbri.c: revision 1.42 sys/dev/pci/auvia.c: revision 1.85 sys/dev/pci/auacer.c: revision 1.39 Drop 512 bytes limit on auvia_round_blocksize(). This fixes attach on 6 channels device. PR kern/55017. round_blocksize must return a multiple of the framesize. aucc(4) supports 3 channels mode. round_blocksize must return a multiple of the framesize even if passed blocksize is greater than the upper limit. round_blocksize must return a multiple of the framesize. It's not divisible when blk=GCSCAUDI_PRD_SIZE_MAX and channels=4. round_blocksize must return a multiple of the framesize even if 6 channels mode. I believe that keeping "good alignment" is just a wish, not constraint.
1.2	07-Feb-2024	msaitoh	Remove ryo@'s mail addresses.
1.1	28-Dec-2008	jmcneill	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; 1.1.16; PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.1.16.2	21-Apr-2010	matt	sync to netbsd-5
1.1.16.1	28-Dec-2008	matt	file gcscaudioreg.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.1.10.2	04-May-2009	yamt	sync with head.
1.1.10.1	28-Dec-2008	yamt	file gcscaudioreg.h was added on branch yamt-nfs-mp on 2009-05-04 08:12:55 +0000
1.1.8.2	01-May-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #260): sys/dev/pci/files.pci: revision 1.309 sys/dev/pci/gcscaudio.c: revision 1.1 sys/dev/pci/gcscaudioreg.h: revision 1.1 PR# port-i386/40284: add AMD Geode CS5536 audio driver; written by SHIMIZU Ryo <ryo@nerv.org>
1.1.8.1	28-Dec-2008	snj	file gcscaudioreg.h was added on branch netbsd-5 on 2009-05-01 01:19:09 +0000
1.1.4.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.4.1	28-Dec-2008	skrll	file gcscaudioreg.h was added on branch nick-hppapmap on 2009-01-19 13:18:25 +0000
1.1.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.2.1	28-Dec-2008	mjf	file gcscaudioreg.h was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.45	02-Oct-2022	rin	PR kern/57035 s/lower/higher/ in comment; for little endian, 2nd word is higher both in numeric and address. It seems that I'm too absorbed into big endian machines these days ;) No binary changes.
1.44	01-Oct-2022	rin	Use appropriate macros. No binary changes for GCC/amd64.
1.43	01-Oct-2022	rin	Remove confusing braces. No binary changes.
1.42	01-Oct-2022	rin	trailing whitespace
1.41	01-Oct-2022	rin	PR kern/57035 genfb(4): Fix blank screen after switching to WSDISPLAYIO_MODE_EMUL for some GPU models like Ice Lake integrated GPU: - Restore lower words of 64-bit BARs correctly. - Do not disable expansion ROM if exists. Thanks ryoon@ for testing!
1.40	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.39	24-Apr-2021	thorpej	branches: 1.39.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.38	07-Jul-2016	msaitoh	branches: 1.38.34; KNF. Remove extra spaces. No functional change.
1.37	24-Jul-2014	riastradh	branches: 1.37.4; Add enable/disable polling hooks to genfb_ops.
1.36	27-Jan-2014	macallan	branches: 1.36.2; back out 1.35 since it completely breaks the mechanism for wsdisplay drivers to determine wether to become console or not. See http://mail-index.netbsd.org/tech-kern/2014/01/15/msg016421.html for fixing this properly.
1.35	16-Jan-2014	jakllsch	Fix PR kern/46376 with Nat Sloss's patch (with slight modification). Serial console now works on x86 with genfb enabled.
1.34	13-Mar-2012	elad	branches: 1.34.2; Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.33	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.32	08-Mar-2011	macallan	branches: 1.32.4; 1.32.8; only try to map the framebuffer if we don't already have an address With this radeondrm works with genfb and r6xx.
1.31	13-Feb-2011	phx	Determine BAR type with pci_mapreg_probe() instead of pci_mapreg_type(), to skip unimplemented BARs. Do not probe non-BAR registers (like subsystem id). This fixes potential panics in pci_mem_find().
1.30	10-Feb-2011	jmcneill	add support for 64-bit BARs
1.29	10-Feb-2011	jmcneill	Use BUS_SPACE_MAP_PREFETCHABLE when mapping the framebuffer.
1.28	09-Feb-2011	jmcneill	switch genfb to polling mode at shutdown
1.27	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.26	16-Dec-2010	cegger	branches: 1.26.2; 1.26.4; ioctl: KNF switch-case, remove useless brackets
1.25	07-Oct-2010	macallan	finish device_t-ification
1.24	25-Feb-2010	macallan	branches: 1.24.2; add a comment on PCI_MAGIC_IO_RANGE
1.23	24-Aug-2009	jmcneill	branches: 1.23.2; Move softc definition to genfb_pcivar.h
1.22	24-Aug-2009	jmcneill	Expose genfb_restore_palette and allow MD code to specify PMF callbacks.
1.21	24-Aug-2009	jmcneill	Register NULL pmf handlers, and note that firmware & MD code is responsible for restoring the display device.
1.20	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.19	06-May-2009	elad	Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.18	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.17	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.16	23-Feb-2009	jmcneill	Turn "bogus parameters" aprint_error into aprint_debug, and change the message to say that the framebuffer was not configured by the firmware.
1.15	17-Feb-2009	jmcneill	Use aprint_*
1.14	16-Feb-2009	jmcneill	Provide callbacks so MD code can explicitly prevent genfb from matching; works around vga/genfb deathmatches.
1.13	15-Feb-2009	jmcneill	Introduce genfb_ops genfb_borrow callback to allow the bus frontend to lend mappings to drm.
1.12	14-Feb-2009	jmcneill	Provide a stub genfb_cnattach for md code to call to give a hint to the bus driver's match function to return a higher confidence (so it can beat out vga(4), for example). genfb needs-flag so md code can determine if the genfb driver is present.
1.11	05-May-2008	jmcneill	branches: 1.11.8; 1.11.14; Change autoconf message 'direct rendering on dev unsupported' to the more accurate 'drm at dev not configured'.
1.10	29-Apr-2008	martin	Convert to new 2 clause license
1.9	10-Apr-2008	cegger	branches: 1.9.2; 1.9.4; use aprint_*_dev and device_xname
1.8	01-Apr-2008	mrg	if genfb_pci can't map regs, don't panic() just return.
1.7	09-Mar-2008	phx	Do not attach genfb when the frame buffer is not flagged as being console (e.g. when a serial console is used). Approved by macallan.
1.6	27-Feb-2008	macallan	get rid of the last #ifdef macppc and use PCI_MAGIC_IO_RANGE to select at which offset we mmap PCI IO space
1.5	21-Dec-2007	macallan	branches: 1.5.2; 1.5.6; only enable the fb mapping at offset 0 after attempting to mmap page 0 This way we avoid a conflict with PCI resources on some Sun graphics cards which have firmware that likes to put memory resources at low addresses. For instance, if we have 1MB framebuffer and registers at 0x30000 we'll get the registers when mmap()ing 0x30000 until someone tries to mmap(0), then we'll get framebuffer memory for the entire 1st MB. This needs an ioctl() to switch between fb-at-0 and PCI ranges.
1.4	01-Dec-2007	ad	branches: 1.4.2; 1.4.6; Back out previous.
1.3	01-Dec-2007	ad	Noisy printfs.
1.2	19-Nov-2007	macallan	various fixes and additions: - deal with bogus parameters - just return instead of panic()ing - try to attach a drm - save and restore BARs in case X messes them up
1.1	10-Apr-2007	macallan	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.14; 1.1.16; 1.1.18; 1.1.22; 1.1.24; move genfb's PCI frontend to dev/pci
1.1.24.2	27-Dec-2007	mjf	Sync with HEAD.
1.1.24.1	08-Dec-2007	mjf	Sync with HEAD.
1.1.22.1	21-Nov-2007	bouyer	Sync with HEAD
1.1.18.5	17-Mar-2008	yamt	sync with head.
1.1.18.4	21-Jan-2008	yamt	sync with head
1.1.18.3	07-Dec-2007	yamt	sync with head
1.1.18.2	03-Sep-2007	yamt	sync with head.
1.1.18.1	10-Apr-2007	yamt	file genfb_pci.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:50 +0000
1.1.16.2	23-Mar-2008	matt	sync with HEAD
1.1.16.1	09-Jan-2008	matt	sync with HEAD
1.1.14.1	21-Nov-2007	joerg	Sync with HEAD.
1.1.8.2	11-Jul-2007	mjf	Sync with head.
1.1.8.1	10-Apr-2007	mjf	file genfb_pci.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:29 +0000
1.1.6.2	15-Apr-2007	yamt	sync with head.
1.1.6.1	10-Apr-2007	yamt	file genfb_pci.c was added on branch yamt-idlelwp on 2007-04-15 16:03:24 +0000
1.1.2.2	10-Apr-2007	ad	Sync with head.
1.1.2.1	10-Apr-2007	ad	file genfb_pci.c was added on branch vmlocking on 2007-04-10 13:24:23 +0000
1.4.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.4.2.1	26-Dec-2007	ad	Sync with head.
1.5.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.5.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.5.2.1	24-Mar-2008	keiichi	sync with head.
1.9.4.6	09-Oct-2010	yamt	sync with head
1.9.4.5	11-Mar-2010	yamt	sync with head
1.9.4.4	16-Sep-2009	yamt	sync with head
1.9.4.3	16-May-2009	yamt	sync with head
1.9.4.2	04-May-2009	yamt	sync with head.
1.9.4.1	16-May-2008	yamt	sync with head.
1.9.2.1	18-May-2008	yamt	sync with head.
1.11.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.11.8.1	03-Mar-2009	skrll	Sync with HEAD.
1.23.2.2	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.23.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.24.2.2	21-Apr-2011	rmind	sync with head
1.24.2.1	05-Mar-2011	rmind	sync with head
1.26.4.2	17-Feb-2011	bouyer	Sync with HEAD
1.26.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.26.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.32.8.2	05-Apr-2012	mrg	sync to latest -current.
1.32.8.1	18-Feb-2012	mrg	merge to -current.
1.32.4.1	17-Apr-2012	yamt	sync with head
1.34.2.2	03-Dec-2017	jdolecek	update from HEAD
1.34.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.36.2.1	10-Aug-2014	tls	Rebase.
1.37.4.1	09-Jul-2016	skrll	Sync with HEAD
1.38.34.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.39.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	18-Oct-2011	dyoung	Add an #include for the genfb_softc type.
1.1	24-Aug-2009	jmcneill	branches: 1.1.2; Move softc definition to genfb_pcivar.h
1.1.2.2	16-Sep-2009	yamt	sync with head
1.1.2.1	24-Aug-2009	yamt	file genfb_pcivar.h was added on branch yamt-nfs-mp on 2009-09-16 13:37:50 +0000
1.25	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.24	31-Jul-2012	bouyer	branches: 1.24.2; 1.24.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.23	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.22	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.21	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.20	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.19	04-Apr-2011	dyoung	branches: 1.19.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.18	13-Nov-2010	uebayasi	branches: 1.18.2; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.17	05-Nov-2010	jakllsch	Pave the way for detachment of pciide(4)-family ATA controllers.
1.16	19-Oct-2009	bouyer	branches: 1.16.2; 1.16.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.15	18-Mar-2008	cube	branches: 1.15.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.14	09-Feb-2007	ad	branches: 1.14.20; 1.14.36; 1.14.40; Merge newlock2 to head.
1.13	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	06-Jul-2005	thorpej	branches: 1.10.2; - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.9	25-Jun-2005	fair	According to FreeBSD and other references, the cs5530 IDE controller, while capable of UDMA mode 2, is swamped if you actually go that fast, which is not good for the other functions on this multifunction southbridge chip, so limit UDMA to mode 1.
1.8	27-Feb-2005	perry	branches: 1.8.2; nuke trailing whitespace
1.7	21-Aug-2004	thorpej	branches: 1.7.4; 1.7.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.6	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.5	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.4	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.3	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.2	27-Jul-2004	rumble	branches: 1.2.2; Add support for the controller found on the AMD (previously NatSemi) Geode SC1100 controller (as found on the Soekris NET4801). The chip has two bugs: the first requires dword alignment, and the second cannot handle exact 64K transfers. Also, fix a few typos while we're here. Timings from FreeBSD. Reviewed by Manuel Bouyer.
1.1	09-Jul-2004	bouyer	branches: 1.1.2; Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.1.2.3	28-Jul-2004	tron	branches: 1.1.2.3.2; Pull up revision 1.2 (requested by rumble in ticket #718): Add support for the controller found on the AMD (previously NatSemi) Geode SC1100 controller (as found on the Soekris NET4801). The chip has two bugs: the first requires dword alignment, and the second cannot handle exact 64K transfers. Also, fix a few typos while we're here. Timings from FreeBSD. Reviewed by Manuel Bouyer.
1.1.2.2	14-Jul-2004	tron	Pull up revision 1.1 (requested by bouyer in ticket #644): Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.1.2.1	09-Jul-2004	tron	file geodeide.c was added on branch netbsd-2-0 on 2004-07-14 09:08:28 +0000
1.1.2.3.2.1	18-Jul-2005	riz	Pull up revision 1.10 (requested by thorpej in ticket #2072): - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.2.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.2.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.2.2.2	03-Aug-2004	skrll	Sync with HEAD
1.2.2.1	27-Jul-2004	skrll	file geodeide.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.7.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.4.1	29-Apr-2005	kent	sync with -current
1.8.2.2	12-Jul-2005	tron	Pull up revision 1.9 (requested by fair in ticket #555): According to FreeBSD and other references, the cs5530 IDE controller, while capable of UDMA mode 2, is swamped if you actually go that fast, which is not good for the other functions on this multifunction southbridge chip, so limit UDMA to mode 1.
1.8.2.1	06-Jul-2005	tron	Pull up revision 1.10 (requested by thorpej in ticket #553): - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.10.2.3	24-Mar-2008	yamt	sync with head.
1.10.2.2	26-Feb-2007	yamt	sync with head.
1.10.2.1	30-Dec-2006	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.14.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.14.36.1	24-Mar-2008	keiichi	sync with head.
1.14.20.1	23-Mar-2008	matt	sync with HEAD
1.15.4.1	11-Mar-2010	yamt	sync with head
1.16.4.2	21-Apr-2011	rmind	sync with head
1.16.4.1	05-Mar-2011	rmind	sync with head
1.16.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.18.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.19.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.19.4.1	30-Oct-2012	yamt	sync with head
1.24.4.1	18-May-2014	rmind	sync with head
1.24.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.24.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.29	14-Oct-2025	macallan	support hardware rendering for mono fonts
1.28	06-Oct-2025	macallan	add backlight control for FX 5200 Go
1.27	30-Aug-2025	macallan	add backlight control
1.26	30-Aug-2025	macallan	put some RAMDAC initialization back, use 8bit writes this time
1.25	29-Aug-2025	macallan	support acceleration on nv3x, aka GeForce FX 5200U and FX Go 5200
1.24	22-Jul-2025	macallan	enable font loading
1.23	21-Jul-2025	macallan	match GeForce FX 5200 Ultra
1.22	25-Sep-2022	thorpej	branches: 1.22.10; Remove unnecessary include of <sys/malloc.h>.
1.21	15-Aug-2022	macallan	nuke pasto in gffb_init()
1.20	10-Mar-2022	andvar	s/woud/would/ in comment.
1.19	10-Mar-2022	riastradh	gffb(4): Nix membar_sync and explain what's going on here.
1.18	10-Mar-2022	riastradh	gffb(4): Use bus_space_barrier, not membar_sync.
1.17	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.16	24-Apr-2021	thorpej	branches: 1.16.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.15	25-Feb-2021	macallan	branches: 1.15.2; match Geforce FX5200 Go found in some iMac G5 models
1.14	21-May-2020	macallan	branches: 1.14.2; prune unneeded includes
1.13	26-Jul-2018	macallan	first step to support chips newer than geforce 2MX
1.12	20-Jan-2017	maya	branches: 1.12.12; 1.12.14; Check pmf_device_register return value. NFC Appeases static analyzers.
1.11	11-Jul-2016	msaitoh	branches: 1.11.2; 1.11.4; KNF. No functional change.
1.10	16-Sep-2015	macallan	prop_dictionary_get_bool() doesn't return FALSE if the property doesn't exist, so initialize things first. No more console stealing on sparc64.
1.9	29-Jun-2014	tsutsui	branches: 1.9.4; 1.9.6; Appease a printf format warning on LP64 machines.
1.8	14-Mar-2014	mrg	branches: 1.8.2; 1.8.4; 1.8.6; remove various set-but-unused variables.
1.7	23-Oct-2013	macallan	fix copyright year
1.6	23-Oct-2013	macallan	minor cleanup: - don't map more PCI space than we have video memory - clarify some comments
1.5	09-Oct-2013	macallan	support WSDISPLAYIO_GET_FBINFO
1.4	09-Oct-2013	macallan	some cleanup: - use macros to shorten a while lot of bus_space_write_4() lines - update some comments - remove debug goop and cargo culted leftovers
1.3	09-Oct-2013	macallan	add hardware acceleration For now this supports only GeForce2 MX cards but most NV1x chips should work once their PCI iDs are added to gffb_match() Tested only on macppc.
1.2	02-Oct-2013	macallan	one step closer to graphics acceleration: - probe and report actual amount of video memory - add DMA machinery to feed commands to the graphics processor - adapt a whole bunch of magic number initialization from xf86-video-nv, now the chip actually reads commands from a buffer in video memory
1.1	18-Sep-2013	macallan	a preliminary driver for nvidia geforce graphics chips so far it only supports the GeForce 2MX, tested on macppc only no acceleration yet, just some DAC setup the main advantage over genfb is that this driver knows how to setup the palette registers for the 2nd output
1.8.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.8.6.1	14-Mar-2014	yamt	file gffb.c was added on branch yamt-pagecache on 2014-05-22 11:40:24 +0000
1.8.4.2	18-May-2014	rmind	sync with head
1.8.4.1	14-Mar-2014	rmind	file gffb.c was added on branch rmind-smpnet on 2014-05-18 17:45:39 +0000
1.8.2.1	10-Aug-2014	tls	Rebase.
1.9.6.3	05-Feb-2017	skrll	Sync with HEAD
1.9.6.2	05-Oct-2016	skrll	Sync with HEAD
1.9.6.1	22-Sep-2015	skrll	Sync with HEAD
1.9.4.3	03-Dec-2017	jdolecek	update from HEAD
1.9.4.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.4.1	29-Jun-2014	tls	file gffb.c was added on branch tls-maxphys on 2014-08-20 00:03:42 +0000
1.11.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.11.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.12.14.1	10-Jun-2019	christos	Sync with HEAD
1.12.12.1	28-Jul-2018	pgoyette	Sync with HEAD
1.14.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.15.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.16.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.22.10.1	02-Aug-2025	perseant	Sync with HEAD
1.6	06-Oct-2025	macallan	add backlight control for FX 5200 Go
1.5	16-Apr-2020	rin	Revert previous for now: http://mail-index.netbsd.org/source-changes/2020/04/16/msg116278.html The reasoning turned out to be wrong; __KERNEL_RCSID() in header files does not overwrite RCSID in main source files. The real problem is that it inserts its RCSID into every object files. However, it can be still useful even if heavily duplicated.
1.4	16-Apr-2020	rin	Stop using __KERNEL_RCSID() in header files; it confuses ident(1) by overwriting RCSID in main source files. XXX The first argument of __KERNEL_RCSID() is neglected for ELF. If we wish to have RCSID of header files in kernel binary, we need something like __FBSDID() macro in FreeBSD.
1.3	23-Oct-2013	macallan	branches: 1.3.4; 1.3.6; 1.3.10; 1.3.48; fix copyright year
1.2	02-Oct-2013	macallan	add a bunch more constants from xf86-video-nv
1.1	18-Sep-2013	macallan	a preliminary driver for nvidia geforce graphics chips so far it only supports the GeForce 2MX, tested on macppc only no acceleration yet, just some DAC setup the main advantage over genfb is that this driver knows how to setup the palette registers for the 2nd output
1.3.48.1	20-Apr-2020	bouyer	Sync with HEAD
1.3.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.3.10.1	23-Oct-2013	tls	file gffbreg.h was added on branch tls-maxphys on 2014-08-20 00:03:42 +0000
1.3.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.6.1	23-Oct-2013	yamt	file gffbreg.h was added on branch yamt-pagecache on 2014-05-22 11:40:24 +0000
1.3.4.2	18-May-2014	rmind	sync with head
1.3.4.1	23-Oct-2013	rmind	file gffbreg.h was added on branch rmind-smpnet on 2014-05-18 17:45:39 +0000
1.20	11-Jul-2016	msaitoh	KNF. No functional change.
1.19	27-Oct-2012	chs	branches: 1.19.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.18	30-Jan-2012	drochner	branches: 1.18.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.17	12-May-2009	cegger	branches: 1.17.12; 1.17.16; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.16	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.15	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.14	19-Oct-2007	ad	branches: 1.14.20; 1.14.34; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	29-Mar-2006	thorpej	branches: 1.11.8; 1.11.10; Use device_cfdata().
1.10	11-Dec-2005	christos	branches: 1.10.4; 1.10.6; 1.10.8; 1.10.10; 1.10.12; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.7	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.6	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.5	31-Jan-2003	thorpej	branches: 1.5.2; Use aprint_*().
1.4	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	03-Sep-2002	augustss	branches: 1.1.2; 1.1.4; Replace the mr driver with the gtp driver. From OpenBSD.
1.1.4.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.2	17-Sep-2002	nathanw	Catch up to -current.
1.1.4.1	03-Sep-2002	nathanw	file gtp.c was added on branch nathanw_sa on 2002-09-17 21:20:16 +0000
1.1.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.1.2.1	03-Sep-2002	jdolecek	file gtp.c was added on branch kqueue on 2002-09-06 08:45:13 +0000
1.5.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.5.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.9.2.3	27-Oct-2007	yamt	sync with head.
1.9.2.2	30-Dec-2006	yamt	sync with head.
1.9.2.1	21-Jun-2006	yamt	sync with head.
1.10.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.10.10.1	19-Apr-2006	elad	sync with head.
1.10.8.1	01-Apr-2006	yamt	sync with head.
1.10.6.1	22-Apr-2006	simonb	Sync with head.
1.10.4.1	09-Sep-2006	rpaulo	sync with head
1.11.10.2	10-Dec-2006	yamt	sync with head.
1.11.10.1	22-Oct-2006	yamt	sync with head
1.11.8.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.14.34.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.14.20.1	16-May-2009	yamt	sync with head
1.17.16.1	18-Feb-2012	mrg	merge to -current.
1.17.12.2	30-Oct-2012	yamt	sync with head
1.17.12.1	17-Apr-2012	yamt	sync with head
1.18.6.2	03-Dec-2017	jdolecek	update from HEAD
1.18.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.19.14.1	05-Oct-2016	skrll	Sync with HEAD
1.13	13-Sep-2022	msaitoh	Add Intel 600 series PCH-LP devices support.
1.12	21-Mar-2022	jmcneill	hdaudio: pci: PCIe config space fixes and Intel PCH snoop support The HD audio specification does not cover PCI config space, and this driver was unconditionally writing to a vendor specific register. Reduce scope of config space accesses based on PCI IDs. With this cleaned up, add support for Intel PCH devices which require some additional vendor specific configuration to bypass no snoop mode.
1.11	28-Oct-2021	msaitoh	Support devices that the PCI subclass is not PCI_SUBCLASS_MULTIMEDIA_HDAUDIO. - Not tested well. - If audioX is not attached, add HDAUDIO_ENABLE_DISPLAYPORT and HDAUDIO_ENABLE_HDMI to your kernel config.
1.10	12-Sep-2018	mrg	use pa_dmat64 if pci_dma64_available().
1.9	12-Sep-2018	mrg	use pci_mapreg_type() instead of assuming PCI_MAPREG_TYPE_MEM. suggested by riastradh@. fixes hdaudio on thinkpad P51 where the registers are mapped over 4GB range. probably fixes intel X10-based systems that map hdaudio high as well, but i was able to work around that in the BIOS. not so for the P51. XXX: pullup-7, pullup-8.
1.8	04-Jun-2017	pgoyette	branches: 1.8.4; 1.8.6; 1.8.8; Move the hdaudio_cd out of the hdaudio_pci module and into the main hdaudio module. Rearrange the inter-module dependencies accordingly. Should resolve (my) PR kern/49836 XXX Pull-ups needed for netbsd-8 branch
1.7	27-Apr-2017	msaitoh	branches: 1.7.2; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.6	31-Jan-2017	khorben	branches: 1.6.4; Always allow hdaudio(4) to suspend Sometimes hdaudio(4) can fail to attach, but the device remains in the tree. Even though the PMF registration is performed, it is explicitly de-registered in the code path for errors - therefore preventing suspending later. This patch makes sure PMF registration is performed accordingly upon errors as well. No objection on tech-kern@. Tested on a Lenovo ThinkPad T440s (amd64). XXX pull-up to netbsd-7
1.5	16-Dec-2016	nonaka	branches: 1.5.2; hdaudio(4): Use pci_intr_alloc(9)/pci_intr_release(9).
1.4	14-Jul-2016	msaitoh	branches: 1.4.2; KNF. No functional change.
1.3	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.2	05-Jul-2015	nonaka	Deregister pmf hook when attach failed.
1.1	28-Mar-2015	jmcneill	branches: 1.1.2; Split hdaudio and PCI attach glue. Even though the driver was written with this separation in mind, all of the code lived in sys/dev/pci/hdaudio. Move common parts to sys/dev/hdaudio and PCI attach glue to sys/dev/pci.
1.1.2.7	28-Aug-2017	skrll	Sync with HEAD
1.1.2.6	05-Feb-2017	skrll	Sync with HEAD
1.1.2.5	05-Oct-2016	skrll	Sync with HEAD
1.1.2.4	09-Jul-2016	skrll	Sync with HEAD
1.1.2.3	22-Sep-2015	skrll	Sync with HEAD
1.1.2.2	06-Apr-2015	skrll	Sync with HEAD
1.1.2.1	28-Mar-2015	skrll	file hdaudio_pci.c was added on branch nick-nhusb on 2015-04-06 15:18:10 +0000
1.4.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.4.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.5.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.6.4.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.7.2.2	23-Sep-2018	martin	Pull up following revision(s) (requested by mrg in ticket #1022): sys/dev/pci/hdaudio_pci.c: revision 1.9 sys/dev/pci/hdaudio_pci.c: revision 1.10 use pci_mapreg_type() instead of assuming PCI_MAPREG_TYPE_MEM. suggested by riastradh@. fixes hdaudio on thinkpad P51 where the registers are mapped over 4GB range. probably fixes intel X10-based systems that map hdaudio high as well, but i was able to work around that in the BIOS. not so for the P51. XXX: pullup-7, pullup-8. use pa_dmat64 if pci_dma64_available().
1.7.2.1	05-Jun-2017	snj	Pull up following revision(s) (requested by pgoyette in ticket #6): sys/dev/hdaudio/hdaudio.c: revision 1.5 sys/dev/pci/hdaudio_pci.c: revision 1.8 Move the hdaudio_cd out of the hdaudio_pci module and into the main hdaudio module. Rearrange the inter-module dependencies accordingly. Should resolve (my) PR kern/49836
1.8.8.1	10-Jun-2019	christos	Sync with HEAD
1.8.6.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.8.4.2	03-Dec-2017	jdolecek	update from HEAD
1.8.4.1	04-Jun-2017	jdolecek	file hdaudio_pci.c was added on branch tls-maxphys on 2017-12-03 11:37:07 +0000
1.2	21-Mar-2022	jmcneill	hdaudio: pci: PCIe config space fixes and Intel PCH snoop support The HD audio specification does not cover PCI config space, and this driver was unconditionally writing to a vendor specific register. Reduce scope of config space accesses based on PCI IDs. With this cleaned up, add support for Intel PCH devices which require some additional vendor specific configuration to bypass no snoop mode.
1.1	28-Mar-2015	jmcneill	branches: 1.1.2; 1.1.18; Split hdaudio and PCI attach glue. Even though the driver was written with this separation in mind, all of the code lived in sys/dev/pci/hdaudio. Move common parts to sys/dev/hdaudio and PCI attach glue to sys/dev/pci.
1.1.18.2	03-Dec-2017	jdolecek	update from HEAD
1.1.18.1	28-Mar-2015	jdolecek	file hdaudio_pci.h was added on branch tls-maxphys on 2017-12-03 11:37:07 +0000
1.1.2.2	06-Apr-2015	skrll	Sync with HEAD
1.1.2.1	28-Mar-2015	skrll	file hdaudio_pci.h was added on branch nick-nhusb on 2015-04-06 15:18:10 +0000
1.82	04-Aug-2023	riastradh	entropy(9): Simplify stages. Split interrupt vs non-interrupt paths. - Nix the entropy stage (cold, warm, hot). Just use the usual kernel `cold' (cold: single-core, single-thread; interrupts may happen), and don't make any three-way distinction about whether interrupts or threads or other CPUs can be running. Instead, while cold, use splhigh/splx or forbid paths to come from interrupt context, and while warm, use mutex or the per-CPU hard and soft interrupt paths for low latency. This comes at a small cost to some interrupt latency, since we may stir the pool in interrupt context -- but only for a very short window early at boot between configure and configure2, so it's hard to imagine it matters much. - Allow rnd_add_uint32 to run in hard interrupt context or with spin locks held, but defer processing to softint and drop samples on the floor if buffer is full. This is mainly used for cheaply tossing samples from drivers for non-HWRNG devices into the entropy pool, so it is often used from interrupt context and/or under spin locks. - New rnd_add_data_intr provides the interrupt-like data entry path for arbitrary buffers and driver-specified entropy estimates: defer processing to softint and drop samples on the floor if buffer is full. - Document that rnd_add_data is forbidden under spin locks outside interrupt context (will crash in LOCKDEBUG), and inadvisable in interrupt context (but technically permitted just in case there are compatibility issues for now); later we can forbid it altogether in interrupt context or under spin locks. - Audit all uses of rnd_add_data to use rnd_add_data_intr where it might be used in interrupt context or under a spin lock. This fixes a regression from last year when the global entropy lock was changed from IPL_VM (spin) to IPL_SOFTSERIAL (adaptive). Thought I'd caught all the problems from that, but another one bit three different people this week, presumably because of recent changes that led to more non-HWRNG drivers entering the entropy consolidation path from rnd_add_uint32. In my attempt to preserve the rnd(9) API for the (now long-since abandoned) prospect of pullup to netbsd-9 in my rewrite of the entropy subsystem in 2020, I didn't introduce a separate entry point for entering entropy from interrupt context or equivalent, i.e., spin locks held, and instead made rnd_add_data rely on cpu_intr_p() to decide whether to process the whole sample under a lock or only take as much as there's buffer space for before scheduling a softint. In retrospect, that was a mistake (though perhaps not as much of a mistake as other entropy API decisions...), a mistake which is finally getting rectified now by rnd_add_data_intr. XXX pullup-10
1.81	06-May-2023	andvar	s/Regiser/Register/ and s/regester/register/ in comments.
1.80	22-May-2022	riastradh	branches: 1.80.4; opencrypto: Make freesession callback return void. No functional change intended: all drivers already return zero unconditionally.
1.79	22-May-2022	riastradh	hifn(4): Prune dead branches. Assert session id validity.
1.78	22-May-2022	riastradh	hifn(4): Make ERESTART return more obvious.
1.77	22-May-2022	riastradh	hifn(4): Nix dead code. crp and callback are guaranteed nonnull.
1.76	22-May-2022	riastradh	hifn(4): hifn_compression always returns zero. Prune dead branches.
1.75	22-May-2022	riastradh	hifn(4): Fix error branches to do crypto_done, not return error.
1.74	17-May-2020	riastradh	pool_cache_prime, not pool_cache_lowat, as suggested by nick.
1.73	17-May-2020	riastradh	Missed a spot.
1.72	17-May-2020	riastradh	Rework hifn(4) RNG logic. Previously we would read 34 bits of entropy into the pool and stop, never to try again. This change finishes the conversion to an on-demand source, although we still use a callout to (a) try again in another second if the 7811 FIFO was emptied, and (b) wait at least 0.4 sec after we start the RNG to read anything from it as the alleged documentation allegedly recommends.
1.71	17-May-2020	riastradh	Tweak locking and use a pool cache for commands and dmamaps. This is enough to get the crypto decelerator working in LOCKDEBUG; previously it would crash the moment you looked at it.
1.70	17-May-2020	riastradh	Merge updates from OpenBSD's hifn(4) crypto decelerator driver. Highlights: - Apply a ten-year-old fix to the driver's misuse of previously exposed ciphertext as IV for the next ciphertext with CBC. (Someone^TM should make sure there are no other drivers still doing this...) - Various zeroing of buffers. - Various minor cosmetic changes.
1.69	17-May-2020	riastradh	Remove bogus comment.
1.68	17-May-2020	riastradh	De-#ifdef __OpenBSD__ify hifn7751.c. Not really helpful; we can just use diff3 to merge updates.
1.67	30-Apr-2020	riastradh	rnd_attach_source calls the callback itself now. No need for every driver to explicitly call it to prime the pool. Eliminate now-unused <sys/rndpool.h>.
1.66	30-Apr-2020	riastradh	Don't attach rndsource until it's actually ready to run.
1.65	29-Feb-2020	mlelstv	Fix signed/unsigned comparisons. Send data little endian.
1.64	18-Feb-2020	msaitoh	No functional change: - u_int_t -> uint_t - whitespace fixes. - KNF.
1.63	27-Dec-2018	maxv	branches: 1.63.6; Remove M_COPY_PKTHDR, M_MOVE_PKTHDR, M_ALIGN and MH_ALIGN.
1.62	05-Oct-2018	msaitoh	Use pci_intr_establish_xname().
1.61	14-May-2015	maxv	branches: 1.61.16; 1.61.18; Use-after-free. Found by Brainy. ok christos@
1.60	09-May-2015	riastradh	Defer to a callout in hifn_rng_get. Fixes lock-against-self. Still can't call rnd_add_data directly from an rndsource callback. (Still plan to fix this with further rnd rototilling.) XXX pullup netbsd-7
1.59	09-May-2015	mrg	initialise sc_mtx slightly earlier, before it is used on systems with rng. fixes problem reported by msaitoh@.
1.58	27-Apr-2015	christos	fix mbuf leak on failure (Brainy)
1.57	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.56	10-Aug-2014	tls	branches: 1.56.2; 1.56.4; Merge tls-earlyentropy branch into HEAD.
1.55	03-Jun-2014	msaitoh	Avoid buffer overflow which was added in rev. 1.52. This change fix a panic at boot time (PR#47950 Reported by Bert Kiers). OK'd by tls.
1.54	29-Mar-2014	christos	branches: 1.54.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.53	03-Jan-2014	pgoyette	Modularize
1.52	13-Jun-2013	tls	branches: 1.52.2; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.51	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.50	02-Feb-2012	tls	branches: 1.50.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.49	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.48	29-Nov-2011	tls	branches: 1.48.2; Remove rnd_extract_data from the public kernel API (it is for use by the stream generators only). Clean up some related minor issues.
1.47	19-Nov-2011	tls	First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.46	13-Nov-2010	uebayasi	branches: 1.46.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.45	16-Jun-2010	hubertf	Make debug message print more than the device name OK mbalmer, jmcneill
1.44	01-Feb-2010	hubertf	branches: 1.44.2; 1.44.4; Use uppercase for acronyms on output: DRAM, SRAM
1.43	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.42	28-Apr-2009	dyoung	Use device_t, cfdata_t, and device_private().
1.41	18-Apr-2009	tsutsui	Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.40	18-Mar-2009	cegger	bcopy -> memcpy
1.39	18-Mar-2009	cegger	bzero -> memset
1.38	18-Mar-2009	cegger	bcmp -> memcmp
1.37	10-Apr-2008	cegger	branches: 1.37.4; 1.37.12; 1.37.18; use aprint_*_dev and device_xname
1.36	02-Feb-2008	tls	branches: 1.36.6; From Darran Hunt at Coyote Point: don't truncate HMAC to 96 bits unless actually asked to. Fixed in FreeBSD a while ago, discussed on tech-kern and tech-crypto.
1.35	09-Jul-2007	ad	branches: 1.35.8; 1.35.14; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.34	04-Mar-2007	christos	branches: 1.34.2; 1.34.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.33	16-Nov-2006	christos	branches: 1.33.4; __unused removal on arguments; approved by core.
1.32	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.31	30-Aug-2006	christos	branches: 1.31.2; 1.31.4; static goes first
1.30	28-Mar-2006	thorpej	Use device_unit().
1.29	25-Nov-2005	thorpej	branches: 1.29.4; 1.29.6; 1.29.8; 1.29.10; 1.29.12; - De-couple the software crypto implementation from the rest of the framework. There is no need to waste the space if you are only using algoritms provided by hardware accelerators. To get the software implementations, add "pseudo-device swcr" to your kernel config. - Lazily initialize the opencrypto framework when crypto drivers (either hardware or swcr) register themselves with the framework.
1.28	16-Oct-2005	tls	branches: 1.28.6; From Gilles Roy <groy@qnx.com>: don't increment buffer pointer twice (once before callback, once in callback) for MAC transforms. How did this driver ever work at all without this?
1.27	16-Oct-2005	tls	HAVE_CRYPTO_LZS, not HAVE_CRYPTO_LSZ (worse, it was inconsistently used)
1.26	16-Oct-2005	tls	Don't use rnd_extract_data to set up IVs for new sessions; it is very expensive, and pointless. As elsewhere in the kernel (and as approved under FIPS-140-2 by multiple test labs, incidentally) we use arc4 to generate IVs here; there is no benefit to their being cryptographically strong so long as there is a sufficient Hamming distance between them.
1.25	15-Oct-2005	tls	Add support for the hardware random number generator on the Hifn 7811 and 795x. This was present in the driver before, but disabled due to problems with the actual randomness of generated numbers on the ubiquitious 7900-series parts. The code here is far, far more conservative than anyone else's driver for this RNG is -- but I believe that conservatism is called for, because the 79xx RNG design is missing a number of pieces from Hifn's "reference" 6500 RNG, and thus the numbers it generates must be treated with some care. Support for the 7811 RNG (which is a full-fledged 6500 type generator) is pretty much the same here as in other variants of this driver, except that it uses Hifn's "worst case" estimate of actual entropy per output bit, so it will accumulate bits much more slowly. The 7811 support is untested.
1.24	15-Oct-2005	tls	From OpenBSD hifn7751.c revision 1.149: enable the command-wait interrupt even if there's only one request in the queue. Sometimes we seem to miss multiple interrupts, and command-wait saves us.
1.23	28-Jun-2005	thorpej	branches: 1.23.2; Use ANSI function decls and static.
1.22	27-Feb-2005	perry	nuke trailing whitespace
1.21	29-Apr-2004	jonathan	branches: 1.21.4; 1.21.6; Identify AES-capable chips (7955, 7956) in the driver attach message.
1.20	16-Nov-2003	jonathan	branches: 1.20.2; Support symmetric crypto ops, including AES, on hifn 7955 and 7956. Not tested on 795[56], but compiles and matches FreeBSD-tested diffs. Suggested by Sam Leffler <sam@errno.com>, as imported into FreeBSD by Sam. Submitted to FreeBSD by Rajesh Vaidyanath <RVaidyanath@hifn.com>.
1.19	28-Aug-2003	thorpej	Remove an unnececssary PCI config read.
1.18	28-Aug-2003	thorpej	Hi/Fn is now spelled Hifn.
1.17	28-Aug-2003	thorpej	ANSI'ify (matches change in openbsd v1.140 of this file).
1.16	28-Aug-2003	thorpej	Print out a bit more information about which part is present at attach time, generally tidy up autoconfiguration messages.
1.15	30-Jul-2003	jonathan	Discard previous driver; it didnt really do anything. Replace with a new from-scratch port of recent OpenBSD hifn7751 source, with changes to the crypto-framework calls to fit Sam Leffler's rework. (Some of the attach-time changes from the previous driver were carried over). Many thanks to Quentin Garnier (cube@cubidou.net) for testing an pre-release of the driver and finding a couple of missing initializations. Note that RNG support is temporarily disabled, until a potential issue with RNG quality is investigated and resolved.
1.14	03-May-2003	wiz	branches: 1.14.2; DMA, not dma nor Dma.
1.13	31-Jan-2003	thorpej	Use aprint_*().
1.12	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.11	06-Jan-2003	wiz	interrupt with two rs.
1.10	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.9	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.8	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.7	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.6	13-Nov-2001	lukem	add RCSID
1.5	07-Jul-2001	thorpej	branches: 1.5.2; bcmp -> memcmp
1.4	07-Jul-2001	thorpej	bzero -> memset
1.3	07-Jul-2001	thorpej	bcopy -> memcpy
1.2	28-Dec-2000	sommerfeld	branches: 1.2.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.1	12-Oct-2000	itojun	branches: 1.1.2; hifn7751 crypto card driver. from openbsd. does nothing useful at this moment - initailize the chip, that's all. TODO: crypto logic framework in kernel (see openbsd sys/crypto) TODO: ipsec frontend (need major rework in ipsec tree - should start with busywait) TODO: character device frontend
1.1.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.1.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	12-Oct-2000	bouyer	file hifn7751.c was added on branch thorpej_scsipi on 2000-11-20 11:42:20 +0000
1.2.2.5	07-Jan-2003	thorpej	Sync with HEAD.
1.2.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.2.2.3	28-Feb-2002	nathanw	Catch up to -current.
1.2.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.5.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.5.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.5.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.14.2.6	11-Dec-2005	christos	Sync with head.
1.14.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.2.1	05-May-2004	tron	Pull up revision 1.21 (requested by jonathan in ticket #239): Identify AES-capable chips (7955, 7956) in the driver attach message.
1.21.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.23.2.4	04-Feb-2008	yamt	sync with head.
1.23.2.3	03-Sep-2007	yamt	sync with head.
1.23.2.2	30-Dec-2006	yamt	sync with head.
1.23.2.1	21-Jun-2006	yamt	sync with head.
1.28.6.1	29-Nov-2005	yamt	sync with head.
1.29.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.29.10.1	19-Apr-2006	elad	sync with head.
1.29.8.2	03-Sep-2006	yamt	sync with head.
1.29.8.1	01-Apr-2006	yamt	sync with head.
1.29.6.1	22-Apr-2006	simonb	Sync with head.
1.29.4.1	09-Sep-2006	rpaulo	sync with head
1.31.4.2	10-Dec-2006	yamt	sync with head.
1.31.4.1	22-Oct-2006	yamt	sync with head
1.31.2.1	18-Nov-2006	ad	Sync with head.
1.33.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.34.4.1	11-Jul-2007	mjf	Sync with head.
1.34.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.35.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.35.8.1	23-Mar-2008	matt	sync with HEAD
1.36.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.37.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.37.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.37.4.3	11-Aug-2010	yamt	sync with head.
1.37.4.2	11-Mar-2010	yamt	sync with head
1.37.4.1	04-May-2009	yamt	sync with head.
1.44.4.2	05-Mar-2011	rmind	sync with head
1.44.4.1	03-Jul-2010	rmind	sync with head
1.44.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.46.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.46.8.2	30-Oct-2012	yamt	sync with head
1.46.8.1	17-Apr-2012	yamt	sync with head
1.48.2.1	18-Feb-2012	mrg	merge to -current.
1.50.6.4	03-Dec-2017	jdolecek	update from HEAD
1.50.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.50.6.2	23-Jun-2013	tls	resync from head
1.50.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.52.2.1	18-May-2014	rmind	sync with head
1.54.2.2	10-Aug-2014	tls	Rebase.
1.54.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.56.4.1	06-Jun-2015	skrll	Sync with HEAD
1.56.2.3	15-May-2015	snj	Pull up following revision(s) (requested by maxv in ticket #763): sys/dev/pci/hifn7751.c: revisions 1.58, 1.61 fix mbuf leak on failure (Brainy) -- Use-after-free. Found by Brainy.
1.56.2.2	11-May-2015	msaitoh	Pull up following revision(s) (requested by mrg in ticket #750): sys/dev/pci/hifn7751.c: revision 1.59 initialise sc_mtx slightly earlier, before it is used on systems with rng. fixes problem reported by msaitoh@.
1.56.2.1	11-May-2015	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #748): sys/dev/pci/hifn7751.c: revision 1.60 Defer to a callout in hifn_rng_get. Fixes lock-against-self. Still can't call rnd_add_data directly from an rndsource callback. (Still plan to fix this with further rnd rototilling.) XXX pullup netbsd-7
1.61.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.61.18.1	10-Jun-2019	christos	Sync with HEAD
1.61.16.2	18-Jan-2019	pgoyette	Synch with HEAD
1.61.16.1	20-Oct-2018	pgoyette	Sync with head
1.63.6.1	29-Feb-2020	ad	Sync with head.
1.80.4.1	11-Aug-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #319): sys/dev/pci/ubsec.c: revision 1.64 sys/dev/pci/hifn7751.c: revision 1.82 lib/libc/gen/getentropy.3: revision 1.5 lib/libc/gen/getentropy.3: revision 1.6 share/man/man4/rnd.4: revision 1.41 lib/libc/sys/getrandom.2: revision 1.2 lib/libc/sys/getrandom.2: revision 1.3 share/man/man5/rc.conf.5: revision 1.193 share/man/man7/entropy.7: revision 1.5 share/man/man7/entropy.7: revision 1.6 share/man/man7/entropy.7: revision 1.7 share/man/man7/entropy.7: revision 1.8 etc/security: revision 1.130 share/man/man7/entropy.7: revision 1.9 etc/security: revision 1.131 sys/crypto/cprng_fast/cprng_fast.c: revision 1.19 sys/sys/rndio.h: revision 1.3 tests/lib/libc/sys/t_getrandom.c: revision 1.5 etc/defaults/rc.conf: revision 1.164 etc/defaults/rc.conf: revision 1.165 sys/sys/rndsource.h: revision 1.10 sys/kern/kern_entropy.c: revision 1.62 sys/kern/kern_entropy.c: revision 1.63 sys/kern/kern_entropy.c: revision 1.64 sys/kern/subr_cprng.c: revision 1.44 sys/kern/kern_entropy.c: revision 1.65 sys/kern/kern_clock.c: revision 1.149 sys/dev/pci/viornd.c: revision 1.22 share/man/man9/rnd.9: revision 1.32 sys/kern/subr_prf.c: revision 1.202 sys/sys/rndsource.h: revision 1.8 sys/sys/rndsource.h: revision 1.9 share/man/man7/entropy.7: revision 1.10 1. Reinstate netbsd<=9 entropy estimator to unblock /dev/random, in parallel with assessment of only confident entropy sources (seed, HWRNG) for security warnings like sshd keys in motd and daily insecurity report. 2. Make multiuser boot wait for first /dev/random output soon after loading a seed and configuring rndctl, so that getentropy(3) meets its contract starting early at boot without introducing blocking paths that could cause hangs in init(8) or single-user mode. Operators can choose to disable this wait in rc.conf. 3. Fix some bugs left over from reducing the global entropy lock from a spin lock at IPL_VM to an adaptive lock at IPL_SOFTSERIAL. 4. Update man pages.
1.9	17-May-2020	riastradh	Reduce diffs from OpenBSD hifn(4) header files.
1.8	11-Dec-2005	christos	merge ktrace-lwp.
1.7	27-Feb-2005	perry	nuke trailing whitespace
1.6	13-Feb-2004	wiz	branches: 1.6.8; 1.6.10; Spell length with h after t. Inspired by a commit by brad@openbsd.
1.5	16-Nov-2003	jonathan	Support symmetric crypto ops, including AES, on hifn 7955 and 7956. Not tested on 795[56], but compiles and matches FreeBSD-tested diffs. Suggested by Sam Leffler <sam@errno.com>, as imported into FreeBSD by Sam. Submitted to FreeBSD by Rajesh Vaidyanath <RVaidyanath@hifn.com>.
1.4	30-Jul-2003	jonathan	Discard previous driver; it didnt really do anything. Replace with a new from-scratch port of recent OpenBSD hifn7751 source, with changes to the crypto-framework calls to fit Sam Leffler's rework. (Some of the attach-time changes from the previous driver were carried over). Many thanks to Quentin Garnier (cube@cubidou.net) for testing an pre-release of the driver and finding a couple of missing initializations. Note that RNG support is temporarily disabled, until a potential issue with RNG quality is investigated and resolved.
1.3	13-May-2003	wiz	branches: 1.3.2; Typos from Jean-Francois Brousseau, via jmc@openbsd.
1.2	06-Jan-2003	wiz	descriptor with a c.
1.1	12-Oct-2000	itojun	branches: 1.1.2; 1.1.4; hifn7751 crypto card driver. from openbsd. does nothing useful at this moment - initailize the chip, that's all. TODO: crypto logic framework in kernel (see openbsd sys/crypto) TODO: ipsec frontend (need major rework in ipsec tree - should start with busywait) TODO: character device frontend
1.1.4.1	07-Jan-2003	thorpej	Sync with HEAD.
1.1.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	12-Oct-2000	bouyer	file hifn7751reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:20 +0000
1.3.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.3.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.3.2.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.6.8.1	29-Apr-2005	kent	sync with -current
1.18	03-Dec-2021	andvar	fix various typos in comments, log messages and documentation.
1.17	17-May-2020	riastradh	Rework hifn(4) RNG logic. Previously we would read 34 bits of entropy into the pool and stop, never to try again. This change finishes the conversion to an on-demand source, although we still use a callout to (a) try again in another second if the 7811 FIFO was emptied, and (b) wait at least 0.4 sec after we start the RNG to read anything from it as the alleged documentation allegedly recommends.
1.16	17-May-2020	riastradh	Tweak locking and use a pool cache for commands and dmamaps. This is enough to get the crypto decelerator working in LOCKDEBUG; previously it would crash the moment you looked at it.
1.15	17-May-2020	riastradh	Merge updates from OpenBSD's hifn(4) crypto decelerator driver. Highlights: - Apply a ten-year-old fix to the driver's misuse of previously exposed ciphertext as IV for the next ciphertext with CBC. (Someone^TM should make sure there are no other drivers still doing this...) - Various zeroing of buffers. - Various minor cosmetic changes.
1.14	17-May-2020	riastradh	Reduce diffs from OpenBSD hifn(4) header files.
1.13	17-May-2020	riastradh	De-#ifdef __OpenBSD__ify hifn7751.c. Not really helpful; we can just use diff3 to merge updates.
1.12	14-Apr-2015	riastradh	Include <sys/rndsource.h> where it is actually used. I had removed <sys/rnd.h> from files that didn't mention anything of the rnd(9) API. But they included other files which assumed <sys/rnd.h> had already been included.
1.11	03-Jan-2014	pgoyette	branches: 1.11.6; Modularize
1.10	13-Jun-2013	tls	branches: 1.10.2; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.9	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.8	19-Nov-2011	tls	branches: 1.8.8; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.7	11-Dec-2005	christos	branches: 1.7.110; merge ktrace-lwp.
1.6	15-Oct-2005	tls	Add support for the hardware random number generator on the Hifn 7811 and 795x. This was present in the driver before, but disabled due to problems with the actual randomness of generated numbers on the ubiquitious 7900-series parts. The code here is far, far more conservative than anyone else's driver for this RNG is -- but I believe that conservatism is called for, because the 79xx RNG design is missing a number of pieces from Hifn's "reference" 6500 RNG, and thus the numbers it generates must be treated with some care. Support for the 7811 RNG (which is a full-fledged 6500 type generator) is pretty much the same here as in other variants of this driver, except that it uses Hifn's "worst case" estimate of actual entropy per output bit, so it will accumulate bits much more slowly. The 7811 support is untested.
1.5	27-Feb-2005	perry	branches: 1.5.4; nuke trailing whitespace
1.4	16-Nov-2003	jonathan	branches: 1.4.8; 1.4.10; Support symmetric crypto ops, including AES, on hifn 7955 and 7956. Not tested on 795[56], but compiles and matches FreeBSD-tested diffs. Suggested by Sam Leffler <sam@errno.com>, as imported into FreeBSD by Sam. Submitted to FreeBSD by Rajesh Vaidyanath <RVaidyanath@hifn.com>.
1.3	30-Jul-2003	jonathan	Discard previous driver; it didnt really do anything. Replace with a new from-scratch port of recent OpenBSD hifn7751 source, with changes to the crypto-framework calls to fit Sam Leffler's rework. (Some of the attach-time changes from the previous driver were carried over). Many thanks to Quentin Garnier (cube@cubidou.net) for testing an pre-release of the driver and finding a couple of missing initializations. Note that RNG support is temporarily disabled, until a potential issue with RNG quality is investigated and resolved.
1.2	13-May-2003	wiz	branches: 1.2.2; Typos from Jean-Francois Brousseau, via jmc@openbsd.
1.1	12-Oct-2000	itojun	branches: 1.1.2; hifn7751 crypto card driver. from openbsd. does nothing useful at this moment - initailize the chip, that's all. TODO: crypto logic framework in kernel (see openbsd sys/crypto) TODO: ipsec frontend (need major rework in ipsec tree - should start with busywait) TODO: character device frontend
1.1.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	12-Oct-2000	bouyer	file hifn7751var.h was added on branch thorpej_scsipi on 2000-11-20 11:42:20 +0000
1.2.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.1	03-Aug-2004	skrll	Sync with HEAD
1.4.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.8.1	29-Apr-2005	kent	sync with -current
1.5.4.1	21-Jun-2006	yamt	sync with head.
1.7.110.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.110.2	30-Oct-2012	yamt	sync with head
1.7.110.1	17-Apr-2012	yamt	sync with head
1.8.8.4	03-Dec-2017	jdolecek	update from HEAD
1.8.8.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.8.8.2	23-Jun-2013	tls	resync from head
1.8.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.10.2.1	18-May-2014	rmind	sync with head
1.11.6.1	06-Jun-2015	skrll	Sync with HEAD
1.35	28-May-2022	andvar	fix various typos in comments.
1.34	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.33	31-Jul-2012	bouyer	branches: 1.33.2; 1.33.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.32	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.31	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.30	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.29	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.28	04-Apr-2011	dyoung	branches: 1.28.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.27	05-Nov-2010	jakllsch	branches: 1.27.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.26	19-Oct-2009	bouyer	branches: 1.26.2; 1.26.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.25	18-Mar-2008	cube	branches: 1.25.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.24	09-Feb-2007	ad	branches: 1.24.20; 1.24.36; 1.24.40; Merge newlock2 to head.
1.23	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.22	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.21	05-Aug-2006	xtraeme	branches: 1.21.4; 1.21.6; HPT368 support. Patch from Jonathan A. Kollasch via PR kern/34097. Thanks.
1.20	16-Jan-2006	bouyer	branches: 1.20.2; 1.20.6; properly use ata_channel->ch_ndrive: - initialize it properly in the bus front-ends (all 2, exept in wdc_pcmcia.c for the "Sandisk CompactFlash Card" where it's set to 1) - remplace hardcoded '2' by ata_channel->ch_ndrive in MI IDE drivers. From Christos Zoulas in kern/32501.
1.19	11-Dec-2005	christos	branches: 1.19.2; merge ktrace-lwp.
1.18	24-May-2005	lukem	branches: 1.18.2; add __KERNEL_RCSID
1.17	27-Feb-2005	perry	branches: 1.17.2; nuke trailing whitespace
1.16	21-Aug-2004	thorpej	branches: 1.16.4; 1.16.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.15	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.14	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.13	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.12	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.11	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.10	02-Aug-2004	bouyer	branches: 1.10.2; Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. Fix kern/26502 from Christian Biere.
1.9	03-Jan-2004	thorpej	branches: 1.9.2; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.8	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.7	28-Nov-2003	chs	add support for HPT302 and HPT371. simplify timing selection. adapted from OpenBSD. tested on an HPT302, review by bouyer@.
1.6	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.5	29-Oct-2003	mycroft	Rework slightly to remove bogus initializer.
1.4	25-Oct-2003	christos	Fix uninitialized variable warnings
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.9.2.1	11-Aug-2004	jmc	Pullup rev 1.10 (requested by bouyer in ticket #734) Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. PR#26502
1.10.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.10.2.2	03-Aug-2004	skrll	Sync with HEAD
1.10.2.1	02-Aug-2004	skrll	file hptide.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.16.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.16.4.1	29-Apr-2005	kent	sync with -current
1.17.2.1	06-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #1455): sys/dev/pci/hptide.c: revision 1.21 sys/dev/pci/pciide_hpt_reg.h: revision 1.14 HPT368 support. Patch from Jonathan A. Kollasch via PR kern/34097. Thanks.
1.18.2.4	24-Mar-2008	yamt	sync with head.
1.18.2.3	26-Feb-2007	yamt	sync with head.
1.18.2.2	30-Dec-2006	yamt	sync with head.
1.18.2.1	21-Jun-2006	yamt	sync with head.
1.19.2.1	01-Feb-2006	yamt	sync with head.
1.20.6.1	11-Aug-2006	yamt	sync with head
1.20.2.1	09-Sep-2006	rpaulo	sync with head
1.21.6.2	10-Dec-2006	yamt	sync with head.
1.21.6.1	22-Oct-2006	yamt	sync with head
1.21.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.21.4.1	18-Nov-2006	ad	Sync with head.
1.24.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.24.36.1	24-Mar-2008	keiichi	sync with head.
1.24.20.1	23-Mar-2008	matt	sync with HEAD
1.25.4.1	11-Mar-2010	yamt	sync with head
1.26.4.2	21-Apr-2011	rmind	sync with head
1.26.4.1	05-Mar-2011	rmind	sync with head
1.26.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.27.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.28.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.28.4.1	30-Oct-2012	yamt	sync with head
1.33.4.1	18-May-2014	rmind	sync with head
1.33.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.33.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.5	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.4	19-Jan-2001	martin	Nuke all uses of splimp() in the isdn subsystem, replace by splnet().
1.3	11-Jan-2001	martin	Move interrupt mapping before initialization of the card (which might cause interrupts). This works around a problem on sparc64, where interrupts without installed handler seem to cause heavy lossage (and interrupts are already enabled at this stage of autoconfiguration).
1.2	10-Jan-2001	martin	Make the driver try mapping memory first, the try io.
1.1	05-Jan-2001	martin	branches: 1.1.1; Initial revision
1.1.1.1	05-Jan-2001	martin	branches: 1.1.1.1.2; Initial import of ISDN4BSD release 0.96
1.1.1.1.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.1.1.1.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.1.1.1.2.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.1.1.1.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.1.1.2.1	05-Jan-2001	bouyer	file i4b_avm_fritz_pci.c was added on branch thorpej_scsipi on 2001-01-05 17:36:04 +0000
1.3	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.2	10-Jan-2001	martin	Prefer memory mapping over io mapping for this card too, although it doesn't have a chance to work on a no-io system because of an io based BAR lacking a mem counterpart.
1.1	05-Jan-2001	martin	branches: 1.1.1; Initial revision
1.1.1.1	05-Jan-2001	martin	branches: 1.1.1.1.2; Initial import of ISDN4BSD release 0.96
1.1.1.1.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.1.1.1.2.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.1.1.1.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.1.1.2.1	05-Jan-2001	bouyer	file i4b_elsa_qs1p.c was added on branch thorpej_scsipi on 2001-01-05 17:36:04 +0000
1.33	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.32	17-Sep-2009	tsutsui	branches: 1.32.12; 1.32.22; Misc KNF and cosmetics, and remove unnecessary casts etc.
1.31	14-Sep-2009	tsutsui	Use device_private(), proper types or variables for device_t and softc. (not yet split though) No crash on gxemul emulating hpcmips mobilepro.
1.30	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.29	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.28	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.27	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.26	26-Jun-2008	drochner	branches: 1.26.4; 1.26.10; constify the pcmcia/cardbus method tables
1.25	10-Apr-2008	cegger	branches: 1.25.4; 1.25.6; 1.25.8; use aprint_*_dev and device_xname
1.24	16-Nov-2006	christos	branches: 1.24.48; __unused removal on arguments; approved by core.
1.23	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.22	30-Aug-2006	christos	branches: 1.22.2; 1.22.4; add missing initializers.
1.21	11-Dec-2005	christos	branches: 1.21.4; 1.21.8; merge ktrace-lwp.
1.20	30-May-2005	christos	branches: 1.20.2; - const poisoning - avoid variable shadowing.
1.19	04-Feb-2005	perry	de-__P
1.18	11-Aug-2004	mycroft	branches: 1.18.4; 1.18.6; Rather than having a call up from the low-level driver to get the card type, instead have a call down from the PCMCIA mid-layer to set it. Use this from pcmcia_function_enable(). (Currently the policy is the same, but this would allow for more flexibility in deciding which mode to use.) Now it is safe to hold the socket enabled during attach, so do that. Only one enable/disable cycle to attach a card now!
1.17	31-Jan-2003	thorpej	branches: 1.17.2; Use aprint_*().
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.12	13-Nov-2001	lukem	add RCSID
1.11	24-Feb-2000	itohy	branches: 1.11.6; 1.11.8; Delete a superfluous newline in attach message.
1.10	22-Feb-2000	enami	Include dev/isa/isavar.h so that this file compiles again.
1.9	22-Feb-2000	thorpej	Don't use void * for chipset tags. Bad hacker, no biscuit.
1.8	01-Feb-2000	chopps	implement probing for available irqs on non-cirrus pcmcia controllers including cardbus controllers running in pcic mode
1.7	20-Dec-1998	nathanw	branches: 1.7.8; Fix up support for PCI attachments of pcic devices on i386 in general, and the Cirrus Logic CL-PD6729 in particular. From danw@mit.edu in PR port-i386/6436.
1.6	17-Dec-1998	msaitoh	fix an invalid panic message.
1.5	30-Nov-1998	jtk	Per PR/6172 (Johnny Lam, lamj@stat.cmu.edu), fix pcidevs identification of CL-PD6729. See <http://www.cirrus.com/products/overviews/pd6729.html>.
1.4	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.3	23-May-1998	matt	Changes so that these compile on NetBSD/alpha as well as NetBSD/i386.
1.2	16-Oct-1997	thorpej	Pull marc-pcmcia branch down from trunk.
1.1	28-Aug-1997	marc	branches: 1.1.2; file i82365_pci.c was initially added on branch marc-pcmcia.
1.1.2.4	16-Oct-1997	thorpej	Copyright/license update.
1.1.2.3	16-Oct-1997	thorpej	KNF.
1.1.2.2	27-Sep-1997	marc	add io bus info. this whole file is a hack, really, so it's ok that this is, too.
1.1.2.1	28-Aug-1997	marc	pci-specific code from i82365 driver bus split. This code needs some serious help before it's likely to work, let alone be reasonable.
1.7.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.11.8.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.11.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.11.6.3	18-Oct-2002	nathanw	Catch up to -current.
1.11.6.2	08-Jan-2002	nathanw	Catch up to -current.
1.11.6.1	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.17.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.17.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.17.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.17.2.1	12-Aug-2004	skrll	Sync with HEAD.
1.18.6.1	12-Feb-2005	yamt	sync with head.
1.18.4.1	29-Apr-2005	kent	sync with -current
1.20.2.1	30-Dec-2006	yamt	sync with head.
1.21.8.1	03-Sep-2006	yamt	sync with head.
1.21.4.1	09-Sep-2006	rpaulo	sync with head
1.22.4.2	10-Dec-2006	yamt	sync with head.
1.22.4.1	22-Oct-2006	yamt	sync with head
1.22.2.1	18-Nov-2006	ad	Sync with head.
1.24.48.2	29-Jun-2008	mjf	Sync with HEAD.
1.24.48.1	02-Jun-2008	mjf	Sync with HEAD.
1.25.8.1	27-Jun-2008	simonb	Sync with head.
1.25.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.25.4.4	11-Mar-2010	yamt	sync with head
1.25.4.3	16-Sep-2009	yamt	sync with head
1.25.4.2	16-May-2009	yamt	sync with head
1.25.4.1	04-May-2009	yamt	sync with head.
1.26.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.26.4.1	28-Apr-2009	skrll	Sync with HEAD.
1.32.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.32.12.1	30-Oct-2012	yamt	sync with head
1.7	17-Sep-2009	tsutsui	Misc KNF and cosmetics, and remove unnecessary casts etc.
1.6	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.5	11-Dec-2005	christos	branches: 1.5.70; 1.5.72; 1.5.74; merge ktrace-lwp.
1.4	27-Feb-2005	perry	nuke trailing whitespace
1.3	04-Feb-2005	perry	de-__P
1.2	22-Feb-2000	thorpej	branches: 1.2.28; 1.2.36; 1.2.38; Don't use void * for chipset tags. Bad hacker, no biscuit.
1.1	20-Dec-1998	nathanw	branches: 1.1.8; Fix up support for PCI attachments of pcic devices on i386 in general, and the Cirrus Logic CL-PD6729 in particular. From danw@mit.edu in PR port-i386/6436.
1.1.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.38.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.38.1	12-Feb-2005	yamt	sync with head.
1.2.36.1	29-Apr-2005	kent	sync with -current
1.2.28.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.28.1	04-Feb-2005	skrll	Sync with HEAD.
1.5.74.2	11-Mar-2010	yamt	sync with head
1.5.74.1	16-May-2008	yamt	sync with head.
1.5.72.1	18-May-2008	yamt	sync with head.
1.5.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.18	21-Sep-2018	maxv	Remove iavc(4).
1.17	11-Jul-2016	msaitoh	branches: 1.17.16; 1.17.18; KNF. No functional change.
1.16	29-Mar-2014	christos	branches: 1.16.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.15	25-Sep-2013	martin	Avoid using unmaped PCI registers. From Wolfgang Stukenbrock in PR kern/48236.
1.14	27-Oct-2012	chs	branches: 1.14.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.13	06-Dec-2009	dyoung	branches: 1.13.12; 1.13.22; Delete dead code: a device-activation hook.
1.12	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.11	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.10	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.9	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.8	10-Apr-2008	cegger	branches: 1.8.4; 1.8.18; use aprint_*_dev and device_xname
1.7	19-Oct-2007	ad	branches: 1.7.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.6	16-Nov-2006	christos	branches: 1.6.8; 1.6.22; 1.6.24; 1.6.28; __unused removal on arguments; approved by core.
1.5	16-Oct-2006	pooka	* __unused games + other relevant fixes * give me ANSI C or give me death (at least where it could be easily spotted)
1.4	11-Dec-2005	christos	branches: 1.4.20; 1.4.22; merge ktrace-lwp.
1.3	11-May-2005	martin	branches: 1.3.2; From Wolfgang Stukenbrock in PR kern/30197: wrong type in CFATTACH_DECL.
1.2	03-Oct-2003	pooka	branches: 1.2.4; 1.2.8; 1.2.16; Ready the kernel side of i4b for primary rate interface support by removing assumptions that there are only two B channels and by adding support for a varying number of channels. Due to this, rename previously used isdn identified "bri" to "isdnif", which better describes the current situation.
1.1	25-Sep-2003	pooka	Add a driver for the CAPI-compliant AVM B1/T1 cards. The driver backend is capable of supporting also ISA cards (no DMA) and primary rate (PRI) cards in addition to the basic rate ones, but I don't any to test on right now, so we don't support those currently. This code was originally written by Juha-Matti Liukkonen <jml@cubical.fi> of Cubical Solutions Ltd. for FreeBSD, and was ported to NetBSD by myself for the same company.
1.2.16.1	11-May-2005	tron	Pull up revision 1.3 (requested by martin in ticket #295): From Wolfgang Stukenbrock in PR kern/30197: wrong type in CFATTACH_DECL.
1.2.8.1	16-May-2005	riz	Pull up revision 1.3 (requested by martin in ticket #1513): From Wolfgang Stukenbrock in PR kern/30197: wrong type in CFATTACH_DECL.
1.2.4.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.4.2	03-Aug-2004	skrll	Sync with HEAD
1.2.4.1	03-Oct-2003	skrll	file iavc_pci.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.3.2.2	27-Oct-2007	yamt	sync with head.
1.3.2.1	30-Dec-2006	yamt	sync with head.
1.4.22.2	10-Dec-2006	yamt	sync with head.
1.4.22.1	22-Oct-2006	yamt	sync with head
1.4.20.1	18-Nov-2006	ad	Sync with head.
1.6.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.6.24.1	06-Nov-2007	matt	sync with HEAD
1.6.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.6.8.1	23-Oct-2007	ad	Sync with head.
1.7.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.8.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.8.4.2	11-Mar-2010	yamt	sync with head
1.8.4.1	16-May-2009	yamt	sync with head
1.13.22.3	03-Dec-2017	jdolecek	update from HEAD
1.13.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.13.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.13.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.13.12.1	30-Oct-2012	yamt	sync with head
1.14.2.1	18-May-2014	rmind	sync with head
1.16.6.1	05-Oct-2016	skrll	Sync with HEAD
1.17.18.1	10-Jun-2019	christos	Sync with HEAD
1.17.16.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.5	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.4	24-Apr-2021	thorpej	branches: 1.4.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.3	20-Jan-2017	maya	branches: 1.3.30; Check pmf_device_register return value. NFC Appeases static analyzers.
1.2	14-Jul-2016	msaitoh	branches: 1.2.2; 1.2.4; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.1	17-Dec-2012	mbalmer	branches: 1.1.2; 1.1.6; 1.1.18; Adding ibmcd(4), a device driver for the IBM 4810 BSP cash drawer port as found e.g. in SurePOS 300 series point of sale terminals. The driver provides a gpio(4) device with three pins: pin 0 to open drawer, pin 1 to read the status and pin 2 to read whether a cash drawer is connected or not.
1.1.18.2	05-Feb-2017	skrll	Sync with HEAD
1.1.18.1	05-Oct-2016	skrll	Sync with HEAD
1.1.6.3	03-Dec-2017	jdolecek	update from HEAD
1.1.6.2	25-Feb-2013	tls	resync with head
1.1.6.1	17-Dec-2012	tls	file ibmcd.c was added on branch tls-maxphys on 2013-02-25 00:29:17 +0000
1.1.2.2	23-Jan-2013	yamt	sync with head
1.1.2.1	17-Dec-2012	yamt	file ibmcd.c was added on branch yamt-pagecache on 2013-01-23 00:06:08 +0000
1.2.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.2.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.3.30.1	23-Mar-2021	thorpej	Convert config_found_ia() call sites where the device only carries a single interface attribute to bare config_found() calls.
1.4.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	26-Aug-2007	xtraeme	Move the ICH SMB defs to dev/ic/i82801lpcreg.h, it's where it belongs. This is to avoid code duplication.
1.1	28-Jul-2007	kiyohara	branches: 1.1.4; 1.1.6; 1.1.8; 1.1.10; Add support for Intel ICH SMBus controller.
1.1.10.2	28-Jul-2007	kiyohara	Add support for Intel ICH SMBus controller.
1.1.10.1	28-Jul-2007	kiyohara	file ichreg.h was added on branch matt-mips64 on 2007-07-28 10:51:57 +0000
1.1.8.3	09-Oct-2007	ad	Sync with head.
1.1.8.2	20-Aug-2007	ad	Sync with HEAD.
1.1.8.1	28-Jul-2007	ad	file ichreg.h was added on branch vmlocking on 2007-08-20 22:07:04 +0000
1.1.6.3	03-Sep-2007	skrll	Sync with HEAD.
1.1.6.2	15-Aug-2007	skrll	Sync with HEAD.
1.1.6.1	28-Jul-2007	skrll	file ichreg.h was added on branch nick-csl-alignment on 2007-08-15 13:48:29 +0000
1.1.4.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.87	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.86	18-Jan-2025	mrg	match meteor lake's smbus.
1.85	07-Aug-2023	msaitoh	branches: 1.85.6; ichsmb(4),ismt(4): Add Snow Ridge support.
1.84	23-Jul-2023	msaitoh	Add support Intel 700 series chipset and Alder Lake-N devices.
1.83	30-May-2023	msaitoh	Use device_printf() instead of aprint_error_dev().
1.82	12-Apr-2023	riastradh	ichsmb(4), tco(4): Add support for TCO on newer Intel chipsets. TCO (`Total Cost of Ownership', Intel's bizarre name for a watchdog timer) used to hang off the Intel I/O platform controller hub's (ICH) low-pin-count interface bridge (LPC IB), or ichlpcib(4). On newer devices, it hangs off the ICH SMBus instead. Tested on INTEL 100SERIES_SMB (works) and INTEL 100SERIES_LP_SMB (doesn't work, still not sure why). XXX kernel revbump: This breaks the module ABI -- tco(4) modules older than the change to make ta_has_rcba into ta_version will incorrectly attach at buses they do not understand. (However, the tco(4) driver is statically built into GENERIC, so maybe it's safe for pullup since the module wouldn't have worked anyway.)
1.81	22-Sep-2022	riastradh	branches: 1.81.4; ichsmb(4): Remove confusing `lpcib_' prefix on register names.
1.80	22-Sep-2022	riastradh	ichsmb(4): Only rescan i2cbus child if requested. This will let us rescan tcoichbus later too on devices where the TCO (Intel platform controller hub watchdog timer) hangs off ichsmb(4) instead of ichlpcib(4).
1.79	22-Sep-2022	riastradh	ichsmb(4): Use config_detach_children to simplify.
1.78	22-Sep-2022	riastradh	ichsmb(4): Attach i2c bus only once. The child could be detached, e.g. with drvctl, and then the bus rescanned, at which point it would reinitialize a mutex without destroying it.
1.77	13-Sep-2022	msaitoh	Add Intel 600 series PCH-LP devices support.
1.76	25-Jan-2022	msaitoh	Add Intel 600 Series PCH (desktop) support.
1.75	27-Oct-2021	msaitoh	Add Elkhart Lake support.
1.74	12-Oct-2021	msaitoh	Add Intel 400 Series PCH-V devices.
1.73	11-Oct-2021	msaitoh	Add Intel 500 Series PCH-H SMBus.
1.72	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.71	12-Jul-2021	msaitoh	branches: 1.71.2; Add Jasper Lake support.
1.70	24-Apr-2021	thorpej	branches: 1.70.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.69	15-Jan-2021	thorpej	branches: 1.69.2; - Match 400-, 495-, and 500-series controllers. - Apply OpenBSD rev 1.38, which ignores the SMBALERT# interrupt that fires spuriously on some systems (ignored by Linux and FreeBSD, as well).
1.68	22-Apr-2020	msaitoh	branches: 1.68.2; Add Whiskey Lake U and Amber Lake Y support.
1.67	16-Apr-2020	msaitoh	Add Comet Lake support.
1.66	18-Feb-2020	msaitoh	branches: 1.66.4; Whitespace fix. No functional change.
1.65	24-Dec-2019	thorpej	branches: 1.65.2; Make ichsmb and piixpm MP-safe: - Synchronize with the interrupt handler using a mutex. - Use a condvar to wait for completion, rather than tsleep(). - Mark our interrupt handler as such. Also, other general correctness fixes: - Loop around testing the completion condition to protect aginst spurious wakes. - The "i2c exec" function returns an error code, so actually do so.
1.64	23-Dec-2019	thorpej	ichsmb_i2c_exec(): No need to check 'cold' to force I2C_F_POLL; the i2c upper layer does it for us.
1.63	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.62	09-Dec-2019	ad	Fix LOCKDEBUG panic on detach.
1.61	21-Nov-2019	ad	If attach fails, don't panic on detach.
1.60	10-Dec-2018	jdolecek	branches: 1.60.4; call pci_intr_release() when pci_intr_establish_xname() fails; I got confused by the opposite conditionals compared to other drivers found by Masanobu SAITOH
1.59	09-Dec-2018	jdolecek	try to detach on shutdown; for now the only practical reason is to test that the driver detach paths work
1.58	09-Dec-2018	jdolecek	convert to pci_intr_alloc() to possibly allow MSI/MSI-X; has no effect on my system, where the device only supports INTx
1.57	09-Apr-2018	msaitoh	branches: 1.57.2; Add 300 series chipset support.
1.56	02-Mar-2018	msaitoh	branches: 1.56.2; Add Apollo Lake and Gemini Lake devices.
1.55	28-Feb-2018	pgoyette	Teach the ichsmb(4) driver how to detach.
1.54	27-Feb-2018	pgoyette	Add dependency on iic module to get access to iicbus_print()
1.53	26-Feb-2018	pgoyette	Modularize the ichsmb(4) driver. No functional changes.
1.52	22-Feb-2018	msaitoh	Add C620 devices.
1.51	17-Aug-2017	msaitoh	Add C3000 devices.
1.50	31-Mar-2017	msaitoh	branches: 1.50.6; Remove extra 0x. This bug was added when replacing bitmask_snprintf(9) with snprintb(3) (in between NetBSD 5 and 6). Old bitmask_snprint(9) didn't add 0x" automatically for hexadecimal value, so old code used it with "0x%s".
1.49	15-Mar-2017	msaitoh	Add another 100 Series SMBus (100SERIES_LP_SMB)
1.48	15-Jan-2017	msaitoh	branches: 1.48.2; Add Intel 200 series devices.
1.47	13-Oct-2016	jdolecek	provide intr xname
1.46	07-Jul-2016	msaitoh	branches: 1.46.2; KNF. Remove extra spaces. No functional change.
1.45	10-Dec-2015	pgoyette	Initialize the mutex before calling xxx_rescan(). Otherwise if we load/attach an i2cbus at a later time (ie, load the iic module), we'll end-up re-initializing the mutex. (Caught with a LOCKDEBUG kernel)
1.44	01-Dec-2015	msaitoh	Add Intel 100 Series chipset's devices.
1.43	16-Nov-2015	msaitoh	Add Braswell PCU SMBus.
1.42	03-May-2015	pgoyette	Teach a couple of i2cbus controllers how to rescan. This enables {,un}loading and {at,de}taching of the iic(4) driver/module at a later time. Tested piixpm on QEMU, and ichsmb on my live server.
1.41	02-Apr-2015	tnn	attach Mobile 5th Gen. Core SMBus
1.40	20-Mar-2015	msaitoh	Add Intel C61x and X99 devices.
1.39	18-Mar-2015	msaitoh	Add 9 Series support.
1.38	15-Dec-2014	msaitoh	- Rename PCI_PRODUCT_INTEL_DH89XX_SMB to PCI_PRODUCT_INTEL_DH89XXCC_SMB - Add PCI_PRODUCT_INTEL_DH89XXCL_SMB
1.37	02-Jul-2014	riastradh	branches: 1.37.2; 1.37.4; Register a null pmf handler even if we failed to attach. Keeps ichsmb(4) from preventing suspend even if it's broken because of ichlpcib(4) grodiness.
1.36	09-Jun-2014	msaitoh	Add Tolapai, Coleto Creek and Bay Trail.
1.35	29-Mar-2014	christos	branches: 1.35.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.34	28-Dec-2013	msaitoh	Add Intel C2000 PCU SMBus
1.33	07-Nov-2013	msaitoh	Add some Mobile 4th generation Intel Core Processor devices (Lynx Point-LP).
1.32	18-Jul-2013	soren	Clear errors before beginning a transfer. Closes PR port-i386/46792.
1.31	18-Jul-2013	msaitoh	Add Intel 8 Series SMBus.
1.30	12-Jan-2013	riastradh	branches: 1.30.2; 1.30.8; Match the C600's other smbus controller.
1.29	29-Nov-2012	msaitoh	Add C600(and X79) SMBus devices.
1.28	26-Nov-2012	riastradh	Match the Intel 7 Series SMBus Controller in ichsmb(4). Tested on a Thinkpad W530.
1.27	14-Feb-2012	pgoyette	branches: 1.27.2; 1.27.6; Replace the xxx_acquire()/xxx_release() rwlocks with mutexes. There are only RW_WRITERs for these, and no RW_READERs, so no need to incur the extra overhead of allowing for both. As discussed on tech-kern. For piixpm and ichsmb, the acquire/release protocol needs to be used, even if the request is I2C_F_POLL'd (or if the device supports only polled mode). Otherwise multiple requests can be running at the same time, and they stomp on each other and create anomolous results. Part 2 addresses my PR kern/45889 3 ACKs from releng
1.26	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.25	15-May-2011	msaitoh	branches: 1.25.4; 1.25.8; Add support for Intel 6 series.
1.24	10-Feb-2011	hannken	Make the SMbus Send/Receive functions work. If cmdlen = 0 and len = 1 use the XXX_HC_CMD_BYTE protocol, send data through XXX_SMB_HCMD or receive data from XXX_SMB_HD0. Ok: Paul Goyette
1.23	30-Jul-2010	njoly	branches: 1.23.2; 1.23.4; Adjust for Intel ICH10 devices update.
1.22	06-Feb-2010	tnn	branches: 1.22.2; 1.22.4; Attach to SMBus on Intel P55 chipset.
1.21	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.20	18-Mar-2009	cegger	bzero -> memset
1.19	03-Feb-2009	pgoyette	branches: 1.19.2; Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.18	16-Dec-2008	christos	replace bitmask_snprintf(9) with snprintb(3)
1.17	01-Oct-2008	christos	branches: 1.17.2; ICH10 support from Brad du Plessis
1.16	08-Sep-2008	njoly	Add missing newline in debug message.
1.15	08-Sep-2008	njoly	Improve message for SMI interrupts.
1.14	08-Sep-2008	njoly	Fix compilation with ICHIIC_DEBUG.
1.13	10-Apr-2008	cegger	branches: 1.13.4; 1.13.6; 1.13.10; use aprint_*_dev and device_xname
1.12	28-Mar-2008	kiyohara	Split device_t/softc.
1.11	09-Dec-2007	jmcneill	branches: 1.11.10; Merge jmcneill-pm branch.
1.10	19-Oct-2007	ad	branches: 1.10.4; 1.10.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.9	01-Sep-2007	riz	branches: 1.9.2; 1.9.6; For SMBus, add the ability to enumerate devices on the bus. This does NOT identify the devices, merely indicates the presence of devices at certain addresses. Tested on ichsmb and nfsmb - other SMBus devices will need to ensure the proper bus type is set. (I2C_TYPE_SMBUS) From Nicolas Joly, via Paul Goyette, in PR#36744.
1.8	27-Aug-2007	xtraeme	branches: 1.8.2; Use rwlock(9) rather than lockmgr(9).
1.7	27-Aug-2007	xtraeme	Attach to the ICH9 SMBus device too.
1.6	27-Aug-2007	xtraeme	Add missing __KERNEL_RCSID().
1.5	26-Aug-2007	xtraeme	Move the ICH SMB defs to dev/ic/i82801lpcreg.h, it's where it belongs. This is to avoid code duplication.
1.4	09-Aug-2007	kiyohara	branches: 1.4.2; 1.4.4; Use bitmask_snprintf().
1.3	05-Aug-2007	xtraeme	branches: 1.3.2; Repeat after me: %zu for size_t, %zd for ssize_t. Reminded by yamt@.
1.2	05-Aug-2007	xtraeme	Use %zd to printf size_t.
1.1	28-Jul-2007	kiyohara	branches: 1.1.4; Add support for Intel ICH SMBus controller.
1.1.4.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.1.4.5	23-Nov-2007	joerg	Register with PM framework.
1.1.4.4	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.1.4.3	03-Sep-2007	jmcneill	Sync with HEAD.
1.1.4.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.1.4.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.3.2.2	05-Aug-2007	xtraeme	Repeat after me: %zu for size_t, %zd for ssize_t. Reminded by yamt@.
1.3.2.1	05-Aug-2007	xtraeme	file ichsmb.c was added on branch matt-mips64 on 2007-08-05 23:05:03 +0000
1.4.4.4	23-Oct-2007	ad	Sync with head.
1.4.4.3	09-Oct-2007	ad	Sync with head.
1.4.4.2	20-Aug-2007	ad	Sync with HEAD.
1.4.4.1	09-Aug-2007	ad	file ichsmb.c was added on branch vmlocking on 2007-08-20 22:07:04 +0000
1.4.2.3	03-Sep-2007	skrll	Sync with HEAD.
1.4.2.2	15-Aug-2007	skrll	Sync with HEAD.
1.4.2.1	09-Aug-2007	skrll	file ichsmb.c was added on branch nick-csl-alignment on 2007-08-15 13:48:30 +0000
1.8.2.2	09-Jan-2008	matt	sync with HEAD
1.8.2.1	06-Nov-2007	matt	sync with HEAD
1.9.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.9.2.4	21-Jan-2008	yamt	sync with head
1.9.2.3	27-Oct-2007	yamt	sync with head.
1.9.2.2	03-Sep-2007	yamt	sync with head.
1.9.2.1	01-Sep-2007	yamt	file ichsmb.c was added on branch yamt-lazymbuf on 2007-09-03 14:36:51 +0000
1.10.6.1	11-Dec-2007	yamt	sync with head.
1.10.4.1	26-Dec-2007	ad	Sync with head.
1.11.10.5	17-Jan-2009	mjf	Sync with HEAD.
1.11.10.4	05-Oct-2008	mjf	Sync with HEAD.
1.11.10.3	28-Sep-2008	mjf	Sync with HEAD.
1.11.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.11.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.10.1	19-Oct-2008	haad	Sync with HEAD.
1.13.6.2	10-Oct-2008	skrll	Sync with HEAD.
1.13.6.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.13.4.4	11-Aug-2010	yamt	sync with head.
1.13.4.3	11-Mar-2010	yamt	sync with head
1.13.4.2	16-May-2009	yamt	sync with head
1.13.4.1	04-May-2009	yamt	sync with head.
1.17.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.17.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.17.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.19.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.22.4.2	31-May-2011	rmind	sync with head
1.22.4.1	05-Mar-2011	rmind	sync with head
1.22.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.23.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.23.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.25.8.1	18-Feb-2012	mrg	merge to -current.
1.25.4.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.25.4.3	23-Jan-2013	yamt	sync with head
1.25.4.2	16-Jan-2013	yamt	sync with (a bit old) head
1.25.4.1	17-Apr-2012	yamt	sync with head
1.27.6.3	03-Dec-2017	jdolecek	update from HEAD
1.27.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.27.6.1	25-Feb-2013	tls	resync with head
1.27.2.1	29-Sep-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1123): sys/dev/pci/ichsmb.c: revision 1.37 Register a null pmf handler even if we failed to attach. Keeps ichsmb(4) from preventing suspend even if it's broken because of ichlpcib(4) grodiness.
1.30.8.1	23-Jul-2013	riastradh	sync with HEAD
1.30.2.2	18-May-2014	rmind	sync with head
1.30.2.1	28-Aug-2013	rmind	sync with head
1.35.2.1	10-Aug-2014	tls	Rebase.
1.37.4.7	28-Aug-2017	skrll	Sync with HEAD
1.37.4.6	05-Feb-2017	skrll	Sync with HEAD
1.37.4.5	05-Dec-2016	skrll	Sync with HEAD
1.37.4.4	09-Jul-2016	skrll	Sync with HEAD
1.37.4.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.37.4.2	06-Jun-2015	skrll	Sync with HEAD
1.37.4.1	06-Apr-2015	skrll	Sync with HEAD
1.37.2.3	09-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1294): sys/dev/pci/ichsmb.c: revisions 1.43-1.44 ichsmb(4): Add Braswell PCU SMBus and Intel 100 Series SMBus.
1.37.2.2	30-Apr-2015	snj	branches: 1.37.2.2.4; Pull up following revision(s) (requested by msaitoh in ticket #725): sys/arch/x86/pci/ichlpcib.c: revisions 1.47, 1.48 sys/dev/pci/ichsmb.c: revisions 1.39, 1.40, 1.41 sys/dev/pci/pucdata.c: revision 1.94 Add 9 Series support. -- Add Intel C61x and X99 devices. -- attach Mobile 5th Gen. Core SMBus
1.37.2.1	17-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #528): sys/dev/pci/ichsmb.c: revision 1.38 sys/dev/pci/pcidevs: revision 1.1206 sys/dev/pci/pcidevs: revision 1.1207 sys/arch/x86/pci/ichlpcib.c: revision 1.44 Change Intel 0x0434 entry: - Rename DH89XX_QA to DH89XXCC_IQIA - Modify the description to DH89xxCC PCIe Endpoint and QuickAssist (include typo fix) - Rename DH89xxCC's names from DH89XX_ to DH89XXCC_. - Add some DH89xxCC's devices. - Add DH89XXCL's devices. - Rename PCI_PRODUCT_INTEL_DH89XX_SMB to PCI_PRODUCT_INTEL_DH89XXCC_SMB - Add PCI_PRODUCT_INTEL_DH89XXCL_SMB Add DH89xxC[CL] LPC devices.
1.37.2.2.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.46.2.3	26-Apr-2017	pgoyette	Sync with HEAD
1.46.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.46.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.48.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.50.6.8	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1891): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.50.6.7	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #1856: sys/dev/pci/ichsmb.c 1.83-1.84 - Add support Intel 700 series chipset and Alder Lake-N devices. - Use device_printf() instead of aprint_error_dev() in ichsmb_i2c_exec().
1.50.6.6	15-Oct-2022	martin	Pull up the following (via patch), requested by msaitoh in ticket #1776: sys/dev/pci/ichsmb.c 1.76-1.77 Add Intel 600 Series PCH support.
1.50.6.5	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1711: sys/dev/pci/ichsmb.c 1.69, 1.71, 1.73-1.75 via patch - Add Intel 400, 495, and 500 series support. - Add Intel Jasper Lake and Elkhart Lake support. - Ignores the SMBALERT# interrupt. Same as other OSes.
1.50.6.4	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1586: sys/dev/pci/ichsmb.c 1.66-1.68 via patch - Add Comet Lake, Whiskey Lake U and Amber Lake Y support. - Whitespace fix.
1.50.6.3	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #926): sys/dev/pci/ichsmb.c: 1.52, 1.56 Add C620 devices. -- Add Apollo Lake and Gemini Lake devices.
1.50.6.2	16-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #763): sys/dev/pci/ichsmb.c: revision 1.57 sys/dev/pci/pucdata.c: revision 1.101 Add 300 series chipset support.
1.50.6.1	26-Feb-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #565): sys/dev/pci/ichsmb.c: 1.51 sys/dev/pci/ismt.c: 1.6 Add C3000 devices.
1.56.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.56.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.57.2.3	21-Apr-2020	martin	Sync with HEAD
1.57.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.57.2.1	10-Jun-2019	christos	Sync with HEAD
1.60.4.6	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1723): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.60.4.5	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #1672: sys/dev/pci/ichsmb.c 1.83-1.84 - Add support Intel 700 series chipset and Alder Lake-N devices. - Use device_printf() instead of aprint_error_dev() in ichsmb_i2c_exec().
1.60.4.4	15-Oct-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1544: sys/dev/pci/ichsmb.c 1.76-1.77 Add Intel 600 Series PCH support.
1.60.4.3	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1381: sys/dev/pci/ichsmb.c 1.69, 1.71, 1.73-1.75 via patch - Add Intel 400, 495, and 500 series support. - Add Intel Jasper Lake and Elkhart Lake support. - Ignore the SMBALERT# interrupt. Same as other OSes.
1.60.4.2	10-Jul-2020	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #996: sys/dev/pci/ichsmb.c 1.62, 1.66-1.68 - Fix LOCKDEBUG panic on detach when attach failed. - Add Comet Lake, Whiskey Lake U and Amber Lake Y support. - Whitespace fix.
1.60.4.1	01-Mar-2020	martin	Pull up following revision(s) (requested by riastradh in ticket #748): sys/dev/pci/ichsmb.c: revision 1.61 If attach fails, don't panic on detach.
1.65.2.1	29-Feb-2020	ad	Sync with head.
1.66.4.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.66.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.68.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.69.2.1	24-Mar-2021	thorpej	Don't filter interface attributes in rescan functions for devices that carry only a single interface attribute. The autoconfiguration machinery already considers interface attributes when searching for possible parents.
1.70.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.71.2.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.81.4.3	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #338): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.81.4.2	01-Aug-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #282): sys/dev/pci/ichsmb.c: revision 1.82 sys/arch/amd64/conf/GENERIC: revision 1.602 sys/arch/x86/pci/tco.c: revision 1.10 sys/arch/x86/pci/tco.h: revision 1.5 sys/arch/x86/pci/ichlpcib.c: revision 1.59 sys/dev/ic/i82801lpcreg.h: revision 1.17 sys/arch/x86/pci/files.pci: revision 1.27 sys/dev/pci/files.pci: revision 1.446 ichsmb(4), tco(4): Add support for TCO on newer Intel chipsets. TCO (`Total Cost of Ownership', Intel's bizarre name for a watchdog timer) used to hang off the Intel I/O platform controller hub's (ICH) low-pin-count interface bridge (LPC IB), or ichlpcib(4). On newer devices, it hangs off the ICH SMBus instead. Tested on INTEL 100SERIES_SMB (works) and INTEL 100SERIES_LP_SMB (doesn't work, still not sure why). XXX kernel revbump: This breaks the module ABI -- tco(4) modules older than the change to make ta_has_rcba into ta_version will incorrectly attach at buses they do not understand. (However, the tco(4) driver is statically built into GENERIC, so maybe it's safe for pullup since the module wouldn't have worked anyway.)
1.81.4.1	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #253: sys/dev/pci/ichsmb.c 1.83-1.84 - Add support Intel 700 series chipset and Alder Lake-N devices. - Use device_printf() instead of aprint_error_dev() in ichsmb_i2c_exec().
1.85.6.1	02-Aug-2025	perseant	Sync with HEAD
1.25	05-Jan-2022	andvar	fix typos, mainly s/comand/command/
1.24	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.23	27-Sep-2016	pgoyette	branches: 1.23.14; 1.23.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.22	29-Mar-2014	christos	branches: 1.22.6; 1.22.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.21	27-Oct-2012	chs	branches: 1.21.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.20	13-Nov-2010	uebayasi	branches: 1.20.8; 1.20.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.19	26-Nov-2009	njoly	branches: 1.19.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.18	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.17	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.16	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.15	28-Apr-2008	martin	branches: 1.15.14; Remove clause 3 and 4 from TNF licenses
1.14	19-Oct-2007	ad	branches: 1.14.16; 1.14.18; 1.14.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	27-Feb-2005	perry	branches: 1.10.4; nuke trailing whitespace
1.9	29-Jun-2003	thorpej	branches: 1.9.2; 1.9.10; 1.9.12; Only need to report the PCI product ID in the management ioctls. From Achim Leubner @ ICP.
1.8	13-May-2003	thorpej	Add support for the user-generated command interface and event logging required by the ICP-Vortex management tools. Many thanks to Achim Leubner at ICP-Vortex for a few bug fixes and for testing.
1.7	31-Jan-2003	thorpej	Use aprint_*().
1.6	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.5	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	24-Apr-2002	ad	branches: 1.2.6; 1.2.8; Match anything with PCI_VENDOR_VORTEX, then use a heuristic to determine the board's interface, and list only exceptions in icp_pci_ident[].
1.1	22-Apr-2002	ad	Add a driver for ICP-Vortex GDT and Intel Storage RAID controllers. Parts taken from OpenBSD. Test hardware kindly provided by Intel. This still needs management bits, and doesn't support older controllers, but that shouldn't be hard to fix.
1.2.8.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.8.1	24-Apr-2002	jdolecek	file icp_pci.c was added on branch kqueue on 2002-06-23 17:47:38 +0000
1.2.6.3	18-Oct-2002	nathanw	Catch up to -current.
1.2.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.2.6.1	24-Apr-2002	nathanw	file icp_pci.c was added on branch nathanw_sa on 2002-06-20 03:45:23 +0000
1.9.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.10.1	29-Apr-2005	kent	sync with -current
1.9.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.4.2	27-Oct-2007	yamt	sync with head.
1.10.4.1	30-Dec-2006	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.14.20.3	11-Mar-2010	yamt	sync with head
1.14.20.2	16-May-2009	yamt	sync with head
1.14.20.1	16-May-2008	yamt	sync with head.
1.14.18.1	18-May-2008	yamt	sync with head.
1.14.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.15.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.19.4.1	05-Mar-2011	rmind	sync with head
1.20.18.3	03-Dec-2017	jdolecek	update from HEAD
1.20.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.20.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.20.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.20.8.1	30-Oct-2012	yamt	sync with head
1.21.2.1	18-May-2014	rmind	sync with head
1.22.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.22.6.1	05-Oct-2016	skrll	Sync with HEAD
1.23.16.1	10-Jun-2019	christos	Sync with HEAD
1.23.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.74	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.73	17-Sep-2022	thorpej	age_start(): Remove redundant IFQ_IS_EMPTY(). This very same condition is checked as soon as we enter the loop 2 statements later.
1.72	22-Aug-2022	thorpej	age_start(): Replace "IFQ_DEQUEUE() -> IF_PREPEND() on failure" with "IFQ_POLL() -> IFQ_DEQUEUE() on success (or fatal error)".
1.71	22-Aug-2022	thorpej	age_encap(): Use m_defrag(), not m_pullup(), to compact an mbuf chain if it has too many DMA segments. Let the caller handle dispose of the packet on fatal errors.
1.70	31-May-2022	andvar	fix various typos in comments, documentation and messages.
1.69	01-Mar-2020	thorpej	- Use the 64-bit PCI DMA tag if available. - In age_dma_alloc(), ensure that all of the control data end up in the same 4GB segment, because the hardware requires this.
1.68	01-Mar-2020	thorpej	It doesn't make any sense to pass ETHER_ALIGN as the alignment constraint to bus_dmamem_alloc(). Use PAGE_SIZE instead.
1.67	04-Feb-2020	thorpej	Use ifmedia_fini().
1.66	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.65	01-Dec-2019	msaitoh	branches: 1.65.2; Use unsigned to avoid undefined behavoir. Found by kUBSan.
1.64	21-Nov-2019	msaitoh	Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt.
1.63	21-Nov-2019	msaitoh	Fix multicast handling. All Atheros controllers use big-endian form when computing multicast hash. Same as OpenBSD.
1.62	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.61	20-Sep-2019	maxv	Fix direction of the loop. Found by the lgtm bot.
1.60	09-Jul-2019	msaitoh	branches: 1.60.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.59	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.58	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.57	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.56	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.55	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.54	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.53	26-Jun-2018	msaitoh	branches: 1.53.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.52	01-Jun-2018	maxv	Fix M_PKTHDR use in if_alc, if_age and if_ena. if_alc and if_age always put in _rxhead a M_PKTHDR-flagged mbuf, so the flag must always be present. Instead of manually adding the flag, add a KASSERT to ensure it is already there. If it weren't, there would be memory corruptions. Same in if_ena, but this one does not compile so we don't really care. Also, use m_remove_pkthdr to remove the flag, instead of doing it manually. This ensures the tags get freed (even though these drivers don't seem to be using mtags).
1.51	26-Sep-2017	knakahara	branches: 1.51.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.50	15-Dec-2016	ozaki-r	branches: 1.50.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.49	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.48	10-Jun-2016	ozaki-r	branches: 1.48.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.47	17-Feb-2016	christos	PR/50817: David Binderman: Delete dead code (copy from if_et.c)
1.46	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.45	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.44	29-Mar-2014	christos	branches: 1.44.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.43	21-Feb-2014	christos	PR/48606: Lloyd Parkes: Drivers not using ifp->if_input but using ether_input directly.
1.42	30-Mar-2013	christos	branches: 1.42.4; remove trailing whitespace
1.41	22-Jul-2012	matt	branches: 1.41.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.40	25-Oct-2011	bouyer	branches: 1.40.2; age_init() is called from age_watchdog() which is interrupt context, we can't sleep here or we get a DIAGNOSTIC panic when age_watchdog() fires. More correct bus_dma(9) usage in age_encap() Introduce a age_shutdown() to be called by pmf(9) at shutdown time, to stop the DMA engine. My system would't properly reboot without this. Be consistent in WAIT/NOWAIT use in init routines Use BUS_DMA_COHERENT where appropriate Rework the interrupt routine a bit, and ACK but do not disable interrupts here. There seems to be a race where interrupts would not be properly reenabled after this, leading do watchdog timeouts. With these changes, the on-board age(4) on ftp.fr.netbsd.org seems to be finally working properly.
1.39	20-Jul-2010	cegger	PR kern/38871: Fix VLAN
1.38	05-Apr-2010	joerg	branches: 1.38.2; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.37	24-Feb-2010	dyoung	branches: 1.37.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.36	19-Jan-2010	pooka	branches: 1.36.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.35	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.34	08-Oct-2009	cegger	Use m_pullup to handle defragmentation. Reinject mbuf into TX queue when it couldn't be loaded. diff for age(4) reviewed/discussed with plunky@ several weeks ago.
1.33	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.32	05-Aug-2009	cegger	make RX hw checksum available by default. TX hw checksum locks up interface.
1.31	05-Aug-2009	cegger	fix typo
1.30	05-Aug-2009	cegger	- Don't access VPD even if hardware advertised the capability. It seems that some revisions of the controllers hang while accessing the VPD. Because VPD access routine is now unused, nuke it. - Let TWSI reload EEPROM if VPD capability is detected. Reloading the EEPROM will also set the Ethernet address, so age(4) now reads AGE_PAR0 and AGE_PAR1 register to get the Ethernet address. This removes removes a lot of hacks and enhance readability a lot. - Double PHY reset timeout as it takes more time to take the PHY out of power-saving state. - Explicitly check power-saving state by checking undocumented PHY registers. If link is not up, poke undocumented registers to take PHY out of power-saving state. This is the same thing done by the Linux driver. - Don't rely on auto-clearing feature of master reset bit, just wait 1ms and check idle status of MAC. From FreeBSD via OpenBSD.
1.29	04-Aug-2009	cegger	support flow control. from OpenBSD. tested by hans@
1.28	28-Apr-2009	cegger	branches: 1.28.2; 1.28.4; on detach, deregister from pmf and free resource
1.27	26-Apr-2009	cegger	age_intr: don't return w/o ack. the interrupt. Add a comment when this special case can happen.
1.26	21-Apr-2009	tsutsui	Fix wrong arguments of bus_dmamem_free(9). XXX: not all resouces are freed properly on error path in age_dma_alloc().
1.25	03-Mar-2009	cegger	mimic fix from nfe(4): when bus_dmamap_create fails, then the map pointer is undefined. age_dma_free() assumes, the map pointer is NULL. Make the assumption always true by setting the map pointer to NULL in bus_dmamap_create() error path in order to prevent age_dma_free() to call bus_dmamap_destroy() on an invalid map.
1.24	03-Mar-2009	cegger	make multicast working. Diff from OpenBSD's if_age.c rev. 1.2.
1.23	23-Feb-2009	cegger	when attach routine fails, also free DMA memory and interrupt mapping. Diff from OpenBSD's if_age.c rev. 1.2. ok cube@
1.22	16-Feb-2009	cegger	make this compile w/o VLAN
1.21	11-Feb-2009	cegger	age_dma_free: set smb pointers to NULL for consistency age_stop: call mii_down() ok cube@
1.20	05-Feb-2009	dyoung	branches: 1.20.2; If the bus_dmamap_load_mbuf() status was non-zero, then the DMA map is undefined, so do not try to unload it.
1.19	05-Feb-2009	dyoung	Initialize ec_mii, and re-use more ethernet common code for media handling. Replace an anonymous constant by PCI_COMMAND_INTERRUPT_DISABLE. These changes have been compile-tested, only.
1.18	03-Feb-2009	cegger	Adapt age_stop() as our ethernet stack expects it. This makes 'ifconfig age0 down' working. Ok cube@
1.17	30-Jan-2009	cegger	age_dma_alloc(): Don't allocate more than really required. It is less likely to fail with ENOMEM when we are short on dma-safe memory. This happens more likely when running NetBSD as Xen Dom0.
1.16	30-Jan-2009	cegger	age_dma_alloc(): print error code to figure out why it failed.
1.15	30-Jan-2009	cegger	remove brackets from return statements. No functional change.
1.14	29-Jan-2009	cegger	fix error path: don't crash when age_dma_alloc() fails.
1.13	24-Jan-2009	cegger	- fix test for IPv4 csum. - No need to double negate TCP/UDP csum test
1.12	23-Jan-2009	cegger	Improve handling of transmitting large packets. Sending large packets works better, but still needs investigations.
1.11	22-Jan-2009	cegger	struct device -> device_t
1.10	22-Jan-2009	cegger	fix crash when bringing up the device and no cable is plugged in
1.9	21-Jan-2009	cegger	Roll our own ETHER_ALIGN #define like many other driver do. Use ETHER_ALIGN to calculate number of segments correctly (actually the same way as FreeBSD and OpenBSD do). Also use ETHER_ALIGN to correctly adjust new mbufs. This makes the driver finally work. Tested by Jochen Kunz: ping -nf 192.168.2.1 PING 192.168.2.1 (192.168.2.1): 56 data bytes ----192.168.2.1 PING Statistics---- 28949 packets transmitted, 28949 packets received, 0.0% packet loss round-trip min/avg/max/stddev = 0.157/0.364/2.272/0.023 ms 2710.6 packets/sec sent, 2710.6 packets/sec received XXX Shouldn't if_ether.h provide ETHER_ALIGN instead of defining it in all drivers?
1.8	18-Jan-2009	cegger	branches: 1.8.2; fix botched age_ioctl(): don't try to roll its own handling when ether_ioctl() is doing all of it right. Tested by Jochen Kunz.
1.7	18-Jan-2009	cegger	beautify dmesg output
1.6	17-Jan-2009	cegger	branches: 1.6.2; Until we do WOL, the PCI power management suspend code is a noop.
1.5	16-Jan-2009	cegger	don't call age_stop and age_init in suspend/resume. pmf network class is already doing that. pointed out by jmcneill
1.4	16-Jan-2009	cegger	rename PCI_PMCSR_PME to PCI_PMCSR_PME_STS. Per request from jmcneill
1.3	16-Jan-2009	cegger	- fix age_read_vpd_word(). - add power management support. From FreeBSD.
1.2	16-Jan-2009	cegger	add RCSID
1.1	16-Jan-2009	cegger	Driver for Attansic L1 gigabit ethernet driver. Written by Pyun YongHyeon for FreeBSD, ported to DragonFlyBSD by Sepherosa Ziehau, ported to OpenBSD by Kevin Lo and then ported to NetBSD by me. XXX Driver needs testing.
1.6.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.6.2.1	17-Jan-2009	mjf	file if_age.c was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.8.2.4	28-Apr-2009	skrll	Sync with HEAD.
1.8.2.3	03-Mar-2009	skrll	Sync with HEAD.
1.8.2.2	19-Jan-2009	skrll	Sync with HEAD.
1.8.2.1	18-Jan-2009	skrll	file if_age.c was added on branch nick-hppapmap on 2009-01-19 13:18:25 +0000
1.20.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.28.4.6	11-Aug-2010	yamt	sync with head.
1.28.4.5	11-Mar-2010	yamt	sync with head
1.28.4.4	16-Sep-2009	yamt	sync with head
1.28.4.3	19-Aug-2009	yamt	sync with head.
1.28.4.2	04-May-2009	yamt	sync with head.
1.28.4.1	28-Apr-2009	yamt	file if_age.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:55 +0000
1.28.2.6	18-Nov-2011	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1694): sys/dev/pci/if_age.c: revision 1.40 via patch - age_init() is called from age_watchdog() which is in interrupt context; we can't sleep here or we get a DIAGNOSTIC panic when age_watchdog() fires. - More correct bus_dma(9) usage in age_encap() - Introduce a age_shutdown() to be called by pmf(9) at shutdown time, to stop the DMA engine. - Be consistent in WAIT/NOWAIT use in init routines - Use BUS_DMA_COHERENT where appropriate - Rework the interrupt routine a bit, and ACK but do not disable interrupts here. There seems to be a race where interrupts would not be properly reenabled after this, leading do watchdog timeouts.
1.28.2.5	08-Nov-2009	snj	Pull up following revision(s) (requested by cegger in ticket #1124): sys/dev/pci/if_age.c: revision 1.34 sys/dev/pci/if_ale.c: revision 1.9 Use m_pullup to handle defragmentation. Reinject mbuf into TX queue when it couldn't be loaded. diff for age(4) reviewed/discussed with plunky@ several weeks ago.
1.28.2.4	08-Nov-2009	snj	Pull up following revision(s) (requested by cegger in ticket #1124): sys/dev/pci/if_age.c: revision 1.30 - Don't access VPD even if hardware advertised the capability. It seems that some revisions of the controllers hang while accessing the VPD. Because VPD access routine is now unused, nuke it. - Let TWSI reload EEPROM if VPD capability is detected. Reloading the EEPROM will also set the Ethernet address, so age(4) now reads AGE_PAR0 and AGE_PAR1 register to get the Ethernet address. This removes removes a lot of hacks and enhance readability a lot. - Double PHY reset timeout as it takes more time to take the PHY out of power-saving state. - Explicitly check power-saving state by checking undocumented PHY registers. If link is not up, poke undocumented registers to take PHY out of power-saving state. This is the same thing done by the Linux driver. - Don't rely on auto-clearing feature of master reset bit, just wait 1ms and check idle status of MAC. From FreeBSD via OpenBSD.
1.28.2.3	08-Nov-2009	snj	Pull up following revision(s) (requested by cegger in ticket #1124): sys/dev/pci/if_age.c: revision 1.29 support flow control. from OpenBSD. tested by hans@
1.28.2.2	03-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #735): sys/dev/pci/if_age.c: revisions 1.1-1.28 sys/dev/pci/if_agereg.h: revisions 1.1-1.2 sys/dev/pci/if_ale.c: revisions 1.1-1.3 sys/dev/pci/if_alereg.h: revisions 1.1-1.2 sys/dev/pci/files.pci: revisions 1.310-1.313 sys/dev/mii/atphy.c: revisions 1.1-1.5 sys/dev/mii/files.mii: revision 1.42 share/man/man4/age.4: revisions 1.1-1.3 share/man/man4/ale.4: revisions 1.1-1.2 share/man/man4/atphy.4: revisions 1.1-1.2 share/man/man4/Makefile: revisions 1.486, 1.489 distrib/sets/lists/man/mi: revisions 1.1118, 1.1132 sys/arch/i386/conf/GENERIC: revisions 1.928, 1.933 sys/arch/i386/conf/XEN2_DOM0: revisions 1.57, 1.60 sys/arch/amd64/conf/GENERIC: revisions 1.239, 1.241 sys/arch/amd64/conf/XEN3_DOM0: revisions 1.38, 1.42 Add the age(4) and ale(4) drivers for Attansic L1 and Atheros AR8121/AR8113/AR8114 devices respectively.
1.28.2.1	28-Apr-2009	snj	file if_age.c was added on branch netbsd-5 on 2009-05-03 23:45:47 +0000
1.36.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.36.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.37.2.2	05-Mar-2011	rmind	sync with head
1.37.2.1	30-May-2010	rmind	sync with head
1.38.2.2	21-Apr-2010	matt	sync to netbsd-5
1.38.2.1	05-Apr-2010	matt	file if_age.c was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.40.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.40.2.1	30-Oct-2012	yamt	sync with head
1.41.2.3	03-Dec-2017	jdolecek	update from HEAD
1.41.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.41.2.1	23-Jun-2013	tls	resync from head
1.42.4.1	18-May-2014	rmind	sync with head
1.44.6.4	05-Feb-2017	skrll	Sync with HEAD
1.44.6.3	09-Jul-2016	skrll	Sync with HEAD
1.44.6.2	19-Mar-2016	skrll	Sync with HEAD
1.44.6.1	06-Jun-2015	skrll	Sync with HEAD
1.48.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.50.8.3	05-Dec-2019	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1462): sys/dev/pci/if_age.c: revision 1.61, 1.63-1.64 via patch - Fix direction of the loop in age_get_macaddr(). - Fix multicast handling. All Atheros controllers use big-endian form when computing multicast hash. - Fix a bug that IFF_ALLMULTI is almost always set.
1.50.8.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.50.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.51.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.51.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.51.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.51.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.53.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.53.2.1	10-Jun-2019	christos	Sync with HEAD
1.60.2.5	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.60.2.4	26-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #483): sys/dev/pci/if_cas.c: revision 1.36 sys/dev/pci/if_alc.c: revision 1.44 sys/dev/pci/if_ale.c: revision 1.35 sys/dev/pci/if_ale.c: revision 1.36 sys/dev/pci/if_age.c: revision 1.64 Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt. Remove accidentally committed debug code. Sorry.
1.60.2.3	26-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #482): sys/dev/pci/if_age.c: revision 1.63 Fix multicast handling. All Atheros controllers use big-endian form when computing multicast hash. Same as OpenBSD.
1.60.2.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.60.2.1	22-Sep-2019	martin	Pull up following revision(s) (requested by maxv in ticket #216): sys/dev/pci/if_age.c: revision 1.61 Fix direction of the loop.
1.65.2.1	29-Feb-2020	ad	Sync with head.
1.3	28-Jul-2009	cegger	branches: 1.3.6; add TWSI registers. from OpenBSD.
1.2	23-Feb-2009	cegger	branches: 1.2.2; 1.2.4; remove unused age_if_flags softc member
1.1	16-Jan-2009	cegger	branches: 1.1.2; 1.1.4; 1.1.6; Driver for Attansic L1 gigabit ethernet driver. Written by Pyun YongHyeon for FreeBSD, ported to DragonFlyBSD by Sepherosa Ziehau, ported to OpenBSD by Kevin Lo and then ported to NetBSD by me. XXX Driver needs testing.
1.1.6.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.4.3	03-Mar-2009	skrll	Sync with HEAD.
1.1.4.2	19-Jan-2009	skrll	Sync with HEAD.
1.1.4.1	16-Jan-2009	skrll	file if_agereg.h was added on branch nick-hppapmap on 2009-01-19 13:18:25 +0000
1.1.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.1.2.1	16-Jan-2009	mjf	file if_agereg.h was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.2.4.3	19-Aug-2009	yamt	sync with head.
1.2.4.2	04-May-2009	yamt	sync with head.
1.2.4.1	23-Feb-2009	yamt	file if_agereg.h was added on branch yamt-nfs-mp on 2009-05-04 08:12:55 +0000
1.2.2.3	08-Nov-2009	snj	Pull up following revision(s) (requested by cegger in ticket #1124): sys/dev/pci/if_agereg.h: revision 1.3 add TWSI registers. from OpenBSD.
1.2.2.2	03-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #735): sys/dev/pci/if_age.c: revisions 1.1-1.28 sys/dev/pci/if_agereg.h: revisions 1.1-1.2 sys/dev/pci/if_ale.c: revisions 1.1-1.3 sys/dev/pci/if_alereg.h: revisions 1.1-1.2 sys/dev/pci/files.pci: revisions 1.310-1.313 sys/dev/mii/atphy.c: revisions 1.1-1.5 sys/dev/mii/files.mii: revision 1.42 share/man/man4/age.4: revisions 1.1-1.3 share/man/man4/ale.4: revisions 1.1-1.2 share/man/man4/atphy.4: revisions 1.1-1.2 share/man/man4/Makefile: revisions 1.486, 1.489 distrib/sets/lists/man/mi: revisions 1.1118, 1.1132 sys/arch/i386/conf/GENERIC: revisions 1.928, 1.933 sys/arch/i386/conf/XEN2_DOM0: revisions 1.57, 1.60 sys/arch/amd64/conf/GENERIC: revisions 1.239, 1.241 sys/arch/amd64/conf/XEN3_DOM0: revisions 1.38, 1.42 Add the age(4) and ale(4) drivers for Attansic L1 and Atheros AR8121/AR8113/AR8114 devices respectively.
1.2.2.1	23-Feb-2009	snj	file if_agereg.h was added on branch netbsd-5 on 2009-05-03 23:45:47 +0000
1.3.6.2	21-Apr-2010	matt	sync to netbsd-5
1.3.6.1	28-Jul-2009	matt	file if_agereg.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.56	09-Mar-2025	mlelstv	Add MBUFTRACE
1.55	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.54	29-Jun-2024	riastradh	branches: 1.54.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.53	17-Sep-2022	thorpej	alc_start(): Remove redundant IFQ_IS_EMPTY(). This very same condition is checked as soon as we enter the loop.
1.52	01-Mar-2020	thorpej	Upon further reflection, we actually do need to ensure that the Rx ring and the Rx Return ring are both in the same 4GB segment.
1.51	01-Mar-2020	thorpej	- Use the 64-bit PCI DMA tag if available. - Add a comment explaining that if we were to use more than one Rx ring / Rx Return ring, all of the Rx-related control data would need to be in the same 4G segment.
1.50	01-Mar-2020	thorpej	It doesn't make any sense to pass ETHER_ALIGN as the alignment constraint to bus_dmamem_alloc(). Use PAGE_SIZE instead.
1.49	08-Feb-2020	maxv	constify
1.48	04-Feb-2020	thorpej	Use ifmedia_fini().
1.47	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.46	18-Dec-2019	msaitoh	branches: 1.46.2; Cast to uint32_t to avoid undefined behavior. Found by kUBSan.
1.45	01-Dec-2019	msaitoh	Use unsigned to avoid undefined behavoir. Found by kUBSan.
1.44	21-Nov-2019	msaitoh	Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt.
1.43	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.42	17-Oct-2019	msaitoh	Add support for Killer E2400 and E2500.
1.41	17-Oct-2019	msaitoh	Some alc(4) fixes: - Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com ) - Apply FreeBSD r304574: > Correct DMA channel number selection on AR816x family of > controllers. For Gigabit Ethernet version of AR816x, AR813x/AR815x > except L1D controller, use vendor recommended ASPM parameters. > While here, increase alc_dma_burst array size. Broken H/W can > return bogus value in theory. - Use static. - Whitespace fix. Remove extra backslash.
1.40	15-Oct-2019	msaitoh	Move PCIe initiaization code from alc_attach() to alc_init_pcie(). No functional change.
1.39	15-Oct-2019	msaitoh	Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com )
1.38	09-Jul-2019	msaitoh	branches: 1.38.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.37	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.36	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.35	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.34	01-May-2019	msaitoh	Add ether_ioctl() call for SIOCSIFADDR and SIOCSIFFLAGS to make behavior the same as before. XXX need cleanup.
1.33	01-May-2019	msaitoh	Fix a wrong and extra ether_ioctl() call in alc_ioctl(). _IOWR types call broke ifreq by the first ether_ioct() and the second ether_ioctl() caused unexpected behavior. One of bug example is that SIOCGIFMEDIA always failed. This bug was added in if_alc.c rev. 1.16. XXX pullup-8
1.32	22-Apr-2019	msaitoh	This driver uses MII(4) and have hook SIOC[GS]IFMEDIA which just pass to ifmedia_ioctl(), the hook is not required because ether_ioctl has it (if_ethersubr.c rev. 1.160). This driver might require some additional fixes for SIOCSIFMTU and other ioctl()s.
1.31	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.30	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.29	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.28	26-Jun-2018	msaitoh	branches: 1.28.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.27	25-Jun-2018	msaitoh	Remove duplicated inclusion of net/bpf.h.
1.26	01-Jun-2018	maxv	Fix M_PKTHDR use in if_alc, if_age and if_ena. if_alc and if_age always put in _rxhead a M_PKTHDR-flagged mbuf, so the flag must always be present. Instead of manually adding the flag, add a KASSERT to ensure it is already there. If it weren't, there would be memory corruptions. Same in if_ena, but this one does not compile so we don't really care. Also, use m_remove_pkthdr to remove the flag, instead of doing it manually. This ensures the tags get freed (even though these drivers don't seem to be using mtags).
1.25	26-Sep-2017	knakahara	branches: 1.25.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.24	21-Dec-2016	maya	branches: 1.24.8; Less magic. 0x00000c00 is equal to PCIE_LCAP_ASPM.
1.23	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.22	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.21	10-Jun-2016	ozaki-r	branches: 1.21.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.20	17-Feb-2016	christos	PR/50818: David Binderman: Delete dead code (copied from if_et.c)
1.19	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.18	26-Jan-2016	christos	PR/50694: David Binderman: Fix wrong pasto constant.
1.17	24-Nov-2015	leot	Add NetBSD RCS Id.
1.16	24-Nov-2015	christos	kern/50456: Freddy DISSAUX: Fix packets lost with tcpdump.
1.15	08-Sep-2015	leot	Add some missing fixes to alc(4) to sync with the latest FreeBSD and OpenBSD alc(4)s. Fix a long standing bug in MAC statistics register access. One additional register was erroneously added in the MAC register set such that 7 TX statistics counters were wrong. (from OpenBSD, if_alc.c 1.29) Remove setting an initial assumed baudrate upon driver attach which is not necessarily correct, there might not even be a link when attaching. (from OpenBSD, if_alc.c 1.22) Add some missing bus_dmamap_sync()'s and sync the others with the FreeBSD code. (from OpenBSD, if_alc.c 1.21) Help with the watchdog timeouts seen when unplugging the cable from the alc(4) NIC while running or the NIC not working if the cable is not plugged in upon boot up. (from OpenBSD, if_alc.c 1.16) ok mrg@
1.14	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.13	19-Jan-2015	christos	CID 1245802: Result always 0.
1.12	18-Jan-2015	christos	PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.11	29-Mar-2014	christos	branches: 1.11.4; 1.11.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.10	21-Feb-2014	christos	PR/48606: Lloyd Parkes: Drivers not using ifp->if_input but using ether_input directly.
1.9	21-Apr-2013	msaitoh	branches: 1.9.4; Delete "PCI_" from PCIX and PICE capability registers.
1.8	30-Mar-2013	christos	remove trailing whitespace
1.7	09-Dec-2012	mrg	avoid illegal sleeps in the softint routine. XXX could probably do better by creating a watchdog thread.
1.6	22-Jul-2012	matt	branches: 1.6.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.5	29-Aug-2011	jmcneill	branches: 1.5.2; 1.5.8; have pci register itself as a module, and make PCI driver modules depend on it
1.4	23-Jun-2011	matt	When printing the interrupt string, be like the other driver and do "alc0: interrupting at msi 1" instead of "alc: msi 1".
1.3	08-Apr-2011	sborrill	branches: 1.3.4; PR kern/38871 Fix LAN on bge(4), alc(4). Flag VLAN capability in ec_capenable as used by network card drivers.
1.2	23-Feb-2011	jmcneill	branches: 1.2.2; 1.2.4; catch up with FreeBSD driver, adds support for AR815x chips
1.1	23-Feb-2011	jmcneill	add network driver for Atheros AR813x/AR815x ethernet controllers, based on a patch from fire crow on tech-net with additional bpf & detach fixes, module support, and a match for 8152 v2.0 devices. alc0 at pci3 dev 0 function 0: Attansic/Atheros L1C/L2C Ethernet alc0: ioapic0 pin 17 alc0: Ethernet address 00:26:6c:9e:d4:c1 ukphy0 at alc0 phy 0: L2 10/100 PHY (OUI 0x00c82e, model 0x0002), rev. 5 ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT-FDX, auto
1.2.4.3	21-Apr-2011	rmind	sync with head
1.2.4.2	05-Mar-2011	rmind	sync with head
1.2.4.1	23-Feb-2011	rmind	file if_alc.c was added on branch rmind-uvmplock on 2011-03-05 20:53:38 +0000
1.2.2.2	05-Mar-2011	bouyer	Sync with HEAD
1.2.2.1	23-Feb-2011	bouyer	file if_alc.c was added on branch bouyer-quota2 on 2011-03-05 15:10:23 +0000
1.3.4.2	06-Jun-2011	jruoho	Sync with HEAD.
1.3.4.1	08-Apr-2011	jruoho	file if_alc.c was added on branch jruoho-x86intr on 2011-06-06 09:08:10 +0000
1.5.8.1	13-Jun-2013	msaitoh	Pull up following revision(s) (requested by mrg in ticket #898): sys/dev/pci/if_alc.c: revision 1.7 avoid illegal sleeps in the softint routine. XXX could probably do better by creating a watchdog thread.
1.5.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.5.2.2	16-Jan-2013	yamt	sync with (a bit old) head
1.5.2.1	30-Oct-2012	yamt	sync with head
1.6.2.4	03-Dec-2017	jdolecek	update from HEAD
1.6.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.2.2	23-Jun-2013	tls	resync from head
1.6.2.1	25-Feb-2013	tls	resync with head
1.9.4.1	18-May-2014	rmind	sync with head
1.11.6.7	05-Feb-2017	skrll	Sync with HEAD
1.11.6.6	09-Jul-2016	skrll	Sync with HEAD
1.11.6.5	19-Mar-2016	skrll	Sync with HEAD
1.11.6.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.11.6.3	22-Sep-2015	skrll	Sync with HEAD
1.11.6.2	06-Jun-2015	skrll	Sync with HEAD
1.11.6.1	06-Apr-2015	skrll	Sync with HEAD
1.11.4.1	04-Feb-2015	snj	Pull up following revision(s) (requested by nonaka in ticket #473): share/man/man4/lii.4: revision 1.3 sys/dev/pci/if_alc.c: revision 1.12 sys/dev/pci/pcidevs: revision 1.1213 sys/dev/pci/if_alcreg.h: revision 1.3 share/man/man4/atphy.4: revision 1.3 share/man/man4/alc.4: revision 1.4 PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.21.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.24.8.5	05-Dec-2019	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1464): sys/dev/pci/if_alc.c 1.44 via patch sys/dev/pci/if_ale.c 1.35-1.36 via patch sys/dev/pci/if_cas.c 1.36 via patch - Fix a bug that IFF_ALLMULTI is almost always set. - Whitespace fixes.
1.24.8.4	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.24.8.3	14-May-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1268): sys/dev/pci/if_alc.c: revision 1.33 sys/dev/pci/if_alc.c: revision 1.34 Fix a wrong and extra ether_ioctl() call in alc_ioctl(). _IOWR types call broke ifreq by the first ether_ioct() and the second ether_ioctl() caused unexpected behavior. One of bug example is that SIOCGIFMEDIA always failed. This bug was added in if_alc.c rev. 1.16. XXX pullup-8 - Add ether_ioctl() call for SIOCSIFADDR and SIOCSIFFLAGS to make behavior the same as before. XXX need cleanup.
1.24.8.2	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.24.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.25.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.25.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.25.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.25.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.28.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.28.2.1	10-Jun-2019	christos	Sync with HEAD
1.38.2.4	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.38.2.3	26-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #483): sys/dev/pci/if_cas.c: revision 1.36 sys/dev/pci/if_alc.c: revision 1.44 sys/dev/pci/if_ale.c: revision 1.35 sys/dev/pci/if_ale.c: revision 1.36 sys/dev/pci/if_age.c: revision 1.64 Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt. Remove accidentally committed debug code. Sorry.
1.38.2.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.38.2.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #375): sys/dev/pci/if_alc.c: revision 1.40 sys/dev/pci/if_alc.c: revision 1.41 sys/dev/pci/if_alc.c: revision 1.42 sys/dev/pci/if_alc.c: revision 1.39 sys/dev/pci/if_alcreg.h: revision 1.8 share/man/man4/alc.4: revision 1.6 Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com ) Move PCIe initialization code from alc_attach() to alc_init_pcie(). No functional change. Some alc(4) fixes: - Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com ) - Apply FreeBSD r304574: > Correct DMA channel number selection on AR816x family of > controllers. For Gigabit Ethernet version of AR816x, AR813x/AR815x > except L1D controller, use vendor recommended ASPM parameters. > While here, increase alc_dma_burst array size. Broken H/W can > return bogus value in theory. - Use static. - Whitespace fix. Remove extra backslash. Add support for Killer E2400 and E2500.
1.46.2.1	29-Feb-2020	ad	Sync with head.
1.54.2.1	02-Aug-2025	perseant	Sync with HEAD
1.10	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.9	08-Feb-2020	maxv	constify
1.8	17-Oct-2019	msaitoh	branches: 1.8.2; Add support for Killer E2400 and E2500.
1.7	03-Sep-2019	msaitoh	No functional change to reduce diff against {Free,Open}BSD: - Fix typo in comment. - Modify white spaces.
1.6	29-Dec-2016	leot	branches: 1.6.16; 1.6.22; Do not access `sc' directly via the CSR_{READ,WRITE}_[124] macros. Patch provided by Amol via PR kern/51708. ok <Riastradh>
1.5	24-Nov-2015	leot	branches: 1.5.2; Add NetBSD RCS Id.
1.4	08-Sep-2015	leot	Add some missing fixes to alc(4) to sync with the latest FreeBSD and OpenBSD alc(4)s. Fix a long standing bug in MAC statistics register access. One additional register was erroneously added in the MAC register set such that 7 TX statistics counters were wrong. (from OpenBSD, if_alc.c 1.29) Remove setting an initial assumed baudrate upon driver attach which is not necessarily correct, there might not even be a link when attaching. (from OpenBSD, if_alc.c 1.22) Add some missing bus_dmamap_sync()'s and sync the others with the FreeBSD code. (from OpenBSD, if_alc.c 1.21) Help with the watchdog timeouts seen when unplugging the cable from the alc(4) NIC while running or the NIC not working if the cable is not plugged in upon boot up. (from OpenBSD, if_alc.c 1.16) ok mrg@
1.3	18-Jan-2015	christos	PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.2	23-Feb-2011	jmcneill	branches: 1.2.2; 1.2.4; 1.2.8; 1.2.20; 1.2.36; 1.2.38; catch up with FreeBSD driver, adds support for AR815x chips
1.1	23-Feb-2011	jmcneill	add network driver for Atheros AR813x/AR815x ethernet controllers, based on a patch from fire crow on tech-net with additional bpf & detach fixes, module support, and a match for 8152 v2.0 devices. alc0 at pci3 dev 0 function 0: Attansic/Atheros L1C/L2C Ethernet alc0: ioapic0 pin 17 alc0: Ethernet address 00:26:6c:9e:d4:c1 ukphy0 at alc0 phy 0: L2 10/100 PHY (OUI 0x00c82e, model 0x0002), rev. 5 ukphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT-FDX, auto
1.2.38.4	05-Feb-2017	skrll	Sync with HEAD
1.2.38.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.2.38.2	22-Sep-2015	skrll	Sync with HEAD
1.2.38.1	06-Apr-2015	skrll	Sync with HEAD
1.2.36.1	04-Feb-2015	snj	Pull up following revision(s) (requested by nonaka in ticket #473): share/man/man4/lii.4: revision 1.3 sys/dev/pci/if_alc.c: revision 1.12 sys/dev/pci/pcidevs: revision 1.1213 sys/dev/pci/if_alcreg.h: revision 1.3 share/man/man4/atphy.4: revision 1.3 share/man/man4/alc.4: revision 1.4 PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.2.20.1	03-Dec-2017	jdolecek	update from HEAD
1.2.8.2	06-Jun-2011	jruoho	Sync with HEAD.
1.2.8.1	23-Feb-2011	jruoho	file if_alcreg.h was added on branch jruoho-x86intr on 2011-06-06 09:08:11 +0000
1.2.4.2	05-Mar-2011	rmind	sync with head
1.2.4.1	23-Feb-2011	rmind	file if_alcreg.h was added on branch rmind-uvmplock on 2011-03-05 20:53:38 +0000
1.2.2.2	05-Mar-2011	bouyer	Sync with HEAD
1.2.2.1	23-Feb-2011	bouyer	file if_alcreg.h was added on branch bouyer-quota2 on 2011-03-05 15:10:23 +0000
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.6.22.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #375): sys/dev/pci/if_alc.c: revision 1.40 sys/dev/pci/if_alc.c: revision 1.41 sys/dev/pci/if_alc.c: revision 1.42 sys/dev/pci/if_alc.c: revision 1.39 sys/dev/pci/if_alcreg.h: revision 1.8 share/man/man4/alc.4: revision 1.6 Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com ) Move PCIe initialization code from alc_attach() to alc_init_pcie(). No functional change. Some alc(4) fixes: - Apply FreeBSD r218141: > alc_rev was used without initialization such that it failed to > apply AR8152 v1.0 specific initialization code. Fix this bug by > explicitly reading PCI device revision id via PCI accessor. > > Reported by: Gabriel Linder ( linder.gabriel <> gmail dot com ) - Apply FreeBSD r304574: > Correct DMA channel number selection on AR816x family of > controllers. For Gigabit Ethernet version of AR816x, AR813x/AR815x > except L1D controller, use vendor recommended ASPM parameters. > While here, increase alc_dma_burst array size. Broken H/W can > return bogus value in theory. - Use static. - Whitespace fix. Remove extra backslash. Add support for Killer E2400 and E2500.
1.6.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.6.16.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.8.2.1	29-Feb-2020	ad	Sync with head.
1.44	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.43	22-Aug-2022	thorpej	ale_encap(): Fix error logic in previous change for when the defrag works, but DMA mapping fails to use DMA resource shortage.
1.42	22-Aug-2022	thorpej	ale_start(): Replace "IFQ_DEQUEUE() -> IF_PREPEND() on failure" with "IFQ_POLL() -> IFQ_DEQUEUE() on success (or fatal error)".
1.41	22-Aug-2022	thorpej	ale_encap(): Use m_defrag(), not m_pullup(), to compact an mbuf chain if it has too many DMA segments. Let the caller handle dispose of the packet on fatal errors.
1.40	01-Mar-2020	thorpej	It doesn't make any sense to pass ETHER_ALIGN as the alignment constraint to bus_dmamem_alloc(). Use PAGE_SIZE instead.
1.39	04-Feb-2020	thorpej	Use ifmedia_fini().
1.38	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.37	01-Dec-2019	msaitoh	branches: 1.37.2; Use unsigned to avoid undefined behavoir. Found by kUBSan.
1.36	21-Nov-2019	msaitoh	Remove accidentally committed debug code. Sorry.
1.35	21-Nov-2019	msaitoh	Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt.
1.34	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.33	09-Jul-2019	msaitoh	branches: 1.33.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.32	28-May-2019	msaitoh	Add missing ETHER_LOCK().
1.31	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.30	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.29	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.28	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.27	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.26	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.25	15-Nov-2018	maxv	Remove the 'copy' argument from m_devget(), unused. While here rename off0->off.
1.24	26-Jun-2018	msaitoh	branches: 1.24.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.23	26-Sep-2017	knakahara	branches: 1.23.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.22	23-May-2017	ozaki-r	branches: 1.22.2; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.21	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.20	09-Feb-2016	ozaki-r	branches: 1.20.2; Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.19	26-Jan-2016	christos	PR/50693: David Binderman: Fix constant pasto.
1.18	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.17	29-Mar-2014	christos	branches: 1.17.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.16	21-Feb-2014	christos	PR/48606: Lloyd Parkes: Drivers not using ifp->if_input but using ether_input directly.
1.15	30-Mar-2013	christos	branches: 1.15.4; remove trailing whitespace
1.14	22-Jul-2012	matt	branches: 1.14.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.13	22-Jan-2011	cegger	branches: 1.13.4; Fixes from PR kern/44395 Masanori Kanaoka: - Fix register address in ale_phy_reset(). from linux - Fix mask value in ale_stop_mac(). - Fix multicast handling. from openbsd - Fix phy reset handling.
1.12	20-Jul-2010	cegger	branches: 1.12.2; 1.12.4; PR kern/38871: Fix VLAN
1.11	05-Apr-2010	joerg	branches: 1.11.2; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.10	19-Jan-2010	pooka	branches: 1.10.2; 1.10.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.9	08-Oct-2009	cegger	Use m_pullup to handle defragmentation. Reinject mbuf into TX queue when it couldn't be loaded. diff for age(4) reviewed/discussed with plunky@ several weeks ago.
1.8	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.7	18-Aug-2009	cegger	fix multicast.
1.6	18-Aug-2009	cegger	don't read/write 1000baseT specific MII registers on 10/100M only device. from OpenBSD.
1.5	05-Aug-2009	cegger	support flow control. from OpenBSD. tested by Kevin Lahey
1.4	28-Jul-2009	cegger	print chipset name. From OpenBSD.
1.3	28-Apr-2009	cegger	branches: 1.3.2; 1.3.4; 1.3.6; on detach, deregister from pmf
1.2	25-Apr-2009	tsutsui	branches: 1.2.2; Add RCS Id.
1.1	20-Apr-2009	cegger	Add ale(4), a driver for Attansic/Atheros L1E gigabit ethernet device. Ported by me from OpenBSD, improvements from Kevin Lahey, successfully tested by Kevin Lahey and Stephen Borrill.
1.2.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.2.2.1	25-Apr-2009	skrll	file if_ale.c was added on branch nick-hppapmap on 2009-04-28 07:35:56 +0000
1.3.6.2	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.3.6.1	28-Apr-2009	jym	file if_ale.c was added on branch jym-xensuspend on 2009-05-13 17:20:24 +0000
1.3.4.6	11-Aug-2010	yamt	sync with head.
1.3.4.5	11-Mar-2010	yamt	sync with head
1.3.4.4	16-Sep-2009	yamt	sync with head
1.3.4.3	19-Aug-2009	yamt	sync with head.
1.3.4.2	04-May-2009	yamt	sync with head.
1.3.4.1	28-Apr-2009	yamt	file if_ale.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:55 +0000
1.3.2.4	08-Nov-2009	snj	Pull up following revision(s) (requested by cegger in ticket #1124): sys/dev/pci/if_age.c: revision 1.34 sys/dev/pci/if_ale.c: revision 1.9 Use m_pullup to handle defragmentation. Reinject mbuf into TX queue when it couldn't be loaded. diff for age(4) reviewed/discussed with plunky@ several weeks ago.
1.3.2.3	10-Sep-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #941): sys/dev/pci/if_ale.c: revision 1.4 sys/dev/pci/if_ale.c: revision 1.5 sys/dev/pci/if_ale.c: revision 1.6 sys/dev/pci/if_ale.c: revision 1.7 print chipset name. From OpenBSD. support flow control. from OpenBSD. tested by Kevin Lahey don't read/write 1000baseT specific MII registers on 10/100M only device. from OpenBSD. fix multicast.
1.3.2.2	03-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #735): sys/dev/pci/if_age.c: revisions 1.1-1.28 sys/dev/pci/if_agereg.h: revisions 1.1-1.2 sys/dev/pci/if_ale.c: revisions 1.1-1.3 sys/dev/pci/if_alereg.h: revisions 1.1-1.2 sys/dev/pci/files.pci: revisions 1.310-1.313 sys/dev/mii/atphy.c: revisions 1.1-1.5 sys/dev/mii/files.mii: revision 1.42 share/man/man4/age.4: revisions 1.1-1.3 share/man/man4/ale.4: revisions 1.1-1.2 share/man/man4/atphy.4: revisions 1.1-1.2 share/man/man4/Makefile: revisions 1.486, 1.489 distrib/sets/lists/man/mi: revisions 1.1118, 1.1132 sys/arch/i386/conf/GENERIC: revisions 1.928, 1.933 sys/arch/i386/conf/XEN2_DOM0: revisions 1.57, 1.60 sys/arch/amd64/conf/GENERIC: revisions 1.239, 1.241 sys/arch/amd64/conf/XEN3_DOM0: revisions 1.38, 1.42 Add the age(4) and ale(4) drivers for Attansic L1 and Atheros AR8121/AR8113/AR8114 devices respectively.
1.3.2.1	28-Apr-2009	snj	file if_ale.c was added on branch netbsd-5 on 2009-05-03 23:45:47 +0000
1.10.4.2	05-Mar-2011	rmind	sync with head
1.10.4.1	30-May-2010	rmind	sync with head
1.10.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.10.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.11.2.2	21-Apr-2010	matt	sync to netbsd-5
1.11.2.1	05-Apr-2010	matt	file if_ale.c was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.12.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.12.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.13.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.13.4.1	30-Oct-2012	yamt	sync with head
1.14.2.3	03-Dec-2017	jdolecek	update from HEAD
1.14.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.14.2.1	23-Jun-2013	tls	resync from head
1.15.4.1	18-May-2014	rmind	sync with head
1.17.6.4	28-Aug-2017	skrll	Sync with HEAD
1.17.6.3	05-Feb-2017	skrll	Sync with HEAD
1.17.6.2	19-Mar-2016	skrll	Sync with HEAD
1.17.6.1	06-Jun-2015	skrll	Sync with HEAD
1.20.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.22.2.3	05-Dec-2019	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1464): sys/dev/pci/if_alc.c 1.44 via patch sys/dev/pci/if_ale.c 1.35-1.36 via patch sys/dev/pci/if_cas.c 1.36 via patch - Fix a bug that IFF_ALLMULTI is almost always set. - Whitespace fixes.
1.22.2.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.22.2.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.23.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.23.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.23.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.23.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.24.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.24.2.1	10-Jun-2019	christos	Sync with HEAD
1.33.2.3	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.33.2.2	26-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #483): sys/dev/pci/if_cas.c: revision 1.36 sys/dev/pci/if_alc.c: revision 1.44 sys/dev/pci/if_ale.c: revision 1.35 sys/dev/pci/if_ale.c: revision 1.36 sys/dev/pci/if_age.c: revision 1.64 Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt. Remove accidentally committed debug code. Sorry.
1.33.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.37.2.1	29-Feb-2020	ad	Sync with head.
1.4	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.3	03-Sep-2019	msaitoh	No functional change to reduce diff against {Free,Open}BSD: - Fix typo in comment. - Modify white spaces.
1.2	25-Apr-2009	tsutsui	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.14; 1.2.74; Add RCS Id.
1.1	20-Apr-2009	cegger	Add ale(4), a driver for Attansic/Atheros L1E gigabit ethernet device. Ported by me from OpenBSD, improvements from Kevin Lahey, successfully tested by Kevin Lahey and Stephen Borrill.
1.2.74.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.2.14.2	21-Apr-2010	matt	sync to netbsd-5
1.2.14.1	25-Apr-2009	matt	file if_alereg.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:40 +0000
1.2.8.2	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.2.8.1	25-Apr-2009	jym	file if_alereg.h was added on branch jym-xensuspend on 2009-05-13 17:20:24 +0000
1.2.6.2	04-May-2009	yamt	sync with head.
1.2.6.1	25-Apr-2009	yamt	file if_alereg.h was added on branch yamt-nfs-mp on 2009-05-04 08:12:56 +0000
1.2.4.2	03-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #735): sys/dev/pci/if_age.c: revisions 1.1-1.28 sys/dev/pci/if_agereg.h: revisions 1.1-1.2 sys/dev/pci/if_ale.c: revisions 1.1-1.3 sys/dev/pci/if_alereg.h: revisions 1.1-1.2 sys/dev/pci/files.pci: revisions 1.310-1.313 sys/dev/mii/atphy.c: revisions 1.1-1.5 sys/dev/mii/files.mii: revision 1.42 share/man/man4/age.4: revisions 1.1-1.3 share/man/man4/ale.4: revisions 1.1-1.2 share/man/man4/atphy.4: revisions 1.1-1.2 share/man/man4/Makefile: revisions 1.486, 1.489 distrib/sets/lists/man/mi: revisions 1.1118, 1.1132 sys/arch/i386/conf/GENERIC: revisions 1.928, 1.933 sys/arch/i386/conf/XEN2_DOM0: revisions 1.57, 1.60 sys/arch/amd64/conf/GENERIC: revisions 1.239, 1.241 sys/arch/amd64/conf/XEN3_DOM0: revisions 1.38, 1.42 Add the age(4) and ale(4) drivers for Attansic L1 and Atheros AR8121/AR8113/AR8114 devices respectively.
1.2.4.1	25-Apr-2009	snj	file if_alereg.h was added on branch netbsd-5 on 2009-05-03 23:45:47 +0000
1.2.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.2.2.1	25-Apr-2009	skrll	file if_alereg.h was added on branch nick-hppapmap on 2009-04-28 07:35:56 +0000
1.38	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.37	18-Feb-2020	msaitoh	Whitespace fix. No functional change.
1.36	09-Dec-2018	jdolecek	branches: 1.36.6; use pci_intr_establish_xname() everywhere
1.35	29-Mar-2014	christos	branches: 1.35.28; 1.35.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.34	23-Sep-2012	chs	branches: 1.34.2; match some more devices.
1.33	30-Jan-2012	drochner	branches: 1.33.2; 1.33.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.32	26-Jul-2011	dyoung	branches: 1.32.2; 1.32.6; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.31	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.30	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.29	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.28	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.27	03-Jul-2008	drochner	branches: 1.27.10; split device/softc (tested with an@pcmcia)
1.26	28-Apr-2008	martin	branches: 1.26.2; 1.26.4; Remove clause 3 and 4 from TNF licenses
1.25	10-Apr-2008	cegger	branches: 1.25.2; 1.25.4; use aprint_*_dev and device_xname
1.24	09-Dec-2007	jmcneill	branches: 1.24.10; Merge jmcneill-pm branch.
1.23	19-Oct-2007	ad	branches: 1.23.4; 1.23.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.22	06-Apr-2007	rumble	branches: 1.22.8; 1.22.10; 1.22.14; If we fail to attach, use the size determined by pci_mapreg_map() rather than a constant, since this will be incorrect on some chips (MPI350) and lead to a panic.
1.21	16-Nov-2006	christos	branches: 1.21.4; 1.21.8; 1.21.10; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	11-Dec-2005	christos	branches: 1.19.20; 1.19.22; merge ktrace-lwp.
1.18	22-Jun-2005	dyoung	branches: 1.18.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.17	27-Feb-2005	perry	nuke trailing whitespace
1.16	21-Aug-2004	thorpej	branches: 1.16.4; 1.16.6; Use ANSI function decls and make use of static.
1.15	05-Aug-2004	mycroft	Fix an Aironet product name, and normalize the usage of "Mini-PCI".
1.14	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.13	29-Jan-2004	martin	Fix copy&pasto in last revision that (luckily) broke compilation on sparc64. Noticed by MOCHIZUKI Tomomitsu.
1.12	28-Jan-2004	onoe	Adopt to new 802.11 framework. Add support of version 5.30.17 firmware of PCMCIA 350 series. MPI350 mini-PCI is NOT YET supported. MONITOR mode is still not yet working.
1.11	13-Oct-2003	dyoung	Switch an(4) to using the header files in the new 802.11 layer.
1.10	31-Jan-2003	thorpej	branches: 1.10.2; Use aprint_*().
1.9	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.8	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.7	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.6	13-Nov-2001	lukem	add RCSID
1.5	08-Jul-2001	thorpej	branches: 1.5.2; Remove unnecessary include files. From Onno van der Linden.
1.4	18-Jun-2001	onoe	Add Cisco/Aironet 350 Reported by Wolfgang Rupprecht <wolfgang@wsrcc.com>. This doesn't fix PR 13181, which says the card doesn't work well.
1.3	12-Jun-2001	thorpej	Don't need NS-related headers here.
1.2	28-Dec-2000	sommerfeld	branches: 1.2.2; 1.2.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.1	14-Dec-2000	onoe	Add PCI version of Aironet, not tested at all. ISA/ISApnp version won't come since I don't know about isa...
1.2.4.4	18-Oct-2002	nathanw	Catch up to -current.
1.2.4.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.2.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.2.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.1	28-Dec-2000	bouyer	file if_an_pci.c was added on branch thorpej_scsipi on 2001-01-05 17:36:04 +0000
1.5.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.5.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.10.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.10.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.16.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.16.4.1	29-Apr-2005	kent	sync with -current
1.18.2.4	21-Jan-2008	yamt	sync with head
1.18.2.3	27-Oct-2007	yamt	sync with head.
1.18.2.2	03-Sep-2007	yamt	sync with head.
1.18.2.1	30-Dec-2006	yamt	sync with head.
1.19.22.2	10-Dec-2006	yamt	sync with head.
1.19.22.1	22-Oct-2006	yamt	sync with head
1.19.20.1	18-Nov-2006	ad	Sync with head.
1.21.10.1	11-Jul-2007	mjf	Sync with head.
1.21.8.2	23-Oct-2007	ad	Sync with head.
1.21.8.1	10-Apr-2007	ad	Sync with head.
1.21.4.1	15-Apr-2007	yamt	sync with head.
1.22.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.22.10.2	09-Jan-2008	matt	sync with HEAD
1.22.10.1	06-Nov-2007	matt	sync with HEAD
1.22.8.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.22.8.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.22.8.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.22.8.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.22.8.1	04-Sep-2007	joerg	Convert an(4) to new style power management.
1.23.6.1	11-Dec-2007	yamt	sync with head.
1.23.4.1	26-Dec-2007	ad	Sync with head.
1.24.10.2	28-Sep-2008	mjf	Sync with HEAD.
1.24.10.1	02-Jun-2008	mjf	Sync with HEAD.
1.25.4.5	11-Mar-2010	yamt	sync with head
1.25.4.4	16-Sep-2009	yamt	sync with head
1.25.4.3	16-May-2009	yamt	sync with head
1.25.4.2	04-May-2009	yamt	sync with head.
1.25.4.1	16-May-2008	yamt	sync with head.
1.25.2.1	18-May-2008	yamt	sync with head.
1.26.4.1	18-Jul-2008	simonb	Sync with head.
1.26.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.27.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.32.6.1	18-Feb-2012	mrg	merge to -current.
1.32.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.32.2.2	30-Oct-2012	yamt	sync with head
1.32.2.1	17-Apr-2012	yamt	sync with head
1.33.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.33.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.33.2.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #691): sys/dev/pci/aac_pci.c: revision 1.34 sys/dev/pci/if_sip.c: revision 1.155 sys/dev/pci/if_tlp_pci.c: revision 1.122 sys/dev/pci/if_an_pci.c: revision 1.34 match some more devices.
1.34.2.1	18-May-2014	rmind	sync with head
1.35.30.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.35.30.1	10-Jun-2019	christos	Sync with HEAD
1.35.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.36.6.1	29-Feb-2020	ad	Sync with head.
1.50	26-Feb-2025	andvar	Fix typos in comments, mainly s/calcurate/calculate/.
1.49	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.48	29-Jun-2024	riastradh	branches: 1.48.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.47	09-Mar-2024	mrg	aq(4): always poll for link status some devices don't have working link status and rather than have a likely incomplete list of issues, always poll as well as use the interrupt if possible. fixes link status on this device: aq0 at pci5 dev 0 function 0: Aquantia AQC107 10 Gigabit Network Adapter (rev. 0x02) aq0: Atlantic revision B1, F/W version 3.1.88 (was otherwise functional, just didn't report status, which likely meant eg, dhcpcd would be upset?) idea via mlelstv@ from linux. remove sc_detect_linkstat and rename sc_poll_linkstat to sc_no_link_intr, as the meaning has changed. simplify the signature for aq_setup_msix() and aq_establish_msix_intr(), removing forward decls that aren't required. obsolete AQ_FORCE_POLL_LINKSTAT.
1.46	07-Feb-2024	msaitoh	Remove ryo@'s mail addresses.
1.45	29-May-2023	rin	aq(4): if_transmit: Invoke softint_schedule(9) with kpreempt disabled. XXX Pull up to netbsd-10 and netbsd-9.
1.44	26-Jan-2023	ryo	- Don't discard a packet even if the TX descriptor is temporarily not enough, and error messages are not output. - The argument to aq_encap_txring() need not be an mbuf reference, pass an entity. pointed out by nisimura@. thanks
1.43	14-Jan-2023	ryo	Added support for the Aquantia (Marvell) AQC113 10G Network Adapter and the variants, to aq(4)
1.42	14-Jan-2023	ryo	fix indentation
1.41	14-Jan-2023	ryo	- avoid panic when failing during attach or detach with modload/drvctl. - free workqueue resources when detaching. - remove debug message.
1.40	14-Jan-2023	ryo	fix build error in sys/module/if_aq/ avoid error of "comparison of integer expressions of different signedness" and -Werror=sign-compare
1.39	02-Nov-2022	andvar	branches: 1.39.2; fix various typos in comments and messages.
1.38	21-Oct-2022	riastradh	aq(4): Remove incorrect ASSERT_SLEEPABLE introduced in 1.33. It is true that aq_stop must be sleepable, and that aq_stop_locked may sleep, but aq_stop_locked will release sc->sc_mutex when it sleeps, which ASSERT_SLEEPABLE doesn't know about. Since that is a spin lock, ASSERT_SLEEPABLE trips over it and crashes. This is not quite right: aq_stop should really only take the lock over specific things that need the lock to synchronize with other threads, like mii_down; aq_init and aq_stop are already serialized by IFNET_LOCK. sc->sc_mutex is used for too much and should have its scope narrowed like I did recently in usbnet(9). But this small change will at least remove a source of crashes for now.
1.37	17-Oct-2022	riastradh	aq(4): Annotate boolean parameters with names for legibility. No functional change intended.
1.36	17-Oct-2022	riastradh	aq(4): Unconditionally halt callout in aq_stop. Fixes panic with callout still running on detach after we destroy the lock, reported by andvar@: fatal page fault in supervisor mode trap type 6 code 0 rip 0xffffffff80dfafec cs 0x8 rflags 0x10286 cr2 0xfffffffffffffff0 ilevel 0x2 rsp 0xffffcd085b291ee0 Skipping crash dump on recursive panic panic: trap cpu0: Begin traceback... vpanic() at netbsd:vpanic+0x183 panic() at netbsd:panic+0x3c trap() at netbsd:trap+0xb27 --- trap (number 6) --- mutex_oncpu() at netbsd:mutex_oncpu+0x1e mutex_vector_enter() at netbsd:mutex_vector_enter+0xb7 aq_tick() at netbsd:aq_tick+0x23 callout_softclock() at netbsd:callout_softclock+0xbd softint_dispatch() at netbsd:softint_dispatch+0xf9 DDB lost frame for netbsd:Xsoftintr+0x4c, trying 0xffffcd085b2920f0 Xsoftintr() at netbsd:Xsoftintr+0x4c --- interrupt --- fa0b2181724b21c1: cpu0: End traceback...
1.35	22-Sep-2022	skrll	Take and release IFNET_LOCK around aq_stop in aq_detach. Also, call with the disable flag set so that interrupts and the callout are halted.
1.34	22-Sep-2022	riastradh	aq(4): Don't schedule tick callout on interrupt if stopping. Make sure to take the lock around access to sc_detect_linkstat too.
1.33	16-Sep-2022	skrll	Some MP improvements - Remove use of IFF_OACTIVE - Remove use of if_timer and provide an MP safe multiqueue watchdog - Sprinkle some lock assertions. Tested by ryo@. Thanks.
1.32	08-Sep-2022	skrll	Sprinkle const. Same binary before and after.
1.31	13-Nov-2021	ryo	firmware ver1 doesn't seem to work with linkstat interrupt, so in this case also set it to polling mode. - Make the version acquisition (aq_fw_reset) timing earlier. - Unmap registers when terminating with attach after failing to set interrupt
1.30	11-Nov-2021	ryo	Fixed a panic problem at attach with aq(4) F/W version 1. - Don't use cprng(9) when attaching, since it's not available yet. - Wait up to 10 seconds because delay is not enough. - Even if the delay is not enough, it will fail attach without panic. - Checked on the actual card D107 (F/W version 1.5.58)
1.29	11-Oct-2021	msaitoh	Add pmf_device_register().
1.28	05-Oct-2021	ryo	fix a panic "m_verify_packet: inconsistent mbuf length" on aq(4). - If mbuf cannot be allocated or some errors occur when receiving a jumboframe, it is necessary to free mbuf chains of the packet being received, and ignore the subsequent segments that make up the packet. - Even if aq_rx_intr() is completed in the middle of the jumboframe reception process, it will resume normally at the next aq_rx_intr().
1.27	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.26	13-Jun-2021	mlelstv	Don't ignore errors of sysmon_envsys_register().
1.25	16-Apr-2021	ryo	branches: 1.25.2; fix typo
1.24	16-Apr-2021	ryo	when the packet size was just n*MCLBYTES, there was an inconsistency in the length of mbuf.
1.23	15-Apr-2021	ryo	mtu for revision B0 and B1 is supported up to 16338 bytes
1.22	15-Apr-2021	ryo	fixed the problem of instability when changing mtu in IFF_UP state
1.21	15-Apr-2021	ryo	add support hardware tcp/udp rx checksum offloading
1.20	18-Feb-2021	ryo	branches: 1.20.2; - use if_initialize() and if_register() instead of if_attach() - add IFEF_MPSAFE to if_extflags pointed out by msaitoh@, thanks
1.19	24-Sep-2020	ryo	branches: 1.19.2; add __aligned() to all __packed structures. suppress warnings, and allow to compiler to generate more efficent code.
1.18	24-Sep-2020	ryo	no need to swap 32bit-packed macaddr on big endian
1.17	14-May-2020	msaitoh	branches: 1.17.2; Remove extra semicolon.
1.16	24-Apr-2020	ryo	fix typo
1.15	24-Apr-2020	ryo	add support AQC100S and D100. not tested, but they are probably the same as the AQC100.
1.14	23-Apr-2020	ryo	sort product table, and tabify
1.13	23-Apr-2020	ryo	On FIBRE devices, there are times when linkstat interrupt doesn't occur? reported from Andrius V. thanks. - use polling instead of linkstat interrupt when FIBRE - add AQ_FORCE_POLL_LINKSTAT options (not by default)
1.12	22-Apr-2020	christos	Add AQUANTIA AQC100 from Andrius V.
1.11	15-Feb-2020	ryo	branches: 1.11.4; 1.11.6; return the ifmedia active status correctly even while the link is not up after attach. pointed out by msaitoh@. thanks.
1.10	10-Feb-2020	ryo	add support VLAN HW filter
1.9	10-Feb-2020	ryo	fix descriptions of register map in comment
1.8	08-Feb-2020	maxv	constify
1.7	04-Feb-2020	thorpej	Use ifmedia_fini().
1.6	31-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.5	25-Jan-2020	msaitoh	Found by kUBSan: - Use unsigned to avoid undefined behavior in aq_hw_init(). - Cast to unsigned to avoid undefined behavior in aq_set_mac_addr().
1.4	17-Jan-2020	ryo	support internal PHY temperature sensor
1.3	17-Jan-2020	ryo	make counters per queue
1.2	17-Jan-2020	ryo	set/clear IFF_OACTIVE flag only on txring 0
1.1	01-Jan-2020	ryo	branches: 1.1.2; add support Aquantia AQC seriese 10G network adapters. this driver is based on the FreeBSD version https://github.com/Aquantia/aqtion-freebsd , but drastically rewritten for NetBSD.
1.1.2.3	29-Feb-2020	ad	Sync with head.
1.1.2.2	25-Jan-2020	ad	Sync with head.
1.1.2.1	17-Jan-2020	ad	Sync with head.
1.11.6.1	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.11.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.11.4.1	15-Feb-2020	martin	file if_aq.c was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.17.2.4	03-Jun-2023	martin	Pull up following revision(s) (requested by rin in ticket #1637): sys/dev/pci/if_aq.c: revision 1.45 aq(4): if_transmit: Invoke softint_schedule(9) with kpreempt disabled.
1.17.2.3	07-Jul-2020	martin	Also pull up more cosmetic revisions to minimize diffs with HEAD, requested by jmcneill in ticket #980: sys/dev/pci/if_aq.c 1.8,1.17 constify, remove duplicate ;
1.17.2.2	07-Jul-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #980): sys/dev/pci/if_aq.c: revision 1.4 sys/dev/pci/if_aq.c: revision 1.5 sys/arch/amd64/conf/GENERIC: revision 1.553 sys/dev/pci/files.pci: revision 1.419 sys/arch/amd64/conf/XEN3_DOM0: revision 1.170 sys/dev/pci/if_aq.c: revision 1.9 share/man/man4/Makefile: revision 1.693 sys/dev/pci/pcidevs: revision 1.1411 share/man/man4/aq.4: revision 1.1 share/man/man4/aq.4: revision 1.3 sys/arch/i386/conf/ALL: revision 1.479 share/man/man4/aq.4: revision 1.4 sys/dev/pci/if_aq.c: revision 1.10 sys/dev/pci/files.pci: revision 1.421 sys/dev/pci/if_aq.c: revision 1.11 sys/dev/pci/if_aq.c: revision 1.12 sys/dev/pci/if_aq.c: revision 1.13 sys/dev/pci/if_aq.c: revision 1.14 sys/dev/pci/if_aq.c: revision 1.15 sys/dev/pci/if_aq.c: revision 1.16 sys/dev/pci/pcidevs: revision 1.1408 sys/arch/amd64/conf/ALL: revision 1.135 sys/net/ethertypes.h: revision 1.19 sys/arch/i386/conf/GENERIC: revision 1.1218 distrib/sets/lists/man/mi: revision 1.1668 sys/dev/pci/if_aq.c: revision 1.1 sys/dev/pci/if_aq.c: revision 1.2 sys/dev/pci/pcidevs: revision 1.1395 sys/dev/pci/if_aq.c: revision 1.3 sys/arch/evbarm/conf/GENERIC64: revision 1.125 Add the ETHERTYPE_QINQ for 802.1ad VLAN stacking add Aquantia AQC 10G network adapters add support Aquantia AQC seriese 10G network adapters. this driver is based on the FreeBSD version https://github.com/Aquantia/aqtion-freebsd , but drastically rewritten for NetBSD. add aq(4) Add Aquantia AQC100, AQC100S and D100. add support VLAN HW filter set/clear IFF_OACTIVE flag only on txring 0 make counters per queue support internal PHY temperature sensor Found by kUBSan: - Use unsigned to avoid undefined behavior in aq_hw_init(). - Cast to unsigned to avoid undefined behavior in aq_set_mac_addr(). fix descriptions of register map in comment return the ifmedia active status correctly even while the link is not up after attach. pointed out by msaitoh@. thanks. On FIBRE devices, there are times when linkstat interrupt doesn't occur? reported from Andrius V. thanks. - use polling instead of linkstat interrupt when FIBRE - add AQ_FORCE_POLL_LINKSTAT options (not by default) sort product table, and tabify add support AQC100S and D100. not tested, but they are probably the same as the AQC100.
1.17.2.1	14-May-2020	martin	file if_aq.c was added on branch netbsd-9 on 2020-07-07 10:29:05 +0000
1.19.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.20.2.1	17-Apr-2021	thorpej	Sync with HEAD.
1.25.2.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.39.2.2	03-Jun-2023	martin	Pull up following revision(s) (requested by rin in ticket #187): sys/dev/pci/if_aq.c: revision 1.45 aq(4): if_transmit: Invoke softint_schedule(9) with kpreempt disabled.
1.39.2.1	30-Jan-2023	martin	Pull up following revision(s) (requested by ryo in ticket #60): sys/dev/pci/if_aq.c: revision 1.40 sys/dev/pci/if_aq.c: revision 1.41 sys/dev/pci/if_aq.c: revision 1.42 sys/dev/pci/if_aq.c: revision 1.43 sys/dev/pci/if_aq.c: revision 1.44 share/man/man4/aq.4: revision 1.5 fix build error in sys/module/if_aq/ avoid error of "comparison of integer expressions of different signedness" and -Werror=sign-compare - avoid panic when failing during attach or detach with modload/drvctl. - free workqueue resources when detaching. - remove debug message. fix indentation Added support for the Aquantia (Marvell) AQC113 10G Network Adapter and the variants, to aq(4) - add AQC113 and the variants - fix first appeared version to 9.1. it was backported. - Don't discard a packet even if the TX descriptor is temporarily not enough, and error messages are not output. - The argument to aq_encap_txring() need not be an mbuf reference, pass an entity. pointed out by nisimura@. thanks
1.48.2.1	02-Aug-2025	perseant	Sync with HEAD
1.50	09-Dec-2018	jdolecek	put back the intr establish to ath_pci_resume() to avoid possibility of functionality change, but add XXX comment that it shouldn't be necessary
1.49	09-Dec-2018	jdolecek	use pci_intr_establish_xname() upon code inspection, remove intr re-establish from ath_pci_resume() - not needed to re-establish interrupt there, no other driver does this, and very likely it would actually fail nowadays
1.48	29-Mar-2014	christos	branches: 1.48.28; 1.48.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.47	30-Mar-2013	christos	branches: 1.47.4; remove trailing whitespace
1.46	04-Aug-2012	riastradh	branches: 1.46.2; Fix error branches in ath pci attachment. Should fix double pci_intr_disestablish if ath_attach fails. (Still not sure why ath_attach sometimes fails here.)
1.45	18-Oct-2011	dyoung	branches: 1.45.2; 1.45.8; Reduce differences from if_ath_cardbus.c.
1.44	07-Oct-2011	dyoung	Stop calling ATH_LOCK_{INIT,DESTROY}(), they don't exist any more.
1.43	29-Aug-2011	jmcneill	have pci register itself as a module, and make PCI driver modules depend on it
1.42	26-Jul-2011	dyoung	Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.41	21-Feb-2011	jmcneill	print the device name at attach, like so: ath0 at pci4 dev 0 function 0: Atheros 9280
1.40	20-Feb-2011	jmcneill	support building as a module
1.39	26-Jan-2011	dyoung	Improve error reporting during attachment to PCI ath(4).
1.38	28-Apr-2010	dyoung	branches: 1.38.2; 1.38.4; To reduce differences between ath(4)'s PCI attachment and its CardBus attachment, delete several unnecessary #includes, re-order subroutines, move around line breaks, change to different names for the same constant, et cetera. Comparing this version of if_ath_pci.c with my uncommitted version of if_ath_cardbus.c, only about five lines are different.
1.37	24-Feb-2010	dyoung	branches: 1.37.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.36	08-Jan-2010	dyoung	branches: 1.36.2; Expand PMF_FN_* macros.
1.35	16-Sep-2009	dyoung	In pmf(9), improve the implementation of device self-suspension and make suspension by self, by drvctl(8), and by ACPI system sleep play nice together. Start solidifying some temporary API changes. 1. Extract a new header file, <sys/device_if.h>, from <sys/device.h> and #include it from <sys/pmf.h> instead of <sys/device.h> to break the circular dependency between <sys/device.h> and <sys/pmf.h>. 2. Introduce pmf_qual_t, an aggregate of qualifications on a PMF suspend/resume call. Start to replace instances of PMF_FN_PROTO, PMF_FN_ARGS, et cetera, with a pmf_qual_t. 3. Introduce the notion of a "suspensor," an entity that holds a device in suspension. More than one suspensor may hold a device at once. A device stays suspended as long as at least one suspensor holds it. A device resumes when the last suspensor releases it. Currently, the kernel defines three suspensors, 3a the system-suspensor: for system suspension, initiated by 'sysctl -w machdep.sleep_state=3', by lid closure, by power-button press, et cetera, 3b the drvctl-suspensor: for device suspension by /dev/drvctl ioctl, e.g., drvctl -S sip0. 3c the system self-suspensor: for device drivers that suspend themselves and their children. Several drivers for network interfaces put the network device to sleep while it is not administratively up, that is, after the kernel calls if_stop(, 1). The self-suspensor should not be used directly. See the description of suspensor delegates, below. A suspensor can have one or more "delegates". A suspensor can release devices that its delegates hold suspended. Right now, only the system self-suspensor has delegates. For each device that a self-suspending driver attaches, it creates the device's self-suspensor, a delegate of the system self-suspensor. Suspensors stop a system-wide suspend/resume cycle from waking devices that the operator put to sleep with drvctl before the cycle. They also help self-suspension to work more simply, safely, and in accord with expectations. 4. Add the notion of device activation level, devact_level_t, and a routine for checking the current activation level, device_activation(). Current activation levels are DEVACT_LEVEL_BUS, DEVACT_LEVEL_DRIVER, and DEVACT_LEVEL_CLASS, which respectively indicate that the device's bus is active, that the bus and device are active, and that the bus, device, and the functions of the device's class (network, audio) are active. Suspend/resume calls can be qualified with a devact_level_t. The power-management framework treats a devact_level_t that qualifies a device suspension as the device's current activation level; it only runs hooks to reduce the activation level from the presumed current level to the fully suspended state. The framework treats a devact_level_t qualifying device resumption as the target activation level; it only runs hooks to raise the activation level to the target. 5. Use pmf_qual_t, devact_level_t, and self-suspensors in several drivers. 6. Temporarily add an unused power-management workqueue that I will remove or replace, soon.
1.34	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	11-Dec-2008	alc	branches: 1.32.2; Sync ath(4) with the new HAL, mostly based on <jmcneill>'s patches. Everything should be fine again !
1.31	09-Jul-2008	joerg	branches: 1.31.2; 1.31.4; 1.31.10; - device/softc split for ath(4)
1.30	12-Mar-2008	dyoung	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.10; Use device_t and its accessors throughout. Use aprint_*_dev(). Improve PMF-ability. Add a 'flags' argument to suspend/resume handlers and callers such as pmf_system_suspend(). Define a flag, PMF_F_SELF, which indicates to PMF that a device is suspending/resuming itself. Add helper routines, pmf_device_suspend_self(dev) and pmf_device_resume_self(dev), that call pmf_device_suspend(dev, PMF_F_SELF) and pmf_device_resume(dev, PMF_F_SELF), respectively. Use PMF_F_SELF to suspend/resume self in ath(4), audio(4), rtw(4), and sip(4). In ath(4) and in rtw(4), replace the icky sc_enable/sc_disable callbacks, provided by the bus front-end, with self-suspension/resumption. Also, clean up the bus front-ends. Make sure that the interrupt handler is disestablished during suspension. Get rid of driver-private flags (e.g., RTW_F_ENABLED, ath_softc->sc_invalid); use device_is_active()/device_has_power() calls, instead. In the network-class suspend handler, call if_stop(, 0) instead of if_stop(, 1), because the latter is superfluous (bus- and driver-suspension hooks will 'disable' the NIC), and it may cause recursion. In the network-class resume handler, prevent infinite recursion through if_init() by getting out early if we are self-suspending (PMF_F_SELF). rtw(4) improvements: Destroy rtw(4) callouts when we detach it. Make rtw at pci detachable. Print some more information with the "rx frame too long" warning. Remove activate() methods: Get rid of rtw_activate() and ath_activate(). The device activate() methods are not good for much these days. Make ath at cardbus resume with crypto functions intact: Introduce a boolean device property, "pmf-powerdown". If pmf-powerdown is present and false, it indicates that a bus back-end should not remove power from a device. Honor this property in cardbus_child_suspend(). Set this property to 'false' in ath_attach(), since removing power from an ath at cardbus seems to lobotomize the WPA crypto engine. XXX Should the pmf-powerdown property propagate toward the root of the device tree? Miscellaneous ath(4) changes: Warn if ath(4) tries to write crypto keys to suspended hardware. Reduce differences between FreeBSD and NetBSD in ath(4) multicast filter setup. Make ath_printrxbuf() print an rx descriptor's status & key index, to help debug crypto errors. Shorten a staircase in ath_ioctl(). Don't check for ieee80211_ioctl() return code ERESTART, it never happens.
1.29	07-Mar-2008	dyoung	In ath_pci_detach(), only disestablish the interrupt handler if it is established.
1.28	28-Jan-2008	dyoung	branches: 1.28.2; 1.28.6; Remove dead code and associated comments.
1.27	21-Dec-2007	dyoung	Initialize and destroy the ATH_LOCK.
1.26	15-Dec-2007	dyoung	In ath_pci_detach(), unmap bus memory.
1.25	14-Dec-2007	dyoung	Comment-out calls to pci_disable_retry() and cardbus_disable_retry(). They do not seem well-justified according to anyone's understanding of what they really do, and it seems especially inappropriate to call them at attach- and resume-time.
1.24	14-Dec-2007	mjf	Reorder the previous commit so we remember to disestablish pci interrupts if we fail to attach ath(4). Thanks dyoung!
1.23	14-Dec-2007	mjf	Fix a crash at shutdown/reboot that was caused by ath(4) registering with pmf's network class even though the ath device may not have attached properly. OK jmcneill
1.22	09-Dec-2007	jmcneill	branches: 1.22.2; Merge jmcneill-pm branch.
1.21	17-Apr-2007	dyoung	branches: 1.21.6; 1.21.8; 1.21.14; 1.21.16; 1.21.18; Cosmetic: fix whitespace, join two lines, remove extraneous #includes. #Define a bitmask using __BITS() instead of a hex constant, for easy comparison with paper documentation.
1.20	17-Apr-2007	dyoung	Add prototype for ath_pci_detach(). Alphabetize prototypes. Extract the code for disabling PCI retries, creating ath_disable_retry(), and call it both during device initialization and on ->PWR_RESUME transitions. Comment out the code in ath_disable_retry() for now, because it is not clear what problem it solves, if any. Set a reasonable PCI Latency Timer. Apparently, it is too much to ask any BIOS to do. The PCI framework ought to set latency timers, instead of individual drivers setting latency timers.
1.19	16-Nov-2006	christos	branches: 1.19.4; 1.19.8; 1.19.10; __unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.16	01-Aug-2006	martin	branches: 1.16.4; 1.16.6; If the card presents 64bit memory, map it as that. Fixes PR kern/34045.
1.15	14-Jul-2006	seanb	Consistently establish / disestablish shutdown hook in the bus layer and remove from common ath_attach(). Having it in both layers (on some bus architectures) was causing a double call to ath_stop() on shutdown which in turn was tickling the bus lockup described therin on slower machines.
1.14	20-Jun-2006	perry	Power hook for ath pci front end. Written and tested by George Michaelson [ggm at apnic.net]
1.13	05-Jun-2006	gdamore	branches: 1.13.2; Import new HAL 0.9.17.2. Approved by sam@ New HAL includes some driver changes to register accesses. Adds support for WLAN devices on AR5312 family devices. Adds support 32-bit SPARC ath devices (untested). ath enabled in SPARC64 GENERIC builds. This HAL is tested and known to work for i386 PCI devices, SPARC64 PCI devices, and AR5312 WiSoC devices. MIPS PCI devices appear to be busted (possibly only on Alchemy hardware, unconfirmed), and cardbus support is untested due to lack of test hardware. Please report any new problems with this import to garrett@.
1.12	02-Apr-2006	gdamore	branches: 1.12.2; Reorganize ath layout as requested by sam@ and suggested by dyoung@ in http://mail-index.netbsd.org/tech-net/2006/03/15/0000.html. The new layout almost precisely matches FreeBSD, and should make future imports much easier. At the same time, import the current 0.9.16.16 HAL from FreeBSD. According to sam@, this is the proper version we should be using.
1.11	11-Dec-2005	christos	branches: 1.11.4; 1.11.6; 1.11.8; 1.11.10; 1.11.12; merge ktrace-lwp.
1.10	22-Jun-2005	martin	branches: 1.10.2; Remove unused <atomic.h> include; add <sys/device.h> includes where needed.
1.9	22-Jun-2005	dyoung	Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.8	27-Feb-2005	perry	nuke trailing whitespace
1.7	30-Jun-2004	mycroft	branches: 1.7.2; 1.7.6; 1.7.8; This should be done in the HAL, but until then... Hack the match function to accept an alternate Atheros vendor ID by munging it to look like the expected one. This is needed for DWL-G520 cards (which otherwise "just work").
1.6	30-Apr-2004	dyoung	Synchronize with FreeBSD by adding two comments.
1.5	16-Dec-2003	dyoung	branches: 1.5.2; Synchronize with FreeBSD: * add constants to enums * HAL update: stops panics, fixes ad hoc-mode beacons, some API changes * get and use mac/phy/rf front-end revision codes * add a custom ath(4) ic_node_getrssi callback which does RSSI averaging * do not immediately scan, but re-associate after missing beacons. * bug fix: don't if_init after detach. * HAL diagnostics ioctl, SIOCGATHDIAG * send DS parameters element in beacons * const-ify some pointers * consolidate rx-filter settings into ath_calcrxfilter * abstract FreeBSD `ticks', NetBSD `hardclock_ticks' with ATH_TICKS() * misc. other changes
1.4	15-Oct-2003	itojun	$NetBSD$
1.3	15-Oct-2003	itojun	translate pci device ID until HAL gets updated. suggested by sam leffler
1.2	13-Oct-2003	dyoung	ath(4) PCI front-end for NetBSD, heavily modified from FreeBSD's. Substantially influenced by the patch from Bruce J.A. Nourish.
1.1	07-Oct-2003	dyoung	branches: 1.1.1; Initial revision
1.1.1.4	21-Jun-2005	dyoung	Import FreeBSD's ath(4) of 2005-05-18
1.1.1.3	29-Apr-2004	dyoung	Import FreeBSD's ath of 28-apr-2004
1.1.1.2	15-Dec-2003	dyoung	Import FreeBSD's ath of 12-dec-2003
1.1.1.1	07-Oct-2003	dyoung	Initial import of ath(4) from FreeBSD.
1.5.2.1	02-Jul-2004	he	Pull up revision 1.7 (requested by mycroft in ticket #582): Accept an alternate Atheros vendor ID by munging it to look like the expected one. This is needed for DWL-G520 cards which otherwise ``just work''. This should eventually be dealt with in the HAL.
1.7.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.6.1	29-Apr-2005	kent	sync with -current
1.7.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.2	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	30-Jun-2004	skrll	file if_ath_pci.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.10.2.6	17-Mar-2008	yamt	sync with head.
1.10.2.5	04-Feb-2008	yamt	sync with head.
1.10.2.4	21-Jan-2008	yamt	sync with head
1.10.2.3	03-Sep-2007	yamt	sync with head.
1.10.2.2	30-Dec-2006	yamt	sync with head.
1.10.2.1	21-Jun-2006	yamt	sync with head.
1.11.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.11.10.1	19-Apr-2006	elad	sync with head.
1.11.8.3	11-Aug-2006	yamt	sync with head
1.11.8.2	26-Jun-2006	yamt	sync with head.
1.11.8.1	11-Apr-2006	yamt	sync with head
1.11.6.2	07-Jun-2006	kardel	Sync with head.
1.11.6.1	22-Apr-2006	simonb	Sync with head.
1.11.4.1	09-Sep-2006	rpaulo	sync with head
1.12.2.1	19-Jun-2006	chap	Sync with head.
1.13.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.16.6.2	10-Dec-2006	yamt	sync with head.
1.16.6.1	22-Oct-2006	yamt	sync with head
1.16.4.1	18-Nov-2006	ad	Sync with head.
1.19.10.1	11-Jul-2007	mjf	Sync with head.
1.19.8.1	27-May-2007	ad	Sync with head.
1.19.4.1	07-May-2007	yamt	sync with head.
1.21.18.1	11-Dec-2007	yamt	sync with head.
1.21.16.1	26-Dec-2007	ad	Sync with head.
1.21.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.21.8.2	23-Mar-2008	matt	sync with HEAD
1.21.8.1	09-Jan-2008	matt	sync with HEAD
1.21.6.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.21.6.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.21.6.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.21.6.1	23-Aug-2007	joerg	Convert ath(4) to new style power management. Remove the resetting of the PCI retry register on resume, the restored configuration should have it already.
1.22.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.28.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.28.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.28.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.28.2.1	24-Mar-2008	keiichi	sync with head.
1.30.10.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.30.10.1	19-Oct-2008	haad	Sync with HEAD.
1.30.8.1	18-Jul-2008	simonb	Sync with head.
1.30.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.30.4.5	11-Aug-2010	yamt	sync with head.
1.30.4.4	11-Mar-2010	yamt	sync with head
1.30.4.3	16-Sep-2009	yamt	sync with head
1.30.4.2	16-May-2009	yamt	sync with head
1.30.4.1	04-May-2009	yamt	sync with head.
1.31.10.1	21-Apr-2010	matt	sync to netbsd-5
1.31.4.1	07-Aug-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #903): sys/dev/cardbus/if_ath_cardbus.c: revision 1.32 sys/dev/pci/if_ath_pci.c: revision 1.32 Sync ath(4) with the new HAL, mostly based on <jmcneill>'s patches. Everything should be fine again !
1.31.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.32.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.36.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.37.2.2	05-Mar-2011	rmind	sync with head
1.37.2.1	30-May-2010	rmind	sync with head
1.38.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.38.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.38.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.45.8.1	12-Aug-2012	martin	Pull up following revision(s) (requested by riastradh in ticket #476): sys/dev/pci/if_ath_pci.c: revision 1.46 Fix error branches in ath pci attachment. Should fix double pci_intr_disestablish if ath_attach fails. (Still not sure why ath_attach sometimes fails here.)
1.45.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.45.2.1	30-Oct-2012	yamt	sync with head
1.46.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.46.2.1	23-Jun-2013	tls	resync from head
1.47.4.1	18-May-2014	rmind	sync with head
1.48.30.1	10-Jun-2019	christos	Sync with HEAD
1.48.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.14	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.13	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.12	24-Nov-2015	jakllsch	branches: 1.12.16; 1.12.18; Revert previous. At least AR9280 and AR9285 apparently do not have working-by-default MSI functionality, despite advertising such. Fixes PR kern/50465.
1.11	17-Nov-2015	jakllsch	Switch PCI athn(4) attachment from pci_intr_map() to pci_intr_alloc()/pci_intr_release(), this enables MSI where available.
1.10	29-Mar-2014	christos	branches: 1.10.4; 1.10.8; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.9	08-Dec-2013	martin	Cosmectic: use the same define in all places when referencing the BAR. No functional change.
1.8	21-Apr-2013	msaitoh	branches: 1.8.4; 1.8.6; Delete "PCI_" from PCIX and PICE capability registers.
1.7	06-Apr-2013	martin	Prefer this driver over the old ath(4)
1.6	05-Apr-2013	martin	Move interrupt printout past main attach message
1.5	03-Apr-2013	christos	instead of ifnet use ethercom.
1.4	31-Mar-2013	martin	setup interface backpointer
1.3	31-Mar-2013	martin	Initialize sc_dev
1.2	31-Mar-2013	martin	Fix obvious copy&pasto
1.1	30-Mar-2013	christos	This is the initial port of OpenBSD's athn(4) driver. It supports quite a few Atheros 802.11n devices. See the athn(4) manpage for a list. This port has only been tested with a TP-LINK TL-WN722N USB adapter which has an AR9271 chipset (VENDOR: 0x0cf3, PRODUCT: 0x9271). The BSS, Monitor, and HostAP modes all seem to work on that adapter, though the later has not been tested much. The driver also supports PCI and CardBUS devices, but those interfaces are completely untested and probably won't work at this point. If someone can provide me with other adapters that should be supported, particularly PCI or CardBUS adapters, or tell me where I can get one cheaply, I would be happy to try to get the driver working on those devices as well.
1.8.6.1	18-May-2014	rmind	sync with head
1.8.4.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.8.4.2	23-Jun-2013	tls	resync from head
1.8.4.1	21-Apr-2013	tls	file if_athn_pci.c was added on branch tls-maxphys on 2013-06-23 06:20:18 +0000
1.10.8.1	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.10.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.10.4.1	29-Mar-2014	yamt	file if_athn_pci.c was added on branch yamt-pagecache on 2014-05-22 11:40:24 +0000
1.12.18.2	10-Jun-2019	christos	Sync with HEAD
1.12.18.1	12-Jul-2018	phil	State save. New kernel config for this brach only. TESTWIFI does produce a kernel. It is not working. athn files not compiling yet and commented out of the TESTWIFI kernel, which only has urtwn 802.11 driver enabled. ieee80211_alq.c and ieee80211_ddb.c not compiling yet.
1.12.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.29	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.28	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.27	14-Jul-2016	msaitoh	branches: 1.27.16; 1.27.18; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.26	29-Mar-2014	christos	branches: 1.26.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.25	26-Jul-2011	dyoung	branches: 1.25.2; 1.25.12; 1.25.16; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.24	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.23	08-Jan-2010	dyoung	branches: 1.23.2; Expand PMF_FN_* macros.
1.22	16-Sep-2009	dyoung	In pmf(9), improve the implementation of device self-suspension and make suspension by self, by drvctl(8), and by ACPI system sleep play nice together. Start solidifying some temporary API changes. 1. Extract a new header file, <sys/device_if.h>, from <sys/device.h> and #include it from <sys/pmf.h> instead of <sys/device.h> to break the circular dependency between <sys/device.h> and <sys/pmf.h>. 2. Introduce pmf_qual_t, an aggregate of qualifications on a PMF suspend/resume call. Start to replace instances of PMF_FN_PROTO, PMF_FN_ARGS, et cetera, with a pmf_qual_t. 3. Introduce the notion of a "suspensor," an entity that holds a device in suspension. More than one suspensor may hold a device at once. A device stays suspended as long as at least one suspensor holds it. A device resumes when the last suspensor releases it. Currently, the kernel defines three suspensors, 3a the system-suspensor: for system suspension, initiated by 'sysctl -w machdep.sleep_state=3', by lid closure, by power-button press, et cetera, 3b the drvctl-suspensor: for device suspension by /dev/drvctl ioctl, e.g., drvctl -S sip0. 3c the system self-suspensor: for device drivers that suspend themselves and their children. Several drivers for network interfaces put the network device to sleep while it is not administratively up, that is, after the kernel calls if_stop(, 1). The self-suspensor should not be used directly. See the description of suspensor delegates, below. A suspensor can have one or more "delegates". A suspensor can release devices that its delegates hold suspended. Right now, only the system self-suspensor has delegates. For each device that a self-suspending driver attaches, it creates the device's self-suspensor, a delegate of the system self-suspensor. Suspensors stop a system-wide suspend/resume cycle from waking devices that the operator put to sleep with drvctl before the cycle. They also help self-suspension to work more simply, safely, and in accord with expectations. 4. Add the notion of device activation level, devact_level_t, and a routine for checking the current activation level, device_activation(). Current activation levels are DEVACT_LEVEL_BUS, DEVACT_LEVEL_DRIVER, and DEVACT_LEVEL_CLASS, which respectively indicate that the device's bus is active, that the bus and device are active, and that the bus, device, and the functions of the device's class (network, audio) are active. Suspend/resume calls can be qualified with a devact_level_t. The power-management framework treats a devact_level_t that qualifies a device suspension as the device's current activation level; it only runs hooks to reduce the activation level from the presumed current level to the fully suspended state. The framework treats a devact_level_t qualifying device resumption as the target activation level; it only runs hooks to raise the activation level to the target. 5. Use pmf_qual_t, devact_level_t, and self-suspensors in several drivers. 6. Temporarily add an unused power-management workqueue that I will remove or replace, soon.
1.21	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.20	09-Jul-2008	joerg	branches: 1.20.8; - device/softc split
1.19	28-Apr-2008	martin	branches: 1.19.2; 1.19.4; 1.19.6; Remove clause 3 and 4 from TNF licenses
1.18	21-Mar-2008	dyoung	branches: 1.18.2; 1.18.4; Use aprint_*_dev() instead of accessing struct device member dv_xname directly.
1.17	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.16	07-Mar-2008	dyoung	Use device_t and accessors.
1.15	19-Oct-2007	ad	branches: 1.15.12; 1.15.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.14	16-Nov-2006	christos	branches: 1.14.8; 1.14.22; 1.14.24; 1.14.28; __unused removal on arguments; approved by core.
1.13	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.12	17-Jun-2006	christos	branches: 1.12.4; 1.12.6; re-factor the pci powestate api. reviewed by gimpy
1.11	11-Dec-2005	christos	branches: 1.11.4; 1.11.8; 1.11.14; 1.11.16; merge ktrace-lwp.
1.10	22-Jun-2005	dyoung	branches: 1.10.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.9	27-Feb-2005	perry	nuke trailing whitespace
1.8	21-Aug-2004	thorpej	branches: 1.8.4; 1.8.6; Use ANSI function decls and make use of static.
1.7	23-Jul-2004	dyoung	branches: 1.7.2; Print and store Cardbus/PCI revision number. Begin conditioning device configuration on revision number. Four revisions are known: 1.1/1.5 -> ADM8211A, 2.0 -> ADM8211B, 3.0 -> ADM8211C. The B and C parts, which are not supported yet, have AP capability.
1.6	17-Feb-2004	dyoung	Move the RF Microdevices RF3000 & Silicon Laboratories SI4126/SI4136 register sets into their own header files for re-use by future drivers.
1.5	29-Jan-2004	dyoung	Get rid of __P.
1.4	29-Jan-2004	dyoung	Add atw_pci_enable, atw_pci_disable. Fixes a panic at "ifconfig down" that was reported by Darren Reed.
1.3	16-Nov-2003	dyoung	Add data-link type DLT_IEEE802_11_RADIO to wi and atw. DLT_IEEE802_11_RADIO lets you monitor radio stats like received signal strength, which diversity antenna was used, channel/frequency, modulation, and data rate.
1.2	13-Oct-2003	dyoung	Adapt atw(4) to the new 802.11 layer. Simplify atw_start, atw_newstate. Synchronize access to atw_start by bracketing the call to ieee80211_next_scan in atw_next_scan with splnet()/splx().
1.1	06-Jul-2003	dyoung	Oops. Add the atw(4) sources, too.
1.7.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.7.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.7.2.2	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	23-Jul-2004	skrll	file if_atw_pci.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.8.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.2.5	24-Mar-2008	yamt	sync with head.
1.10.2.4	17-Mar-2008	yamt	sync with head.
1.10.2.3	27-Oct-2007	yamt	sync with head.
1.10.2.2	30-Dec-2006	yamt	sync with head.
1.10.2.1	21-Jun-2006	yamt	sync with head.
1.11.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.11.14.1	19-Jun-2006	chap	Sync with head.
1.11.8.1	26-Jun-2006	yamt	sync with head.
1.11.4.1	09-Sep-2006	rpaulo	sync with head
1.12.6.2	10-Dec-2006	yamt	sync with head.
1.12.6.1	22-Oct-2006	yamt	sync with head
1.12.4.1	18-Nov-2006	ad	Sync with head.
1.14.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.14.24.2	23-Mar-2008	matt	sync with HEAD
1.14.24.1	06-Nov-2007	matt	sync with HEAD
1.14.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.14.8.1	23-Oct-2007	ad	Sync with head.
1.15.16.3	28-Sep-2008	mjf	Sync with HEAD.
1.15.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.15.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.12.1	24-Mar-2008	keiichi	sync with head.
1.18.4.4	11-Mar-2010	yamt	sync with head
1.18.4.3	16-May-2009	yamt	sync with head
1.18.4.2	04-May-2009	yamt	sync with head.
1.18.4.1	16-May-2008	yamt	sync with head.
1.18.2.1	18-May-2008	yamt	sync with head.
1.19.6.1	19-Oct-2008	haad	Sync with HEAD.
1.19.4.1	18-Jul-2008	simonb	Sync with head.
1.19.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.20.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.23.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.25.16.1	18-May-2014	rmind	sync with head
1.25.12.2	03-Dec-2017	jdolecek	update from HEAD
1.25.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.25.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.26.6.1	05-Oct-2016	skrll	Sync with HEAD
1.27.18.1	10-Jun-2019	christos	Sync with HEAD
1.27.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.65	10-Feb-2024	andvar	s/alloted/allotted/ in comments.
1.64	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.63	27-Jun-2023	nisimura	correct register function annotation as it's for receive coalesce.
1.62	25-Sep-2022	andvar	s/inerrupt/interrupt/ and s/intrrupt/interrupt/ in comments.
1.61	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.60	22-Jan-2022	skrll	Ensure bus_dmatag_subregion is called with an inclusive max_addr everywhere.
1.59	10-Sep-2021	jakllsch	No need to #include "vlan.h", as NVLAN is not used in these drivers now.
1.58	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.57	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.56	18-Oct-2019	msaitoh	branches: 1.56.2; From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.55	18-Oct-2019	msaitoh	- Add missing splnet()/splx() around mii_tick(). Same as OpenBSD rev. 1.23 - Use device_printf() instead of aprint_error_dev)() in bce_watchdog(). - Remove unneccesary inclusion.
1.54	18-Oct-2019	msaitoh	s/initalize/initialize/ in comment or printf message.
1.53	17-Oct-2019	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.52	30-May-2019	msaitoh	branches: 1.52.2; Simplify MII structure initialization and reference. No functional change.
1.51	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.50	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.49	10-Jan-2019	msaitoh	Remove extra register read in bce_mii_write().
1.48	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.47	26-Jun-2018	msaitoh	branches: 1.47.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.46	19-Apr-2017	msaitoh	branches: 1.46.4; 1.46.10; Add another BCM4401 chip support.
1.45	19-Apr-2017	msaitoh	Use macro. No binary change.
1.44	15-Dec-2016	ozaki-r	branches: 1.44.2; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.43	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.42	10-Jun-2016	ozaki-r	branches: 1.42.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.41	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.40	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.39	10-Aug-2014	tls	branches: 1.39.4; Merge tls-earlyentropy branch into HEAD.
1.38	29-Mar-2014	christos	branches: 1.38.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.37	22-Jul-2012	matt	branches: 1.37.2; 1.37.4; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.36	20-Feb-2012	mrg	convert to device_t. been running on my inspiron 8500 for a couple of months.
1.35	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.34	19-Nov-2011	tls	branches: 1.34.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.33	13-Nov-2010	uebayasi	branches: 1.33.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.32	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.31	24-Feb-2010	dyoung	branches: 1.31.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.30	19-Jan-2010	pooka	branches: 1.30.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.29	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.28	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.27	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.26	16-Jul-2009	jakllsch	Correctly bus_dmamap_sync() the transmit descriptors. Previously the sync could run off the end and not sync the beginning of the ring. Logic gleaned from if_nfe.
1.25	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.24	18-Jan-2009	mrg	branches: 1.24.2; Enable flow control and poll the status appropriately. contributed anonymously.
1.23	11-Mar-2008	dyoung	branches: 1.23.4; 1.23.12; Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.22	22-Feb-2008	dyoung	use device_t and accessors. Use aprint__dev(). Do not cast the softc to 'struct device ', but use &sc->bce_dev instead. Use PMF_FN_*.
1.21	30-Jan-2008	simonb	branches: 1.21.2; 1.21.6; Add a simple resume hook that just calls bce_reset(), and register this with pmf_device_register(). Makes suspend/resume with bce happy. Much thanks to Jared for help here.
1.20	30-Jan-2008	simonb	WextraRemove unused DPRINTF/DPRINTFN macros. Remove NetBSD 1.6 compatibility checks (way out of date). Be somewhat consistent (and closer to KNF) with whitespace. Use uintN_t instead of u_intN_t. No functional changes.
1.19	19-Jan-2008	dyoung	Use the common code for ethernet media handling. Compiles on i386.
1.18	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.17	19-Jul-2007	dsl	branches: 1.17.6; 1.17.12; 1.17.14; 1.17.16; 1.17.18; 1.17.20; include sys/cdefs.h
1.16	09-Jul-2007	ad	branches: 1.16.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.15	21-Mar-2007	dan	teach bce(4) about being an rnd(4) source
1.14	04-Mar-2007	christos	branches: 1.14.2; 1.14.4; 1.14.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.13	21-Feb-2007	mrg	add a pair of new bus_dma(9) functions: int _bus_dmatag_subregion(bus_dma_tag_t tag, bus_addr_t min_addr, bus_addr_t max_addr, bus_dma_tag_t *newtag, int flags) void _bus_dmatag_destroy(bus_dma_tag_t tag) that allow a (normally broken/limited) device to restrict the bus address range it can talk to. this is used by bce(4) to limit DMA addresses to 1GB range, the maximum the chip can address. all this is from Yorick Hardy <yhardy@uj.ac.za> with input from several people on tech-kern. XXX: bus_dma(9) needs an update still.
1.12	16-Nov-2006	christos	branches: 1.12.4; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	07-Jul-2006	simonb	branches: 1.10.4; 1.10.6; s/writting/writing/.
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.8; 1.9.16; merge ktrace-lwp.
1.8	02-Feb-2005	christos	branches: 1.8.6; PR/29196: Robert Elz: Add support for Broadcom BCM4401-B0 ethernet
1.7	30-Jan-2005	thorpej	Eliminate use of M_HASFCS.
1.6	21-Aug-2004	thorpej	branches: 1.6.4; 1.6.6; Use ANSI function decls and make use of static.
1.5	09-Jul-2004	mycroft	branches: 1.5.2; Keep track of what we set the interrupt mask to, rather than reading it from the chip.
1.4	23-Apr-2004	joda	fix obvious past-o where the activity led register gets overwritten by the traffic meter led register
1.3	29-Sep-2003	mrg	branches: 1.3.2; - convert to new aprint*() autoconfig print mechanism - clean up some magic numbers - make 1.6 friendly
1.2	28-Sep-2003	mrg	from Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>: - if_bcereg.h should only have register definitions, so declarations of softc structure etc. should be in if_bce.c. - The types of DMA descriptors should be u_int32_t, not unsigned long. - netinet headers are not required here. - Values passed via bce_tx_ring should also be byte-swapped. - byte-swapping is not needed for bus_space access. from me: - KNF - mark all magic numbers with /* MAGIC / so they can be identified at some point. there are 33 in total though many appear to be the same value, or related (eg, the 6 emac points.) leaves these issues: - RX pre-packet headers need to be byte-swapped or not? - PAGE_SIZE bytes are allocated for both TX and RX DMA ring descriptors, but they should be 1024 (== sizeof(struct bce_dma_slot) N[TR]XDESC). - the mcast filter.
1.1	27-Sep-2003	mrg	add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.3.2.4	05-Oct-2003	tron	Pull up revision 1.3 (requested by mrg in ticket #1486): - convert to new aprint*() autoconfig print mechanism - clean up some magic numbers - make 1.6 friendly
1.3.2.3	05-Oct-2003	tron	Pull up revision 1.2 (requested by mrg in ticket #1486): from Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>: - if_bcereg.h should only have register definitions, so declarations of softc structure etc. should be in if_bce.c. - The types of DMA descriptors should be u_int32_t, not unsigned long. - netinet headers are not required here. - Values passed via bce_tx_ring should also be byte-swapped. - byte-swapping is not needed for bus_space access. from me: - KNF - mark all magic numbers with /* MAGIC / so they can be identified at some point. there are 33 in total though many appear to be the same value, or related (eg, the 6 emac points.) leaves these issues: - RX pre-packet headers need to be byte-swapped or not? - PAGE_SIZE bytes are allocated for both TX and RX DMA ring descriptors, but they should be 1024 (== sizeof(struct bce_dma_slot) N[TR]XDESC). - the mcast filter.
1.3.2.2	05-Oct-2003	tron	Pull up revision 1.1 (requested by mrg in ticket #1486): add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.3.2.1	29-Sep-2003	tron	file if_bce.c was added on branch netbsd-1-6 on 2003-10-05 11:47:15 +0000
1.5.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.5.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.5.2.2	03-Aug-2004	skrll	Sync with HEAD
1.5.2.1	09-Jul-2004	skrll	file if_bce.c was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.6.6.1	12-Feb-2005	yamt	sync with head.
1.6.4.1	29-Apr-2005	kent	sync with -current
1.8.6.7	17-Mar-2008	yamt	sync with head.
1.8.6.6	27-Feb-2008	yamt	sync with head.
1.8.6.5	04-Feb-2008	yamt	sync with head.
1.8.6.4	21-Jan-2008	yamt	sync with head
1.8.6.3	03-Sep-2007	yamt	sync with head.
1.8.6.2	26-Feb-2007	yamt	sync with head.
1.8.6.1	30-Dec-2006	yamt	sync with head.
1.9.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.9.8.1	11-Aug-2006	yamt	sync with head
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.6.2	10-Dec-2006	yamt	sync with head.
1.10.6.1	22-Oct-2006	yamt	sync with head
1.10.4.1	18-Nov-2006	ad	Sync with head.
1.12.4.3	24-Mar-2007	yamt	sync with head.
1.12.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.12.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.14.6.1	29-Mar-2007	reinoud	Pullup to -current
1.14.4.1	11-Jul-2007	mjf	Sync with head.
1.14.2.3	20-Aug-2007	ad	Sync with HEAD.
1.14.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.14.2.1	10-Apr-2007	ad	Sync with head.
1.16.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.17.20.2	19-Jul-2007	dsl	include sys/cdefs.h
1.17.20.1	19-Jul-2007	dsl	file if_bce.c was added on branch matt-mips64 on 2007-07-19 22:04:23 +0000
1.17.18.2	20-Jan-2008	bouyer	Sync with HEAD
1.17.18.1	13-Dec-2007	bouyer	Sync with HEAD
1.17.16.1	11-Dec-2007	yamt	sync with head.
1.17.14.1	26-Dec-2007	ad	Sync with head.
1.17.12.1	18-Feb-2008	mjf	Sync with HEAD.
1.17.6.2	23-Mar-2008	matt	sync with HEAD
1.17.6.1	09-Jan-2008	matt	sync with HEAD
1.21.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.21.2.1	24-Mar-2008	keiichi	sync with head.
1.23.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.23.4.6	11-Aug-2010	yamt	sync with head.
1.23.4.5	11-Mar-2010	yamt	sync with head
1.23.4.4	16-Sep-2009	yamt	sync with head
1.23.4.3	18-Jul-2009	yamt	sync with head.
1.23.4.2	16-May-2009	yamt	sync with head
1.23.4.1	04-May-2009	yamt	sync with head.
1.24.2.2	23-Jul-2009	jym	Sync with HEAD.
1.24.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.30.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.31.2.2	05-Mar-2011	rmind	sync with head
1.31.2.1	30-May-2010	rmind	sync with head
1.33.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.33.8.2	30-Oct-2012	yamt	sync with head
1.33.8.1	17-Apr-2012	yamt	sync with head
1.34.2.2	24-Feb-2012	mrg	sync to -current.
1.34.2.1	18-Feb-2012	mrg	merge to -current.
1.37.4.1	18-May-2014	rmind	sync with head
1.37.2.2	03-Dec-2017	jdolecek	update from HEAD
1.37.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.38.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.39.4.5	28-Aug-2017	skrll	Sync with HEAD
1.39.4.4	05-Feb-2017	skrll	Sync with HEAD
1.39.4.3	09-Jul-2016	skrll	Sync with HEAD
1.39.4.2	19-Mar-2016	skrll	Sync with HEAD
1.39.4.1	06-Jun-2015	skrll	Sync with HEAD
1.42.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.42.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.44.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.46.10.4	26-Jan-2019	pgoyette	Sync with HEAD
1.46.10.3	18-Jan-2019	pgoyette	Synch with HEAD
1.46.10.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.46.10.1	28-Jul-2018	pgoyette	Sync with HEAD
1.46.4.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1415): sys/dev/pci/if_bce.c: revision 1.55 sys/dev/pci/if_bce.c: revision 1.56 sys/dev/pci/if_bcereg.h: revision 1.5 - Add missing splnet()/splx() around mii_tick(). Same as OpenBSD rev. 1.23 - Use device_printf() instead of aprint_error_dev)() in bce_watchdog(). - Remove unnecessary inclusion. - From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.47.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.47.2.1	10-Jun-2019	christos	Sync with HEAD
1.52.2.2	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.52.2.1	23-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #370): sys/dev/pci/if_bce.c: revision 1.55 sys/dev/pci/if_bce.c: revision 1.56 sys/dev/pci/if_bcereg.h: revision 1.5 - Add missing splnet()/splx() around mii_tick(). Same as OpenBSD rev. 1.23 - Use device_printf() instead of aprint_error_dev)() in bce_watchdog(). - Remove unnecessary inclusion. From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.56.2.1	29-Feb-2020	ad	Sync with head.
1.6	27-Jun-2023	nisimura	correct register function annotation as it's for receive coalesce.
1.5	18-Oct-2019	msaitoh	From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.4	11-Dec-2005	christos	branches: 1.4.156; 1.4.164; 1.4.170; merge ktrace-lwp.
1.3	29-Sep-2003	mrg	branches: 1.3.2; 1.3.6; - convert to new aprint*() autoconfig print mechanism - clean up some magic numbers - make 1.6 friendly
1.2	28-Sep-2003	mrg	from Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>: - if_bcereg.h should only have register definitions, so declarations of softc structure etc. should be in if_bce.c. - The types of DMA descriptors should be u_int32_t, not unsigned long. - netinet headers are not required here. - Values passed via bce_tx_ring should also be byte-swapped. - byte-swapping is not needed for bus_space access. from me: - KNF - mark all magic numbers with /* MAGIC / so they can be identified at some point. there are 33 in total though many appear to be the same value, or related (eg, the 6 emac points.) leaves these issues: - RX pre-packet headers need to be byte-swapped or not? - PAGE_SIZE bytes are allocated for both TX and RX DMA ring descriptors, but they should be 1024 (== sizeof(struct bce_dma_slot) N[TR]XDESC). - the mcast filter.
1.1	27-Sep-2003	mrg	add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.3.6.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.6.3	18-Sep-2004	skrll	Sync with HEAD.
1.3.6.2	03-Aug-2004	skrll	Sync with HEAD
1.3.6.1	29-Sep-2003	skrll	file if_bcereg.h was added on branch ktrace-lwp on 2004-08-03 10:49:07 +0000
1.3.2.4	05-Oct-2003	tron	Pull up revision 1.3 (requested by mrg in ticket #1486): - convert to new aprint*() autoconfig print mechanism - clean up some magic numbers - make 1.6 friendly
1.3.2.3	05-Oct-2003	tron	Pull up revision 1.2 (requested by mrg in ticket #1486): from Izumi Tsutsui <tsutsui@ceres.dti.ne.jp>: - if_bcereg.h should only have register definitions, so declarations of softc structure etc. should be in if_bce.c. - The types of DMA descriptors should be u_int32_t, not unsigned long. - netinet headers are not required here. - Values passed via bce_tx_ring should also be byte-swapped. - byte-swapping is not needed for bus_space access. from me: - KNF - mark all magic numbers with /* MAGIC / so they can be identified at some point. there are 33 in total though many appear to be the same value, or related (eg, the 6 emac points.) leaves these issues: - RX pre-packet headers need to be byte-swapped or not? - PAGE_SIZE bytes are allocated for both TX and RX DMA ring descriptors, but they should be 1024 (== sizeof(struct bce_dma_slot) N[TR]XDESC). - the mcast filter.
1.3.2.2	05-Oct-2003	tron	Pull up revision 1.1 (requested by mrg in ticket #1486): add new driver for broadcom BCM4401 chipset (as seen on recent dell laptops) written by Cliff Wright <cliff@snipe444.org> and tested by yours truly. XXX: missing mcast filter support. thanks cliff!
1.3.2.1	29-Sep-2003	tron	file if_bcereg.h was added on branch netbsd-1-6 on 2003-10-05 11:47:15 +0000
1.4.170.1	23-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #370): sys/dev/pci/if_bce.c: revision 1.55 sys/dev/pci/if_bce.c: revision 1.56 sys/dev/pci/if_bcereg.h: revision 1.5 - Add missing splnet()/splx() around mii_tick(). Same as OpenBSD rev. 1.23 - Use device_printf() instead of aprint_error_dev)() in bce_watchdog(). - Remove unnecessary inclusion. From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.4.164.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.4.156.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1415): sys/dev/pci/if_bce.c: revision 1.55 sys/dev/pci/if_bce.c: revision 1.56 sys/dev/pci/if_bcereg.h: revision 1.5 - Add missing splnet()/splx() around mii_tick(). Same as OpenBSD rev. 1.23 - Use device_printf() instead of aprint_error_dev)() in bce_watchdog(). - Remove unnecessary inclusion. - From OpenBSD: - Mark ETHERCAP_VLAN_MTU. - Clear the powerdown mode. Fixes PR kern/24911 reported by Werner Backes. - Set proper LED modes.
1.399	04-Oct-2025	thorpej	Add a shared function to query the common properties used for configuring an Ethernet address.
1.398	26-May-2025	bouyer	Mark the interrupt handler MPSAFE; the driver was made MPSAFE 2 years ago but it seems that pci_intr_setattr(PCI_INTR_MPSAFE) was missed. OK skrll@
1.397	10-Nov-2024	mlelstv	Add MBUFTRACE
1.396	14-Sep-2024	skrll	Update bge_if_flags in bge_ifflags_cb before calling bge_setmulti so the new setting of IFF_PROMISC is used by bge_setmulti
1.395	07-Sep-2024	skrll	Conditionally call mutex_obj_free for the mutexes in bge_release_resources as there are code path where they might not have been allocated (yat).
1.394	28-Aug-2024	skrll	Apply changes from PR/58584 after testing (and fixing). Tested on an Apple M1.
1.393	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.392	29-Jun-2024	riastradh	branches: 1.392.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.391	01-Apr-2024	andvar	s/Doucments/Documents/ in comment.
1.390	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.389	01-Aug-2023	andvar	s/diabled/disabled/ in comments.
1.388	11-Oct-2022	andvar	branches: 1.388.2; fix typos in log messages s/bus_dmamem_create/bus_dmamap_create/ and s/bus_dmamem_load/bus_dmamap_load/. Inspired by recent similar fixes in OpenBSD.
1.387	04-Sep-2022	skrll	bge(4): Use BUS_DMA_WAITOK in ring initilization now that it's safe
1.386	04-Sep-2022	skrll	bge(4): fix the MP improvements and improve some more. - Have two locks sc_core_lock at IPL_NONE and sc_intr_lock at IPL_NET and use appropriately. - Use stopping flags instead of bge_if_flags so that bge_if_flags only needs to be protected by the sc_core_lock - Use ifmedia_init_with_lock and provide the sc_intr_lock. mii operatiions are done from the interrupt handler. - Fixup locking in bge_detach. - Rename bge_watchdog to bge_watchdog_tick to avoid confusion with the if_watchdog method. - Sprinkle some more asserts.
1.385	04-Sep-2022	skrll	bge(4): use a local variable to make bge_ifmedia initialization pretty. NFC
1.384	27-Aug-2022	skrll	Convert bge_pending_rxintr_change to a bool (from int)
1.383	27-Aug-2022	skrll	Include headers once (and don't pretend this compiles without INET)
1.382	27-Aug-2022	skrll	Correct the abbreviation of approximately to 'approx.'
1.381	25-Aug-2022	skrll	G/C
1.380	19-Aug-2022	skrll	Make bge_detaching a bool. NFC.
1.379	19-Aug-2022	skrll	softc member naming consistency with wm(4). (Taylor made me do it... not really)
1.378	19-Aug-2022	skrll	Should be checking ifp->if_flags in bge_init and not sc->bge_if_flags. Spotted by Taylor.
1.377	14-Aug-2022	skrll	bge: Use BUS_DMA_WAITOK in attach routines.
1.376	14-Aug-2022	skrll	bge: Mirror the bus_dma RX buffer changes in the OpenBSD driver This change reduces the amount of work done in the interrupt handler.
1.375	14-Aug-2022	skrll	Make bge(4) MP safe This started out as a fix so that LOCKDEBUG wouldn't explode with kernel lock spinout. LOCKDEBUG is too aggressive now and really should be relaxed.
1.374	14-Aug-2022	skrll	bge: Fix bge_ring_map bus_dmamap_sync POST{READ,WRITE} operations. When the range of ring entries wraps around sync from the start of the ring to the last entry.
1.373	07-Aug-2022	skrll	Provide and use (when appropriate) a bge_free_jumbo_mem.
1.372	07-Aug-2022	skrll	Misc tidyup. NFC.
1.371	07-Aug-2022	skrll	Unwrap a long line and remove unecessary brackets. NFC.
1.370	07-Aug-2022	skrll	Fix the KNF. oops.
1.369	07-Aug-2022	skrll	whitespace
1.368	07-Aug-2022	skrll	KNF
1.367	26-Jul-2022	skrll	Whitespace
1.366	22-Jul-2022	skrll	malloc -> kmem.
1.365	03-Jul-2022	skrll	* empty log message *
1.364	03-Jul-2022	skrll	KNF
1.363	03-Jul-2022	skrll	Perform a BUS_DMASYNC_POSTREAD on the bge_status_block before reading from it as required by bus_dma(9).
1.362	02-Jul-2022	skrll	KNF. Remove brackets around return values.
1.361	02-Jul-2022	skrll	One timeout message (the write one) in bge_miibus_writereg is enough.
1.360	02-Jul-2022	skrll	Modernise bge_update_all_threshes. Same code before and after.
1.359	02-Jul-2022	skrll	Simplification. Same code before and after.
1.358	02-Jul-2022	skrll	More const. Remove strange ifp assignments in bge_init.
1.357	30-Jun-2022	skrll	KNF some comments
1.356	30-Jun-2022	skrll	Remove strange (duplicate) assignment.
1.355	30-Jun-2022	skrll	KNF
1.354	28-Jun-2022	skrll	Sprinkle some const. Same code before and after.
1.353	19-May-2022	buhrow	For chips which contain an ASF/IPMI firmware, instruct the chip to shut the host ASF firmware down when attaching the device so the IPMI BMC can use the same physical port even when NetBSD doesn't have a network configuration on the device. By contrast, when the device gets a network configuration assigned to it and bge_init() is called, the host ASF firmware is brought up so both NetBSD and the IPMI BMc can use the same physical port. This now matches FreeBSD behavior, as well as behavior from NetBSD-5.2. Tested on a Sunfire X2200-M2 system with the following chip: bge1 at pci7 dev 4 function 1: Broadcom BCM5715 Gigabit Ethernet bge1: interrupting at ioapic0 pin 11 bge1: HW config 000000d4, 00000014, 00000000, 00000000 00000000 bge1: ASIC BCM5715 A3 (0x9003), Ethernet address 00:1e:68:XX:XX:XX bge1: setting short Tx thresholds brgphy1 at bge1 phy 1: BCM5714 1000BASE-T/X media interface, rev. 0 brgphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto
1.352	07-Apr-2022	andvar	fix various typos in comments.
1.351	22-Jan-2022	martin	Explicitly cast a 40-bit mask to bus_addr_t in a dead branch (gcc apparently is not clever enough and warns about the truncation)
1.350	22-Jan-2022	skrll	Ensure bus_dmatag_subregion is called with an inclusive max_addr everywhere.
1.349	16-Sep-2021	andvar	fix various typos, mainly in comments.
1.348	03-Sep-2021	andvar	fix typos in comments, mainly s/extention/extension/ and s/sufficent/sufficient/
1.347	30-Aug-2021	jmcneill	bge_txeof/bge_rxeof may be called with no pending descriptors, so skip bus_dmamap_sync with len=0
1.346	02-Jul-2020	msaitoh	prop_data_data_nocopy -> prop_data_value
1.345	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.344	04-Feb-2020	thorpej	Use ifmedia_fini().
1.343	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.342	25-Nov-2019	msaitoh	branches: 1.342.2; Modify PHY initialization code. This change fix a bug that SK-9D21 doesn't detect MII PHY. - Add error check to bge_miibus_writereg(). - Change return value of bge_miibus_readreg() when a read error occurred. It also add error message using with aprint_debug_dev(). This error occurs on some devices while detecting MII devices. - Move the location of BGE_MI_MODE register's initialization to next to bge_chipinit(). - Set BGE_MAC_MODE before calling ifmedia_init() and/or mii_attach(). - Add retry code for mii_attach() failed. Same as FreeBSD.
1.341	25-Nov-2019	msaitoh	- Use *_FLUSH() more. The main purpose is to wait following delay() correctly. - Add missing DELAY(80) after writing BGE_MI_MODE register.
1.340	25-Nov-2019	msaitoh	Fix a bug that SK-9D41 can't detect fiber media. Check the subsystem ID correctly. This bug was added in if_bge.c rev. 1.161.
1.339	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.338	23-Sep-2019	maxv	Remove (unused) reference to m_pktdat.
1.337	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.336	30-Jul-2019	msaitoh	- Avoid undefined behavior in bge_setmulti(). found by kUBSan. - Avoid undefined behavior when setting the MAC address in bge_init(). found by kUBSan.
1.335	09-Jul-2019	msaitoh	branches: 1.335.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.334	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.333	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.332	24-May-2019	msaitoh	No functional change: - ac -> ec - KNF.
1.331	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.330	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.329	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.328	05-Apr-2019	bouyer	Fix ASIC check, BGE_ASICREV(sc->bge_chipid) is not a chipid. Fixes bogus input error reporting in stat. This is cosmetic, exept for routed(8) which thinks that the interface is broken and stops accepting routes from it.
1.327	20-Feb-2019	msaitoh	- Add support for BCM5762 ASIC devices. - Add BCM5762, BCM5725, BCM5727, BCM57764, BCM57767 and BCM57787.
1.326	20-Feb-2019	msaitoh	Add BCM5702FE, another BCM5704S and SysKonnect SK-9Mxx support.
1.325	20-Feb-2019	msaitoh	Simplify PCI device table. No functional change.
1.324	20-Feb-2019	msaitoh	Add BCM5717's another device ID support.
1.323	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.322	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.321	16-Jan-2019	msaitoh	Remove the definition of BRGPHY_MII_1000CTL and use MII_GTCR.
1.320	02-Dec-2018	bouyer	Don't destroy the dma maps if we're not disabling the adapter, avoids a KASSERT() when bus_dmamap_destroy() is called from interrupt context via bge_watchdog() Set IFF_OACTIVE only when bge_encap() fails on adapter ressource shortage. Otherwise we may set IFF_OACTIVE while no transmit is in progress, and nothing will clear it. If bus_dmamap_load_mbuf() fails with EFBIG, m_defrag() the chain and retry. Refine the check for the 4GB boundary workaround (a fragment should also not cross the boundary), and do it only for TSO. If bge_encap() fails and didn't set IFF_OACTIVE, drop the packet. Bring in more hardware bug workarounds from freebsd. With these it seems that a BCM5720 A0 can survive a few hours of internet load with TSO4 enabled.
1.319	30-Nov-2018	jdolecek	allow MSI-X if supported by the hardware, discussed with Manuel
1.318	30-Nov-2018	jdolecek	simplify intr establish code - rely on pci_intr_alloc() to return interrupt types which are possible for pci_intr_establish(); remove fallbacks to retry with MSI explicitly disabled bge(4) specifically needs to disable MSI on some boards which are known to have broken MSI support, so this can't use pci_intr_alloc() with just NULL counts discussed on tech-kern@, and specifically bge(4) with Manuel https://mail-index.netbsd.org/tech-kern/2018/11/27/msg024240.html
1.317	27-Nov-2018	bouyer	More TSO4 fixes, from the freebsd driver: - the chip doens't want the lenght of options, but the complete lenght of ip headers (ip + tcp + options). Fix this for the BGE_IS_5717_PLUS() and BGE_IS_5705_PLUS() cases; FreeBSD doens't cover the last case so leave it as is for now. This fixes checksum failures for heavy transfers. - It looks like the transmit engine hangs if the TCP segment crosses a 4GB boundary. FreeBSD fixes it by mapping everything below 4GB; instead try detect when this happens and do the bounce only when needed. With these fixes I could transfers 3GB images over ftp at gigabit speed (112MB/s with wget) without problems. Tested on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0
1.316	24-Nov-2018	bouyer	5717 and newer handle TSO in a sighly different way. From FreeBSD. While there, use different lowat/hiwat values if MTU is larger than ETHERMTU, from FreeBSD. Tested at LIP6 on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0 with all ipv4 checksums and TSO4 enabled.
1.315	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.314	26-Jun-2018	msaitoh	branches: 1.314.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.313	22-Nov-2017	msaitoh	branches: 1.313.2; Revert part of if_bge.c 1.312. It's not reqired to mask other than VLAN ID bits in VLAN tag.
1.312	28-Sep-2017	christos	- only get the vtag if we have it like the other drivers - mask the hardware vlan tag
1.311	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.310	27-Apr-2017	msaitoh	branches: 1.310.2; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.309	25-Apr-2017	snj	branches: 1.309.2; fix typo
1.308	20-Apr-2017	msaitoh	5776[0156], 5778[02568] and 5779[015] are not Fast Ethernet but Gigabit Ethernet.
1.307	20-Apr-2017	msaitoh	Fix device timeout problem when MSI is used on BCM57762, BCM57765 and BCM57785. Check correctly for BGEF_TAGGED_STATUS. These devices' ASCI core is BCM5776[56]. Other BSD's BGE_IS_5717_PLUS() include BCM5776[56] but ours doesn't (because Linux tg3 does so). Tested with snj's MacBook (BCM57765 (BCM57765 B0 ASIC)) and my Acer Aspire One 756 (BCM57785(BCM57765 B0 ASIC)). Also Thunderbolt Ethernet (BCM57762 (BCM57766 A0 ASIC)) should work.
1.306	13-Apr-2017	msaitoh	Add forgotten BGE_EVCNT_INCR() for spurious interrupt.
1.305	13-Apr-2017	msaitoh	- Read chip ID correctly on BCM57782 and BCM57786. - Add BXCM57766 A0 string. - Add comment for BCM5776[56]'s workaround.
1.304	12-Apr-2017	msaitoh	Add a workaround for BCM57780 to prevent device timeout. This change prevent L1PM feature do wrongy. Tested with Dell latitude 2120.
1.303	12-Apr-2017	msaitoh	Fix a bug that BCM5785 write wrongly to the PCI config area.
1.302	12-Apr-2017	msaitoh	Add counters for spurious interrupts.
1.301	12-Apr-2017	msaitoh	Detach sysctl and event counters in bge_release_resources().
1.300	15-Dec-2016	ozaki-r	branches: 1.300.2; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.299	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.298	11-Jul-2016	knakahara	branches: 1.298.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.297	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.296	12-May-2016	ozaki-r	Protect ifnet list with psz and psref The change ensures that ifnet objects in the ifnet list aren't freed during list iterations by using pserialize(9) and psref(9). Note that the change adds a pslist(9) for ifnet but doesn't remove the original ifnet list (ifnet_list) to avoid breaking kvm(3) users. We shouldn't use the original list in the kernel anymore.
1.295	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.294	18-Nov-2015	msaitoh	Delete #ifdef __HAVE_PCI_MSI_MSIX and use pci_intr_allo()/pci_intr_release() on any systems. The third argument of pci_intr_alloc() is left because we will use the code to support MSI-X in future.
1.293	21-Jul-2015	knakahara	mofity for if_wm and if_bge to use pci_intr_alloc()
1.292	14-Jun-2015	martin	Make sure not to re-arm the statistics callout when we are about to detach.
1.291	23-May-2015	msaitoh	Remove debug printf().
1.290	18-May-2015	msaitoh	Add missing pci_intr_release() for the detach.
1.289	17-May-2015	msaitoh	Fix compile error on non-__HAVE_PCI_MSI_MSIX machines.
1.288	17-May-2015	msaitoh	- Add MSI support. - Use tagged status function for 5717 and newer devices. All controllers except BCM5700 support tagged status but we use tagged status only for MSI case on BCM5717. Otherwise MSI on BCM5717 does not work. Same as other *BSDs.
1.287	01-May-2015	msaitoh	Add some workaround code for BGE_ASICREV_BCM5784 from Linux.
1.286	01-May-2015	msaitoh	Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. Same as OpenBSD and Linux.
1.285	30-Apr-2015	msaitoh	Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657.
1.284	30-Apr-2015	msaitoh	- BGE_CHIPID_BCM5784_A1 is not 0x5784100 but 0x5784001. Only dmesg output is affected. - Add BGE_CHIPREV_5784_AX and BGE_CHIPID_BCM5784_B0.
1.283	27-Apr-2015	christos	fix mbuf leak on failure (Brainy)
1.282	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.281	06-Apr-2015	martin	Make sure to halt (not just stop) the bge_tick callout during detach.
1.280	17-Feb-2015	enami	Make this file compiles again with BGE_DEBUG defined.
1.279	17-Feb-2015	msaitoh	Print bit setting of bge_asf_mode if BGE_DEBUG is set.
1.278	11-Feb-2015	msaitoh	Fix three bugs reported by enami@: - bge_miibus_writereg(): Fix a bug that BCM5906 may leave an APE lock. - Fix hwcfg4 isn't printed correctly. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices.
1.277	10-Aug-2014	tls	branches: 1.277.2; 1.277.4; Merge tls-earlyentropy branch into HEAD.
1.276	05-Aug-2014	msaitoh	In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver.
1.275	24-Jul-2014	msaitoh	Yet another change for PR#48451. Don't use prefetchable memory for BAR0, too. Reported by taca@.
1.274	24-Jul-2014	msaitoh	- Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). See also: http://mail-index.netbsd.org/port-sparc64/2013/10/28/msg002036.html - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436).
1.273	22-Jul-2014	msaitoh	Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... The way to not to use prefetchable memory was adviced by Riastradh. Tested with HP MicroServer Gen8.
1.272	02-Jul-2014	msaitoh	Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. Thanks zafer@ for testing.
1.271	23-Jun-2014	msaitoh	Print APE firmware correctly.
1.270	17-Jun-2014	msaitoh	Add BGEPHYF_NO_WIRESPEED flag for fiber devices.
1.269	17-Jun-2014	msaitoh	- Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). The auto negotiation may have some bugs. - Add 2500SX support (not tested). - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD.
1.268	17-Jun-2014	msaitoh	Fix a variable name in last commit.
1.267	17-Jun-2014	msaitoh	- Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET.
1.266	29-Mar-2014	christos	branches: 1.266.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.265	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.264	07-Nov-2013	msaitoh	Revert previos. Sorry.
1.263	07-Nov-2013	msaitoh	Add Quark X1000 devices.
1.262	31-Oct-2013	msaitoh	Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670). This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion().
1.261	31-Oct-2013	msaitoh	Cleanup flag setting. No functional change. - Split flags into bge_flags and bge_phy_flags. - Rename flags. PHY related flags are prefixed with BGEPHYF_. Other flags are prefixed with BGEF_.
1.260	28-Oct-2013	msaitoh	Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. Same as Linux tg3 and FreeBSD (part of FreeBSD r204978).
1.259	12-Sep-2013	martin	Remove unused variable
1.258	08-Jul-2013	msaitoh	- Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document.
1.257	08-Jul-2013	msaitoh	57765 series is not based on 5717 series. 5717 series is based on 57765 series.
1.256	05-Jul-2013	msaitoh	Fix a link detect bug on non-autopoll systems. Same as OpenBSD (rev.1.329 and 1.336) and FreeBSD (r213710).
1.255	03-Jul-2013	msaitoh	Fix the setting of sc->bge_flags for 5717 and newer devices. Reported by Mike Belopuhov.
1.254	03-Jul-2013	msaitoh	Add BGE_JUMBO_CAPABLE flag to some chips. Before this commit, the following chips support the jumbo frame function: 5700 5701 5702 5703 5704 With this commit, the following chip support it, too: 5714 5780 5717 5718 5719 (exclude rev. A0) 5720 57765 57766
1.253	31-May-2013	msaitoh	branches: 1.253.2; - Revert rev. 1.250. I really misunderstood. Add some comments. - Set DMA watermark depend on the PCI max payload size.
1.252	30-May-2013	msaitoh	0x4B657654(BGE_SRAM_FW_MB_MAGIC) == ~0xB49A89AB(BGE_SRAM_FW_MB_RESET_MAGIC), so the same code I added in rev. 1.223 exists in bge_poll_fw(), too. Remove it. And then, call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver.
1.251	29-May-2013	msaitoh	Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits.
1.250	28-May-2013	msaitoh	Fix a bug that the MAC_MODE register is incorrectly set in bge_reset().
1.249	28-May-2013	msaitoh	Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly.
1.248	24-May-2013	msaitoh	On sparc64, the memory arbiter may be disabled by OF. Move the HW config checking stuff to after bge_reset(). bge_reset() initializes both byte order settings and the memory arbiter. The hwcfg stuff will be required in bge_reset() in future, so we should rethink about hwconfig and bge_reset().
1.247	24-May-2013	msaitoh	Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD.
1.246	24-May-2013	msaitoh	bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86.
1.245	21-May-2013	martin	Stopgap: disable access to BGE_SRAM_DATA_SIG, it is not implemented in all variants and access may crash the kernel. msaitoh: please fix properly, I have no idea how.
1.244	13-May-2013	msaitoh	For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved.
1.243	13-May-2013	msaitoh	Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. Re-enable NVRAM lock stuff again (i.e. revert rev. 1.233).
1.242	10-May-2013	christos	no need for vlan.h here.
1.241	08-May-2013	msaitoh	Check the hardware config words and print them. This change only read them and print the values.
1.240	08-May-2013	msaitoh	Add missing bus_dmamap_sync() in bge_intr(). The status word is in DMAed area. Same as *BSDs.
1.239	05-May-2013	msaitoh	Check whether sc->bge_dmatag != NULL or not. Fix a bug that bge_release_resources() panics when bge_attach failed.
1.238	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.237	12-Apr-2013	msaitoh	Cleanup. No functional change.
1.236	11-Apr-2013	msaitoh	Update comments from 57XX-PG105-R.pdf and 5718-PG106-R.pdf. No functional change.
1.235	08-Apr-2013	msaitoh	Fix the setting of Tx Random Backoff Register.
1.234	08-Apr-2013	msaitoh	Check BGE_SGDIG_STS when the chip is NOT a 5717 A0, like freebsd (OpenBSD rev. 1.327).
1.233	08-Apr-2013	msaitoh	Temporary disable the NVRAM locking stuff for PR#47716. I'll get PowerMac G5 in a few days and will fix correctly.
1.232	02-Apr-2013	msaitoh	Fix a bug that the NVRAM lock timeout occured on sparc64 onboard bge. Tested by martin. - Don't acquire NVRAM lock if no EEPROM(or NVRAM?). - Reset NVRAM before bge_reset(). It's required to acquire NVRAM lock in bge_reset().
1.231	02-Apr-2013	msaitoh	Fix APE lock bug for PHYs (FreeBSD r248993)
1.230	30-Mar-2013	christos	remove trailing whitespace
1.229	27-Mar-2013	msaitoh	Revert 1.228 except BGE_NO_EEPROM flag for BCM5906. Sorry.
1.228	27-Mar-2013	msaitoh	Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly.
1.227	24-Mar-2013	msaitoh	- Add detach function. - Remove old unused bge_vpd_*.
1.226	24-Mar-2013	msaitoh	Back out rev. 1.210. The PCI config registers of bge are shadowed in the device register window, so don't use pci_conf_*() again because those access are slow.
1.225	23-Mar-2013	msaitoh	Change the order of register settings a little to match both Broadcom's document and Linux tg3.
1.224	23-Mar-2013	msaitoh	5718 and 57785 document say we should wait 100us.
1.223	21-Mar-2013	msaitoh	Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section 7". 57781's document has the same note. I verified the loop really looped on those chips. But, I don't know why tg3 doesn't it.
1.222	21-Mar-2013	msaitoh	- Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Add/sync some of the comments - Use macro. - Remove unused code.
1.221	19-Mar-2013	msaitoh	Fix compile error if BGE_DEBUG is set.
1.220	19-Mar-2013	msaitoh	- Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD.
1.219	19-Mar-2013	msaitoh	Fix three bugs: - An workaround for TX data corruption is only for 5719 "A0". Fix the wrong evaluation. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay().
1.218	19-Mar-2013	msaitoh	Fix Buffer Manager Mbuf Pool settings which was modified in rev. 1.202.
1.217	17-Mar-2013	msaitoh	No functional change: - Fix comment - Use macro - Remove unused macro
1.216	17-Mar-2013	msaitoh	- ADD support for BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Use macros. - Remove unused code. - Sync with the latest {Free,Open}BSD's bge.
1.215	14-Mar-2013	msaitoh	Sync with *BSD. No functional change: - make bge_chipid() and use it. - use switch() instead of a lot of if()s. - sort lines.
1.214	13-Mar-2013	msaitoh	- Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Use macro.
1.213	07-Mar-2013	msaitoh	Add DELAY(40) after turning on write DMA state machine.
1.212	07-Mar-2013	msaitoh	BCM5720 is not supported yet.
1.211	07-Mar-2013	msaitoh	Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before.
1.210	07-Mar-2013	msaitoh	Fix a bug that some PCI config registers are accessed not using pci_conf_{read,write}. Registers named BGE_PCI_* are PCI config registers.
1.209	07-Mar-2013	msaitoh	Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G and BCM5787F.
1.208	03-Mar-2013	msaitoh	Style fix. No functional change.
1.207	27-Feb-2013	msaitoh	No functional change: - Fix comments. - Change macro name. - Change sysctl related functions for consistency.
1.206	26-Feb-2013	msaitoh	Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220)
1.205	26-Feb-2013	msaitoh	Use macro. Remove extra semicolon. No functional change.
1.204	25-Feb-2013	msaitoh	Use macro. Remove duplicated macro. Remove unused variable. No functional Change.
1.203	22-Feb-2013	msaitoh	Style change. It's mainly intended to reduce the diff between Fx and Nx.
1.202	17-Sep-2012	tsutsui	Add support for BCM57762, found in Apple's Thunderbolt to Gigabit Ethernet Adapter. PR kern/46961 from Ryo ONODERA. Also revert weird BGE_MAX_FRAMELEN macro definition to sync its usage with the original FreeBSD's bge.
1.201	22-Jul-2012	matt	branches: 1.201.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.200	02-Feb-2012	tls	branches: 1.200.2; 1.200.6; 1.200.8; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.199	02-Nov-2011	yamt	branches: 1.199.2; 1.199.6; allow disabling interrupt mitigation
1.198	09-Jun-2011	cegger	the BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. Tested with root on nfs. patch from msaitoh@
1.197	03-Jun-2011	cegger	fix crash by NULL pointer dereferencing I get everytime right after this message: bge0: discarding oversize frame (len=-4)
1.196	22-May-2011	mrg	branches: 1.196.2; match some fujitsu bge interfaces seen on PRIMEPOWER 250/450 systems. from Michael Moll <kvedulv@kvedulv.de>, taken from openbsd.
1.195	02-May-2011	jym	Fix BCM5709 PHY detection for ethernet PHYs (the SerDes case being already handled): - export bge(4) and bnx(4) CHIP ID and PHY flags to brgphy(4). Move to "unsigned int" rather than "int", and reuse the same softc members for chipid and phyflags (behavior controlled by the sc_isbge/isbnx boolean). - apply bug fix for revisions A and B, so that autonegotiation can complete (from OpenBSD). Bug reported by Rivo Nurges via private mail, patch tested and confirmed working by him (with thanks!)
1.194	18-Apr-2011	buhrow	Fixes for kern/40018. Our driver initializes the Broadcom hardware to peform a tcp and udp checksum on only the payload of the tcp or udp packet, rather than the entire packet. The FreeBSD, OpenBSD and Linux drivers instruct the hardware to compute the checksum for the entire packet. I believe the bug is that some revisions of the BCM hardware, under certain circumstances, revert to doing the complete checksum calculation, as the FreeBSD, OpenBSD and Linux drivers request, while things are running. As a result, when we pull the computed checksum from the hardware and pass it up to the upper layers, we assume the checksum is the more minimal one, and the upper layers perform the appropriate checks, which, when this happens, cause the packet to be rejected because the resultant checksum is decidedly incorrect. This patch changes the driver to instruct the hardware to perform the checksum over the entire packet, just as the FreeBSD, OpenBSD and Linux drivers do, and to notify the upper layers appropriately. This patch appears to work on all revisions of the hardware that have been tested. (See the list in the bug report.) this patch is approved by tls.
1.193	08-Apr-2011	sborrill	Revert previous couple of commits. bge_setvlan() is only called by bge_init() and so does not get called when a vlan child is added. Calling bge_setvlan() from bge_ifflags_cb() doesn't work either as it appears that the callback is called before ec_capenable is updated by if_vlan.c.
1.192	08-Apr-2011	sborrill	Fix typo in previous commit that would have stopped relevant bit being set.
1.191	08-Apr-2011	sborrill	PR kern/38871 Fix LAN on bge(4), alc(4). Flag VLAN capability in ec_capenable as used by network card drivers.
1.190	09-Jan-2011	jruoho	branches: 1.190.2; Don't use SYSCTL_SETUP in a driver.
1.189	05-Jan-2011	sketch	Use aprint_verbose_dev.
1.188	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.187	16-Oct-2010	taca	Remove a variable which assigned but never used, I found while reading if_bge.c yesterday.
1.186	21-Jul-2010	msaitoh	Use ether_set_ifflags_cb()
1.185	03-Jun-2010	msaitoh	Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. Sorry...
1.184	11-May-2010	njoly	Fix some typos in comments.
1.183	25-Apr-2010	ad	console spam
1.182	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.181	01-Apr-2010	msaitoh	Fix the code for 5822 B0 errata. Use pci_conf_read() and access at 32bit alignment.
1.180	03-Feb-2010	msaitoh	branches: 1.180.2; 1.180.4; - Make the initialize sequence close to the document. - Fix ASF heartbeat sending bug (FreeBSD rev. 1.271 (r202821)) - Use new handshake command for BCM5750 or new controllers (FreeBSD rev. 1.272 (r202822))
1.179	01-Feb-2010	msaitoh	Fix the bug that unaligned access occurs on amd64. It also fixes the bug that error bits aren't cleard because these bits are W2C (in other word, W1C). Reported by Michael van Elst.
1.178	28-Jan-2010	msaitoh	ANSI C, KNF and tabify
1.177	28-Jan-2010	msaitoh	- Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - remove an unused structure. - KNF
1.176	25-Jan-2010	martin	Back out part of my device property change: the mac-address part was already done better in another place which I overlooked when merging local changes.
1.175	24-Jan-2010	martin	Do not try to read the EEPROM if we already know we don't have one.
1.174	24-Jan-2010	martin	Add support for device properties to override the mac address and set the BGE_NO_EEPROM flag if we have an onboard device w/o SEEPROM.
1.173	24-Jan-2010	msaitoh	Fix a printf() message in debug code reported by Bert Kiers.
1.172	24-Jan-2010	msaitoh	Many changes from FreeBSD and OpenBSD: - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Many other bugfixes. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags.
1.171	24-Jan-2010	msaitoh	Fix the bug that both BGE_PCIE and BGE_PCIX are set on PCI-Express devices.
1.170	24-Jan-2010	msaitoh	KNF & fix typo in comment. No functional change.
1.169	19-Jan-2010	pooka	Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.168	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.167	25-Jul-2009	msaitoh	Call prop_dictionary_set_uint32() before mii_attach(), so that brgphy can do prop_dictionary_get_uint32() correctly.
1.166	23-Jul-2009	msaitoh	Reduce the difference against FreeBSD and OpenBSD Define BGE_IS_JUMBO_CAPABLE and use it. No functional change. Cleanup the code for BGE_PCI_DMA_RW_CTL regster setting. No functional change. Fix a printf message.
1.165	05-May-2009	msaitoh	Doesn't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers.
1.164	23-Apr-2009	msaitoh	use proplib
1.163	23-Apr-2009	msaitoh	remove obsolete code.
1.162	19-Apr-2009	msaitoh	Some fixes and enhancements: Both if_bge* and brgphy.[ch]: Check the parent's MAC and use the quirk code for the bug like OpenBSD and FreeBSD. Some bugs can't identify the PHY ID. For example, 5704 Ax has the ADC bug, but 5704 A3 and 5704 B0 have the same PHY ID and revision. Add BGE_PHY_CRC_BUG, BGE_PHY_ADC_BUG, BGE_PHY_5704_A0_BUG, BGE_PHY_JITTER_BUG, BGE_PHY_ADJUST_TRIM and BGE_PHY_BER_BUG for this flag. Some of the DSP patches are newly taken from OpenBSD and FreeBSD. if_bge*: Remove duplicated BGE_CHIPID_BCM5714_A0 entry in the known CHIPID table. Fix obsolete comments. brgphy.[ch]: Add some PHY IDs. TODO: Add more three quirk code into bge and brgphy (the brgphy side's are currently #if0'ed). Add support for bnx into brgphy (currently #if0'ed)
1.161	16-Apr-2009	msaitoh	Take some code from FreeBSD and OpenBSD: Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) Add some fiber support (OpenBSD 1.220) Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) Add some known CHIPID entries.
1.160	16-Apr-2009	msaitoh	Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL (on conventional PCI) again.
1.159	22-Mar-2009	msaitoh	Take some code from FreeBSD and OpenBSD -Add some known CHIPID entries. -Add a code for BCM5906 into bge_chipinit(). We have to do more work for BCM5906. -Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL (on conventional PCI).
1.158	22-Mar-2009	msaitoh	Make bge(4) close to FreBSD and OpenBSD. NO FUNCTIONAL CHANGE BY THIS COMMIT.
1.157	22-Mar-2009	msaitoh	Make bge(4) close to FreBSD and OpenBSD. NO FUNCTIONAL CHANGE BY THIS COMMIT.
1.156	22-Mar-2009	msaitoh	eliminate PHY_1 quirk (like OpenBSD (rev. 1.110) and FreeBSD (rev. 1.44))
1.155	30-Jan-2009	he	branches: 1.155.2; Make fibre cards, at least 5701 models, work, by initializing the bge_ifmedia.ifm_media field, which is not done by ifmedia_set(). Patch lifted from FreeBSD's revision 1.71.
1.154	10-Nov-2008	cegger	PR kern/39608: bge(4) constant freeze on BCM5906M: Correct DMA values. Patch from Karl Uwe Lockhoff via tech-kern.
1.153	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.152	31-Aug-2008	tron	branches: 1.152.2; 1.152.4; Make multicast setup work again which got broken by the last commit. Patch provided by Matthias Drochner. This fixes PR kern/39439.
1.151	25-Aug-2008	cegger	PR 39241: Add support for Broadcom BCM5906(M) from Karl Uwe Lockhoff
1.150	25-Jul-2008	dsl	Add {} around VLAN_INPUT_TAG() - VLAN_INPUT_TAG() can't be a do ... while (0).
1.149	24-Jul-2008	sborrill	Add support for Broadcom 5722 and identify 5755 ASICs.
1.148	17-Jun-2008	mlelstv	branches: 1.148.2; add rnd(4) hooks. Note that interrupts on bge hardware may occur with certain patterns, especially when the interrupt mitigation logic kicks in. So this might be a very weak entropy source.
1.147	17-Jun-2008	mlelstv	No longer try to start sending packets from status change handler. This is already handled by the general status change code. For non-TBI devices always call mii_tick(). This fixes ifwatchd signalling.
1.146	08-Jun-2008	mlelstv	branches: 1.146.2; really set Tx thresholds
1.145	07-Feb-2008	dyoung	branches: 1.145.6; 1.145.8; 1.145.10; 1.145.12; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.144	04-Feb-2008	mlelstv	Check wether interrupts originate from the bge hardware. Fixes PR kern/33433.
1.143	02-Feb-2008	tron	Avoid using a stale PCI chipset tag pointer. This avoids panics while attaching a 3Com 3C996B-T on my ULTRA60. Reviewed by Matthew Green.
1.142	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.141	09-Dec-2007	jmcneill	branches: 1.141.2; Merge jmcneill-pm branch.
1.140	07-Nov-2007	ad	branches: 1.140.2; 1.140.4; Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.139	28-Sep-2007	msaitoh	branches: 1.139.2; 1.139.4; Fix three bugs. 1) Move a magic number writing code. Fix a watchdog timeout and "block failed to stop" problem on BCM5701 B5 card. 2) Fix incorrect register write. Fix a watchdog timeout "block failed to stop" problem on BCM5700 B2 card. 3) Wait more long time for PCISTATE register to return to its original pre-reset state. Fix "pcistate failed to revert" problem on BCM5703 A2 card.
1.138	24-Sep-2007	joerg	Convert bge(4) to device/softc separation. Use aprint_* instead of printf all over the place. Only the debug macros are kept. Place the TSO message about the hard case under debug.
1.137	29-Aug-2007	dyoung	branches: 1.137.2; Constify: LLADDR() -> CLLADDR().
1.136	26-Aug-2007	dyoung	branches: 1.136.2; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.135	10-Aug-2007	taca	Add support for BCM5786 with simply adding its PCI product id. Testing on Fujitsu FMV-D5240 and works fine.
1.134	07-Aug-2007	markd	branches: 1.134.2; Disable firmware fastboot on 5754 and 5755. From FreeBSD by way of Phil Nelson.
1.133	06-Aug-2007	markd	Add support for BCM5755 and BCM5754/5787. Suggestions from Jonathan Stone and fixes from OpenBSD for various hardware bugs/errata.
1.132	09-Jul-2007	ad	branches: 1.132.2; 1.132.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.131	09-Jun-2007	mlelstv	Do not check link state before starting transmitter. Otherwise mbufs will just pile up in the interface queue.
1.130	21-May-2007	cube	Add support for BCM5715. Works seemingly well on hardware provided by pavel@. Thanks, Pavel!
1.129	25-Apr-2007	joerg	Fix a NULL reference on failing mbuf allocation. In bge_start return if IFF_OACTIVE is set in combination with IFF_RUNNING, drop obscure check for length of interface queue. Remove message about failing bge_encap, it can happen just too easily because of full descriptor rings.
1.128	16-Apr-2007	tron	Add support for Broadcom BCM5753 and Broadcom BCM5753M Gigabit Ethernet. Patch supplied by Mustafa Dogan in PR kern/36139, confirmed to work by Tonnerre Lombard on #NetBSD IRC channel.
1.127	05-Mar-2007	tsutsui	branches: 1.127.2; 1.127.4; Pass (char *) to mtod(9) for pointer arith (in #ifndef __NO_STRICT_ALIGNMENT part).
1.126	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.125	19-Feb-2007	bouyer	We have to bus_mapmap_load() after m_adj, otherwise the map will point at the wrong place in the buffer. Fix an issue in the !sc->bge_rx_alignment_bug case reported by Nicolas Joly. While there remove a duplicate m_adj(), and add a missing bus_dmamap_unload().
1.124	17-Feb-2007	bouyer	bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet).
1.123	09-Feb-2007	ad	branches: 1.123.2; Merge newlock2 to head.
1.122	26-Nov-2006	tsutsui	branches: 1.122.2; Sync two alternative IDs with pcidevs.
1.121	26-Nov-2006	tsutsui	Fix a typo.
1.120	26-Nov-2006	tsutsui	Add and fix some BGE_CHIPIDs. From FreeBSD and OpenBSD.
1.119	26-Nov-2006	tsutsui	Add support for BCM5752M. from Matt Kolb on tech-kern and OpenBSD.
1.118	26-Nov-2006	tsutsui	Pull some bge_encap() fixes from OpenBSD: > - Move TX ring full sanity check further up and check the number of DMA > segments from the DMA map, instead of counting the DMA segments in the > for loop and breaking out later. > - Unload the DMA map if encountering an error condition.
1.117	25-Nov-2006	tsutsui	BCM5701 B2 requires BGE_QUIRK_ONLY_PHY_1 quirk. From Brian Baird in PR kern/23778.
1.116	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.115	28-Oct-2006	tsutsui	- bcopy -> strcpy (in debug message) - bcopy -> memcpy
1.114	28-Oct-2006	tsutsui	- use continue for an empty statement in for() loop - remove an empty statement in if() clause by inverting logic - use KDASSERT(9) rather than #ifdef DEBUG + KASSERT(9) - replace commented out M_WRITABLE() with !M_READONLY(9)
1.113	28-Oct-2006	tsutsui	M_TRALINGSPACE(m) returns 0 if M_READONLY(m) is true, so no need to call both. Pointed out by Pavel Cahyna.
1.112	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.111	29-Sep-2006	christos	Coverity CID 3666: Remove debug to expose KASSERT's (from Arnaud Lacombe)
1.110	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.109	01-Jun-2006	jonathan	branches: 1.109.6; 1.109.8; Rework bge_reset() to more closely follow the Broadcom-supplied Linux driver: 1. Instead of enabling the buffer manager hardware-FSM only on pre-5705 devices, initialize the buffer manager on all bge devices. Modelled on the Broadcom-authored Linux driver. 2. Instead of enabling the memory arbiter hardware-FSM only on pre-5705 devices, initialize the memory arbiter on all bge devices. Modelled on the Broadcom-authored Linux driver. 3. Ditto the second copy of code which enables the memory arbiter. Also, add an XXX asking, why taunt the chip this way a second time? (The most plausible explanatoin is that enabling the memory arbiter twice is my own darn fault, likely a cut-and-paste glitch I made many moons ago, when re-ordering hardware enables to match the Linux drivers.)
1.108	01-Jun-2006	jonathan	Fix a minor thinko in ascertaining whether or not a bge(4) device is attached via PCI-Express: The previous code first checked that the bge ASIC-revision matched the 5750 ASIC-revisoin (the bcm5721 has the same ASIC revision). However, the bcm5752 is also a PCI-Express device, but has a different ASIC revision. Thus, we were setting sc->bge_pcie to zero for bcm5752s, which in turn causes bge_reset() to not perform required PCI-Express setup. The test for a 5750 ASIC revision may (or may not) have been carried across from the FreeBSD bge(4) driver. FreeBSD's bge(4) does not properly detect or handle post-5750 bge devices. Instead, FreeBSD's bge(4) keeps a sofc copy of the ASIC revision, and for post-5750 devices (5752, 5714, ...) overwrites that softc copy of the ASIC-revision with the 5750 ASIC revision. Thus, the test (mutatis mutandis, using FreeBDS's softc asic-revision field), was correct for FreeBSD; but manifestly incorrect for NetBSD. Mark Davies (mark at mcs.vuw.ac..nz) has confirmed via private email that this change fixes PR kern/kern/33509: his bcm5752 now works.
1.107	28-May-2006	blymn	Clean up bogus whitespace
1.106	27-Apr-2006	jonathan	branches: 1.106.2; Add an default entry to bge_majorrevs[] for bge chips with the 5780 ASIC revision. Add (accidentally omitted) PCI product-ID entries for the BCM5780 and BCM5780 (Serverworks HT-2000 to bge_products[], so that bge_probe() will attach bcm5780 variants. Tested on a bcm5780. Handling for the bcm5780S with 1000base-X PHY almost certainly has the wrong quirks; our quirk entries are driven off ASIC ID. However, the 5780 (usually integral copper) and 5780S (TBI?) have the same ASIC ID. The Linux drivers ascertain whether to use integral copper phy or external PHY via EEPROM bits. I suspect the only way we're going to handle the 5780 vs. 5780S properly is to add exact chip-id matches for all 5780/5780S variants to bge_revisions[]. However, I don't know those ids until other users with bcm5780s hardware can get this revision, try it, and report the output. So, here goes.
1.105	18-Apr-2006	christos	PR/33828: Pawel Chwalowski: Add support for Broadcom BCM5789 chip
1.104	04-Mar-2006	thorpej	branches: 1.104.2; 1.104.4; 1.104.6; Static'ify.
1.103	22-Feb-2006	rpaulo	In bge_ioctl(), don't call bge_init() if the interface is already up. Makes ifwatchd work on this interface.
1.102	24-Dec-2005	perry	branches: 1.102.2; 1.102.4; 1.102.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.101	19-Dec-2005	skrll	Typo in comment.
1.100	12-Dec-2005	jonathan	Fix silly cut-and-paste typo (CSR_READ(sc, BGE_MARBMODE_ENABLE), should obviously have been CSR_READ(sc, BGE_MARB_MODE) incurred during my backport of 5714/5715 support from my netbsd-3 development branch, back into -current.
1.99	12-Dec-2005	jonathan	Add support for bcm5714 and (possibly incorrect) support for the bcm5715 PCI-Express dual-port NICs. Taken from the Broadcom-supplied Linux driver, bcm-8.3.13a. Add nascent support for the bcm5780, but (since I have no bcm5780 to test), don't yet add an entry which allows us to match or attach to a 5780. Change 5752 support: Follow the lead of the Broadcom-supplied Linux driver (bcm) and the Linux tg3 driver, and treat the 5752 like a 5750.
1.98	07-Dec-2005	jonathan	Add PCI device-ids to sys/dev/pci/pcidevs for newer Broadcom gigabit ethernet devices: bcm5714, bcm5752 (previously the 5789, but fvdl committed that whilst I was musing). Add definitions to sys/dev/pci/if_bgereg.h for the Broadcom 57xx-family ASIC revisions on these newer chips. Add entries to the PCI-device-version and bge-internal-asic-revision tables in sys/dev/pci/if_bge.c to use these new devices. Pending further information, follow the lead of FreeBSD's if_bge.c driver and, pending further info, treat these new chips as we do the 5750.
1.97	06-Dec-2005	fvdl	Match the 5789.
1.96	24-Nov-2005	martin	Use %zd for size_t. Noticed by Kurt Schreiner on current-users.
1.95	24-Nov-2005	jonathan	Add TSO support to bge(4) for PCI-Express bge devices, which are shipped from the factory with TSO-capable firmware. The TSO support here may also work on 5705 chips, but that is (so far) untested. TSO support written after careful reading of the Linux tg3 driver, and (after attempting to deconstruct the cut-and-paste mess therein) very close reading of the Broadcom-supplied Linux driver, particularly the building of Tx-DMA buffer descriptors (bds). The TSO code herein was then rewritten from scratch, circa 4am local time, October 27 2005. (In other words: this is 4am software; caveat emptor.) Other magic register settings in this patch are required; without them, attepmting to use TSO locks up the chip. The required register settings were extracted from the cited Linux drivers. Note that TSO-capable firmware for the 5703/5704 is distributed in non-GPL form with the aforementioned Linux drivers. Once the 5705 case is debugged, (particularly the pseudo-header checksum precalculation flagged with an XXX) downloading that TSO-capable firmware to the 5703/5704 should, in principle, enable TSO support on all but the original bcm5700 (I forget if the 5701 can support TSO, or not). Note also that the ``hard case '' of IP/TCP headers spanning more than one mbuf is not handled; I haven't been able to trigger it. In any case, since TSO applies only to packets generated by the local TCP, and our TCP always leaves space for TCP headers and a normal IP header, TSO on an IP/TCP header spanning multiple headers can only arise due to insertion of IP options. I beleive that we are clearly better off outlawing that case, and requiring ip_insertoptions() to pull-up TCP headers on any packets with M_CSUM_TSOv4 set. As far as I know, bge hardwar does not support TSO for IPv6.
1.94	15-Nov-2005	jonathan	Correct a performance bug from Bill Paul's original FreeBSD bge(4) driver: Each call to the FreeBSD bge_start() routine the transmit producer pointer index from the chip mailbox register BGE_MBX_TX_HOST_PROD0_LO. The local copy of that value is then updated by bge_encap() as bge_encap() encapsulates packets in the Tx ring. If bge_encap() succeds in encpuslating one or more packets, bge_start() tells the chip to start sending the newly-encinitiates writes the new value back to the chip mailbox register. However, comparison of the Linux drivers (Broadcom-supplied and open-source tg3.c) and to the OpenSolaris driver confirms that register BGE_MBX_TX_HOST_PROD0_LO is write-only to software. Thus, we can just keep a copy in the softc, and eliminate the (expensive) PCI register write on each call to bge_start(). ``Make it so''.
1.93	06-Sep-2005	tsarna	branches: 1.93.6; Allow reception of VLAN sized frames. From FreeBSD. Reviewed by cube.
1.92	22-Aug-2005	gavan	Add quirk for BGE_CHIPID_BCM5751_A1. From Kurt Schreiner <ks at ub.uni-mainz dot de> via PR kern/31028.
1.91	21-Aug-2005	gavan	Recognise Broadcom 5751M Gigabit Ethernet, as found on ThinkPad T43p.
1.90	20-Jun-2005	atatat	branches: 1.90.2; Change the rest of the sysctl subsystem to use const consistently. The __UNCONST macro is now used only where necessary and the RW macros are gone. Most of the changes here are consumers of the sysctl_createv(9) interface that now takes a pair of const pointers which used not to be.
1.89	22-May-2005	christos	No 0x in front of %p...
1.88	02-May-2005	yamt	split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.87	27-Feb-2005	perry	branches: 1.87.2; nuke trailing whitespace
1.86	21-Feb-2005	thorpej	Define M_CSUM_DATA_IPv4_IPHL() and M_CSUM_DATA_IPv4_OFFSET() macros to extract data from csum_data, rather than just open-coding it.
1.85	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition also restructure and simplify input VLAN ID code handling somewhat, to not use local have_vlan/vlan_tag variables, and use the appropriate cur_rx condition on the one place where it's needed
1.84	18-Feb-2005	heas	Adjust csum_data usage; mask off the high 16 bits.
1.83	26-Jan-2005	jmcneill	Missed a break in the powerhook (shouldn't matter functionally, but fix anyway for correctness).
1.82	26-Jan-2005	jmcneill	Add powerhook to bge(4)
1.81	24-Jan-2005	matt	branches: 1.81.2; Add IFNET_FOREACH and IFADDR_FOREACH macros and start using them.
1.80	08-Jan-2005	fredb	branches: 1.80.2; Recognize the Broadcom 5721 Gigabit Ethernet, as found on SuperMicro P8SCi motherboards.
1.79	19-Nov-2004	jmmv	Fix typo in previous commit that was breaking the build (hi, tacha@!).
1.78	19-Nov-2004	tacha	Add support for Broadcom BCM5705K.
1.77	30-Oct-2004	thorpej	When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.76	28-Oct-2004	cube	Add support for the BCM575x variants. They're close to the 5705 one, with some special cases for the PCI-Express versions. As there is no documentation available, all of this is inspired from Linux or taken from FreeBSD (itself inspired from Linux). There are a lot of magic numbers. Tested on two BCM5751 devices as found in newer Dell computers.
1.75	29-Sep-2004	yamt	- mbuf external storage is of (caddr_t), not (caddr_t *). - remove a needless cast.
1.74	21-Sep-2004	yamt	introduce M_EXT_RW to allow mbuf external storage R/W. discussed on tech-net@.
1.73	25-May-2004	atatat	Remaining sysctl descriptions under hw subtree (ath and bge)
1.72	15-May-2004	thorpej	Add PAUSE-related event counters. Slightly modified from patch supplied by HITOSHI Osada.
1.71	15-May-2004	thorpej	Patches from HITOSHI Osada: * Set buffer management high water marks for MTU > 1514. * Set BGE_MAX_RX_FRAME_LOWAT (from Linux driver).
1.70	12-May-2004	tron	Add support for Broadcom BCM5788 GbE. Patches supplied by FUKAUMI Naoki in PR kern/25166.
1.69	10-Apr-2004	thorpej	Add 802.3x flow control support. From HITOSHI Osada.
1.68	06-Apr-2004	keihan	s/netbsd.org/NetBSD.org/g
1.67	04-Apr-2004	jonathan	Fix mistaken arguments to bge_set_thresh(); the prior revision would pass the bge_softc * to a function expecting the struct ifnet*, triggering a panic every time the Rx-mitigation value is changed via sysctl. (Testing via kernel debugger before Andrew's recent sysctl(9) changes was done with the exact same mistaken argument.)
1.66	27-Mar-2004	atatat	branches: 1.66.2; GC ath_node_root as well, but modify to work around the single (tightly scoped) reason for recording the node address by recording the assigned number. Dink pci/if_bge.c to match, since ic/ath.c was used as the archetype.
1.65	27-Mar-2004	atatat	GC bge_node_root, since it's not needed and probably wrong.
1.64	27-Mar-2004	jonathan	Update yesterday's if_bge sysctl hooks for hw.bge.* (based on yesterday's sys/dev/ic/ath.c) to match today's ath.c driver. Commit now in the hope that Andrew Brown will pick up this file for any more pending changes.
1.63	20-Mar-2004	jonathan	Add a table of Rx interrupt-threshold register values for mitigating Rx interrupts, functions to post a request for new table entries, and code to apply pending Rx-interrupt control values at the next hardware interrupt. As used in a third-party proprietary tree since at least March 2003. As discussed on tech-kern/tech-net in January 2004 (in the context of NetBSD for packet capture, bpf, and FreeBSD-sylte IFF_POLL), and as posted to tech-net for comments in mid-March 2004. Still missing sysctl or other knobs to acutally change the config-time values, due to my ignorance of any accepted per-device sysctl namespace.
1.62	20-Mar-2004	jonathan	Possibloe fix for some bge chip revisions taking a long time to reset (e.g., polling for a half-second or more at splnet(), blocking most interrupts, durin an ifconfig down/ifconfig up). Appears to help for a 5704C rev A3, which is the only chip I've ever seen that had even a mild version of the reported problem.
1.61	20-Mar-2004	jonathan	Check for BGE_PCI_PCISTATE register failing to revert on reset. if it occurs, print a message indicating why the reset took so long.
1.60	10-Mar-2004	drochner	use "=" instead of "\|=" on first assignment to mbuf.csum_flags (probably unnecessary because zero-initialized on mget, but cheaper anyway)
1.59	06-Mar-2004	martin	Set BGE_JUMBO_RXRING_VALID in bge_flags when allocating jumbo ring buffers, so it can be freed later. From HITOSHI Osada in PR kern/24686.
1.58	22-Jan-2004	jonathan	Add hooks to dynamically change the bge Rx interrupt thresholds. I found empiricaly that (at least on bcm5700s) the Rx coalesce and bd counts cannot be updated on-the-fly; attempting to do so (even at splhigh()) causes weird behaviour. Instead, add a softc flag to record that the desired softc values for Rx-interrupt thresholds have changed; check that boolean in the interrupt routine. If set, apply the new values there and clear the flag.
1.57	14-Dec-2003	jonathan	Do 5704-specific setup of DMA control register for all 5704 variants, including chips that matched only the asic-revision fallback table. When dong DMA setup for 5703 and 5704, use the new BGE_ASICREV() macro to extract the ASIC revision.
1.56	09-Dec-2003	pooka	The highest 16 bits of chipid identify the ASIC, so print only them.
1.55	08-Dec-2003	pooka	Match 5703 ASIC A3. Also, when printing the ASIC ID, print the revision in hex in all cases to avoid extra confusion if only the "superclass" of the ASIC is recognized.
1.54	11-Nov-2003	fvdl	From FreeBSD: * erratum: disable the nocrc RX bit, as it may cause problems on the 570{1-4}. adjust the length of the incoming packet accordingly to trim it. * the 5704 has a smaller MBUF_POOL, so set a smaller value Local change: * Pass the autoneg force flag to mii_attach. Some PHYs need to be kicked out of their falsely autoneged 10baseT state with this.
1.53	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.52	23-Oct-2003	fvdl	Need the NOCRC bit, or my 5704B5 will start acting up.
1.51	23-Oct-2003	fvdl	* match more cards (from FreeBSD) * don't set the NOCRC bit in the mode control register, it can cause problems on some chips (from the broadcom errata via FreeBSD) * implement a fallback quirktable that is searched only using the major asic revision, so that the driver has a shot at supporting newer versions properly without modification * rename asicrev -> chipid, like the FreeBSD driver
1.50	05-Sep-2003	tron	From FreeBSD as suggested by Nicolas Joly in PR kern/22682: Do not initalize the LED's to 0x00. The default configuration the chip comes up in should yeild proper operation of the LED's.
1.49	27-Aug-2003	fvdl	Support the BCM5704C rev 3. From Nicolas Joly, PR 22620
1.48	26-Aug-2003	hannken	Finish last commit (Make this compile with options DEBUG). NetBSD has no M_WRITABLE().
1.47	23-Aug-2003	cjep	Make this compile with options DEBUG (part of PR#22582 from Frank Kardel).
1.46	22-Aug-2003	jonathan	Check in hooks to fix checksum offload on bge devices. Empirical observation is that some 570x devices can get themselves into a state where they miscompute off-loaded TCP or UDP checksums on packets so small that Ethernet padding is required. Further obsevation suggests that the bge checksum-offload hardware is adding those padding bytes into its TCP checksum computation. (Once a 5700 gets in this state, even a warm boot won't fix it: it needs a hard powerdown.) Work around the problem by padding such runts with zeros: even if the checksum-offload adds in extra zeros, the resulting sum will be correct. Also, dont trust the checksum-offload on received packets smaller than the minimum ethernet frame, in case the Rx-side has a similar bug. Finally, on packets where we do trust the outboard Rx-side TCP or UDP checksum, the bge did not include the pseudo-header. Set the M_CSUM_NO_PSEUDOHDR bit as well as M_CSUM_DATA, and rely on udp_input() or tcp_input() adding in the sum via in_cksum_phdr().
1.45	22-Aug-2003	jonathan	Fix a bug in compaction of `DMA runt' fragmented packet chains: if we m_dup() a packet to compactify it, and later run out of DMA descriptors, bge_encap() will return ENOBUFS, hoping the driver will try again later. But we have just m_freem()'d the original chain which was m_dup()'d, leaving a pointer to the just-freed packet header in the tx queue. Fix by always walking the chain, shuffling data towards the head; except if we find a runt in the very last mbuf, we must borrow data from its predecessor. (Patch is verbatim from a third-party tree, apologies for any style woes.)
1.44	17-Jul-2003	hannken	Add support for the BCM5705: - Only BCM5705M asic rev A1 was tested. Thanks to Bill Paul (wpaul@freebsd.org) for help and support. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.43	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.42	30-Jun-2003	ragge	branches: 1.42.2; Set the send queue to max(BGE_TX_RING_CNT - 1, IFQ_MAXLEN); sometimes IFQ_MAXLEN may be choosen to a larger value than number of descriptors. Why this setting from the beginning???
1.41	15-Jun-2003	fvdl	Handle 64bit DMA addresses on PCI for platforms that can (currently only enabled on amd64). Add a dmat64 field to various PCI attach structures, and pass it down where needed. Implement a simple new function called pci_dma64_available(pa) to test if 64bit DMA addresses may be used. This returns 1 iff _PCI_HAVE_DMA64 is defined in <machine/pci_machdep.h>, and there is more than 4G of memory.
1.40	01-Jun-2003	fvdl	Add newer 5704 asic revs, and add them to the table, with the BGE_QUIRK_ONLY_PHY_1 quirk in their entry.
1.39	03-May-2003	wiz	DMA, not dma nor Dma.
1.38	24-Mar-2003	tron	Don't print a kernel message like "bge0: gigabit link up" when a Gigabit link is detected. No other ethernet driver does this and this driver doesn't do it for other modes.
1.37	07-Mar-2003	jonathan	Commit workaround for hardware bug: 5701 chips running in PCI-X mode require that the DMA receive buffers be aligned. The driver was deliberately mis-aligning by 2 bytes, to force the layer-2/3 headers to 32-bit alignment. Workaround: if chip is a 5701, and is in PCI-X mode, leave the DMA buffers aligned. If the host CPU requires alignment, copy the buffer after reception to force aligment. Tested on an i386 in PCI-X bus, with __NO_STRICT_ALIGNMENT forced off. Patch and comments reworked to minimize drift from FreeBSD if_bge.c rev 1.14.
1.36	07-Mar-2003	jonathan	Fix typo: PRODUCER_BUG quirk mask overlapped the PCIX register-bug quirk. (Harmless, since we dont use the latter; but we will use it very soon.)
1.35	06-Mar-2003	jonathan	Commit the FreeBSD alternative of a private patch to check NIC internal ram for using TBI, versus MII/GMII. FreeBSD commit log and versions: Obtain the media type from the shared memory and only use the eeprom as a fallback. if_bge.c: rev 1.30 if_bgereg.h: rev 1.13 Thanks to Paul Saab @mu.org.
1.34	06-Mar-2003	jonathan	FreeBSD fix for problems reported (to FreeBSD) with 16-bit accesses to 32-bit device registers. Add (untested) NetBSD fix for big-endian hosts. From FreeBSD versions: if_bge.c: 1.26 and 1.27 if_bgereg.h: 1.12 FreeBSD (MFC) comment: MFC: Avoid 16-bit accesses to device memory. Use volatile where appropriate in referencing device memory. Parenthesize BGE_HOSTADDR define properly.
1.33	22-Feb-2003	tsutsui	hz -> Hz
1.32	09-Feb-2003	tron	Correct typo in an e-mail address.
1.31	31-Jan-2003	thorpej	Change ext_size to a size_t, and update the signature of ext_free.
1.30	31-Jan-2003	thorpej	Use aprint_*().
1.29	17-Jan-2003	itojun	branches: 1.29.2; correct the use of m_tag. correct indentation (:set sw=8 ts=8 please, jonathan)
1.28	17-Jan-2003	itojun	switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag implementation. it will simplify porting across *bsd (such as kame/altq), and make us more synchronized. from Joel Wilsson
1.27	17-Jan-2003	jonathan	Grr. Fix consistent typo, to match what i actually tested before committing.
1.26	17-Jan-2003	jonathan	Enable the 5700 DMA bug and producer-pointer fixes, by setting the quirk on all 5700 revision B devices. (These fixes have not been tested against NetBSD recently; committing the fix and the enable separately gives us flexibility about which fixes get pulled into the NetBSD-1.6 branch.)
1.25	17-Jan-2003	jonathan	Commit accumulated bge fixes: * Add support for 5704C dual-channel chip with integral copper PHY (tested) and 5704S dual-channel SERDES/TBI gbic (untested). Add PHY DSP patch for 5704. * Update PHY DSP-code patch for bcm5401 to match latest Linux driver. * Add PHY DSP-code patch for 5703 (untested). * Update onchip buffer tunables to recommended values from Linux drivers. * Disable MWI access. This chip family cannot hanlde PCI stalls in the middle of an MWI burst. The driver has heuristics to detect PCI line size, but under load, some PCI bridges may force stalls which the attach-time heruistics do not catch. Some PCI bridges never do this, so maybe it should be a tunable option. * bcm5700 rev Bx chips have a race condition, where updating the Tx producer pointer goes un-noticed by the chip. Workaround is to write the new producer-pointer value twice. * bcm5700 chips rev Bx wedge up if given DMA descriptors of eight bytes or less. Once hit, only reovery is a watchdog timeout/reset. If the offending packet is retransmitted, the chip will wedge again... Check for teeny fragments in a Tx request, and either fold the teeny chunk residue into an adjacent mbuf, or m_dup the entire buffer. (NB: quirk not yet enabled; in-place folding tested only on FreeBSD.) * Add workaround for revision Bx bcm5700: chip bugs in decoding of PCI register writes may leave the hardware in (partial) powersave state, such that writes to "indirect" registers do not work. Explicitly force chip into D0 state at attach time. * Accessing PHY registers with the bge chip in autopoll mode, when link-state is the process of changing, may cause the bge chip to assert PCI errors. Workaround: when doing miibus register access, save autopoll state, disable around access, and restore autopoll state. NB: issuing PHY resets may give a window where the problem still occurs. * Increase Tx interrupt-coalescing thresholds, to reduce Tx-done interrupts.
1.24	26-Dec-2002	matt	Recognize BCM5702X and BCM5703X. Add a ONLY_PHY_1 quirk and use it. Add support for getting the macaddr for 5703 (taken from FreeBSD).
1.23	06-Oct-2002	kristerw	Do not try to print uninitialized variable.
1.22	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.21	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.20	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.19	18-Jul-2002	mjl	branches: 1.19.2; 1.19.4; Correct obviously switched args to memset()
1.18	13-Jul-2002	thorpej	branches: 1.18.2; Add a BGE_QUIRK_CSUM_BROKEN quirk, and give it to BCM5700 B0, rather than testing for the revision all over the driver. If the chip has broken checksums, then don't advertise the capability to the stack.
1.17	13-Jul-2002	thorpej	Give BCM5700 Ax and Bx revs the BGE_QUIRK_LINK_STATE_BROKEN quirk, and test for that, rather than checking revision where the quirk is handled.
1.16	13-Jul-2002	thorpej	Add a lookup table with ASIC revision, name, and any quirks. Print ASIC revision at attach time. Add BCM5703 revisions.
1.15	13-Jul-2002	thorpej	Move the BCM5401 DSP patch out of the bge driver and into the brgphy driver; all users of the BCM5400 and BCM5401 need the DSP patch and the sledgehammer-reset-at-media-set-time. Also add a DSP patch for the BCM5411 gleaned from Apple's GMAC driver for Darwin. Tested with a 3Com 3c996-T (BCM5700 + BCM5401).
1.14	12-Jul-2002	enami	Match Altima AC1001.
1.13	01-Jul-2002	thorpej	Fix multicast filter programming.
1.12	28-Jun-2002	thorpej	Fix a comment.
1.11	28-Jun-2002	thorpej	When stopping the various functional blocks of the chip, clear the bit and then read it back in a loop (with appropriate delays) waiting for it to read back clear. This fixes a problem where the bus would hang when bringing down the interface or changing interface flags on a system with a sufficiently fast CPU (e.g. 2GHz P4 Xeon).
1.10	28-Jun-2002	fvdl	Rearrange alignment code a bit.
1.9	28-Jun-2002	thorpej	Increase previously added delays from 2us to 10us.
1.8	28-Jun-2002	thorpej	* Only set the jumbo frame capability if we are able to allocate jumbo frame buffers. * Garbage-collect some stuff we don't need. * Fix timeout detection in the firmware handshake.
1.7	27-Jun-2002	thorpej	Use a table to match BCM570x products.
1.6	27-Jun-2002	thorpej	Don't call m_aux_find() in the inner loop when creating the Tx descriptor list. Instead, call it once and remember the result, and only call it looking for VLAN tags if VLANs are configured on the interface.
1.5	27-Jun-2002	thorpej	Put some delay in the loops that poll for MII transaction completion. Without this, reading the PHY can hang the bus on a sufficiently fast CPU.
1.4	27-Jun-2002	fvdl	Disable TCP/UDP checksums on the receive side again.
1.3	26-Jun-2002	cjs	Match newer NetGear GA302T cards.
1.2	24-Jun-2002	fvdl	According to the Linux driver from Broadcom, only the 5700 B0 revision had rx checksum problems. So only disable rx checksum for that revision.
1.1	22-Jun-2002	fvdl	Driver for Broadcom 570x / Altima AC 1000 gigabit ethernet cards. Written by Bill Paul for FreeBSD. This port started out with the port done to OpenBSD by nate@openbsd.org, but ended up looking much more like the port of the ti driver done by drochner@netbsd.org (they are similar in structure).
1.18.2.3	20-Jul-2002	gehenna	catch up with -current.
1.18.2.2	15-Jul-2002	gehenna	catch up with -current.
1.18.2.1	13-Jul-2002	gehenna	file if_bge.c was added on branch gehenna-devsw on 2002-07-15 10:35:36 +0000
1.19.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.19.4.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.19.4.1	18-Jul-2002	jdolecek	file if_bge.c was added on branch kqueue on 2002-09-06 08:45:14 +0000
1.19.2.5	17-Jan-2003	thorpej	Sync with HEAD.
1.19.2.4	29-Dec-2002	thorpej	Sync with HEAD.
1.19.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.19.2.2	01-Aug-2002	nathanw	Catch up to -current.
1.19.2.1	18-Jul-2002	nathanw	file if_bge.c was added on branch nathanw_sa on 2002-08-01 02:45:14 +0000
1.29.2.7	15-Mar-2004	jmc	Pullup rev 1.59 (requested by martin in ticket #1622) Set BGE_JUMBO_RXRING_VALID in bge_flags when allocating jumbo ring buffers so it can be freed later. PR#24686
1.29.2.6	20-Oct-2003	cyber	Pull up revision 1.50 (requested by tron in ticket #1447): From FreeBSD as suggested by Nicolas Joly in PR kern/22682: Do not initalize the LED's to 0x00. The default configuration the chip comes up in should yeild proper operation of the LED's.
1.29.2.5	16-Jun-2003	grant	Pull up revision 1.35 (requested by tron in ticket #1250): Commit the FreeBSD alternative of a private patch to check NIC internal ram for using TBI, versus MII/GMII. FreeBSD commit log and versions: Obtain the media type from the shared memory and only use the eeprom as a fallback. if_bge.c: rev 1.30 if_bgereg.h: rev 1.13 Thanks to Paul Saab @mu.org.
1.29.2.4	16-Jun-2003	grant	Pull up revision 1.34 (requested by tron in ticket #1249): FreeBSD fix for problems reported (to FreeBSD) with 16-bit accesses to 32-bit device registers. Add (untested) NetBSD fix for big-endian hosts. From FreeBSD versions: if_bge.c: 1.26 and 1.27 if_bgereg.h: 1.12 FreeBSD (MFC) comment: MFC: Avoid 16-bit accesses to device memory. Use volatile where appropriate in referencing device memory. Parenthesize BGE_HOSTADDR define properly.
1.29.2.3	16-Jun-2003	grant	Pull up revision 1.38 (requested by tron in ticket #1230): Don't print a kernel message like "bge0: gigabit link up" when a Gigabit link is detected. No other ethernet driver does this and this driver doesn't do it for other modes.
1.29.2.2	28-Jan-2003	jmc	Pullup from -current (lots of revisions 1.1-1.19,1.22-1.27 (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.29.2.1	17-Jan-2003	jmc	file if_bge.c was added on branch netbsd-1-6 on 2003-01-28 06:19:13 +0000
1.42.2.12	11-Dec-2005	christos	Sync with head.
1.42.2.11	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.42.2.10	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.42.2.9	04-Feb-2005	skrll	Sync with HEAD.
1.42.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.42.2.7	29-Nov-2004	skrll	Sync with HEAD.
1.42.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.42.2.5	19-Oct-2004	skrll	Sync with HEAD
1.42.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.42.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.42.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.42.2.1	03-Aug-2004	skrll	Sync with HEAD
1.66.2.3	29-May-2004	tron	branches: 1.66.2.3.2; Pull up revision 1.73 (requested by atatat in ticket #392): Remaining sysctl descriptions under hw subtree (ath and bge)
1.66.2.2	13-May-2004	jdc	Pull up revision 1.70 (requested by tron in ticket #316) Add support for Broadcom BCM5788 GbE. Patches supplied by FUKAUMI Naoki in PR kern/25166.
1.66.2.1	05-Apr-2004	tron	Pull up revision 1.67 (requested by jonathan in ticket #59): Fix mistaken arguments to bge_set_thresh(); the prior revision would pass the bge_softc * to a function expecting the struct ifnet*, triggering a panic every time the Rx-mitigation value is changed via sysctl. (Testing via kernel debugger before Andrew's recent sysctl(9) changes was done with the exact same mistaken argument.)
1.66.2.3.2.1	24-Jan-2005	he	Pull up revision 1.77 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.80.2.1	29-Apr-2005	kent	sync with -current
1.81.2.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.81.2.1	12-Feb-2005	yamt	sync with head.
1.87.2.7	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1862): sys/dev/pci/if_bge.c 1.89,1.95,1.97,1.101,1.103, 1.105,1.107,1.117-1.122, 1.128-1.131,1.133-1.135,1.139 via patch sys/dev/pci/if_bgereg.h 1.31,1.34-1.39,1.42-1.43 sys/dev/pci/pcidevs patch Some cleanups(typo, whitespace, debug message). Add TSO support to bge(4) Add support for BCM5715, BCM5752M, BCM5754, BCM5755, BCM5786 and 5787. Add support for BCM5789. Fixes PR/33828 from Pawel Chwalowskibg. Add support for BCM5753 and BCM5753M Fixes PR kern/36139 from Mustafa Dogan. Add and fix some BGE_CHIPIDs. From FreeBSD and OpenBSD. In bge_ioctl(), don't call bge_init() if the interface is already up (Makes ifwatchd work on this interface). BCM5701 B2 requires BGE_QUIRK_ONLY_PHY_1 quirk. From Brian Baird in kern/23778. Move TX ring full sanity check further up and check the number of DMA segments from the DMA map, instead of counting the DMA segments in the for loop and breaking out later. Unload the DMA map if encountering an error condition. Fix a NULL reference on failing mbuf allocation. In bge_start return if IFF_OACTIVE is set in combination with IFF_RUNNING, drop obscure check for length of interface queue. Remove message about failing bge_encap, it can happen just too easily because of full descriptor rings. Fix some watchdog timeout and "block failed to stop" problems on some cards. Fix "pcistate failed to revert" problem on many cards.
1.87.2.6	05-Mar-2007	ghen	Pull up following revision(s) (requested by bouyer in ticket #1671): sys/dev/pci/if_bgereg.h: revision 1.40 sys/dev/pci/if_bge.c: revision 1.124 sys/dev/pci/if_bge.c: revision 1.125 bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet). We have to bus_mapmap_load() after m_adj, otherwise the map will point at the wrong place in the buffer. Fix an issue in the !sc->bge_rx_alignment_bug case reported by Nicolas Joly. While there remove a duplicate m_adj(), and add a missing bus_dmamap_unload().
1.87.2.5	04-Jun-2006	tron	Apply patch (requested by jonathan in ticket #1351): Add support to bge(4) for bcm5714, bcm5715, bcm5780/HT-2000 and bcm5752 chip variants.
1.87.2.4	22-Nov-2005	tron	Pull up following revision(s) (requested by jonathan in ticket #985): sys/dev/pci/if_bgereg.h: revision 1.27 sys/dev/pci/if_bge.c: revision 1.94 Correct a performance bug from Bill Paul's original FreeBSD bge(4) driver: Each call to the FreeBSD bge_start() routine the transmit producer pointer index from the chip mailbox register BGE_MBX_TX_HOST_PROD0_LO. The local copy of that value is then updated by bge_encap() as bge_encap() encapsulates packets in the Tx ring. If bge_encap() succeds in encpuslating one or more packets, bge_start() tells the chip to start sending the newly-encinitiates writes the new value back to the chip mailbox register. However, comparison of the Linux drivers (Broadcom-supplied and open-source tg3.c) and to the OpenSolaris driver confirms that register BGE_MBX_TX_HOST_PROD0_LO is write-only to software. Thus, we can just keep a copy in the softc, and eliminate the (expensive) PCI register write on each call to bge_start(). ``Make it so''.
1.87.2.3	15-Sep-2005	snj	Pull up following revision(s) (requested by tron in ticket #757): sys/dev/pci/if_bge.c: revision 1.93 Allow reception of VLAN sized frames. From FreeBSD. Reviewed by cube.
1.87.2.2	08-Sep-2005	tron	Pull up following revision(s) (requested by gavan in ticket #752): sys/dev/pci/if_bgereg.h: revision 1.25 sys/dev/pci/if_bge.c: revision 1.92 Add quirk for BGE_CHIPID_BCM5751_A1. From Kurt Schreiner <ks at ub.uni-mainz dot de> via PR kern/31028.
1.87.2.1	08-Sep-2005	tron	Pull up following revision(s) (requested by gavan in ticket #752): sys/dev/pci/if_bge.c: revision 1.91 Recognise Broadcom 5751M Gigabit Ethernet, as found on ThinkPad T43p.
1.90.2.9	11-Feb-2008	yamt	sync with head.
1.90.2.8	04-Feb-2008	yamt	sync with head.
1.90.2.7	21-Jan-2008	yamt	sync with head
1.90.2.6	15-Nov-2007	yamt	sync with head.
1.90.2.5	27-Oct-2007	yamt	sync with head.
1.90.2.4	03-Sep-2007	yamt	sync with head.
1.90.2.3	26-Feb-2007	yamt	sync with head.
1.90.2.2	30-Dec-2006	yamt	sync with head.
1.90.2.1	21-Jun-2006	yamt	sync with head.
1.93.6.2	29-Nov-2005	yamt	sync with head.
1.93.6.1	22-Nov-2005	yamt	sync with head.
1.102.6.2	01-Jun-2006	kardel	Sync with head.
1.102.6.1	22-Apr-2006	simonb	Sync with head.
1.102.4.1	09-Sep-2006	rpaulo	sync with head
1.102.2.1	01-Mar-2006	yamt	sync with head.
1.104.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.104.4.2	11-May-2006	elad	sync with head
1.104.4.1	19-Apr-2006	elad	sync with head.
1.104.2.2	26-Jun-2006	yamt	sync with head.
1.104.2.1	24-May-2006	yamt	sync with head.
1.106.2.1	19-Jun-2006	chap	Sync with head.
1.109.8.2	10-Dec-2006	yamt	sync with head.
1.109.8.1	22-Oct-2006	yamt	sync with head
1.109.6.3	06-Feb-2007	ad	Quieten noisy boot messages.
1.109.6.2	12-Jan-2007	ad	Sync with head.
1.109.6.1	18-Nov-2006	ad	Sync with head.
1.122.2.12	13-Jun-2012	sborrill	Pull up the following revisions(s) (requested by buhrow in ticket #1449): sys/dev/pci/if_bge.c: 1.194 via patch Instruct hardware to perform checksumming over the entire packet not just over the payload and notify upper layers appropriately. Fixes PR kern/40018.
1.122.2.11	18-Aug-2009	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1342): sys/dev/pci/if_bge.c: revisions 1.135, 1.146, 1.151-1.152, 1.155-1.167 via patch sys/dev/pci/if_bgereg.h: revisions 1.43, 1.48-1.52 sys/dev/pci/if_bgevar.h: revision 1.1 sys/dev/mii/brgphy.c: revisions 1.36, 1.40, 1.43-1.44 and 1.47-1.48 via patch sys/dev/mii/brgphyreg.h: revisions 1.3-1.4 sys/dev/mii/miidevs: revisions 1.75, 1.77, 1.85 sys/dev/mii/miidevs.h: regen sys/dev/mii/miidevs_data.h: regen sys/dev/mii/miivar.h: revision 1.54 For bge: - Really set Tx threshold. - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the "link changed" bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. - Add BCM5786 support. for brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2, BCM5708C, BCM5722, and BCM5906.
1.122.2.10	20-Aug-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1172): sys/dev/pci/if_bge.c: revision 1.149 sys/dev/pci/if_bgereg.h: revision 1.47 Add support for Broadcom 5722 and identify 5755 ASICs.
1.122.2.9	17-May-2008	bouyer	Pull up following revision(s) (requested by mlelstv in ticket #1144): sys/dev/pci/if_bge.c: revision 1.144 via patch Check wether interrupts originate from the bge hardware. Fixes PR kern/33433.
1.122.2.8	15-Oct-2007	xtraeme	Pull up following revision(s) (requested by msaitoh in ticket #934): sys/dev/pci/if_bge.c: patch Shut up a debug message.
1.122.2.7	07-Oct-2007	xtraeme	Pull up following revision(s) (requested by msaitoh in ticket #918): sys/dev/pci/if_bge.c: revision 1.139 Fix three bugs. 1) Move a magic number writing code. Fix a watchdog timeout and "block failed to stop" problem on BCM5701 B5 card. 2) Fix incorrect register write. Fix a watchdog timeout "block failed to stop" problem on BCM5700 B2 card. 3) Wait more long time for PCISTATE register to return to its original pre-reset state. Fix "pcistate failed to revert" problem on BCM5703 A2 card.
1.122.2.6	31-Aug-2007	xtraeme	Pull up following revision(s) (requested by phil in ticket #819): sys/dev/mii/miidevs.h: regen sys/dev/mii/miidevs: revision 1.71 sys/dev/pci/if_bge.c: revision 1.133 sys/dev/pci/if_bge.c: revision 1.134 sys/dev/pci/if_bgereg.h: revision 1.43 sys/dev/mii/brgphy.c: revision 1.34 sys/dev/mii/miidevs_data.h: regen Add BCM5755 and 5754 Gig-E PHYs. Regen for BCM5755 and 5754. Add support for BCM5754 and BCM5755. From suggestions from Jonathan Stone. Add a couple more BGE_CHIPIDs Add support for BCM5755 and BCM5754/5787. Suggestions from Jonathan Stone and fixes from OpenBSD for various hardware bugs/errata. Disable firmware fastboot on 5754 and 5755. From FreeBSD by way of Phil Nelson.
1.122.2.5	21-Aug-2007	liamjfoy	Pull up following revision(s) (requested by pavel in ticket #821): sys/dev/pci/if_bge.c: revision 1.130 Add support for BCM5715. Works seemingly well on hardware provided by pavel@. Thanks, Pavel!
1.122.2.4	10-Jun-2007	bouyer	Pull up following revision(s) (requested by mlelstv in ticket #714): sys/dev/pci/if_bge.c: revision 1.131 Do not check link state before starting transmitter. Otherwise mbufs will just pile up in the interface queue.
1.122.2.3	28-Apr-2007	bouyer	branches: 1.122.2.3.2; Pull up following revision(s) (requested by joerg in ticket #594): sys/dev/pci/if_bge.c: revision 1.129 Fix a NULL reference on failing mbuf allocation. In bge_start return if IFF_OACTIVE is set in combination with IFF_RUNNING, drop obscure check for length of interface queue. Remove message about failing bge_encap, it can happen just too easily because of full descriptor rings.
1.122.2.2	16-Apr-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #578): sys/dev/pci/if_bge.c: revision 1.128 Add support for Broadcom BCM5753 and Broadcom BCM5753M Gigabit Ethernet. Patch supplied by Mustafa Dogan in PR kern/36139, confirmed to work by Tonnerre Lombard on #NetBSD IRC channel.
1.122.2.1	27-Feb-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #466): sys/dev/pci/if_bgereg.h: revision 1.40 sys/dev/pci/if_bge.c: revision 1.124-1.125 bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet). We have to bus_mapmap_load() after m_adj, otherwise the map will point at the wrong place in the buffer. Fix an issue in the !sc->bge_rx_alignment_bug case reported by Nicolas Joly. While there remove a duplicate m_adj(), and add a missing bus_dmamap_unload().
1.122.2.3.2.4	04-Sep-2008	skrll	Sync with netbsd-4.
1.122.2.3.2.3	03-Jun-2008	skrll	Sync with netbsd-4.
1.122.2.3.2.2	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.122.2.3.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.123.2.3	07-May-2007	yamt	sync with head.
1.123.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.123.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.127.4.1	11-Jul-2007	mjf	Sync with head.
1.127.2.6	09-Oct-2007	ad	Sync with head.
1.127.2.5	01-Sep-2007	ad	Update for pool_cache API changes.
1.127.2.4	20-Aug-2007	ad	Sync with HEAD.
1.127.2.3	15-Jul-2007	ad	Sync with head.
1.127.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.127.2.1	27-May-2007	ad	Sync with head.
1.132.6.10	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.132.6.9	11-Nov-2007	joerg	Sync with HEAD.
1.132.6.8	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.132.6.7	02-Oct-2007	joerg	Sync with HEAD.
1.132.6.6	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.132.6.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.132.6.4	23-Aug-2007	joerg	Introduce pci_net_generic_power, which should be enough for most network drivers and be a good foundation for C&P for the rest. For iwi(4), don't reset the PCI retry register again, pci_conf_restore should take care of that already. For bge(4), add a NetBSD style if_stop.
1.132.6.3	21-Aug-2007	joerg	Make bge(4) a bit more like other drivers in the tree with regard to the data stored in the softc. Convert to pnp_register.
1.132.6.2	16-Aug-2007	jmcneill	Sync with HEAD.
1.132.6.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.132.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.132.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.134.2.2	07-Aug-2007	markd	Disable firmware fastboot on 5754 and 5755. From FreeBSD by way of Phil Nelson.
1.134.2.1	07-Aug-2007	markd	file if_bge.c was added on branch matt-mips64 on 2007-08-07 01:13:57 +0000
1.136.2.4	23-Mar-2008	matt	sync with HEAD
1.136.2.3	09-Jan-2008	matt	sync with HEAD
1.136.2.2	08-Nov-2007	matt	sync with -HEAD
1.136.2.1	06-Nov-2007	matt	sync with HEAD
1.137.2.1	06-Oct-2007	yamt	sync with head.
1.139.4.3	18-Feb-2008	mjf	Sync with HEAD.
1.139.4.2	27-Dec-2007	mjf	Sync with HEAD.
1.139.4.1	19-Nov-2007	mjf	Sync with HEAD.
1.139.2.1	13-Nov-2007	bouyer	Sync with HEAD
1.140.4.1	11-Dec-2007	yamt	sync with head.
1.140.2.1	26-Dec-2007	ad	Sync with head.
1.141.2.1	20-Jan-2008	bouyer	Sync with HEAD
1.145.12.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.145.12.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.145.10.6	11-Aug-2010	yamt	sync with head.
1.145.10.5	11-Mar-2010	yamt	sync with head
1.145.10.4	16-Sep-2009	yamt	sync with head
1.145.10.3	19-Aug-2009	yamt	sync with head.
1.145.10.2	16-May-2009	yamt	sync with head
1.145.10.1	04-May-2009	yamt	sync with head.
1.145.8.1	17-Jun-2008	yamt	sync with head.
1.145.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.145.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.145.6.1	29-Jun-2008	mjf	Sync with HEAD.
1.146.2.2	28-Jul-2008	simonb	Sync with head.
1.146.2.1	18-Jun-2008	simonb	Sync with head.
1.148.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.148.2.1	19-Oct-2008	haad	Sync with HEAD.
1.152.4.7	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.152.4.6	24-Apr-2011	riz	Pull up following revision(s) (requested by buhrow in ticket #1601): sys/dev/pci/if_bge.c: revision 1.194 Fixes for kern/40018. Our driver initializes the Broadcom hardware to peform a tcp and udp checksum on only the payload of the tcp or udp packet, rather than the entire packet. The FreeBSD, OpenBSD and Linux drivers instruct the hardware to compute the checksum for the entire packet. I believe the bug is that some revisions of the BCM hardware, under certain circumstances, revert to doing the complete checksum calculation, as the FreeBSD, OpenBSD and Linux drivers request, while things are running. As a result, when we pull the computed checksum from the hardware and pass it up to the upper layers, we assume the checksum is the more minimal one, and the upper layers perform the appropriate checks, which, when this happens, cause the packet to be rejected because the resultant checksum is decidedly incorrect. This patch changes the driver to instruct the hardware to perform the checksum over the entire packet, just as the FreeBSD, OpenBSD and Linux drivers do, and to notify the upper layers appropriately. This patch appears to work on all revisions of the hardware that have been tested. (See the list in the bug report.) this patch is approved by tls.
1.152.4.5	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1359): sys/dev/pci/if_bge.c 1.170-1.181 sys/dev/pci/if_bgereg.h 1.53-1.56 sys/dev/pci/if_bgevar.h 1.2-1.5 sys/dev/pci/pcidevs 1.1017, 1.1021 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/miidevs 1.89, 1.93 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/dev/mii/brgphy.c 1.49, 1.52 - Fix a bug of BGE_RDMA_MODE_FIFO_LONG_BURST definition. - Fix the bug that both BGE_PCIE and BGE_PCIX are set on PCI-Express devices. - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags. - (brgphy(4)) Add support for BCM5461, BCM5464, BCM5784 and BCM5761. - (brgphy(4)) Enable brgphy_jumbo_settings(). - (brgphy(4)) Enable the Ethernet@Wirespeed function. - Add support for device properties to override the mac address and set the BGE_NO_EEPROM flag if we have an onboard device w/o SEEPROM. - Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - Fixes the bug that error bits arent cleard because these bit are W2C (in other words W1C). - Many other bugfixes from FreeBSD and OpenBSD. - remove an unused structure. - ANSI C, KNF and tabify - fix typo in comment.
1.152.4.4	04-Aug-2009	snj	Pull up following revision(s) (requested by msaitoh in ticket #870): sys/dev/pci/if_bge.c: revision 1.167 Call prop_dictionary_set_uint32() before mii_attach() so that brgphy can do prop_dictionary_get_uint32() correctly.
1.152.4.3	04-Aug-2009	snj	Apply patch (requested by msaitoh in ticket #866): For bge: - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. For brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2 and BCM5722.
1.152.4.2	02-Feb-2009	snj	branches: 1.152.4.2.4; Pull up following revision(s) (requested by he in ticket #384): sys/dev/pci/if_bge.c: revision 1.155 Make fibre cards, at least 5701 models, work, by initializing the bge_ifmedia.ifm_media field, which is not done by ifmedia_set(). Patch lifted from FreeBSD's revision 1.71.
1.152.4.1	15-Nov-2008	snj	Pull up following revision(s) (requested by cegger in ticket #44): sys/dev/pci/if_bge.c: revision 1.154 PR kern/39608: bge(4) constant freeze on BCM5906M: Correct DMA values. Patch from Karl Uwe Lockhoff via tech-kern.
1.152.4.2.4.1	21-Apr-2010	matt	sync to netbsd-5
1.152.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.152.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.152.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.155.2.2	23-Jul-2009	jym	Sync with HEAD.
1.155.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.180.4.6	12-Jun-2011	rmind	sync with head
1.180.4.5	31-May-2011	rmind	sync with head
1.180.4.4	21-Apr-2011	rmind	sync with head
1.180.4.3	05-Mar-2011	rmind	sync with head
1.180.4.2	03-Jul-2010	rmind	sync with head
1.180.4.1	30-May-2010	rmind	sync with head
1.180.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.180.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.180.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.190.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.196.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.199.6.1	18-Feb-2012	mrg	merge to -current.
1.199.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.199.2.2	30-Oct-2012	yamt	sync with head
1.199.2.1	17-Apr-2012	yamt	sync with head
1.200.8.1	29-Nov-2014	martin	Pullup revision 1.260 (via patch), requested by msaitoh in #1200: Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. This change fixes a serious stability problem on those chips.
1.200.6.1	29-Nov-2014	martin	Pullup revision 1.260 (via patch), requested by msaitoh in #1200: Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. This change fixes a serious stability problem on those chips.
1.200.2.5	16-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #1256): sys/dev/pci/if_bge.c: revision 1.278 Fix three bugs reported by enami@: - bge_miibus_writereg(): Fix a bug that BCM5906 may leave an APE lock. - Fix hwcfg4 isn't printed correctly. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices.
1.200.2.4	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/mii/brgphy.c 1.69, 1.71-1.76 via patch sys/dev/mii/brgphyreg.h 1.6-1.7, 1.9 sys/dev/pci/if_bge.c 1.259,1.267-1.276 via patch sys/dev/pci/if_bgereg.h 1.87-1.88 sys/dev/pci/if_bnx.c 1.48, 1.49, 1.52-1.57 via patch sys/dev/pci/if_bnxreg.h 1.15-1.18 sys/dev/pci/if_bnxvar.h 1.4-1.6 Changes for bge(4), bnx(4) and brgphy(4): - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719 into pcidevs. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Fix typo in comment. - KNF - brgphy(4): - Add BCM5708S support. - Add 2500SX support. - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. - Fix media detect on some Fiber chips. - Add BCM5706. - bge(4): - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET. - Add BGEPHYF_NO_WIRESPEED flag for fiber devices. - Print APE firmware correctly. - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packets about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... Tested with HP MicroServer Gen8. - Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436). - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver. - bnx(4): - Enclose #ifdef BNX_DEBUG correctly. - Add 2500SX support. - Print some information (ASCI revision, PCI status, etc). From FreeBSD.
1.200.2.3	17-Nov-2013	bouyer	Apply patch, requested by msaitoh in ticket #983: sys/dev/mii/brgphy.c 1.68 via patch sys/dev/pci/if_bge.c 1.260-1.262 via patch sys/dev/pci/if_bgereg.h 1.85-1.86 via patch sys/dev/pci/if_bgevar.h 1.17 via patch Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. Same as Linux tg3 and FreeBSD (part of FreeBSD r204978). Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670). This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion().
1.200.2.2	07-Sep-2013	bouyer	sys/dev/pci/if_bge.c 1.203-1.237, 1.239-1.241, 1.243-1.258 via patch sys/dev/pci/if_bgereg.h 1.58-1.74, 1.76-1.83 via patch sys/dev/pci/if_bgevar.h 1.10-1.16 via patch sys/dev/pci/pcidevs 1.1149 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/brgphy.c 1.61-1.63, 1.65, 1.67 via patch sys/dev/mii/miivar.h 1.61 via patch sys/dev/mii/miidevs 1.112-1.113 via patch sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/bge.4 1.13-1.15 via patch Add some device support, fix a lot of bugs and add some enahcements. - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F, BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - brgphy(4): Add BCM5756, BCM5717C, BCM5719C, BCM5720C and BCM57780. - Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Sync with BSD. No functional change: - make bge_chipid() and use it. - use switch() instead of a lot of if()s. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay(). - Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD. - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section7". 57781's document has the same note. - 5718 and 57785 document say we should wait 100us. - Change the order of register settings a little to match both Broadcom's document and Linux tg3. - Add detach function. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix a bug that the NVRAM lock timeout occured on sparc64 onboard bge. - Check BGE_SGDIG_STS when the chip is NOT a 5717 A0, like freebsd (OpenBSD rev. 1.327). - Fix the setting of Tx Random Backoff Register. - Add BGE_PCIMISCCTL_PCISTATE_RW in BGE_INIT. It's required to set the PCISTATE register correctly. - Update comments from 57XX-PG105-R.pdf and 5718-PG106-R.pdf. - Add missing bus_dmamap_sync() in bge_intr(). The status word is in DMAed area. Same as other BSDs. - Check the hardware config words and print them. This change only read them and print the values. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver. - Set DMA watermark depend on the PCI max payload size. - Add BGE_JUMBO_CAPABLE flag to some chips. With this commit, 5714, 5780, 5717, 5718, 5719 (exclude rev. A0), 5720, 57765 and 57766 are added to support jumbo frame. - Fix the setting of sc->bge_flags for 5717 and newer devices. - Fix a link detect bug on non-autopoll systems. Same as OpenBSD (rev.1.329 and 1.336) and FreeBSD (r213710). - 57765 series is not based on 5717 series. 5717 series is based on 57765 series. - Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document. - Change sysctl related functions for consistency. - Style change. - Use macro. Remove duplicated macro. Remove unused variable. - Fix comments. Add comments. - Remove extra semicolon. Remove unused code. [msaitoh, ticket #939]
1.200.2.1	26-Aug-2013	msaitoh	Pull up following revision(s) (requested by tsutsui in ticket #652): sys/dev/mii/brgphy.c 1.60 sys/dev/pci/if_bge.c 1.202 sys/dev/pci/if_bgereg.h 1.57 Add support for BCM57762 and BCM57765, found in Apple's Thunderbolt to Gigabit Ethernet adapter. PR kern/46961.
1.201.2.5	03-Dec-2017	jdolecek	update from HEAD
1.201.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.201.2.3	23-Jun-2013	tls	resync from head
1.201.2.2	25-Feb-2013	tls	resync with head
1.201.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.253.2.2	18-May-2014	rmind	sync with head
1.253.2.1	28-Aug-2013	rmind	sync with head
1.266.2.2	10-Aug-2014	tls	Rebase.
1.266.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.277.4.10	28-Aug-2017	skrll	Sync with HEAD
1.277.4.9	05-Feb-2017	skrll	Sync with HEAD
1.277.4.8	05-Oct-2016	skrll	Sync with HEAD
1.277.4.7	09-Jul-2016	skrll	Sync with HEAD
1.277.4.6	29-May-2016	skrll	Sync with HEAD
1.277.4.5	19-Mar-2016	skrll	Sync with HEAD
1.277.4.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.277.4.3	22-Sep-2015	skrll	Sync with HEAD
1.277.4.2	06-Jun-2015	skrll	Sync with HEAD
1.277.4.1	06-Apr-2015	skrll	Sync with HEAD
1.277.2.3	30-Jul-2015	martin	Pull up following revision(s) (requested by maxv in ticket #889): sys/dev/pci/if_bge.c: revision 1.283 sys/dev/pci/if_pcn.c: revision 1.60 Fix mbuf leaks in error branches (found by the Brainy code scanner)
1.277.2.2	09-May-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #740): sys/dev/pci/if_bge.c: revision 1.285 sys/dev/pci/if_bgereg.h: revision 1.90 Use another firmware command in bge_asf_driver_up(). Same as Lunux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657.
1.277.2.1	17-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #526): sys/dev/pci/if_bge.c: revision 1.278 Fix three bugs reported by enami@: - bge_miibus_writereg(): Fix a bug that BCM5906 may leave an APE lock. - Fix hwcfg4 isn't printed correctly. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices.
1.298.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.298.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.300.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.309.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.310.2.8	05-Dec-2019	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1462): sys/dev/pci/if_bge.c: revision 1.340 sys/dev/pci/if_bge.c: revision 1.341 sys/dev/pci/if_bge.c: revision 1.336 - Avoid undefined behavior in bge_setmulti(). found by kUBSan. - Avoid undefined behavior when setting the MAC address in bge_init(). found by kUBSan. Fix a bug that SK-9D41 can't detect fiber media. Check the subsystem ID correctly. This bug was added in if_bge.c rev. 1.161. - Use *_FLUSH() more. The main purpose is to wait following delay() correctly. - Add missing DELAY(80) after writing BGE_MI_MODE register.
1.310.2.7	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.310.2.6	07-Apr-2019	martin	Pull up following revision(s) (requested by bouyer in ticket #1228): sys/dev/pci/if_bge.c: revision 1.328 Fix ASIC check, BGE_ASICREV(sc->bge_chipid) is not a chipid. Fixes bogus input error reporting in stat. This is cosmetic, exept for routed(8) which thinks that the interface is broken and stops accepting routes from it.
1.310.2.5	07-Mar-2019	martin	Pull up the following, requested by msaitoh in ticket #1209: sys/dev/pci/if_bge.c 1.323-1.327 sys/dev/pci/if_bgereg.h 1.95 sys/dev/mii/brgphy.c 1.80 share/man/man4/bge.4 1.16 - Add BCM5717's another device ID support. - Add BCM5702FE, another BCM5704S and SysKonnect SK-9Mxx support. - Add support for BCM5762 ASIC devices. - Add BCM5762, BCM5725, BCM5727, BCM57764, BCM57767 and BCM57787. - Simplify PCI device table. - Add /* FALLTHROUGH */
1.310.2.4	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1130): sys/dev/pci/if_bge.c: revision 1.320 sys/dev/pci/if_bgevar.h: revision 1.24 sys/dev/pci/if_bge.c: revision 1.317 sys/dev/pci/if_bgereg.h: revision 1.94 More TSO4 fixes, from the freebsd driver: - the chip doens't want the lenght of options, but the complete lenght of ip headers (ip + tcp + options). Fix this for the BGE_IS_5717_PLUS() and BGE_IS_5705_PLUS() cases; FreeBSD doens't cover the last case so leave it as is for now. This fixes checksum failures for heavy transfers. - It looks like the transmit engine hangs if the TCP segment crosses a 4GB boundary. FreeBSD fixes it by mapping everything below 4GB; instead try detect when this happens and do the bounce only when needed. With these fixes I could transfers 3GB images over ftp at gigabit speed (112MB/s with wget) without problems. Tested on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0 - Don't destroy the dma maps if we're not disabling the adapter, avoids a KASSERT() when bus_dmamap_destroy() is called from interrupt context via bge_watchdog() Set IFF_OACTIVE only when bge_encap() fails on adapter ressource shortage. Otherwise we may set IFF_OACTIVE while no transmit is in progress, and nothing will clear it. If bus_dmamap_load_mbuf() fails with EFBIG, m_defrag() the chain and retry. Refine the check for the 4GB boundary workaround (a fragment should also not cross the boundary), and do it only for TSO. If bge_encap() fails and didn't set IFF_OACTIVE, drop the packet. Bring in more hardware bug workarounds from freebsd. With these it seems that a BCM5720 A0 can survive a few hours of internet load with TSO4 enabled.
1.310.2.3	26-Nov-2018	snj	Pull up following revision(s) (requested by bouyer in ticket #1107): sys/dev/pci/if_bge.c: revision 1.316 5717 and newer handle TSO in a sighly different way. From FreeBSD. While there, use different lowat/hiwat values if MTU is larger than ETHERMTU, from FreeBSD. Tested at LIP6 on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0 with all ipv4 checksums and TSO4 enabled.
1.310.2.2	24-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #389): sys/net/if_ether.h: revision 1.69 sys/net/if_vlan.c: revision 1.108 sys/dev/pci/if_bge.c: revision 1.313 sys/net/if_vlanvar.h: revision 1.11 sys/net/if_vlanvar.h: revision 1.12 sys/net/if_ether.h: revision 1.70 sys/net/if_vlan.c: revision 1.110 sys/dev/pci/if_wm.c: revision 1.544 sys/dev/pci/if_wmreg.h: revision 1.105 Fix a bug that a vlan packet which has priority or CFI bit in the tag causes panic. Revert part of if_bge.c 1.312. It's not required to mask other than VLAN ID bits in VLAN tag. Revert if_wmreg.h 1.104 and if_wm.c 1.542. It's not required to mask other than VLAN ID bits in VLAN tag. No functional change: - u_int16_t -> uint16_t - u_short -> uint16_t - tag_hash_func -> vlan_tag_hash - 0 -> NULL because vlr_parent is a pointer.
1.310.2.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.313.2.6	26-Jan-2019	pgoyette	Sync with HEAD
1.313.2.5	18-Jan-2019	pgoyette	Synch with HEAD
1.313.2.4	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.313.2.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.313.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.313.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.314.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.314.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.314.2.1	10-Jun-2019	christos	Sync with HEAD
1.335.2.2	19-May-2022	martin	Pull up following revision(s) (requested by buhrow in ticket #1456): sys/dev/pci/if_bge.c: revision 1.353 For chips which contain an ASF/IPMI firmware, instruct the chip to shut the host ASF firmware down when attaching the device so the IPMI BMC can use the same physical port even when NetBSD doesn't have a network configuration on the device. By contrast, when the device gets a network configuration assigned to it and bge_init() is called, the host ASF firmware is brought up so both NetBSD and the IPMI BMc can use the same physical port. This now matches FreeBSD behavior, as well as behavior from NetBSD-5.2. Tested on a Sunfire X2200-M2 system with the following chip: bge1 at pci7 dev 4 function 1: Broadcom BCM5715 Gigabit Ethernet bge1: interrupting at ioapic0 pin 11 bge1: HW config 000000d4, 00000014, 00000000, 00000000 00000000 bge1: ASIC BCM5715 A3 (0x9003), Ethernet address 00:1e:68:XX:XX:XX bge1: setting short Tx thresholds brgphy1 at bge1 phy 1: BCM5714 1000BASE-T/X media interface, rev. 0 brgphy1: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto
1.335.2.1	27-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #487): sys/dev/pci/if_bge.c: revision 1.340 sys/dev/pci/if_bge.c: revision 1.341 sys/dev/pci/if_bge.c: revision 1.342 sys/dev/pci/if_bge.c: revision 1.336 - Avoid undefined behavior in bge_setmulti(). found by kUBSan. - Avoid undefined behavior when setting the MAC address in bge_init(). found by kUBSan. Fix a bug that SK-9D41 can't detect fiber media. Check the subsystem ID correctly. This bug was added in if_bge.c rev. 1.161. - Use *_FLUSH() more. The main purpose is to wait following delay() correctly. - Add missing DELAY(80) after writing BGE_MI_MODE register. Modify PHY initialization code. This change fix a bug that SK-9D21 doesn't detect MII PHY. - Add error check to bge_miibus_writereg(). - Change return value of bge_miibus_readreg() when a read error occurred. It also add error message using with aprint_debug_dev(). This error occurs on some devices while detecting MII devices. - Move the location of BGE_MI_MODE register's initialization to next to bge_chipinit(). - Set BGE_MAC_MODE before calling ifmedia_init() and/or mii_attach(). - Add retry code for mii_attach() failed. Same as FreeBSD.
1.342.2.1	29-Feb-2020	ad	Sync with head.
1.388.2.1	13-Oct-2024	martin	Pull up following revision(s) (requested by skrll in ticket #968): sys/dev/pci/if_bgevar.h: revision 1.41 sys/dev/pci/if_bgevar.h: revision 1.42 sys/dev/pci/if_bge.c: revision 1.394 s/jumo/jumbo/ in comments. Apply changes from PR/58584 after testing (and fixing). Tested on an Apple M1.
1.392.2.1	02-Aug-2025	perseant	Sync with HEAD
1.98	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.97	02-Jul-2022	skrll	Use C99 types. same code before and after.
1.96	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.95	20-Feb-2019	msaitoh	- Add support for BCM5762 ASIC devices. - Add BCM5762, BCM5725, BCM5727, BCM57764, BCM57767 and BCM57787.
1.94	02-Dec-2018	bouyer	Don't destroy the dma maps if we're not disabling the adapter, avoids a KASSERT() when bus_dmamap_destroy() is called from interrupt context via bge_watchdog() Set IFF_OACTIVE only when bge_encap() fails on adapter ressource shortage. Otherwise we may set IFF_OACTIVE while no transmit is in progress, and nothing will clear it. If bus_dmamap_load_mbuf() fails with EFBIG, m_defrag() the chain and retry. Refine the check for the 4GB boundary workaround (a fragment should also not cross the boundary), and do it only for TSO. If bge_encap() fails and didn't set IFF_OACTIVE, drop the packet. Bring in more hardware bug workarounds from freebsd. With these it seems that a BCM5720 A0 can survive a few hours of internet load with TSO4 enabled.
1.93	13-Apr-2017	msaitoh	branches: 1.93.4; 1.93.10; 1.93.12; - Read chip ID correctly on BCM57782 and BCM57786. - Add BXCM57766 A0 string. - Add comment for BCM5776[56]'s workaround.
1.92	12-Apr-2017	msaitoh	Add a workaround for BCM57780 to prevent device timeout. This change prevent L1PM feature do wrongy. Tested with Dell latitude 2120.
1.91	17-May-2015	msaitoh	branches: 1.91.2; 1.91.4; - Add MSI support. - Use tagged status function for 5717 and newer devices. All controllers except BCM5700 support tagged status but we use tagged status only for MSI case on BCM5717. Otherwise MSI on BCM5717 does not work. Same as other *BSDs.
1.90	30-Apr-2015	msaitoh	Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657.
1.89	30-Apr-2015	msaitoh	- BGE_CHIPID_BCM5784_A1 is not 0x5784100 but 0x5784001. Only dmesg output is affected. - Add BGE_CHIPREV_5784_AX and BGE_CHIPID_BCM5784_B0.
1.88	17-Jun-2014	msaitoh	branches: 1.88.2; 1.88.4; - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET.
1.87	27-May-2014	msaitoh	Fix typo in a comment.
1.86	31-Oct-2013	msaitoh	branches: 1.86.2; Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670). This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion().
1.85	31-Oct-2013	msaitoh	Cleanup flag setting. No functional change. - Split flags into bge_flags and bge_phy_flags. - Rename flags. PHY related flags are prefixed with BGEPHYF_. Other flags are prefixed with BGEF_.
1.84	13-Jul-2013	msaitoh	Fix the chip revision of 57765_AX.
1.83	08-Jul-2013	msaitoh	- Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document.
1.82	08-Jul-2013	msaitoh	57765 series is not based on 5717 series. 5717 series is based on 57765 series.
1.81	31-May-2013	msaitoh	branches: 1.81.2; Remove unused macros.
1.80	30-May-2013	msaitoh	0x4B657654(BGE_SRAM_FW_MB_MAGIC) == ~0xB49A89AB(BGE_SRAM_FW_MB_RESET_MAGIC), so the same code I added in rev. 1.223 exists in bge_poll_fw(), too. Remove it. And then, call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver.
1.79	28-May-2013	msaitoh	Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly.
1.78	14-May-2013	msaitoh	Add some definitions. Fix compile error.
1.77	08-May-2013	msaitoh	No functional change: - Fix typo - Add comment - Remove duplicated definitions
1.76	08-May-2013	msaitoh	Check the hardware config words and print them. This change only read them and print the values.
1.75	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.74	11-Apr-2013	msaitoh	Fix typo in comment.
1.73	11-Apr-2013	msaitoh	Add BGE_PCIMISCCTL_PCISTATE_RW in BGE_INIT. It's required to set the PCISTATE register correctly.
1.72	08-Apr-2013	msaitoh	Fix the setting of Tx Random Backoff Register.
1.71	27-Mar-2013	msaitoh	Add BGE_SRAM_FW_MB_RESET_MAGIC again.
1.70	27-Mar-2013	msaitoh	Revert rev. 1.68.
1.69	27-Mar-2013	msaitoh	Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly.
1.68	21-Mar-2013	msaitoh	Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section 7". 57781's document has the same note. I verified the loop really looped on those chips. But, I don't know why tg3 doesn't it.
1.67	21-Mar-2013	msaitoh	- Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Add/sync some of the comments - Use macro. - Remove unused code.
1.66	19-Mar-2013	msaitoh	- Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD.
1.65	19-Mar-2013	msaitoh	Fix three bugs: - An workaround for TX data corruption is only for 5719 "A0". Fix the wrong evaluation. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay().
1.64	17-Mar-2013	msaitoh	No functional change: - Fix comment - Use macro - Remove unused macro
1.63	17-Mar-2013	msaitoh	- ADD support for BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Use macros. - Remove unused code. - Sync with the latest {Free,Open}BSD's bge.
1.62	13-Mar-2013	msaitoh	- Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Use macro.
1.61	27-Feb-2013	msaitoh	No functional change: - Fix comments. - Change macro name. - Change sysctl related functions for consistency.
1.60	26-Feb-2013	msaitoh	Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220)
1.59	26-Feb-2013	msaitoh	Use macro. Remove extra semicolon. No functional change.
1.58	25-Feb-2013	msaitoh	Use macro. Remove duplicated macro. Remove unused variable. No functional Change.
1.57	17-Sep-2012	tsutsui	Add support for BCM57762, found in Apple's Thunderbolt to Gigabit Ethernet Adapter. PR kern/46961 from Ryo ONODERA. Also revert weird BGE_MAX_FRAMELEN macro definition to sync its usage with the original FreeBSD's bge.
1.56	03-Feb-2010	msaitoh	branches: 1.56.12; 1.56.18; 1.56.22; - Make the initialize sequence close to the document. - Fix ASF heartbeat sending bug (FreeBSD rev. 1.271 (r202821)) - Use new handshake command for BCM5750 or new controllers (FreeBSD rev. 1.272 (r202822))
1.55	28-Jan-2010	msaitoh	- Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - remove an unused structure. - KNF
1.54	24-Jan-2010	msaitoh	Many changes from FreeBSD and OpenBSD: - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Many other bugfixes. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags.
1.53	04-Aug-2009	taca	Fix a bug of BGE_RDMA_MODE_FIFO_LONG_BURST definition. Confirmed by msaito@.
1.52	23-Apr-2009	msaitoh	use proplib
1.51	19-Apr-2009	msaitoh	Some fixes and enhancements: Both if_bge* and brgphy.[ch]: Check the parent's MAC and use the quirk code for the bug like OpenBSD and FreeBSD. Some bugs can't identify the PHY ID. For example, 5704 Ax has the ADC bug, but 5704 A3 and 5704 B0 have the same PHY ID and revision. Add BGE_PHY_CRC_BUG, BGE_PHY_ADC_BUG, BGE_PHY_5704_A0_BUG, BGE_PHY_JITTER_BUG, BGE_PHY_ADJUST_TRIM and BGE_PHY_BER_BUG for this flag. Some of the DSP patches are newly taken from OpenBSD and FreeBSD. if_bge*: Remove duplicated BGE_CHIPID_BCM5714_A0 entry in the known CHIPID table. Fix obsolete comments. brgphy.[ch]: Add some PHY IDs. TODO: Add more three quirk code into bge and brgphy (the brgphy side's are currently #if0'ed). Add support for bnx into brgphy (currently #if0'ed)
1.50	16-Apr-2009	msaitoh	Take some code from FreeBSD and OpenBSD: Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) Add some fiber support (OpenBSD 1.220) Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) Add some known CHIPID entries.
1.49	22-Mar-2009	msaitoh	Make bge(4) close to FreBSD and OpenBSD. NO FUNCTIONAL CHANGE BY THIS COMMIT.
1.48	25-Aug-2008	cegger	branches: 1.48.2; 1.48.4; 1.48.8; 1.48.12; PR 39241: Add support for Broadcom BCM5906(M) from Karl Uwe Lockhoff
1.47	24-Jul-2008	sborrill	Add support for Broadcom 5722 and identify 5755 ASICs.
1.46	17-Jun-2008	mlelstv	branches: 1.46.2; add rnd(4) hooks. Note that interrupts on bge hardware may occur with certain patterns, especially when the interrupt mitigation logic kicks in. So this might be a very weak entropy source.
1.45	09-Dec-2007	jmcneill	branches: 1.45.10; 1.45.12; 1.45.14; 1.45.16; 1.45.18; Merge jmcneill-pm branch.
1.44	24-Sep-2007	joerg	branches: 1.44.6; 1.44.8; Convert bge(4) to device/softc separation. Use aprint_* instead of printf all over the place. Only the debug macros are kept. Place the TSO message about the hard case under debug.
1.43	06-Aug-2007	markd	branches: 1.43.2; 1.43.4; 1.43.6; Add a couple more BGE_CHIPIDs
1.42	18-May-2007	pavel	branches: 1.42.2; 1.42.6; from FreeBSD's src/sys/dev/bge/if_bgereg.h rev 1.22: Fix a typo. BGE_HCC_TX_MAX_COAL_BDS_INT is at offset 0x3C24, not 0x3C34 Noticed by Patrick Welche on tech-kern@, ok cube.
1.41	04-Mar-2007	christos	branches: 1.41.2; 1.41.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.40	17-Feb-2007	bouyer	bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet).
1.39	26-Nov-2006	tsutsui	branches: 1.39.2; 1.39.4; Revert a CHIPID entry in the previous: In OpenBSD's if_bgereg.h, CHIPID 0x4101000 is defined as BCM5750_B1 but our PR kern/31028 says it's BCM5751_A1 on BCM5751M on IBM T43p, and the value seems reasonable.
1.38	26-Nov-2006	tsutsui	Add and fix some BGE_CHIPIDs. From FreeBSD and OpenBSD.
1.37	26-Nov-2006	tsutsui	TAB/space cleanup.
1.36	26-Nov-2006	tsutsui	Add support for BCM5752M. from Matt Kolb on tech-kern and OpenBSD.
1.35	26-Nov-2006	tsutsui	make DMA descriptor members volatile.
1.34	18-Apr-2006	christos	branches: 1.34.8; 1.34.10; PR/33828: Pawel Chwalowski: Add support for Broadcom BCM5789 chip
1.33	16-Feb-2006	perry	branches: 1.33.2; 1.33.4; 1.33.6; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.32	24-Dec-2005	perry	branches: 1.32.2; 1.32.4; 1.32.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.31	12-Dec-2005	jonathan	Add support for bcm5714 and (possibly incorrect) support for the bcm5715 PCI-Express dual-port NICs. Taken from the Broadcom-supplied Linux driver, bcm-8.3.13a. Add nascent support for the bcm5780, but (since I have no bcm5780 to test), don't yet add an entry which allows us to match or attach to a 5780. Change 5752 support: Follow the lead of the Broadcom-supplied Linux driver (bcm) and the Linux tg3 driver, and treat the 5752 like a 5750.
1.30	07-Dec-2005	jonathan	Add PCI device-ids to sys/dev/pci/pcidevs for newer Broadcom gigabit ethernet devices: bcm5714, bcm5752 (previously the 5789, but fvdl committed that whilst I was musing). Add definitions to sys/dev/pci/if_bgereg.h for the Broadcom 57xx-family ASIC revisions on these newer chips. Add entries to the PCI-device-version and bge-internal-asic-revision tables in sys/dev/pci/if_bge.c to use these new devices. Pending further information, follow the lead of FreeBSD's if_bge.c driver and, pending further info, treat these new chips as we do the 5750.
1.29	25-Nov-2005	tsutsui	Remove unnecessary byteswap code so that bge(4) should work at least on macppc.
1.28	24-Nov-2005	jonathan	Add TSO support to bge(4) for PCI-Express bge devices, which are shipped from the factory with TSO-capable firmware. The TSO support here may also work on 5705 chips, but that is (so far) untested. TSO support written after careful reading of the Linux tg3 driver, and (after attempting to deconstruct the cut-and-paste mess therein) very close reading of the Broadcom-supplied Linux driver, particularly the building of Tx-DMA buffer descriptors (bds). The TSO code herein was then rewritten from scratch, circa 4am local time, October 27 2005. (In other words: this is 4am software; caveat emptor.) Other magic register settings in this patch are required; without them, attepmting to use TSO locks up the chip. The required register settings were extracted from the cited Linux drivers. Note that TSO-capable firmware for the 5703/5704 is distributed in non-GPL form with the aforementioned Linux drivers. Once the 5705 case is debugged, (particularly the pseudo-header checksum precalculation flagged with an XXX) downloading that TSO-capable firmware to the 5703/5704 should, in principle, enable TSO support on all but the original bcm5700 (I forget if the 5701 can support TSO, or not). Note also that the ``hard case '' of IP/TCP headers spanning more than one mbuf is not handled; I haven't been able to trigger it. In any case, since TSO applies only to packets generated by the local TCP, and our TCP always leaves space for TCP headers and a normal IP header, TSO on an IP/TCP header spanning multiple headers can only arise due to insertion of IP options. I beleive that we are clearly better off outlawing that case, and requiring ip_insertoptions() to pull-up TCP headers on any packets with M_CSUM_TSOv4 set. As far as I know, bge hardwar does not support TSO for IPv6.
1.27	15-Nov-2005	jonathan	Correct a performance bug from Bill Paul's original FreeBSD bge(4) driver: Each call to the FreeBSD bge_start() routine the transmit producer pointer index from the chip mailbox register BGE_MBX_TX_HOST_PROD0_LO. The local copy of that value is then updated by bge_encap() as bge_encap() encapsulates packets in the Tx ring. If bge_encap() succeds in encpuslating one or more packets, bge_start() tells the chip to start sending the newly-encinitiates writes the new value back to the chip mailbox register. However, comparison of the Linux drivers (Broadcom-supplied and open-source tg3.c) and to the OpenSolaris driver confirms that register BGE_MBX_TX_HOST_PROD0_LO is write-only to software. Thus, we can just keep a copy in the softc, and eliminate the (expensive) PCI register write on each call to bge_start(). ``Make it so''.
1.26	28-Aug-2005	thorpej	branches: 1.26.6; Remove unused and incorrect macro BGE_HOSTADDR().
1.25	22-Aug-2005	gavan	Add quirk for BGE_CHIPID_BCM5751_A1. From Kurt Schreiner <ks at ub.uni-mainz dot de> via PR kern/31028.
1.24	27-Feb-2005	perry	branches: 1.24.2; 1.24.4; nuke trailing whitespace
1.23	26-Jan-2005	jmcneill	Add powerhook to bge(4)
1.22	28-Oct-2004	cube	branches: 1.22.4; 1.22.6; Add support for the BCM575x variants. They're close to the 5705 one, with some special cases for the PCI-Express versions. As there is no documentation available, all of this is inspired from Linux or taken from FreeBSD (itself inspired from Linux). There are a lot of magic numbers. Tested on two BCM5751 devices as found in newer Dell computers.
1.21	15-May-2004	thorpej	Add PAUSE-related event counters. Slightly modified from patch supplied by HITOSHI Osada.
1.20	15-May-2004	thorpej	Patches from HITOSHI Osada: * Set buffer management high water marks for MTU > 1514. * Set BGE_MAX_RX_FRAME_LOWAT (from Linux driver).
1.19	10-Apr-2004	thorpej	Add 802.3x flow control support. From HITOSHI Osada.
1.18	20-Mar-2004	jonathan	branches: 1.18.4; Possibloe fix for some bge chip revisions taking a long time to reset (e.g., polling for a half-second or more at splnet(), blocking most interrupts, durin an ifconfig down/ifconfig up). Appears to help for a 5704C rev A3, which is the only chip I've ever seen that had even a mild version of the reported problem.
1.17	22-Jan-2004	jonathan	Add hooks to dynamically change the bge Rx interrupt thresholds. I found empiricaly that (at least on bcm5700s) the Rx coalesce and bd counts cannot be updated on-the-fly; attempting to do so (even at splhigh()) causes weird behaviour. Instead, add a softc flag to record that the desired softc values for Rx-interrupt thresholds have changed; check that boolean in the interrupt routine. If set, apply the new values there and clear the flag.
1.16	22-Jan-2004	jonathan	Add some extra register definitions from FreeBSD and Linux: BGE_MISC_TIMER register definition, and more bit definitions for HCC-mode. (I would credit individuals, but the attributions are buried deep in my own tree.)
1.15	08-Dec-2003	pooka	Match 5703 ASIC A3. Also, when printing the ASIC ID, print the revision in hex in all cases to avoid extra confusion if only the "superclass" of the ASIC is recognized.
1.14	02-Nov-2003	wiz	boundary, not boundry. Inspired by Tom Cosgrove.
1.13	23-Oct-2003	fvdl	* match more cards (from FreeBSD) * don't set the NOCRC bit in the mode control register, it can cause problems on some chips (from the broadcom errata via FreeBSD) * implement a fallback quirktable that is searched only using the major asic revision, so that the driver has a shot at supporting newer versions properly without modification * rename asicrev -> chipid, like the FreeBSD driver
1.12	27-Aug-2003	fvdl	Support the BCM5704C rev 3. From Nicolas Joly, PR 22620
1.11	17-Jul-2003	hannken	Add support for the BCM5705: - Only BCM5705M asic rev A1 was tested. Thanks to Bill Paul (wpaul@freebsd.org) for help and support. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.10	01-Jun-2003	fvdl	branches: 1.10.2; Add newer 5704 asic revs, and add them to the table, with the BGE_QUIRK_ONLY_PHY_1 quirk in their entry.
1.9	07-Mar-2003	jonathan	Commit workaround for hardware bug: 5701 chips running in PCI-X mode require that the DMA receive buffers be aligned. The driver was deliberately mis-aligning by 2 bytes, to force the layer-2/3 headers to 32-bit alignment. Workaround: if chip is a 5701, and is in PCI-X mode, leave the DMA buffers aligned. If the host CPU requires alignment, copy the buffer after reception to force aligment. Tested on an i386 in PCI-X bus, with __NO_STRICT_ALIGNMENT forced off. Patch and comments reworked to minimize drift from FreeBSD if_bge.c rev 1.14.
1.8	06-Mar-2003	jonathan	Commit the FreeBSD alternative of a private patch to check NIC internal ram for using TBI, versus MII/GMII. FreeBSD commit log and versions: Obtain the media type from the shared memory and only use the eeprom as a fallback. if_bge.c: rev 1.30 if_bgereg.h: rev 1.13 Thanks to Paul Saab @mu.org.
1.7	06-Mar-2003	jonathan	FreeBSD fix for problems reported (to FreeBSD) with 16-bit accesses to 32-bit device registers. Add (untested) NetBSD fix for big-endian hosts. From FreeBSD versions: if_bge.c: 1.26 and 1.27 if_bgereg.h: 1.12 FreeBSD (MFC) comment: MFC: Avoid 16-bit accesses to device memory. Use volatile where appropriate in referencing device memory. Parenthesize BGE_HOSTADDR define properly.
1.6	17-Jan-2003	fvdl	branches: 1.6.2; Add local_ctl_reg field to softc.
1.5	17-Jan-2003	jonathan	Add register definitions for 5704 internal ASIC revision. Add cpp definitions for the DMA control register fields needed for 5703/5704 configuration on PCI-X. Add softc copy of internal"local control" register clobbered by reset.
1.4	13-Jul-2002	thorpej	branches: 1.4.2; 1.4.4; 1.4.6; Give BCM5700 Ax and Bx revs the BGE_QUIRK_LINK_STATE_BROKEN quirk, and test for that, rather than checking revision where the quirk is handled.
1.3	13-Jul-2002	thorpej	Add a lookup table with ASIC revision, name, and any quirks. Print ASIC revision at attach time. Add BCM5703 revisions.
1.2	24-Jun-2002	fvdl	Add a few revision definitions.
1.1	22-Jun-2002	fvdl	Driver for Broadcom 570x / Altima AC 1000 gigabit ethernet cards. Written by Bill Paul for FreeBSD. This port started out with the port done to OpenBSD by nate@openbsd.org, but ended up looking much more like the port of the ti driver done by drochner@netbsd.org (they are similar in structure).
1.4.6.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.4.6.1	13-Jul-2002	jdolecek	file if_bgereg.h was added on branch kqueue on 2002-09-06 08:45:14 +0000
1.4.4.3	17-Jan-2003	thorpej	Sync with HEAD.
1.4.4.2	01-Aug-2002	nathanw	Catch up to -current.
1.4.4.1	13-Jul-2002	nathanw	file if_bgereg.h was added on branch nathanw_sa on 2002-08-01 02:45:14 +0000
1.4.2.2	15-Jul-2002	gehenna	catch up with -current.
1.4.2.1	13-Jul-2002	gehenna	file if_bgereg.h was added on branch gehenna-devsw on 2002-07-15 10:35:36 +0000
1.6.2.4	16-Jun-2003	grant	Pull up revision 1.8 (requested by tron in ticket #1250): Commit the FreeBSD alternative of a private patch to check NIC internal ram for using TBI, versus MII/GMII. FreeBSD commit log and versions: Obtain the media type from the shared memory and only use the eeprom as a fallback. if_bge.c: rev 1.30 if_bgereg.h: rev 1.13 Thanks to Paul Saab @mu.org.
1.6.2.3	16-Jun-2003	grant	Pull up revision 1.7 (requested by tron in ticket #1249): FreeBSD fix for problems reported (to FreeBSD) with 16-bit accesses to 32-bit device registers. Add (untested) NetBSD fix for big-endian hosts. From FreeBSD versions: if_bge.c: 1.26 and 1.27 if_bgereg.h: 1.12 FreeBSD (MFC) comment: MFC: Avoid 16-bit accesses to device memory. Use volatile where appropriate in referencing device memory. Parenthesize BGE_HOSTADDR define properly.
1.6.2.2	28-Jan-2003	jmc	Pullup from -current (lots of revisions 1.1-1.6 (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.6.2.1	17-Jan-2003	jmc	file if_bgereg.h was added on branch netbsd-1-6 on 2003-01-28 06:20:12 +0000
1.10.2.8	11-Dec-2005	christos	Sync with head.
1.10.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.10.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.10.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.18.4.1	01-Dec-2005	riz	Pull up following revision(s) (requested by tsutsui in ticket #10169): sys/dev/pci/if_bgereg.h: revision 1.29 Remove unnecessary byteswap code so that bge(4) should work at least on macppc.
1.22.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.22.6.1	12-Feb-2005	yamt	sync with head.
1.22.4.1	29-Apr-2005	kent	sync with -current
1.24.4.6	21-Jan-2008	yamt	sync with head
1.24.4.5	27-Oct-2007	yamt	sync with head.
1.24.4.4	03-Sep-2007	yamt	sync with head.
1.24.4.3	26-Feb-2007	yamt	sync with head.
1.24.4.2	30-Dec-2006	yamt	sync with head.
1.24.4.1	21-Jun-2006	yamt	sync with head.
1.24.2.6	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1862): sys/dev/pci/if_bge.c 1.89,1.95,1.97,1.101,1.103, 1.105,1.107,1.117-1.122, 1.128-1.131,1.133-1.135,1.139 via patch sys/dev/pci/if_bgereg.h 1.31,1.34-1.39,1.42-1.43 sys/dev/pci/pcidevs patch Some cleanups(typo, whitespace, debug message). Add TSO support to bge(4) Add support for BCM5715, BCM5752M, BCM5754, BCM5755, BCM5786 and 5787. Add support for BCM5789. Fixes PR/33828 from Pawel Chwalowskibg. Add support for BCM5753 and BCM5753M Fixes PR kern/36139 from Mustafa Dogan. Add and fix some BGE_CHIPIDs. From FreeBSD and OpenBSD. In bge_ioctl(), don't call bge_init() if the interface is already up (Makes ifwatchd work on this interface). BCM5701 B2 requires BGE_QUIRK_ONLY_PHY_1 quirk. From Brian Baird in kern/23778. Move TX ring full sanity check further up and check the number of DMA segments from the DMA map, instead of counting the DMA segments in the for loop and breaking out later. Unload the DMA map if encountering an error condition. Fix a NULL reference on failing mbuf allocation. In bge_start return if IFF_OACTIVE is set in combination with IFF_RUNNING, drop obscure check for length of interface queue. Remove message about failing bge_encap, it can happen just too easily because of full descriptor rings. Fix some watchdog timeout and "block failed to stop" problems on some cards. Fix "pcistate failed to revert" problem on many cards.
1.24.2.5	05-Mar-2007	ghen	Pull up following revision(s) (requested by bouyer in ticket #1671): sys/dev/pci/if_bgereg.h: revision 1.40 sys/dev/pci/if_bge.c: revision 1.124 sys/dev/pci/if_bge.c: revision 1.125 bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet). We have to bus_mapmap_load() after m_adj, otherwise the map will point at the wrong place in the buffer. Fix an issue in the !sc->bge_rx_alignment_bug case reported by Nicolas Joly. While there remove a duplicate m_adj(), and add a missing bus_dmamap_unload().
1.24.2.4	04-Jun-2006	tron	Apply patch (requested by jonathan in ticket #1351): Add support to bge(4) for bcm5714, bcm5715, bcm5780/HT-2000 and bcm5752 chip variants.
1.24.2.3	27-Nov-2005	riz	Pull up following revision(s) (requested by tsutsui in ticket #1004): sys/dev/pci/if_bgereg.h: revision 1.29 Remove unnecessary byteswap code so that bge(4) should work at least on macppc.
1.24.2.2	22-Nov-2005	tron	Pull up following revision(s) (requested by jonathan in ticket #985): sys/dev/pci/if_bgereg.h: revision 1.27 sys/dev/pci/if_bge.c: revision 1.94 Correct a performance bug from Bill Paul's original FreeBSD bge(4) driver: Each call to the FreeBSD bge_start() routine the transmit producer pointer index from the chip mailbox register BGE_MBX_TX_HOST_PROD0_LO. The local copy of that value is then updated by bge_encap() as bge_encap() encapsulates packets in the Tx ring. If bge_encap() succeds in encpuslating one or more packets, bge_start() tells the chip to start sending the newly-encinitiates writes the new value back to the chip mailbox register. However, comparison of the Linux drivers (Broadcom-supplied and open-source tg3.c) and to the OpenSolaris driver confirms that register BGE_MBX_TX_HOST_PROD0_LO is write-only to software. Thus, we can just keep a copy in the softc, and eliminate the (expensive) PCI register write on each call to bge_start(). ``Make it so''.
1.24.2.1	08-Sep-2005	tron	Pull up following revision(s) (requested by gavan in ticket #752): sys/dev/pci/if_bgereg.h: revision 1.25 sys/dev/pci/if_bge.c: revision 1.92 Add quirk for BGE_CHIPID_BCM5751_A1. From Kurt Schreiner <ks at ub.uni-mainz dot de> via PR kern/31028.
1.26.6.2	29-Nov-2005	yamt	sync with head.
1.26.6.1	22-Nov-2005	yamt	sync with head.
1.32.6.1	22-Apr-2006	simonb	Sync with head.
1.32.4.1	09-Sep-2006	rpaulo	sync with head
1.32.2.1	18-Feb-2006	yamt	sync with head.
1.33.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.33.4.1	19-Apr-2006	elad	sync with head.
1.33.2.1	24-May-2006	yamt	sync with head.
1.34.10.1	10-Dec-2006	yamt	sync with head.
1.34.8.1	12-Jan-2007	ad	Sync with head.
1.39.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.39.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.39.2.5	18-Aug-2009	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1342): sys/dev/pci/if_bge.c: revisions 1.135, 1.146, 1.151-1.152, 1.155-1.167 via patch sys/dev/pci/if_bgereg.h: revisions 1.43, 1.48-1.52 sys/dev/pci/if_bgevar.h: revision 1.1 sys/dev/mii/brgphy.c: revisions 1.36, 1.40, 1.43-1.44 and 1.47-1.48 via patch sys/dev/mii/brgphyreg.h: revisions 1.3-1.4 sys/dev/mii/miidevs: revisions 1.75, 1.77, 1.85 sys/dev/mii/miidevs.h: regen sys/dev/mii/miidevs_data.h: regen sys/dev/mii/miivar.h: revision 1.54 For bge: - Really set Tx threshold. - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the "link changed" bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. - Add BCM5786 support. for brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2, BCM5708C, BCM5722, and BCM5906.
1.39.2.4	20-Aug-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1172): sys/dev/pci/if_bge.c: revision 1.149 sys/dev/pci/if_bgereg.h: revision 1.47 Add support for Broadcom 5722 and identify 5755 ASICs.
1.39.2.3	31-Aug-2007	xtraeme	Pull up following revision(s) (requested by phil in ticket #819): sys/dev/mii/miidevs.h: regen sys/dev/mii/miidevs: revision 1.71 sys/dev/pci/if_bge.c: revision 1.133 sys/dev/pci/if_bge.c: revision 1.134 sys/dev/pci/if_bgereg.h: revision 1.43 sys/dev/mii/brgphy.c: revision 1.34 sys/dev/mii/miidevs_data.h: regen Add BCM5755 and 5754 Gig-E PHYs. Regen for BCM5755 and 5754. Add support for BCM5754 and BCM5755. From suggestions from Jonathan Stone. Add a couple more BGE_CHIPIDs Add support for BCM5755 and BCM5754/5787. Suggestions from Jonathan Stone and fixes from OpenBSD for various hardware bugs/errata. Disable firmware fastboot on 5754 and 5755. From FreeBSD by way of Phil Nelson.
1.39.2.2	30-May-2007	riz	Pull up following revision(s) (requested by pavel in ticket #671): sys/dev/pci/if_bgereg.h: revision 1.42 from FreeBSD's src/sys/dev/bge/if_bgereg.h rev 1.22: Fix a typo. BGE_HCC_TX_MAX_COAL_BDS_INT is at offset 0x3C24, not 0x3C34 Noticed by Patrick Welche on tech-kern@, ok cube.
1.39.2.1	27-Feb-2007	riz	branches: 1.39.2.1.2; Pull up following revision(s) (requested by bouyer in ticket #466): sys/dev/pci/if_bgereg.h: revision 1.40 sys/dev/pci/if_bge.c: revision 1.124-1.125 bus_dma fixes for bge(4). It now works on amd64 with more than 4GB physical RAM (jumbo frames untested yet). We have to bus_mapmap_load() after m_adj, otherwise the map will point at the wrong place in the buffer. Fix an issue in the !sc->bge_rx_alignment_bug case reported by Nicolas Joly. While there remove a duplicate m_adj(), and add a missing bus_dmamap_unload().
1.39.2.1.2.3	04-Sep-2008	skrll	Sync with netbsd-4.
1.39.2.1.2.2	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.39.2.1.2.1	04-Jun-2007	wrstuden	Update to today's netbsd-4.
1.41.4.1	11-Jul-2007	mjf	Sync with head.
1.41.2.3	09-Oct-2007	ad	Sync with head.
1.41.2.2	20-Aug-2007	ad	Sync with HEAD.
1.41.2.1	27-May-2007	ad	Sync with head.
1.42.6.4	02-Oct-2007	joerg	Sync with HEAD.
1.42.6.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.42.6.2	21-Aug-2007	joerg	Make bge(4) a bit more like other drivers in the tree with regard to the data stored in the softc. Convert to pnp_register.
1.42.6.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.42.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.43.6.2	06-Aug-2007	markd	Add a couple more BGE_CHIPIDs
1.43.6.1	06-Aug-2007	markd	file if_bgereg.h was added on branch matt-mips64 on 2007-08-06 12:23:09 +0000
1.43.4.1	06-Oct-2007	yamt	sync with head.
1.43.2.2	09-Jan-2008	matt	sync with HEAD
1.43.2.1	06-Nov-2007	matt	sync with HEAD
1.44.8.1	11-Dec-2007	yamt	sync with head.
1.44.6.1	26-Dec-2007	ad	Sync with head.
1.45.18.2	28-Jul-2008	simonb	Sync with head.
1.45.18.1	18-Jun-2008	simonb	Sync with head.
1.45.16.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.45.16.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.45.14.3	11-Mar-2010	yamt	sync with head
1.45.14.2	19-Aug-2009	yamt	sync with head.
1.45.14.1	04-May-2009	yamt	sync with head.
1.45.12.1	17-Jun-2008	yamt	sync with head.
1.45.10.2	28-Sep-2008	mjf	Sync with HEAD.
1.45.10.1	29-Jun-2008	mjf	Sync with HEAD.
1.46.2.1	19-Oct-2008	haad	Sync with HEAD.
1.48.12.1	21-Apr-2010	matt	sync to netbsd-5
1.48.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.48.4.3	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.48.4.2	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1359): sys/dev/pci/if_bge.c 1.170-1.181 sys/dev/pci/if_bgereg.h 1.53-1.56 sys/dev/pci/if_bgevar.h 1.2-1.5 sys/dev/pci/pcidevs 1.1017, 1.1021 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/miidevs 1.89, 1.93 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/dev/mii/brgphy.c 1.49, 1.52 - Fix a bug of BGE_RDMA_MODE_FIFO_LONG_BURST definition. - Fix the bug that both BGE_PCIE and BGE_PCIX are set on PCI-Express devices. - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags. - (brgphy(4)) Add support for BCM5461, BCM5464, BCM5784 and BCM5761. - (brgphy(4)) Enable brgphy_jumbo_settings(). - (brgphy(4)) Enable the Ethernet@Wirespeed function. - Add support for device properties to override the mac address and set the BGE_NO_EEPROM flag if we have an onboard device w/o SEEPROM. - Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - Fixes the bug that error bits arent cleard because these bit are W2C (in other words W1C). - Many other bugfixes from FreeBSD and OpenBSD. - remove an unused structure. - ANSI C, KNF and tabify - fix typo in comment.
1.48.4.1	04-Aug-2009	snj	Apply patch (requested by msaitoh in ticket #866): For bge: - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. For brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2 and BCM5722.
1.48.2.1	28-Apr-2009	skrll	Sync with HEAD.
1.56.22.4	03-Dec-2017	jdolecek	update from HEAD
1.56.22.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.56.22.2	23-Jun-2013	tls	resync from head
1.56.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.56.18.4	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/mii/brgphy.c 1.69, 1.71-1.76 via patch sys/dev/mii/brgphyreg.h 1.6-1.7, 1.9 sys/dev/pci/if_bge.c 1.259,1.267-1.276 via patch sys/dev/pci/if_bgereg.h 1.87-1.88 sys/dev/pci/if_bnx.c 1.48, 1.49, 1.52-1.57 via patch sys/dev/pci/if_bnxreg.h 1.15-1.18 sys/dev/pci/if_bnxvar.h 1.4-1.6 Changes for bge(4), bnx(4) and brgphy(4): - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719 into pcidevs. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Fix typo in comment. - KNF - brgphy(4): - Add BCM5708S support. - Add 2500SX support. - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. - Fix media detect on some Fiber chips. - Add BCM5706. - bge(4): - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET. - Add BGEPHYF_NO_WIRESPEED flag for fiber devices. - Print APE firmware correctly. - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packets about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... Tested with HP MicroServer Gen8. - Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436). - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver. - bnx(4): - Enclose #ifdef BNX_DEBUG correctly. - Add 2500SX support. - Print some information (ASCI revision, PCI status, etc). From FreeBSD.
1.56.18.3	17-Nov-2013	bouyer	Apply patch, requested by msaitoh in ticket #983: sys/dev/mii/brgphy.c 1.68 via patch sys/dev/pci/if_bge.c 1.260-1.262 via patch sys/dev/pci/if_bgereg.h 1.85-1.86 via patch sys/dev/pci/if_bgevar.h 1.17 via patch Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. Same as Linux tg3 and FreeBSD (part of FreeBSD r204978). Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670). This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion().
1.56.18.2	07-Sep-2013	bouyer	sys/dev/pci/if_bge.c 1.203-1.237, 1.239-1.241, 1.243-1.258 via patch sys/dev/pci/if_bgereg.h 1.58-1.74, 1.76-1.83 via patch sys/dev/pci/if_bgevar.h 1.10-1.16 via patch sys/dev/pci/pcidevs 1.1149 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/brgphy.c 1.61-1.63, 1.65, 1.67 via patch sys/dev/mii/miivar.h 1.61 via patch sys/dev/mii/miidevs 1.112-1.113 via patch sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/bge.4 1.13-1.15 via patch Add some device support, fix a lot of bugs and add some enahcements. - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F, BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - brgphy(4): Add BCM5756, BCM5717C, BCM5719C, BCM5720C and BCM57780. - Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Sync with BSD. No functional change: - make bge_chipid() and use it. - use switch() instead of a lot of if()s. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay(). - Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD. - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section7". 57781's document has the same note. - 5718 and 57785 document say we should wait 100us. - Change the order of register settings a little to match both Broadcom's document and Linux tg3. - Add detach function. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix a bug that the NVRAM lock timeout occured on sparc64 onboard bge. - Check BGE_SGDIG_STS when the chip is NOT a 5717 A0, like freebsd (OpenBSD rev. 1.327). - Fix the setting of Tx Random Backoff Register. - Add BGE_PCIMISCCTL_PCISTATE_RW in BGE_INIT. It's required to set the PCISTATE register correctly. - Update comments from 57XX-PG105-R.pdf and 5718-PG106-R.pdf. - Add missing bus_dmamap_sync() in bge_intr(). The status word is in DMAed area. Same as other BSDs. - Check the hardware config words and print them. This change only read them and print the values. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver. - Set DMA watermark depend on the PCI max payload size. - Add BGE_JUMBO_CAPABLE flag to some chips. With this commit, 5714, 5780, 5717, 5718, 5719 (exclude rev. A0), 5720, 57765 and 57766 are added to support jumbo frame. - Fix the setting of sc->bge_flags for 5717 and newer devices. - Fix a link detect bug on non-autopoll systems. Same as OpenBSD (rev.1.329 and 1.336) and FreeBSD (r213710). - 57765 series is not based on 5717 series. 5717 series is based on 57765 series. - Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document. - Change sysctl related functions for consistency. - Style change. - Use macro. Remove duplicated macro. Remove unused variable. - Fix comments. Add comments. - Remove extra semicolon. Remove unused code. [msaitoh, ticket #939]
1.56.18.1	26-Aug-2013	msaitoh	Pull up following revision(s) (requested by tsutsui in ticket #652): sys/dev/mii/brgphy.c 1.60 sys/dev/pci/if_bge.c 1.202 sys/dev/pci/if_bgereg.h 1.57 Add support for BCM57762 and BCM57765, found in Apple's Thunderbolt to Gigabit Ethernet adapter. PR kern/46961.
1.56.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.56.12.1	30-Oct-2012	yamt	sync with head
1.81.2.2	18-May-2014	rmind	sync with head
1.81.2.1	28-Aug-2013	rmind	sync with head
1.86.2.1	10-Aug-2014	tls	Rebase.
1.88.4.2	28-Aug-2017	skrll	Sync with HEAD
1.88.4.1	06-Jun-2015	skrll	Sync with HEAD
1.88.2.1	09-May-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #740): sys/dev/pci/if_bge.c: revision 1.285 sys/dev/pci/if_bgereg.h: revision 1.90 Use another firmware command in bge_asf_driver_up(). Same as Lunux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657.
1.91.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.91.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.93.12.1	10-Jun-2019	christos	Sync with HEAD
1.93.10.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.93.4.2	07-Mar-2019	martin	Pull up the following, requested by msaitoh in ticket #1209: sys/dev/pci/if_bge.c 1.323-1.327 sys/dev/pci/if_bgereg.h 1.95 sys/dev/mii/brgphy.c 1.80 share/man/man4/bge.4 1.16 - Add BCM5717's another device ID support. - Add BCM5702FE, another BCM5704S and SysKonnect SK-9Mxx support. - Add support for BCM5762 ASIC devices. - Add BCM5762, BCM5725, BCM5727, BCM57764, BCM57767 and BCM57787. - Simplify PCI device table. - Add /* FALLTHROUGH */
1.93.4.1	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1130): sys/dev/pci/if_bge.c: revision 1.320 sys/dev/pci/if_bgevar.h: revision 1.24 sys/dev/pci/if_bge.c: revision 1.317 sys/dev/pci/if_bgereg.h: revision 1.94 More TSO4 fixes, from the freebsd driver: - the chip doens't want the lenght of options, but the complete lenght of ip headers (ip + tcp + options). Fix this for the BGE_IS_5717_PLUS() and BGE_IS_5705_PLUS() cases; FreeBSD doens't cover the last case so leave it as is for now. This fixes checksum failures for heavy transfers. - It looks like the transmit engine hangs if the TCP segment crosses a 4GB boundary. FreeBSD fixes it by mapping everything below 4GB; instead try detect when this happens and do the bounce only when needed. With these fixes I could transfers 3GB images over ftp at gigabit speed (112MB/s with wget) without problems. Tested on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0 - Don't destroy the dma maps if we're not disabling the adapter, avoids a KASSERT() when bus_dmamap_destroy() is called from interrupt context via bge_watchdog() Set IFF_OACTIVE only when bge_encap() fails on adapter ressource shortage. Otherwise we may set IFF_OACTIVE while no transmit is in progress, and nothing will clear it. If bus_dmamap_load_mbuf() fails with EFBIG, m_defrag() the chain and retry. Refine the check for the 4GB boundary workaround (a fragment should also not cross the boundary), and do it only for TSO. If bge_encap() fails and didn't set IFF_OACTIVE, drop the packet. Bring in more hardware bug workarounds from freebsd. With these it seems that a BCM5720 A0 can survive a few hours of internet load with TSO4 enabled.
1.42	28-Aug-2024	skrll	Apply changes from PR/58584 after testing (and fixing). Tested on an Apple M1.
1.41	21-Feb-2023	andvar	branches: 1.41.6; s/jumo/jumbo/ in comments.
1.40	04-Sep-2022	skrll	branches: 1.40.4; bge(4): fix the MP improvements and improve some more. - Have two locks sc_core_lock at IPL_NONE and sc_intr_lock at IPL_NET and use appropriately. - Use stopping flags instead of bge_if_flags so that bge_if_flags only needs to be protected by the sc_core_lock - Use ifmedia_init_with_lock and provide the sc_intr_lock. mii operatiions are done from the interrupt handler. - Fixup locking in bge_detach. - Rename bge_watchdog to bge_watchdog_tick to avoid confusion with the if_watchdog method. - Sprinkle some more asserts.
1.39	02-Sep-2022	skrll	Whitespace - mostly align the line continutation characters
1.38	27-Aug-2022	skrll	Convert bge_pending_rxintr_change to a bool (from int)
1.37	26-Aug-2022	skrll	G/C unused softc members
1.36	19-Aug-2022	skrll	Make bge_detaching a bool. NFC.
1.35	19-Aug-2022	skrll	softc member naming consistency with wm(4). (Taylor made me do it... not really)
1.34	14-Aug-2022	skrll	bge: Mirror the bus_dma RX buffer changes in the OpenBSD driver This change reduces the amount of work done in the interrupt handler.
1.33	14-Aug-2022	skrll	Make bge(4) MP safe This started out as a fix so that LOCKDEBUG wouldn't explode with kernel lock spinout. LOCKDEBUG is too aggressive now and really should be relaxed.
1.32	14-Aug-2022	skrll	Whitespace
1.31	07-Aug-2022	skrll	Provide and use (when appropriate) a bge_free_jumbo_mem.
1.30	25-Jul-2022	skrll	Some more BITS(3)
1.29	25-Jul-2022	skrll	G/C the unused bge_rx_mini_chain member from bge_chain_data. It was removed from the FreeBSD driver in 2003.
1.28	02-Jul-2022	skrll	Sprinkle some BITS(3). Same code before and after.
1.27	02-Jul-2022	skrll	Use BUS_ADDR_{LO,HI}32. NFCI.
1.26	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.25	13-Sep-2019	msaitoh	branches: 1.25.2; if_flags is neither int nor short. It's unsigned short.
1.24	27-Nov-2018	bouyer	More TSO4 fixes, from the freebsd driver: - the chip doens't want the lenght of options, but the complete lenght of ip headers (ip + tcp + options). Fix this for the BGE_IS_5717_PLUS() and BGE_IS_5705_PLUS() cases; FreeBSD doens't cover the last case so leave it as is for now. This fixes checksum failures for heavy transfers. - It looks like the transmit engine hangs if the TCP segment crosses a 4GB boundary. FreeBSD fixes it by mapping everything below 4GB; instead try detect when this happens and do the bounce only when needed. With these fixes I could transfers 3GB images over ftp at gigabit speed (112MB/s with wget) without problems. Tested on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0
1.23	12-Apr-2017	msaitoh	branches: 1.23.4; 1.23.10; 1.23.12; Add counters for spurious interrupts.
1.22	12-Apr-2017	msaitoh	Detach sysctl and event counters in bge_release_resources().
1.21	18-Nov-2015	msaitoh	branches: 1.21.2; 1.21.4; Delete #ifdef __HAVE_PCI_MSI_MSIX and use pci_intr_allo()/pci_intr_release() on any systems. The third argument of pci_intr_alloc() is left because we will use the code to support MSI-X in future.
1.20	14-Jun-2015	martin	Make sure not to re-arm the statistics callout when we are about to detach.
1.19	17-May-2015	msaitoh	- Add MSI support. - Use tagged status function for 5717 and newer devices. All controllers except BCM5700 support tagged status but we use tagged status only for MSI case on BCM5717. Otherwise MSI on BCM5717 does not work. Same as other *BSDs.
1.18	14-Apr-2015	riastradh	Include <sys/rndsource.h> where it is actually used. I had removed <sys/rnd.h> from files that didn't mention anything of the rnd(9) API. But they included other files which assumed <sys/rnd.h> had already been included.
1.17	31-Oct-2013	msaitoh	branches: 1.17.6; Cleanup flag setting. No functional change. - Split flags into bge_flags and bge_phy_flags. - Rename flags. PHY related flags are prefixed with BGEPHYF_. Other flags are prefixed with BGEF_.
1.16	31-May-2013	msaitoh	branches: 1.16.2; - Revert rev. 1.250. I really misunderstood. Add some comments. - Set DMA watermark depend on the PCI max payload size.
1.15	24-May-2013	msaitoh	bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86.
1.14	24-Mar-2013	msaitoh	- Add detach function. - Remove old unused bge_vpd_*.
1.13	21-Mar-2013	msaitoh	- Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Add/sync some of the comments - Use macro. - Remove unused code.
1.12	17-Mar-2013	msaitoh	- ADD support for BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Use macros. - Remove unused code. - Sync with the latest {Free,Open}BSD's bge.
1.11	07-Mar-2013	msaitoh	Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before.
1.10	27-Feb-2013	msaitoh	Fix macro definitions. No binary change.
1.9	02-Feb-2012	tls	branches: 1.9.2; 1.9.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.8	19-Nov-2011	tls	branches: 1.8.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.7	18-May-2011	dyoung	branches: 1.7.4; #include <sys/bus.h>, not <machine/bus.h>.
1.6	09-Jan-2011	jruoho	branches: 1.6.2; Don't use SYSCTL_SETUP in a driver.
1.5	03-Feb-2010	msaitoh	branches: 1.5.4; 1.5.6; s/u_int/uint/
1.4	03-Feb-2010	msaitoh	- Make the initialize sequence close to the document. - Fix ASF heartbeat sending bug (FreeBSD rev. 1.271 (r202821)) - Use new handshake command for BCM5750 or new controllers (FreeBSD rev. 1.272 (r202822))
1.3	28-Jan-2010	msaitoh	- Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - remove an unused structure. - KNF
1.2	24-Jan-2010	msaitoh	Many changes from FreeBSD and OpenBSD: - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Many other bugfixes. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags.
1.1	23-Apr-2009	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; use proplib
1.1.10.3	21-Aug-2009	bouyer	The wrong version of this file was added to the branch when processing ticket #1342: use the file provided via patch in #1342 instead of version from HEAD.
1.1.10.2	18-Aug-2009	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1342): sys/dev/pci/if_bge.c: revisions 1.135, 1.146, 1.151-1.152, 1.155-1.167 via patch sys/dev/pci/if_bgereg.h: revisions 1.43, 1.48-1.52 sys/dev/pci/if_bgevar.h: revision 1.1 sys/dev/mii/brgphy.c: revisions 1.36, 1.40, 1.43-1.44 and 1.47-1.48 via patch sys/dev/mii/brgphyreg.h: revisions 1.3-1.4 sys/dev/mii/miidevs: revisions 1.75, 1.77, 1.85 sys/dev/mii/miidevs.h: regen sys/dev/mii/miidevs_data.h: regen sys/dev/mii/miivar.h: revision 1.54 For bge: - Really set Tx threshold. - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the "link changed" bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. - Add BCM5786 support. for brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2, BCM5708C, BCM5722, and BCM5906.
1.1.10.1	23-Apr-2009	bouyer	file if_bgevar.h was added on branch netbsd-4 on 2009-08-18 10:10:19 +0000
1.1.8.4	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.1.8.3	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1359): sys/dev/pci/if_bge.c 1.170-1.181 sys/dev/pci/if_bgereg.h 1.53-1.56 sys/dev/pci/if_bgevar.h 1.2-1.5 sys/dev/pci/pcidevs 1.1017, 1.1021 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/miidevs 1.89, 1.93 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/dev/mii/brgphy.c 1.49, 1.52 - Fix a bug of BGE_RDMA_MODE_FIFO_LONG_BURST definition. - Fix the bug that both BGE_PCIE and BGE_PCIX are set on PCI-Express devices. - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags. - (brgphy(4)) Add support for BCM5461, BCM5464, BCM5784 and BCM5761. - (brgphy(4)) Enable brgphy_jumbo_settings(). - (brgphy(4)) Enable the Ethernet@Wirespeed function. - Add support for device properties to override the mac address and set the BGE_NO_EEPROM flag if we have an onboard device w/o SEEPROM. - Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - Fixes the bug that error bits arent cleard because these bit are W2C (in other words W1C). - Many other bugfixes from FreeBSD and OpenBSD. - remove an unused structure. - ANSI C, KNF and tabify - fix typo in comment.
1.1.8.2	04-Aug-2009	snj	Apply patch (requested by msaitoh in ticket #866): For bge: - Fix the case of setting a flag in BGE_PCI_DMA_RW_CTL on conventional PCI. - Fix some bugs that the driver unable to detect link loss (OpenBSD 1.135, FreeBSD 1.102) - bge_tick(): fix some extra input error (OpenBSD 1.221 FreeBSD 1.212) - Acknowledge link change interrupts by setting the 'link changed' bit in the status register (used to have no effect). FreeBSD 1.44 (OpenBSD 1.21) - Add some fiber support (OpenBSD 1.220) - Set BGE_MISCCFG_KEEP_GPHY_POWER as {Open,Free}BSD and linux. - Add special case code to fix a problem with the BCM5704 in TBI mode (FreeBSD rev. 1.71, OpenBSD rev. 1.35) - Force an interrupt in bge_ifmedia_upd (FreeBSD rev. 1.205, OpenBSD rev. 1.242) - Add missing workaround for 5705 (FreeBSD rev. 1.44, OpenBSD rev. 1.21) - Don't print error message only when the register is BGE_SRS_MODE. It occurs on some environment (and once after boot?). Other *BSDs and Linux don't print the error message for all registers. - Fix comments. - Remove obsolete codes. For brgphy: - Add some quirk codes. - Add support for BCM5462, BCM54K2 and BCM5722.
1.1.8.1	23-Apr-2009	snj	file if_bgevar.h was added on branch netbsd-5 on 2009-08-04 19:46:20 +0000
1.1.6.2	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.6.1	23-Apr-2009	jym	file if_bgevar.h was added on branch jym-xensuspend on 2009-05-13 17:20:24 +0000
1.1.4.3	11-Mar-2010	yamt	sync with head
1.1.4.2	04-May-2009	yamt	sync with head.
1.1.4.1	23-Apr-2009	yamt	file if_bgevar.h was added on branch yamt-nfs-mp on 2009-05-04 08:12:56 +0000
1.1.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.1.2.1	23-Apr-2009	skrll	file if_bgevar.h was added on branch nick-hppapmap on 2009-04-28 07:35:56 +0000
1.5.6.2	21-Apr-2010	matt	sync to netbsd-5
1.5.6.1	03-Feb-2010	matt	file if_bgevar.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:41 +0000
1.5.4.2	31-May-2011	rmind	sync with head
1.5.4.1	05-Mar-2011	rmind	sync with head
1.6.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.7.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.4.1	17-Apr-2012	yamt	sync with head
1.8.2.1	18-Feb-2012	mrg	merge to -current.
1.9.6.3	03-Dec-2017	jdolecek	update from HEAD
1.9.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.6.1	23-Jun-2013	tls	resync from head
1.9.2.2	17-Nov-2013	bouyer	Apply patch, requested by msaitoh in ticket #983: sys/dev/mii/brgphy.c 1.68 via patch sys/dev/pci/if_bge.c 1.260-1.262 via patch sys/dev/pci/if_bgereg.h 1.85-1.86 via patch sys/dev/pci/if_bgevar.h 1.17 via patch Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. For PCI-X BCM5704, set maximum outstanding split transactions to 0. Same as Linux tg3 and FreeBSD (part of FreeBSD r204978). Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD (part of r199670). This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion().
1.9.2.1	07-Sep-2013	bouyer	sys/dev/pci/if_bge.c 1.203-1.237, 1.239-1.241, 1.243-1.258 via patch sys/dev/pci/if_bgereg.h 1.58-1.74, 1.76-1.83 via patch sys/dev/pci/if_bgevar.h 1.10-1.16 via patch sys/dev/pci/pcidevs 1.1149 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/brgphy.c 1.61-1.63, 1.65, 1.67 via patch sys/dev/mii/miivar.h 1.61 via patch sys/dev/mii/miidevs 1.112-1.113 via patch sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/bge.4 1.13-1.15 via patch Add some device support, fix a lot of bugs and add some enahcements. - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F, BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - brgphy(4): Add BCM5756, BCM5717C, BCM5719C, BCM5720C and BCM57780. - Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Sync with BSD. No functional change: - make bge_chipid() and use it. - use switch() instead of a lot of if()s. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay(). - Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD. - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section7". 57781's document has the same note. - 5718 and 57785 document say we should wait 100us. - Change the order of register settings a little to match both Broadcom's document and Linux tg3. - Add detach function. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix a bug that the NVRAM lock timeout occured on sparc64 onboard bge. - Check BGE_SGDIG_STS when the chip is NOT a 5717 A0, like freebsd (OpenBSD rev. 1.327). - Fix the setting of Tx Random Backoff Register. - Add BGE_PCIMISCCTL_PCISTATE_RW in BGE_INIT. It's required to set the PCISTATE register correctly. - Update comments from 57XX-PG105-R.pdf and 5718-PG106-R.pdf. - Add missing bus_dmamap_sync() in bge_intr(). The status word is in DMAed area. Same as other BSDs. - Check the hardware config words and print them. This change only read them and print the values. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver. - Set DMA watermark depend on the PCI max payload size. - Add BGE_JUMBO_CAPABLE flag to some chips. With this commit, 5714, 5780, 5717, 5718, 5719 (exclude rev. A0), 5720, 57765 and 57766 are added to support jumbo frame. - Fix the setting of sc->bge_flags for 5717 and newer devices. - Fix a link detect bug on non-autopoll systems. Same as OpenBSD (rev.1.329 and 1.336) and FreeBSD (r213710). - 57765 series is not based on 5717 series. 5717 series is based on 57765 series. - Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document. - Change sysctl related functions for consistency. - Style change. - Use macro. Remove duplicated macro. Remove unused variable. - Fix comments. Add comments. - Remove extra semicolon. Remove unused code. [msaitoh, ticket #939]
1.16.2.1	18-May-2014	rmind	sync with head
1.17.6.4	28-Aug-2017	skrll	Sync with HEAD
1.17.6.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.17.6.2	22-Sep-2015	skrll	Sync with HEAD
1.17.6.1	06-Jun-2015	skrll	Sync with HEAD
1.21.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.21.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.23.12.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.23.12.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.23.12.1	10-Jun-2019	christos	Sync with HEAD
1.23.10.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.23.4.1	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1130): sys/dev/pci/if_bge.c: revision 1.320 sys/dev/pci/if_bgevar.h: revision 1.24 sys/dev/pci/if_bge.c: revision 1.317 sys/dev/pci/if_bgereg.h: revision 1.94 More TSO4 fixes, from the freebsd driver: - the chip doens't want the lenght of options, but the complete lenght of ip headers (ip + tcp + options). Fix this for the BGE_IS_5717_PLUS() and BGE_IS_5705_PLUS() cases; FreeBSD doens't cover the last case so leave it as is for now. This fixes checksum failures for heavy transfers. - It looks like the transmit engine hangs if the TCP segment crosses a 4GB boundary. FreeBSD fixes it by mapping everything below 4GB; instead try detect when this happens and do the bounce only when needed. With these fixes I could transfers 3GB images over ftp at gigabit speed (112MB/s with wget) without problems. Tested on a bge0 at pci4 dev 0 function 0: Broadcom BCM5720 Gigabit Ethernet bge0: APE firmware NCSI 1.4.22.0 bge0: interrupting at msi1 vec 0 bge0: HW config 002b1194, 00006014, 0002aa38, 00000000 0000000c bge0: ASIC BCM5720 A0 (0x5720000), Ethernet address d0:94:66:8b:9c:18 bge0: setting short Tx thresholds brgphy0 at bge0 phy 1: BCM5720C 1000BASE-T media interface, rev. 0 - Don't destroy the dma maps if we're not disabling the adapter, avoids a KASSERT() when bus_dmamap_destroy() is called from interrupt context via bge_watchdog() Set IFF_OACTIVE only when bge_encap() fails on adapter ressource shortage. Otherwise we may set IFF_OACTIVE while no transmit is in progress, and nothing will clear it. If bus_dmamap_load_mbuf() fails with EFBIG, m_defrag() the chain and retry. Refine the check for the 4GB boundary workaround (a fragment should also not cross the boundary), and do it only for TSO. If bge_encap() fails and didn't set IFF_OACTIVE, drop the packet. Bring in more hardware bug workarounds from freebsd. With these it seems that a BCM5720 A0 can survive a few hours of internet load with TSO4 enabled.
1.25.2.1	29-Feb-2020	ad	Sync with head.
1.40.4.1	13-Oct-2024	martin	Pull up following revision(s) (requested by skrll in ticket #968): sys/dev/pci/if_bgevar.h: revision 1.41 sys/dev/pci/if_bgevar.h: revision 1.42 sys/dev/pci/if_bge.c: revision 1.394 s/jumo/jumbo/ in comments. Apply changes from PR/58584 after testing (and fixing). Tested on an Apple M1.
1.41.6.1	02-Aug-2025	perseant	Sync with HEAD
1.116	10-Nov-2024	mlelstv	Add MBUFTRACE
1.115	29-Jun-2024	riastradh	branches: 1.115.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.114	23-Apr-2024	hans	Sync the statistics DMA memory correctly before it is accessed.
1.113	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.112	01-Aug-2023	andvar	fix various typos in comments.
1.111	04-Aug-2022	andvar	s/bufferred/buffered/ in memory description.L:
1.110	05-Dec-2021	msaitoh	s/exisit/exist/ in comment.
1.109	13-Aug-2021	andvar	s/struture/structure/ s/structre/structure/
1.108	14-Feb-2021	jakllsch	if_bnx.c: fix misleading indentation suggested by rillig@
1.107	13-Feb-2021	jakllsch	Curb aprint_*() abuse in bnx(4)
1.106	13-Feb-2021	jakllsch	Revert bnx(4) to INTx interrupts. Should fix PR kern/55974. This driver does not yet do the special MSI and MSI-X setup that the chip apparently requires.
1.105	17-Jul-2020	jdolecek	branches: 1.105.2; if bnx_tx_encap() fails because mbuf is too fragmented or too long, drop the mbuf instead of wedging the TX queue forever; found by code inspection this is quite unlikely scenario since it requires mbuf chain consisting of more than 8 frags and m_defrag() failing, so probably unrelated to PR kern/47229
1.104	17-Jul-2020	jdolecek	make sure bnx_alloc_pkts() always clears BNX_ALLOC_PKTS_FLAG and IFF_OACTIVE, even on dmamap allocation failure; remove unnecessary IFF_UP condition
1.103	17-Jul-2020	jdolecek	remove mistakely committed in-progress code from previous commit
1.102	17-Jul-2020	jdolecek	re-enable MSI/MSI-X, the TX timeouts were caused by the IFF_OACTIVE handling, which was fixed in previous revision
1.101	16-Jul-2020	jdolecek	need to clear IFF_OACTIVE after allocating more pkts in order to actually start processing the queue in bnx_start()
1.100	16-Jul-2020	jdolecek	disable MSI/MSI-X for time being, it causes interrupt timeouts and needs to be investigated before enabling this again
1.99	14-Jul-2020	jdolecek	make bnx_wk (used to trigger bnx_alloc_pkts()) part of softc instead of using a static variable, so it's independant for each adapter
1.98	14-Jul-2020	jdolecek	remove call to bnx_alloc_pkts() from bnx_init_tx_chain() - it can be called from softint context via bnx_watchdog(), and bnx_alloc_pkts() dmamap allocation can't run in (soft)interrupt context
1.97	14-Jul-2020	jdolecek	only destroy the TX pkgs and dmamaps in bnx_dma_free() called on detach, don't do it anymore in bnx_free_tx_chain(), called from bnx_init() avoid occassional panic triggered by the free code having to allocate memory from interrupt context when bnx_init() is called from bnx_watchdog()
1.96	12-Jul-2020	jdolecek	enable MSI/MSI-X if supported by adapter tested MSI-X with Broadcom NetXtreme II BCM5709 1000Base-T
1.95	18-May-2020	msaitoh	Fix typo in comment.
1.94	28-Feb-2020	msaitoh	Replace bnx_ifmedia_upd() in bnx_init() with mii_ifmedia_change() for future locking changes.
1.93	28-Feb-2020	msaitoh	Replace bnx_ifmedia_upd() in bnx_mgmt_init() with mii_ifmedia_change() for future locking changes. Suggested by thorpej@. XXX Need more work for bnx_ifmedia_upd() itself.
1.92	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.91	04-Feb-2020	thorpej	Use ifmedia_fini().
1.90	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.89	27-Dec-2019	msaitoh	branches: 1.89.2; s/recive/receive/ in comment.
1.88	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.87	26-Jul-2019	msaitoh	Calculate the multicast hash table correctly to avoid undefined behavior. Found by KUBSan.
1.86	09-Jul-2019	msaitoh	Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.85	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.84	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.83	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.82	24-Apr-2019	msaitoh	This driver uses MII(4) and have hook SIOCGIFMEDIA which just pass to ifmedia_ioctl(), the hook is not required because ether_ioctl has it (if_ethersubr.c rev. 1.160). This driver might require some additional fixes for SIOCSIFMTU and other ioctl()s.
1.81	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.80	08-Apr-2019	msaitoh	Make it compilable with BNX_DEBUG.
1.79	05-Apr-2019	msaitoh	BCM5709 is PCIe based. To wait for pending PCI transactions to complete, wait until PCIe's transaction pending bit is cleared.
1.78	05-Apr-2019	msaitoh	- Allow time for the management firmware to enter the running state. - Enable/Disable management frames (NC-SI) to flow to the MCP if MFW exists. - Print the boot code version, management firmware version and some flags. - Gbps -> GT/s for PCIe speed. - Use aprint_*() in bnx_print_adapter_info(). - Use braces for some macors' argument. - Remove unused macros. - KNF.
1.77	05-Apr-2019	msaitoh	Fix a bug that BNX_MISC_ENABLE_SET_BITS is not correctly set on BCM5709.
1.76	04-Apr-2019	msaitoh	0 -> NULL.
1.75	01-Apr-2019	msaitoh	s/adddress/address/
1.74	29-Mar-2019	msaitoh	- Use m_defrag() when bus_dmamap_load_mbuf() returned with EFBIG. - Style change to reduce diff against {Free,Open}BSD. - Make it compilable with BNX_DEBUG. - Use __nothing for null macro.
1.73	29-Mar-2019	msaitoh	Sync with OpenBSD: - Don't need to clear if_timer during attach. (OpenBSD rev. 1.23) - Do the minimal initialization of the firmware so that ASF always works. From ambrisko@FreeBSD (OpenBSD rev. 1.32) - Move the interrupt establishment till after everything in the softc is set up (OpenBSD rev. 1.40) - Let mii_attach() know where the PHY is located instead of scanning for it since we know where it will be anyway and remove the code from the MII bus read/write functions to force reading/writing from the predetermined location. Copied from bge(4) and this is what the upstream FreeBSD bce(4) driver has done once FreBSD gained a mii_attach().
1.72	28-Mar-2019	msaitoh	- Remove extra cast (from OpenBSD rev. 1.95) - Add support flow control (from OpenBSD rev. 1.97 and 1.100). The main difference between NetBSD and other *BSDs is that the flow control does really works. The lower 8bit of BNX_L2CTX_CTX_TYPE should be 0xff (from Linux). Note that I have NetXtremeII-PG203-R.pdf and it says the field is reserved and should be 0.
1.71	27-Mar-2019	msaitoh	- Remove extra check (because the same check is done in if_schedule_deferred_start(). - Remove extra space.
1.70	27-Mar-2019	msaitoh	Apply OpenBSD rev. 1.94: > ido not disable interrupts in the isr and then enable them again > when leaving. when you're handling an interrupt it is masked. > whacking the chip is work for no gain. > > modify the interrupt handler so it only processes the rings once > rather than looping over them until it runs out of work to do > > looping in the isr is bad for several reasons: > > firstly, the chip does interrupt mitigation so you have a > decent/predictable amount of work to do in the isr. your first loop > will do that chunk of work (ie, it pulls off 50ish packets), and > then the successive looping aggressively pull one or two packets > off the rx ring. these extra loops work against the benefit that > interrupt mitigation provides. > > bus space reads are slow. we should avoid doing them where possible > (but we should always do them when necessary). > > doing the loop 5 times per isr works against the mclgeti semantics. > it knows a nic is busy and therefore needs more rx descriptors by > watching to see when the nic uses all of its descriptors between > interrupts. if we're aggressively pulling packets off by looping > in the isr then we're skewing this check. > > ok deraadt@
1.69	27-Mar-2019	msaitoh	Sync with other *BSDs: From OpenBSD's rev. 1.93: > to quote from the gospel of bus_dma.9: > > Synchronization operations are expressed from the perspective of the host > RAM, e.g., a device -> memory operation is a READ and a memory -> device > operation is a WRITE. > > the status block that the isr reads is written to by the device. > the chip writes to memory, it is therefore a READ. > > this also adds the preread sync when the map is set up and the postread > sync when the map is torn down for better symmetry. there are probably > more issues like this in the code, but this is a start. > > discovered while discussing another diff with claudio@
1.68	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.67	08-Jan-2019	msaitoh	Fix a bug that bnx_miibus_read_reg() returns wrong value when BNX_PHY_INT_MODE_AUTO_POLLING_FLAG is set. This bug doesn't cause a real problem because BNX_PHY_INT_MODE_AUTO_POLLING_FLAG is not set.
1.66	08-Jan-2019	msaitoh	KNF. No functional change.
1.65	26-Jun-2018	msaitoh	branches: 1.65.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.64	08-May-2018	msaitoh	- Fix a bug that bnx(4) panics on shutdown. Stop callout before restroy. Reported by Andreas Gustafsson in PR#53265. - Make sure not to re-arm the callout when we are about to detach. Same as if_bge.c rev. 1.292. - Use pci_intr_establish_xname().
1.63	08-Feb-2018	dholland	branches: 1.63.2; Typos.
1.62	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.61	15-Dec-2016	ozaki-r	branches: 1.61.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.60	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.59	10-Jun-2016	ozaki-r	branches: 1.59.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.58	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.57	09-Jul-2014	msaitoh	branches: 1.57.4; Call bnx_print_adapter_info() before mii_attach() to not to split bnx's dmesg into two parts by brgphy's dmesg.
1.56	01-Jul-2014	msaitoh	Print some information (ASIC revision, PCI status, etc). From FreeBSD.
1.55	01-Jul-2014	msaitoh	No functional change. - s/u_int/uint/ - Modify comment a bit (sync with OpenBSD)
1.54	24-Jun-2014	msaitoh	No functional change: - Tabify. - Remove trailing white spaces. - Fix typo in comment.
1.53	17-Jun-2014	msaitoh	- Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). The auto negotiation may have some bugs. - Add 2500SX support (not tested). - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD.
1.52	12-Jun-2014	msaitoh	No functional change: - Remove BRGPHY_SERDES_ANAR_* and BRGPHY_SERDES_ANLPAR_. Those registers are the same as MII_ANAR_ and MII_ANLPAR_*. - Fix typo. - Remove trailing white spaces. - KNF.
1.51	29-Mar-2014	christos	branches: 1.51.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.50	13-Mar-2014	hannken	Add __diagused.
1.49	13-Sep-2013	martin	#ifdef variable declarations/initializations like their use
1.48	30-Mar-2013	christos	branches: 1.48.4; remove trailing whitespace
1.47	22-Jul-2012	matt	branches: 1.47.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.46	06-Mar-2012	bouyer	bnx_start() is not MP-safe, so do not create the bnx_alloc_pkts()'s workqueue MPSAFE. This could lead to bnx_start() being running at the same time on 2 CPUs, with packets being sent twice. Problem found and reported by Beverly Schwartz and Greg Troxel.
1.45	22-Sep-2011	jym	branches: 1.45.2; 1.45.6; 1.45.8; The bnx(4) handler runs at IPL_NET (works with spin mutexes), while allocations can be deferred to a workqueue(9): set its highest IPL to IPL_NET and not IPL_NONE, which is protected by an adaptive/sleeping mutex.
1.44	15-Jul-2011	jym	Use a workqueue(9) to defer allocation of TX packets. Loosely inspired from the OpenBSD bnx(4) driver. Tested by Roger Pau Monné, roger.pau at entel.upc.edu. Fixes kern/45051.
1.43	02-May-2011	jym	Fix BCM5709 PHY detection for ethernet PHYs (the SerDes case being already handled): - export bge(4) and bnx(4) CHIP ID and PHY flags to brgphy(4). Move to "unsigned int" rather than "int", and reuse the same softc members for chipid and phyflags (behavior controlled by the sc_isbge/isbnx boolean). - apply bug fix for revisions A and B, so that autonegotiation can complete (from OpenBSD). Bug reported by Rivo Nurges via private mail, patch tested and confirmed working by him (with thanks!)
1.42	26-Jan-2011	dyoung	Get out of the interrupt handler early if !IFF_RUNNING.
1.41	06-Jan-2011	jym	branches: 1.41.2; 1.41.4; bnx(4) SERDES NICs will definitely work better with the PHY initialization code... rather than without. (forgot to commit it two weeks ago -- should be part of the BCM 5709S patch)
1.40	06-Jan-2011	jym	Reflect 5709S support in bnx(4) (comment fix)
1.39	11-Dec-2010	martin	Sanitize names a bit, still could use more namespace cleanup
1.38	11-Dec-2010	pgoyette	Fix a couple of name-space changes that fell through the cracks. Hello, martin!
1.37	09-Dec-2010	jym	Remove htolexx calls, as bnx(4) NICs are capable of swapping endianess. The macros for PAGE_SIZE were fixed with the previous patch (if_bnxreg.h splitting, rev 1.11). From OpenBSD. Their commit log reads: bnx(4) is a bit special. The chip itself is capable of swapping endianess so there is no need for htoleXX calls. The only thing needed is the correct layout of the DMA-ed structures. Additionally it uses PAGE_SIZE but assumed that it is always 4k. Fix the macros that failed to respect that so that it works on 8k PAGE_SIZE systems. This makes bnx(4) work on sparc64.
1.36	09-Dec-2010	jym	Similar to bge(4), split if_bnxreg.h in two parts: - if_bnxreg.h: general register values, flags... for bnx(4), that can be included in other drivers - if_bnxvar.h: PCI or device specific code (device state data, debug macros etc.), which should remain private to bnx(4) No comments, no objections on current-users@. See discussion: http://mail-index.netbsd.org/current-users/2010/12/01/msg014926.html
1.35	27-Nov-2010	jym	Missing break in switch case. From OpenBSD.
1.34	26-Nov-2010	dholland	Don't test flags with &&. From Henning Petersen in PR 44151.
1.33	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.32	11-Mar-2010	msaitoh	branches: 1.32.2; Add missing ifmedia_delete_instance() in bnx_detach().
1.31	19-Jan-2010	pooka	branches: 1.31.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.30	20-Nov-2009	bouyer	Remove duplicate PCI_PRODUCT_BROADCOM_BCM5709 entries. BNX_MAX_MRU is now the 1500-bytes MTU; use BNX_MAX_JUMBO_MRU instead.
1.29	18-Nov-2009	bouyer	Sync with OpenBSD: if_bnx.c,v 1.85. Major changes: Add support for the Broadcom BCM5709 and BCM5716 chips. This upgrade the B06 firmware, and add the B09 firmware required by the BCM5709. Remplace TX dmamap array with a dynamically-grown list. Tested with: bnx0 at pci4 dev 0 function 0: Broadcom NetXtreme II BCM5708 1000Base-T and bnx0 at pci0 dev 0 function 0: Broadcom NetXtreme II BCM5709 1000Base-SX
1.28	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.27	05-May-2009	cegger	Attach on BCM5709 chip. XXX Firmware fails to complete initialization on this chip.
1.26	17-Apr-2009	dyoung	Bug fix: don't release the Rx DMA maps twice.
1.25	07-Apr-2009	dyoung	Destroy sc->bnx_timeout in bnx_detach().
1.24	03-Apr-2009	dyoung	Fix SIOCSIFFLAGS. Mark as safe to detach at shutdown.
1.23	18-Mar-2009	cegger	bzero -> memset
1.22	18-Mar-2009	cegger	bcmp -> memcmp
1.21	27-Jan-2009	dyoung	branches: 1.21.2; Protect against spurious "bnx0: Double mbuf allocation failure!" panics. Extract a new function, bnx_add_buf(), from bnx_get_buf(). bnx_add_buf() just tries to add an mbuf to the Rx chain. bnx_get_buf() tries to allocate an mbuf, first, and then to add the mbuf to the chain. Use bnx_add_buf() to try to put an mbuf that we took off of the chain back onto the chain, and use bnx_get_buf() everywhere else. This change seems to be in keeping with the author's intent, but it looks like the code broke at some point. Thanks to Jarle Greipsland, David Dudley, and others for reporting this bug and for testing patches.
1.20	09-Sep-2008	mhitch	branches: 1.20.2; 1.20.4; 1.20.10; Add support for SerDes controllers; from the OpenBSD driver. Tested on a Dell Blade server by me, and an HP Blade server by Havard.
1.19	09-Sep-2008	mhitch	In bnx_release_resources(), don't de-reference a NULL pointer if the rx_mbuf_map has not been allocated yet. From Havard Eidnes.
1.18	07-Feb-2008	dyoung	branches: 1.18.6; 1.18.10; 1.18.12; 1.18.16; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.17	06-Feb-2008	joerg	Don't use misaligned PCI config access. Reported by Christoph Egger.
1.16	31-Dec-2007	dyoung	Don't buffer packets while the link is down, but drop them on the floor like most other ethernet drivers do. bnx(4) has "vanilla" MII media-handling, so use ether_mediachange and ether_mediastatus.
1.15	30-Dec-2007	dyoung	For consistency with other softc members, s/ethercom/bnx_ec/.
1.14	29-Dec-2007	dyoung	Get rid of bnx_suspend_resume(), it did redundant resets. Merge bnx_if_stop() and bnx_stop(). Handle bnx_stop(, 1) by doing a reset. Protect the bnx_stop() call in bnx_detach() with splnet(). Thanks to joerg@ for suggesting the above. In bnx_stop(), deactivate the PHYs with mii_down(). Return right away if !IFF_RUNNING. Use LIST_EMPTY(). In bnx_ifmedia_upd(), propagate all mii_mediachg() return codes to the caller except for ENXIO, which just means that the PHY is suspended. In bnx_init(), propagate bnx_ifmedia_upd() return codes to the caller. In bnx_intr(), replace a common subexpression with a variable, 'sblk'. In bnx_tick(), let the MII bus keep ticking after the link is up: check link status and exit after calling mii_tick(), not before.
1.13	26-Dec-2007	dyoung	Add device suspend/resume/detach. (Tested with drvctl -S bnx0, drvctl -Q bnx0, drvctl -d bnx0.) Separate device_t storage from softc storage and use CFATTACH_DECL_NEW().
1.12	15-Dec-2007	perry	__FUNCTION__ -> __func__
1.11	04-Dec-2007	dyoung	branches: 1.11.4; Reprogram multicast filter after SIOCADDMULTI / SIOCDELMULTI. Fixes IPv6 stateless address configuration problems reported on tech-net@. Patch tested and found to work by Martti Kuparinen.
1.10	09-Nov-2007	martti	branches: 1.10.2; Added missing newline.
1.9	29-Aug-2007	dyoung	branches: 1.9.4; 1.9.6; Constify: LLADDR() -> CLLADDR().
1.8	09-Aug-2007	bouyer	branches: 1.8.2; Pass the right value to VLAN_INPUT_TAG(), this makes vlan(4) on bnx(4) work.
1.7	09-Jul-2007	ad	branches: 1.7.2; 1.7.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.6	10-Apr-2007	bouyer	Fix build on 32bit ports; should fix GENERIC_DIAGNOSTIC build problem reported by Hisashi T Fujinaka on current-users.
1.5	09-Apr-2007	bouyer	RX buffers are malloced memory of 9216 bytes. This can require from 1 to 4 DMA memory segments, depending on how the buffer is in memory. When receiving a packet, we allocate a new one to remplace the one we've used. It can need more segments than the one it remplace, leading to corrution of the RX descriptors, and a panic in bus_dmamap_sync() (DIAGNOSTIC kernels) or possibly memory corruption. Fix: - bnx_get_buf() allocates as many buffer as possible, checking the number of free RX descriptors. Because one receive buffer is not guaranteed to be remplaced on receive, call bnx_get_buf() from bnx_tick() too. This also improve error handling from bnx_get_buf(). - use MCLGET() instead of MEXTMALLOC() if we're running with the standard ethernet MTU. This gives us more receive buffers and waste less memory.
1.4	09-Apr-2007	bouyer	Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman
1.3	04-Mar-2007	christos	branches: 1.3.2; 1.3.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.2	15-Feb-2007	bouyer	branches: 1.2.2; Fix 2 panics with DIAGNOSTIC kernels: - remove the code to bring down the PHY in bnx_stop(), it's wrong (ifm_data isn't updated) and lead to a panic in mii_phy_setmedia(), or reading past the end mii_media_table[]. - make sure the dma_map matches the mbuf in the rx structures. We would sync/unload the wrong map, leading to a DIAGNOSTIC panic, or eventually leaking memory when bounce buffers are needed.
1.1	17-Dec-2006	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.1.10.2	12-Jan-2007	ad	Sync with head.
1.1.10.1	17-Dec-2006	ad	file if_bnx.c was added on branch newlock2 on 2007-01-12 00:57:41 +0000
1.1.8.8	11-Feb-2008	yamt	sync with head.
1.1.8.7	21-Jan-2008	yamt	sync with head
1.1.8.6	07-Dec-2007	yamt	sync with head
1.1.8.5	15-Nov-2007	yamt	sync with head.
1.1.8.4	03-Sep-2007	yamt	sync with head.
1.1.8.3	26-Feb-2007	yamt	sync with head.
1.1.8.2	30-Dec-2006	yamt	sync with head.
1.1.8.1	17-Dec-2006	yamt	file if_bnx.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:44 +0000
1.1.6.6	19-Nov-2007	bouyer	Pull up following revision(s) (requested by martti in ticket #1876): sys/dev/pci/if_bnx.c: revision 1.10 Added missing newline.
1.1.6.5	17-Sep-2007	msaitoh	Pull up following revision(s) (requested by bouyer in ticket #1827): sys/dev/pci/if_bnx.c: revision 1.8 Pass the right value to VLAN_INPUT_TAG(), this makes vlan(4) on bnx(4) work.
1.1.6.4	25-Apr-2007	ghen	Pull up following revision(s) (requested by bouyer in ticket #1762): sys/dev/pci/if_bnx.c: revision 1.4 sys/dev/pci/if_bnx.c: revision 1.5 sys/dev/pci/if_bnx.c: revision 1.6 sys/dev/pci/if_bnxreg.h: revision 1.2 Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman RX buffers are malloced memory of 9216 bytes. This can require from 1 to 4 DMA memory segments, depending on how the buffer is in memory. When receiving a packet, we allocate a new one to remplace the one we've used. It can need more segments than the one it remplace, leading to corrution of the RX descriptors, and a panic in bus_dmamap_sync() (DIAGNOSTIC kernels) or possibly memory corruption. Fix: - bnx_get_buf() allocates as many buffer as possible, checking the number of free RX descriptors. Because one receive buffer is not guaranteed to be remplaced on receive, call bnx_get_buf() from bnx_tick() too. This also improve error handling from bnx_get_buf(). - use MCLGET() instead of MEXTMALLOC() if we're running with the standard ethernet MTU. This gives us more receive buffers and waste less memory. Fix build on 32bit ports; should fix GENERIC_DIAGNOSTIC build problem reported by Hisashi T Fujinaka on current-users.
1.1.6.3	19-Feb-2007	tron	Pull up following revision(s) (requested by bouyer in ticket #1667): sys/dev/pci/if_bnx.c: revision 1.2 Fix 2 panics with DIAGNOSTIC kernels: - remove the code to bring down the PHY in bnx_stop(), it's wrong (ifm_data isn't updated) and lead to a panic in mii_phy_setmedia(), or reading past the end mii_media_table[]. - make sure the dma_map matches the mbuf in the rx structures. We would sync/unload the wrong map, leading to a DIAGNOSTIC panic, or eventually leaking memory when bounce buffers are needed.
1.1.6.2	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1608): sys/arch/i386/conf/INSTALL: revision 1.299 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.20 via patch sys/arch/i386/conf/GENERIC: revision 1.803 via patch sys/dev/pci/files.pci: revision 1.275 via patch sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 via patch sys/arch/amd64/conf/GENERIC: revision 1.123 via patch sys/dev/pci/if_bnx.c: revision 1.1 via patch share/man/man4/bnx.4: revision 1.1-1.2 via patch distrib/sets/lists/man/mi: revision 1.959 via patch sys/arch/amd64/conf/INSTALL: revision 1.62 via patch sys/dev/microcode/bnx/bnxfw.h: revision 1.1-1.2 via patch sys/dev/pci/if_bnxreg.h: revision 1.1 via patch share/man/man4/Makefile: revision 1.417 via patch sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 via patch Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64. Add a NetBSD RCS tag. Nx -> Ox; remove trailing whitespace.
1.1.6.1	17-Dec-2006	ghen	file if_bnx.c was added on branch netbsd-3 on 2006-12-23 14:43:47 +0000
1.1.4.7	01-Feb-2008	riz	Pull up following revision(s) (requested by dyoung in ticket #1011): sys/dev/pci/if_bnx.c: revision 1.11 Reprogram multicast filter after SIOCADDMULTI / SIOCDELMULTI. Fixes IPv6 stateless address configuration problems reported on tech-net@. Patch tested and found to work by Martti Kuparinen.
1.1.4.6	25-Nov-2007	xtraeme	Pull up following revision(s) (requested by martti in ticket #986): sys/dev/pci/if_bnx.c: revision 1.10 Added missing newline.
1.1.4.5	24-Aug-2007	liamjfoy	branches: 1.1.4.5.2; Pull up following revision(s) (requested by bouyer in ticket #826): sys/dev/pci/if_bnx.c: revision 1.8 Pass the right value to VLAN_INPUT_TAG(), this makes vlan(4) on bnx(4) work.
1.1.4.4	20-Apr-2007	pavel	branches: 1.1.4.4.2; Pull up following revision(s) (requested by bouyer in ticket #566): sys/dev/pci/if_bnx.c: revision 1.4 sys/dev/pci/if_bnx.c: revision 1.5 sys/dev/pci/if_bnx.c: revision 1.6 sys/dev/pci/if_bnxreg.h: revision 1.2 Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman RX buffers are malloced memory of 9216 bytes. This can require from 1 to 4 DMA memory segments, depending on how the buffer is in memory. When receiving a packet, we allocate a new one to remplace the one we've used. It can need more segments than the one it remplace, leading to corrution of the RX descriptors, and a panic in bus_dmamap_sync() (DIAGNOSTIC kernels) or possibly memory corruption. Fix: - bnx_get_buf() allocates as many buffer as possible, checking the number of free RX descriptors. Because one receive buffer is not guaranteed to be remplaced on receive, call bnx_get_buf() from bnx_tick() too. This also improve error handling from bnx_get_buf(). - use MCLGET() instead of MEXTMALLOC() if we're running with the standard ethernet MTU. This gives us more receive buffers and waste less memory. Fix build on 32bit ports; should fix GENERIC_DIAGNOSTIC build problem reported by Hisashi T Fujinaka on current-users.
1.1.4.3	17-Feb-2007	tron	Pull up following revision(s) (requested by bouyer in ticket #451): sys/dev/pci/if_bnx.c: revision 1.2 Fix 2 panics with DIAGNOSTIC kernels: - remove the code to bring down the PHY in bnx_stop(), it's wrong (ifm_data isn't updated) and lead to a panic in mii_phy_setmedia(), or reading past the end mii_media_table[]. - make sure the dma_map matches the mbuf in the rx structures. We would sync/unload the wrong map, leading to a DIAGNOSTIC panic, or eventually leaking memory when bounce buffers are needed.
1.1.4.2	21-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #287): sys/arch/i386/conf/INSTALL: revision 1.299 sys/arch/i386/conf/XEN2_DOM0: revision 1.20 sys/arch/i386/conf/GENERIC: revision 1.803 sys/dev/pci/files.pci: revision 1.275 sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 sys/arch/amd64/conf/GENERIC: revision 1.123 sys/dev/pci/if_bnx.c: revision 1.1 share/man/man4/bnx.4: revision 1.1 distrib/sets/lists/man/mi: revision 1.959 sys/arch/amd64/conf/INSTALL: revision 1.62 sys/dev/microcode/bnx/bnxfw.h: revision 1.1 sys/arch/i386/conf/ALL: revision 1.76 sys/dev/pci/if_bnxreg.h: revision 1.1 share/man/man4/Makefile: revision 1.417 sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.1.4.1	17-Dec-2006	tron	file if_bnx.c was added on branch netbsd-4 on 2006-12-21 13:44:01 +0000
1.1.4.5.2.1	25-Nov-2007	xtraeme	Pull up following revision(s) (requested by martti in ticket #986): sys/dev/pci/if_bnx.c: revision 1.10 Added missing newline.
1.1.4.4.2.3	03-Jun-2008	skrll	Sync with netbsd-4.
1.1.4.4.2.2	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.1.4.4.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.1.2.2	18-Dec-2006	yamt	sync with head.
1.1.2.1	17-Dec-2006	yamt	file if_bnx.c was added on branch yamt-splraiseipl on 2006-12-18 11:42:13 +0000
1.2.2.2	15-Apr-2007	yamt	sync with head.
1.2.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.3.4.1	11-Jul-2007	mjf	Sync with head.
1.3.2.4	09-Oct-2007	ad	Sync with head.
1.3.2.3	20-Aug-2007	ad	Sync with HEAD.
1.3.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.3.2.1	10-Apr-2007	ad	Sync with head.
1.7.6.4	09-Dec-2007	jmcneill	Sync with HEAD.
1.7.6.3	11-Nov-2007	joerg	Sync with HEAD.
1.7.6.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.7.6.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.7.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.7.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.8.2.3	23-Mar-2008	matt	sync with HEAD
1.8.2.2	09-Jan-2008	matt	sync with HEAD
1.8.2.1	06-Nov-2007	matt	sync with HEAD
1.9.6.4	18-Feb-2008	mjf	Sync with HEAD.
1.9.6.3	27-Dec-2007	mjf	Sync with HEAD.
1.9.6.2	08-Dec-2007	mjf	Sync with HEAD.
1.9.6.1	19-Nov-2007	mjf	Sync with HEAD.
1.9.4.1	13-Nov-2007	bouyer	Sync with HEAD
1.10.2.2	26-Dec-2007	ad	Sync with head.
1.10.2.1	08-Dec-2007	ad	Sync with head.
1.11.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.18.16.1	19-Oct-2008	haad	Sync with HEAD.
1.18.12.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.18.10.5	11-Aug-2010	yamt	sync with head.
1.18.10.4	11-Mar-2010	yamt	sync with head
1.18.10.3	16-Sep-2009	yamt	sync with head
1.18.10.2	16-May-2009	yamt	sync with head
1.18.10.1	04-May-2009	yamt	sync with head.
1.18.6.1	28-Sep-2008	mjf	Sync with HEAD.
1.20.10.1	21-Apr-2010	matt	sync to netbsd-5
1.20.4.3	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.20.4.2	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1268): sys/dev/mii/miidevs: revision 1.90 sys/dev/pci/pcidevs: revision 1.1001 sys/dev/mii/brgphy.c: revision 1.51 sys/dev/microcode/bnx/bnxfw.h: revision 1.3 sys/dev/pci/if_bnx.c: revision 1.29 sys/dev/pci/if_bnxreg.h: revision 1.9 sys/dev/pci/if_bnx.c: revision 1.30 Add support for the Broadcom BCM5709 and BCM5716 chips. Upgrade the B06 firmware and add the B09 firmware required by the BCM5709. Replace TX dmamap array with a dynamically-grown list. Support associated BCM5709CAX and BCM5709C PHYs.
1.20.4.1	18-May-2009	bouyer	Pull up following revision(s) (requested by kefren in ticket #760): sys/dev/pci/if_bnx.c: revision 1.21 Protect against spurious "bnx0: Double mbuf allocation failure!" panics. Extract a new function, bnx_add_buf(), from bnx_get_buf(). bnx_add_buf() just tries to add an mbuf to the Rx chain. bnx_get_buf() tries to allocate an mbuf, first, and then to add the mbuf to the chain. Use bnx_add_buf() to try to put an mbuf that we took off of the chain back onto the chain, and use bnx_get_buf() everywhere else. This change seems to be in keeping with the author's intent, but it looks like the code broke at some point. Thanks to Jarle Greipsland, David Dudley, and others for reporting this bug and for testing patches.
1.20.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.20.2.1	03-Mar-2009	skrll	Sync with HEAD.
1.21.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.31.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.32.2.3	31-May-2011	rmind	sync with head
1.32.2.2	05-Mar-2011	rmind	sync with head
1.32.2.1	30-May-2010	rmind	sync with head
1.41.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.41.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.45.8.2	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/mii/brgphy.c 1.69, 1.71-1.76 via patch sys/dev/mii/brgphyreg.h 1.6-1.7, 1.9 sys/dev/pci/if_bge.c 1.259,1.267-1.276 via patch sys/dev/pci/if_bgereg.h 1.87-1.88 sys/dev/pci/if_bnx.c 1.48, 1.49, 1.52-1.57 via patch sys/dev/pci/if_bnxreg.h 1.15-1.18 sys/dev/pci/if_bnxvar.h 1.4-1.6 Changes for bge(4), bnx(4) and brgphy(4): - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719 into pcidevs. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Fix typo in comment. - KNF - brgphy(4): - Add BCM5708S support. - Add 2500SX support. - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. - Fix media detect on some Fiber chips. - Add BCM5706. - bge(4): - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET. - Add BGEPHYF_NO_WIRESPEED flag for fiber devices. - Print APE firmware correctly. - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packets about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... Tested with HP MicroServer Gen8. - Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436). - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver. - bnx(4): - Enclose #ifdef BNX_DEBUG correctly. - Add 2500SX support. - Print some information (ASCI revision, PCI status, etc). From FreeBSD.
1.45.8.1	07-Mar-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #100): sys/dev/pci/if_bnx.c: revision 1.46 bnx_start() is not MP-safe, so do not create the bnx_alloc_pkts()'s workqueue MPSAFE. This could lead to bnx_start() being running at the same time on 2 CPUs, with packets being sent twice. Problem found and reported by Beverly Schwartz and Greg Troxel.
1.45.6.1	11-Mar-2012	mrg	sync to latest -current
1.45.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.45.2.2	30-Oct-2012	yamt	sync with head
1.45.2.1	17-Apr-2012	yamt	sync with head
1.47.2.3	03-Dec-2017	jdolecek	update from HEAD
1.47.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.47.2.1	23-Jun-2013	tls	resync from head
1.48.4.1	18-May-2014	rmind	sync with head
1.51.2.1	10-Aug-2014	tls	Rebase.
1.57.4.3	05-Feb-2017	skrll	Sync with HEAD
1.57.4.2	09-Jul-2016	skrll	Sync with HEAD
1.57.4.1	19-Mar-2016	skrll	Sync with HEAD
1.59.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.61.8.3	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.61.8.2	09-May-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #814): sys/dev/pci/if_bnxvar.h: revision 1.7 sys/dev/pci/if_bnx.c: revision 1.64 - Fix a bug that bnx(4) panics on shutdown. Stop callout before restroy. Reported by Andreas Gustafsson in PR#53265. - Make sure not to re-arm the callout when we are about to detach. Same as if_bge.c rev. 1.292. - Use pci_intr_establish_xname().
1.61.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.63.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.63.2.3	18-Jan-2019	pgoyette	Synch with HEAD
1.63.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.63.2.1	21-May-2018	pgoyette	Sync with HEAD
1.65.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.65.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.65.2.1	10-Jun-2019	christos	Sync with HEAD
1.89.2.1	29-Feb-2020	ad	Sync with head.
1.105.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.115.2.1	02-Aug-2025	perseant	Sync with HEAD
1.30	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.29	05-Feb-2024	andvar	triple "r" typos, mainly s/interrrupt/interrupt/ in comments and one definition.
1.28	12-Dec-2021	andvar	fix various typos, mainly in comments.
1.27	13-Feb-2021	jakllsch	Curb aprint_*() abuse in bnx(4)
1.26	24-May-2019	msaitoh	branches: 1.26.10; Whilespace fix. No functional change.
1.25	05-Apr-2019	msaitoh	Add BNX_PCICFG_DEVICE_CONTROL.
1.24	05-Apr-2019	msaitoh	- Allow time for the management firmware to enter the running state. - Enable/Disable management frames (NC-SI) to flow to the MCP if MFW exists. - Print the boot code version, management firmware version and some flags. - Gbps -> GT/s for PCIe speed. - Use aprint_*() in bnx_print_adapter_info(). - Use braces for some macors' argument. - Remove unused macros. - KNF.
1.23	04-Apr-2019	msaitoh	No functional change. This change reduces diff against DragonFly: - Fix BNX_PCI_GRC_WINDOW_ADDR_VALUE's name. Note that this macro is not used. - Whitespace change. - Sort some lines.
1.22	29-Mar-2019	msaitoh	- Use m_defrag() when bus_dmamap_load_mbuf() returned with EFBIG. - Style change to reduce diff against {Free,Open}BSD. - Make it compilable with BNX_DEBUG. - Use __nothing for null macro.
1.21	28-Mar-2019	msaitoh	No functional change: - Fix DB_AND_RANDOMFALSE(). It has no real bug because the macro is not used from anywhere. - Wse const. - Whitespace change.
1.20	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.19	03-Jun-2018	maxv	branches: 1.19.2; Constify the microcode variables used by BNX. This moves 38 pages of kernel memory from .data to .rodata.
1.18	01-Jul-2014	msaitoh	branches: 1.18.26; Print some information (ASIC revision, PCI status, etc). From FreeBSD.
1.17	01-Jul-2014	msaitoh	No functional change. - s/u_int/uint/ - Modify comment a bit (sync with OpenBSD)
1.16	24-Jun-2014	msaitoh	No functional change: - Tabify. - Remove trailing white spaces. - Fix typo in comment.
1.15	17-Jun-2014	msaitoh	- Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). The auto negotiation may have some bugs. - Add 2500SX support (not tested). - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD.
1.14	02-May-2011	jym	branches: 1.14.10; 1.14.14; 1.14.28; Fix BCM5709 PHY detection for ethernet PHYs (the SerDes case being already handled): - export bge(4) and bnx(4) CHIP ID and PHY flags to brgphy(4). Move to "unsigned int" rather than "int", and reuse the same softc members for chipid and phyflags (behavior controlled by the sc_isbge/isbnx boolean). - apply bug fix for revisions A and B, so that autonegotiation can complete (from OpenBSD). Bug reported by Rivo Nurges via private mail, patch tested and confirmed working by him (with thanks!)
1.13	11-Dec-2010	martin	branches: 1.13.2; Sanitize names a bit, still could use more namespace cleanup
1.12	11-Dec-2010	martin	Namespace police: don't use very broad names like CTX_MASK (which accidently collide with sparc* specific MD macros), mechanically prefix with BNX_.
1.11	09-Dec-2010	jym	Similar to bge(4), split if_bnxreg.h in two parts: - if_bnxreg.h: general register values, flags... for bnx(4), that can be included in other drivers - if_bnxvar.h: PCI or device specific code (device state data, debug macros etc.), which should remain private to bnx(4) No comments, no objections on current-users@. See discussion: http://mail-index.netbsd.org/current-users/2010/12/01/msg014926.html
1.10	19-Jan-2010	pooka	branches: 1.10.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.9	18-Nov-2009	bouyer	Sync with OpenBSD: if_bnx.c,v 1.85. Major changes: Add support for the Broadcom BCM5709 and BCM5716 chips. This upgrade the B06 firmware, and add the B09 firmware required by the BCM5709. Remplace TX dmamap array with a dynamically-grown list. Tested with: bnx0 at pci4 dev 0 function 0: Broadcom NetXtreme II BCM5708 1000Base-T and bnx0 at pci0 dev 0 function 0: Broadcom NetXtreme II BCM5709 1000Base-SX
1.8	14-Apr-2009	dyoung	Delete two unused data members, the powerhook and the shutdownhook handles.
1.7	24-Jun-2008	gmcgarry	branches: 1.7.4; 1.7.6; 1.7.10; 1.7.14; Replace gcc variadic macros with c99 variadic macros.
1.6	06-Feb-2008	joerg	branches: 1.6.6; 1.6.10; 1.6.12; 1.6.14; Don't use misaligned PCI config access. Reported by Christoph Egger.
1.5	31-Dec-2007	dyoung	Don't buffer packets while the link is down, but drop them on the floor like most other ethernet drivers do. bnx(4) has "vanilla" MII media-handling, so use ether_mediachange and ether_mediastatus.
1.4	30-Dec-2007	dyoung	For consistency with other softc members, s/ethercom/bnx_ec/.
1.3	26-Dec-2007	dyoung	Add device suspend/resume/detach. (Tested with drvctl -S bnx0, drvctl -Q bnx0, drvctl -d bnx0.) Separate device_t storage from softc storage and use CFATTACH_DECL_NEW().
1.2	09-Apr-2007	bouyer	branches: 1.2.10; 1.2.16; 1.2.18; 1.2.22; Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman
1.1	17-Dec-2006	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.12; 1.1.16; 1.1.18; Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.1.18.1	11-Jul-2007	mjf	Sync with head.
1.1.16.1	10-Apr-2007	ad	Sync with head.
1.1.12.1	15-Apr-2007	yamt	sync with head.
1.1.10.2	12-Jan-2007	ad	Sync with head.
1.1.10.1	17-Dec-2006	ad	file if_bnxreg.h was added on branch newlock2 on 2007-01-12 00:57:41 +0000
1.1.8.5	11-Feb-2008	yamt	sync with head.
1.1.8.4	21-Jan-2008	yamt	sync with head
1.1.8.3	03-Sep-2007	yamt	sync with head.
1.1.8.2	30-Dec-2006	yamt	sync with head.
1.1.8.1	17-Dec-2006	yamt	file if_bnxreg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:44 +0000
1.1.6.3	25-Apr-2007	ghen	Pull up following revision(s) (requested by bouyer in ticket #1762): sys/dev/pci/if_bnx.c: revision 1.4 sys/dev/pci/if_bnx.c: revision 1.5 sys/dev/pci/if_bnx.c: revision 1.6 sys/dev/pci/if_bnxreg.h: revision 1.2 Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman RX buffers are malloced memory of 9216 bytes. This can require from 1 to 4 DMA memory segments, depending on how the buffer is in memory. When receiving a packet, we allocate a new one to remplace the one we've used. It can need more segments than the one it remplace, leading to corrution of the RX descriptors, and a panic in bus_dmamap_sync() (DIAGNOSTIC kernels) or possibly memory corruption. Fix: - bnx_get_buf() allocates as many buffer as possible, checking the number of free RX descriptors. Because one receive buffer is not guaranteed to be remplaced on receive, call bnx_get_buf() from bnx_tick() too. This also improve error handling from bnx_get_buf(). - use MCLGET() instead of MEXTMALLOC() if we're running with the standard ethernet MTU. This gives us more receive buffers and waste less memory. Fix build on 32bit ports; should fix GENERIC_DIAGNOSTIC build problem reported by Hisashi T Fujinaka on current-users.
1.1.6.2	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1608): sys/arch/i386/conf/INSTALL: revision 1.299 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.20 via patch sys/arch/i386/conf/GENERIC: revision 1.803 via patch sys/dev/pci/files.pci: revision 1.275 via patch sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 via patch sys/arch/amd64/conf/GENERIC: revision 1.123 via patch sys/dev/pci/if_bnx.c: revision 1.1 via patch share/man/man4/bnx.4: revision 1.1-1.2 via patch distrib/sets/lists/man/mi: revision 1.959 via patch sys/arch/amd64/conf/INSTALL: revision 1.62 via patch sys/dev/microcode/bnx/bnxfw.h: revision 1.1-1.2 via patch sys/dev/pci/if_bnxreg.h: revision 1.1 via patch share/man/man4/Makefile: revision 1.417 via patch sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 via patch Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64. Add a NetBSD RCS tag. Nx -> Ox; remove trailing whitespace.
1.1.6.1	17-Dec-2006	ghen	file if_bnxreg.h was added on branch netbsd-3 on 2006-12-23 14:43:47 +0000
1.1.4.3	20-Apr-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #566): sys/dev/pci/if_bnx.c: revision 1.4 sys/dev/pci/if_bnx.c: revision 1.5 sys/dev/pci/if_bnx.c: revision 1.6 sys/dev/pci/if_bnxreg.h: revision 1.2 Protect bnx_tick() with splnet. Bring in fixes and improvements from OpenBSD: revision 1.25 - Simplify the arguments to bnx_tx_encap. - Don't copy the bd_chain head pointers into temporary objects, they are available globally. From scottl@FreeBSD revision 1.26 Overhaul the transmit path: - Eliminate the bnx_dmamap_arg structure. - Refactor the loop that fills the buffer descriptor so that it can be done with a single set of logic in a single loop instead of two sets of logic. - Eliminate the need to cache and pass descriptor indexes between the start loop and the encap function. - Change the start loop to always check the ifnet sendq for more work. From scottl@FreeBSD revision 1.27 make the exit label naming scheme match the current function names, removes a FreeBSD-ism from the original driver. revision 1.28 -> 1.30 - Ensure that at least 16 TX descriptors are kept unused in the ring. - Use more complete error handling for TX load problems. From scottl@FreeBSD revision 1.31 replace a few more instances of hand rolled code with the LIST_FOREACH macro. revision 1.33 In bnx_start, check the used_tx_bd count rather than the descriptors mbuf pointer to see if the transmit ring is full. The mbuf pointer is set only in the last descriptor of a multi-descriptor packet. By relying on the mbuf pointers of the earlier descriptors, the driver would sometimes overwrite a descriptor belonging to a packet that wasn't completed yet. Also, tx_chain_prod wasn't updated inside the loop, causing the wrong descriptor to be checked after the first iteration. The upshot of all this was the loss of some transmitted packets at medium to high packet rates. In bnx_tx_encap, remove a couple of old statements that shuffled around the tx_mbuf_map pointers. These now correspond 1-to-1 with the transmit descriptors, and they are not supposed to be changed. Correct a couple of inaccurate comments. From jdp@FreeBSD revision 1.43 Allow the bnx(4) driver to make use of all of the available hardware multicast hash slots. The bnx(4) hardware supports 8 slots instead of 4 like the bge(4) hardware. From Mike Karels via FreeBSD Tested by Brad, biorn@ and Johan M:son Lindman RX buffers are malloced memory of 9216 bytes. This can require from 1 to 4 DMA memory segments, depending on how the buffer is in memory. When receiving a packet, we allocate a new one to remplace the one we've used. It can need more segments than the one it remplace, leading to corrution of the RX descriptors, and a panic in bus_dmamap_sync() (DIAGNOSTIC kernels) or possibly memory corruption. Fix: - bnx_get_buf() allocates as many buffer as possible, checking the number of free RX descriptors. Because one receive buffer is not guaranteed to be remplaced on receive, call bnx_get_buf() from bnx_tick() too. This also improve error handling from bnx_get_buf(). - use MCLGET() instead of MEXTMALLOC() if we're running with the standard ethernet MTU. This gives us more receive buffers and waste less memory. Fix build on 32bit ports; should fix GENERIC_DIAGNOSTIC build problem reported by Hisashi T Fujinaka on current-users.
1.1.4.2	21-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #287): sys/arch/i386/conf/INSTALL: revision 1.299 sys/arch/i386/conf/XEN2_DOM0: revision 1.20 sys/arch/i386/conf/GENERIC: revision 1.803 sys/dev/pci/files.pci: revision 1.275 sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 sys/arch/amd64/conf/GENERIC: revision 1.123 sys/dev/pci/if_bnx.c: revision 1.1 share/man/man4/bnx.4: revision 1.1 distrib/sets/lists/man/mi: revision 1.959 sys/arch/amd64/conf/INSTALL: revision 1.62 sys/dev/microcode/bnx/bnxfw.h: revision 1.1 sys/arch/i386/conf/ALL: revision 1.76 sys/dev/pci/if_bnxreg.h: revision 1.1 share/man/man4/Makefile: revision 1.417 sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64.
1.1.4.1	17-Dec-2006	tron	file if_bnxreg.h was added on branch netbsd-4 on 2006-12-21 13:44:01 +0000
1.1.2.2	18-Dec-2006	yamt	sync with head.
1.1.2.1	17-Dec-2006	yamt	file if_bnxreg.h was added on branch yamt-splraiseipl on 2006-12-18 11:42:13 +0000
1.2.22.1	02-Jan-2008	bouyer	Sync with HEAD
1.2.18.1	26-Dec-2007	ad	Sync with head.
1.2.16.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.10.2	23-Mar-2008	matt	sync with HEAD
1.2.10.1	09-Jan-2008	matt	sync with HEAD
1.6.14.1	27-Jun-2008	simonb	Sync with head.
1.6.12.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.6.10.2	11-Mar-2010	yamt	sync with head
1.6.10.1	04-May-2009	yamt	sync with head.
1.6.6.1	29-Jun-2008	mjf	Sync with HEAD.
1.7.14.1	21-Apr-2010	matt	sync to netbsd-5
1.7.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.7.6.2	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.7.6.1	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1268): sys/dev/mii/miidevs: revision 1.90 sys/dev/pci/pcidevs: revision 1.1001 sys/dev/mii/brgphy.c: revision 1.51 sys/dev/microcode/bnx/bnxfw.h: revision 1.3 sys/dev/pci/if_bnx.c: revision 1.29 sys/dev/pci/if_bnxreg.h: revision 1.9 sys/dev/pci/if_bnx.c: revision 1.30 Add support for the Broadcom BCM5709 and BCM5716 chips. Upgrade the B06 firmware and add the B09 firmware required by the BCM5709. Replace TX dmamap array with a dynamically-grown list. Support associated BCM5709CAX and BCM5709C PHYs.
1.7.4.1	28-Apr-2009	skrll	Sync with HEAD.
1.10.4.2	31-May-2011	rmind	sync with head
1.10.4.1	05-Mar-2011	rmind	sync with head
1.13.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.14.28.1	10-Aug-2014	tls	Rebase.
1.14.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.14.10.1	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/mii/brgphy.c 1.69, 1.71-1.76 via patch sys/dev/mii/brgphyreg.h 1.6-1.7, 1.9 sys/dev/pci/if_bge.c 1.259,1.267-1.276 via patch sys/dev/pci/if_bgereg.h 1.87-1.88 sys/dev/pci/if_bnx.c 1.48, 1.49, 1.52-1.57 via patch sys/dev/pci/if_bnxreg.h 1.15-1.18 sys/dev/pci/if_bnxvar.h 1.4-1.6 Changes for bge(4), bnx(4) and brgphy(4): - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719 into pcidevs. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Fix typo in comment. - KNF - brgphy(4): - Add BCM5708S support. - Add 2500SX support. - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. - Fix media detect on some Fiber chips. - Add BCM5706. - bge(4): - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET. - Add BGEPHYF_NO_WIRESPEED flag for fiber devices. - Print APE firmware correctly. - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packets about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... Tested with HP MicroServer Gen8. - Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436). - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver. - bnx(4): - Enclose #ifdef BNX_DEBUG correctly. - Add 2500SX support. - Print some information (ASCI revision, PCI status, etc). From FreeBSD.
1.18.26.1	25-Jun-2018	pgoyette	Sync with HEAD
1.19.2.1	10-Jun-2019	christos	Sync with HEAD
1.26.10.1	03-Apr-2021	thorpej	Sync with HEAD.
1.15	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.14	14-Jul-2020	jdolecek	make bnx_wk (used to trigger bnx_alloc_pkts()) part of softc instead of using a static variable, so it's independant for each adapter
1.13	12-Jul-2020	jdolecek	enable MSI/MSI-X if supported by adapter tested MSI-X with Broadcom NetXtreme II BCM5709 1000Base-T
1.12	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.11	11-Apr-2019	kamil	branches: 1.11.6; Fix CVS Id usage
1.10	05-Apr-2019	msaitoh	- Allow time for the management firmware to enter the running state. - Enable/Disable management frames (NC-SI) to flow to the MCP if MFW exists. - Print the boot code version, management firmware version and some flags. - Gbps -> GT/s for PCIe speed. - Use aprint_*() in bnx_print_adapter_info(). - Use braces for some macors' argument. - Remove unused macros. - KNF.
1.9	29-Mar-2019	msaitoh	Sync with OpenBSD: - Don't need to clear if_timer during attach. (OpenBSD rev. 1.23) - Do the minimal initialization of the firmware so that ASF always works. From ambrisko@FreeBSD (OpenBSD rev. 1.32) - Move the interrupt establishment till after everything in the softc is set up (OpenBSD rev. 1.40) - Let mii_attach() know where the PHY is located instead of scanning for it since we know where it will be anyway and remove the code from the MII bus read/write functions to force reading/writing from the predetermined location. Copied from bge(4) and this is what the upstream FreeBSD bce(4) driver has done once FreBSD gained a mii_attach().
1.8	28-Mar-2019	msaitoh	- Remove extra cast (from OpenBSD rev. 1.95) - Add support flow control (from OpenBSD rev. 1.97 and 1.100). The main difference between NetBSD and other *BSDs is that the flow control does really works. The lower 8bit of BNX_L2CTX_CTX_TYPE should be 0xff (from Linux). Note that I have NetXtremeII-PG203-R.pdf and it says the field is reserved and should be 0.
1.7	08-May-2018	msaitoh	branches: 1.7.2; - Fix a bug that bnx(4) panics on shutdown. Stop callout before restroy. Reported by Andreas Gustafsson in PR#53265. - Make sure not to re-arm the callout when we are about to detach. Same as if_bge.c rev. 1.292. - Use pci_intr_establish_xname().
1.6	01-Jul-2014	msaitoh	branches: 1.6.8; 1.6.22; 1.6.28; Print some information (ASIC revision, PCI status, etc). From FreeBSD.
1.5	01-Jul-2014	msaitoh	No functional change. - s/u_int/uint/ - Modify comment a bit (sync with OpenBSD)
1.4	24-Jun-2014	msaitoh	No functional change: - Tabify. - Remove trailing white spaces. - Fix typo in comment.
1.3	26-Apr-2012	gdt	branches: 1.3.2; 1.3.12; Use RX_PAGES for rx_bd_chain_{,r}seq. Before, TX_PAGES was used, but no harm was done because TX_PAGES == RX_PAGES == 2. Found by Frank Kastenholz of BBN.
1.2	15-Jul-2011	jym	branches: 1.2.2; 1.2.6; 1.2.8; Use a workqueue(9) to defer allocation of TX packets. Loosely inspired from the OpenBSD bnx(4) driver. Tested by Roger Pau Monné, roger.pau at entel.upc.edu. Fixes kern/45051.
1.1	09-Dec-2010	jym	branches: 1.1.6; Similar to bge(4), split if_bnxreg.h in two parts: - if_bnxreg.h: general register values, flags... for bnx(4), that can be included in other drivers - if_bnxvar.h: PCI or device specific code (device state data, debug macros etc.), which should remain private to bnx(4) No comments, no objections on current-users@. See discussion: http://mail-index.netbsd.org/current-users/2010/12/01/msg014926.html
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	09-Dec-2010	rmind	file if_bnxvar.h was added on branch rmind-uvmplock on 2011-03-05 20:53:39 +0000
1.2.8.2	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/mii/brgphy.c 1.69, 1.71-1.76 via patch sys/dev/mii/brgphyreg.h 1.6-1.7, 1.9 sys/dev/pci/if_bge.c 1.259,1.267-1.276 via patch sys/dev/pci/if_bgereg.h 1.87-1.88 sys/dev/pci/if_bnx.c 1.48, 1.49, 1.52-1.57 via patch sys/dev/pci/if_bnxreg.h 1.15-1.18 sys/dev/pci/if_bnxvar.h 1.4-1.6 Changes for bge(4), bnx(4) and brgphy(4): - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719 into pcidevs. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Fix typo in comment. - KNF - brgphy(4): - Add BCM5708S support. - Add 2500SX support. - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. - Fix media detect on some Fiber chips. - Add BCM5706. - bge(4): - Print yet another hardware config word in SRAM(BGE_SRAM_DATA_CFG_5) for BCM5717 and newer devices. - Fix a trivial bug in #ifdef NOTYET. - Add BGEPHYF_NO_WIRESPEED flag for fiber devices. - Print APE firmware correctly. - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packets about 50%. Same as {Free,Open}BSD. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems get NMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. I don't know the reason... Tested with HP MicroServer Gen8. - Fix a bug that bge can't be initialized correctly on some sparc64 systems (PR#49018). - Change initializaton order as the same as the document, Linux tg3 and FreeBSD (part of r241436). - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. Same as FreeBSD r204978 and Linux tg3 driver. - bnx(4): - Enclose #ifdef BNX_DEBUG correctly. - Add 2500SX support. - Print some information (ASCI revision, PCI status, etc). From FreeBSD.
1.2.8.1	07-May-2012	riz	Pull up following revision(s) (requested by gdt in ticket #208): sys/dev/pci/if_bnxvar.h: revision 1.3 Use RX_PAGES for rx_bd_chain_{,r}seq. Before, TX_PAGES was used, but no harm was done because TX_PAGES == RX_PAGES == 2. Found by Frank Kastenholz of BBN.
1.2.6.1	29-Apr-2012	mrg	sync to latest -current.
1.2.2.1	23-May-2012	yamt	sync with head.
1.3.12.1	10-Aug-2014	tls	Rebase.
1.3.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.28.1	21-May-2018	pgoyette	Sync with HEAD
1.6.22.1	09-May-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #814): sys/dev/pci/if_bnxvar.h: revision 1.7 sys/dev/pci/if_bnx.c: revision 1.64 - Fix a bug that bnx(4) panics on shutdown. Stop callout before restroy. Reported by Andreas Gustafsson in PR#53265. - Make sure not to re-arm the callout when we are about to detach. Same as if_bge.c rev. 1.292. - Use pci_intr_establish_xname().
1.6.8.2	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.6.8.1	01-Jul-2014	sborrill	file if_bnxvar.h was added on branch netbsd-5 on 2015-11-09 12:35:32 +0000
1.7.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.7.2.1	10-Jun-2019	christos	Sync with HEAD
1.11.6.1	29-Feb-2020	ad	Sync with head.
1.14	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.13	23-May-2022	rin	branches: 1.13.10; Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.12	24-Apr-2022	skrll	Trailing whitespace
1.11	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.10	08-May-2021	thorpej	Use pci_compatible_match().
1.9	30-May-2020	jdolecek	branches: 1.9.6; discussed with thorpej@, and it seems it's better to allocate the bss alignment buffer statically (as part of softc) just to be sure to not affect performance
1.8	30-May-2020	jdolecek	constify, and make static where possible
1.7	25-Mar-2020	thorpej	- Refactor the firmware loading code into a set of common routines in the core driver that understand how to find model-specific files before the generic files. This greatly simplifies the firmware loading procedure in each of the bus front ends. Adding additional firmware file types (such as CLM files) much simpler. - Make firmware image selection table-driven rather than a set of switch() and if() statements. - Add several additional firmware image entries.
1.6	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.5	01-Nov-2018	martin	Make this compilable with 32bit paddr_t.
1.4	15-Oct-2018	maya	branches: 1.4.2; Apply OpenBSD if_bwfm_pci.c:1.25 by patrick: Add bus_dmamap_sync(9) calls to bwfm(4) so that we make sure the data is synced properly before the CPU or the WiFi chip access the supplied memory. Makes PCIe-connected bwfm(4) work on ARM-based machines.
1.3	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.2	01-Sep-2018	riastradh	bwfm has only one tx descriptor; limit mbuf chains to one segment. PR kern/53287 Can't use m_defrag because it always yields two segments. Discussion on tech-net: https://mail-index.netbsd.org/tech-net/2018/09/01/msg007031.html Diagnosed and tested by maya.
1.1	11-May-2018	maya	branches: 1.1.2; 1.1.4; add bwfm pci support, from openbsd Tested on BCM43602.
1.1.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.1.4.1	10-Jun-2019	christos	Sync with HEAD
1.1.2.6	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1.2.5	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.1.2.4	20-Oct-2018	pgoyette	Sync with head
1.1.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1.2.2	21-May-2018	pgoyette	Sync with HEAD
1.1.2.1	11-May-2018	pgoyette	file if_bwfm_pci.c was added on branch pgoyette-compat on 2018-05-21 04:36:06 +0000
1.4.2.3	02-Nov-2018	martin	Additional pullup for ticket #1072: sys/dev/pci/if_bwfm_pci.c 1.5 Make this compilable with 32bit paddr_t.
1.4.2.2	31-Oct-2018	martin	Backport of bfwm code, requested by maya in ticket #1072: sys/dev/ic/bwfm.c 1.14 sys/dev/ic/bwfmreg.h 1.3 sys/dev/ic/bwfmvar.h 1.3 sys/dev/pci/if_bwfm_pci.c 1.4 sys/dev/pci/if_bwfm_pci.h 1.8 sys/dev/usb/if_bwfm_usb.c 1.8 via patch.
1.4.2.1	15-Oct-2018	martin	file if_bwfm_pci.c was added on branch netbsd-8 on 2018-10-31 09:23:00 +0000
1.9.6.1	13-May-2021	thorpej	Sync with HEAD.
1.13.10.1	02-Aug-2025	perseant	Sync with HEAD
1.2	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.1	11-May-2018	maya	branches: 1.1.2; 1.1.8; add bwfm pci support, from openbsd Tested on BCM43602.
1.1.8.2	31-Oct-2018	martin	Backport of bfwm code, requested by maya in ticket #1072: sys/dev/ic/bwfm.c 1.14 sys/dev/ic/bwfmreg.h 1.3 sys/dev/ic/bwfmvar.h 1.3 sys/dev/pci/if_bwfm_pci.c 1.4 sys/dev/pci/if_bwfm_pci.h 1.8 sys/dev/usb/if_bwfm_usb.c 1.8 via patch.
1.1.8.1	11-May-2018	martin	file if_bwfm_pci.h was added on branch netbsd-8 on 2018-10-31 09:23:00 +0000
1.1.2.2	21-May-2018	pgoyette	Sync with HEAD
1.1.2.1	11-May-2018	pgoyette	file if_bwfm_pci.h was added on branch pgoyette-compat on 2018-05-21 04:36:06 +0000
1.18	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.17	08-May-2021	thorpej	Use pci_compatible_match().
1.16	09-Dec-2018	jdolecek	branches: 1.16.16; use pci_intr_establish_xname() everywhere
1.15	07-Jul-2016	msaitoh	branches: 1.15.16; 1.15.18; KNF. Remove extra spaces. No functional change.
1.14	29-Mar-2014	christos	branches: 1.14.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.13	30-Mar-2013	christos	branches: 1.13.4; remove trailing whitespace
1.12	26-Jul-2011	dyoung	branches: 1.12.2; 1.12.12; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.11	18-May-2011	dyoung	#include <sys/bus.h>, not <machine/bus.h>.
1.10	09-Oct-2010	phx	branches: 1.10.2; Missing \n after printing device type.
1.9	19-Jan-2010	pooka	branches: 1.9.2; 1.9.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.8	26-Apr-2009	cegger	branches: 1.8.2; check error code from bwi_attach()
1.7	26-Apr-2009	cegger	fix error handling
1.6	26-Apr-2009	cegger	device_t/softc split
1.5	26-Apr-2009	cegger	prepare device_t/softc split
1.4	23-Apr-2009	kefren	Add power hooks. Proposed a couple of days ago on tech-kern@, no objections received. Tested on Dell Inspiron 2200 with BCM4318
1.3	10-Jan-2009	cegger	branches: 1.3.2; 1.3.4; 1.3.6; Attach on BCM4328. This chip supports 11n.
1.2	09-Jan-2009	macallan	Fix previous screwup. This is the real bwi driver. Ported by Taylor R. Campbell
1.1	09-Jan-2009	macallan	PCI attachment for the bwi driver From Urban Boquist
1.3.6.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.3.4.3	28-Apr-2009	skrll	Sync with HEAD.
1.3.4.2	19-Jan-2009	skrll	Sync with HEAD.
1.3.4.1	10-Jan-2009	skrll	file if_bwi_pci.c was added on branch nick-hppapmap on 2009-01-19 13:18:25 +0000
1.3.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.3.2.1	10-Jan-2009	mjf	file if_bwi_pci.c was added on branch mjf-devfs2 on 2009-01-17 13:28:59 +0000
1.8.2.3	11-Mar-2010	yamt	sync with head
1.8.2.2	04-May-2009	yamt	sync with head.
1.8.2.1	26-Apr-2009	yamt	file if_bwi_pci.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:56 +0000
1.9.4.2	31-May-2011	rmind	sync with head
1.9.4.1	05-Mar-2011	rmind	sync with head
1.9.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.10.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.12.12.3	03-Dec-2017	jdolecek	update from HEAD
1.12.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.12.12.1	23-Jun-2013	tls	resync from head
1.12.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.13.4.1	18-May-2014	rmind	sync with head
1.14.6.1	09-Jul-2016	skrll	Sync with HEAD
1.15.18.1	10-Jun-2019	christos	Sync with HEAD
1.15.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.16.16.1	13-May-2021	thorpej	Sync with HEAD.
1.49	04-Oct-2025	thorpej	Stop using the property dictionary directly.
1.48	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.47	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.46	30-May-2021	joerg	Unindent code to match the logical structure, ignoring the commented out if. This code should be rewritten to use BUS_DMA_WAITOK, but that's a project of its own.
1.45	08-May-2021	thorpej	branches: 1.45.2; Use pci_compatible_match().
1.44	15-Sep-2020	mrg	branches: 1.44.6; mii_detach() must come before ifmedia_fini(), otherwise mii_detach() will have a NULL mutex and fault. this appears to have been introduced in february, and may be a problem in other drivers converted to use ifmedia_fini(). i've only fixed the two that were faulting for me on sunblade 2000 during reboot.
1.43	02-Jul-2020	msaitoh	prop_data_data_nocopy -> prop_data_value
1.42	08-Mar-2020	thorpej	Use BUS_ADDR_{HI,LO}32().
1.41	01-Mar-2020	thorpej	Use the 64-bit PCI DMA tag if avalable.
1.40	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.39	04-Feb-2020	thorpej	Use ifmedia_fini().
1.38	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.37	26-Dec-2019	msaitoh	branches: 1.37.2; Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants.
1.36	21-Nov-2019	msaitoh	Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt.
1.35	28-May-2019	msaitoh	branches: 1.35.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.34	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.33	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.32	06-Feb-2019	msaitoh	Fix a bug that all ports' MAC address become the last port's address on Non-OF environment. The Saturn multi port card has only one VPD ROM and all ports share it. If the card has four port, it has four "local-mac-address" entries. Before this commit, the code keep the last one and use it for all ports. The Saturn four port card has three bridge. e.g.
1.31	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.30	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.29	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.28	15-Nov-2018	maxv	Remove the 'copy' argument from m_devget(), unused. While here rename off0->off.
1.27	26-Jun-2018	msaitoh	branches: 1.27.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.26	15-Dec-2016	ozaki-r	branches: 1.26.8; 1.26.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.25	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.24	09-Feb-2016	ozaki-r	branches: 1.24.2; Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.23	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.22	10-Aug-2014	tls	branches: 1.22.4; Merge tls-earlyentropy branch into HEAD.
1.21	29-Mar-2014	christos	branches: 1.21.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.20	13-Sep-2013	martin	Remove unused variables
1.19	30-Mar-2013	christos	branches: 1.19.4; remove trailing whitespace
1.18	22-Jul-2012	matt	branches: 1.18.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.17	03-Feb-2012	martin	Add missing <sys/rnd.h> include, ack: jdc@
1.16	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.15	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.14	29-Aug-2011	jmcneill	branches: 1.14.2; 1.14.6; have pci register itself as a module, and make PCI driver modules depend on it
1.13	07-Jan-2011	jmcneill	support building cas(4) as a module
1.12	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.11	14-Jul-2010	jnemeth	When searching PROM data to get a MAC address, search for National Semiconductor Saturn data as well as Sun Cassini data.
1.10	14-Jul-2010	jnemeth	Don't leak kernel stack data when an enaddr can't be found.
1.9	17-Jun-2010	mrg	disable interrupts when resetting the device.
1.8	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.7	11-Mar-2010	mrg	branches: 1.7.2; 1.7.4; various aprint_* fixes.
1.6	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.5	22-Jan-2010	jdc	branches: 1.5.2; Remove port-specific code for determining MAC address, and use device properties instead (c.f. gem(4) and hme(4)).
1.4	19-Jan-2010	pooka	Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.3	17-Jan-2010	jdc	Mask out (disable) cas interrupts on detach and suspend. (Re-)enable interrupts on resume.
1.2	09-Jan-2010	martin	adapt to recent pmf changes
1.1	07-Jan-2010	jdc	Driver for Sun Cassini/Cassini+ (GigaSwift) Ethernet (also known as National Semiconductor Saturn). Based on the OpenBSD driver written by Mark Kettenis: detach support added MAC address lookup modified (only tested on sparc64) Tested on single GigaSwift (UTP and MMF) and quad GigaSwift PCI cards.
1.5.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.5.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.7.4.3	05-Mar-2011	rmind	sync with head
1.7.4.2	03-Jul-2010	rmind	sync with head
1.7.4.1	30-May-2010	rmind	sync with head
1.7.2.3	11-Aug-2010	yamt	sync with head.
1.7.2.2	11-Mar-2010	yamt	sync with head
1.7.2.1	11-Mar-2010	yamt	file if_cas.c was added on branch yamt-nfs-mp on 2010-03-11 15:03:45 +0000
1.14.6.1	18-Feb-2012	mrg	merge to -current.
1.14.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.14.2.2	30-Oct-2012	yamt	sync with head
1.14.2.1	17-Apr-2012	yamt	sync with head
1.18.2.3	03-Dec-2017	jdolecek	update from HEAD
1.18.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.18.2.1	23-Jun-2013	tls	resync from head
1.19.4.1	18-May-2014	rmind	sync with head
1.21.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.22.4.3	05-Feb-2017	skrll	Sync with HEAD
1.22.4.2	19-Mar-2016	skrll	Sync with HEAD
1.22.4.1	06-Jun-2015	skrll	Sync with HEAD
1.24.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.26.14.4	26-Jan-2019	pgoyette	Sync with HEAD
1.26.14.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.26.14.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.26.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.26.8.1	05-Dec-2019	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1464): sys/dev/pci/if_alc.c 1.44 via patch sys/dev/pci/if_ale.c 1.35-1.36 via patch sys/dev/pci/if_cas.c 1.36 via patch - Fix a bug that IFF_ALLMULTI is almost always set. - Whitespace fixes.
1.27.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.27.2.1	10-Jun-2019	christos	Sync with HEAD
1.35.2.2	21-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/arch/amd64/conf/GENERIC: revision 1.552 sys/dev/pci/if_casreg.h: revision 1.2 sys/dev/pci/if_cas.c: revision 1.37 share/man/man4/cas.4: revision 1.4 sys/dev/pci/if_casvar.h: revision 1.7 Update supported device list with part number. From FreeBSD. Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants. Add cas(4).
1.35.2.1	26-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #483): sys/dev/pci/if_cas.c: revision 1.36 sys/dev/pci/if_alc.c: revision 1.44 sys/dev/pci/if_ale.c: revision 1.35 sys/dev/pci/if_ale.c: revision 1.36 sys/dev/pci/if_age.c: revision 1.64 Fix a bug that IFF_ALLMULTI is almost always set. OpenBSD's ac_multirangecnt is not NetBSD's ec_multicnt. Remove accidentally committed debug code. Sorry.
1.37.2.1	29-Feb-2020	ad	Sync with head.
1.44.6.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.44.6.1	13-May-2021	thorpej	Sync with HEAD.
1.45.2.1	31-May-2021	cjep	sync with head
1.7	01-Mar-2025	andvar	s/compete/complete/ in comment.
1.6	28-May-2022	andvar	branches: 1.6.10; fix various typos, mainly in comments.
1.5	09-Dec-2021	andvar	s/atached/attached/ and removing double s in state.
1.4	17-Aug-2021	andvar	fix multiplei repetitive typos in comments, messages and documentation. mainly because copy paste code big amount of files are affected.
1.3	29-Jan-2020	msaitoh	u_int{32,64}_t -> uint{32,64}_t. No functional change.
1.2	26-Dec-2019	msaitoh	branches: 1.2.2; Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants.
1.1	07-Jan-2010	jdc	branches: 1.1.4; 1.1.66; 1.1.72; Driver for Sun Cassini/Cassini+ (GigaSwift) Ethernet (also known as National Semiconductor Saturn). Based on the OpenBSD driver written by Mark Kettenis: detach support added MAC address lookup modified (only tested on sparc64) Tested on single GigaSwift (UTP and MMF) and quad GigaSwift PCI cards.
1.1.72.1	21-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/arch/amd64/conf/GENERIC: revision 1.552 sys/dev/pci/if_casreg.h: revision 1.2 sys/dev/pci/if_cas.c: revision 1.37 share/man/man4/cas.4: revision 1.4 sys/dev/pci/if_casvar.h: revision 1.7 Update supported device list with part number. From FreeBSD. Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants. Add cas(4).
1.1.66.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.1.4.2	11-Mar-2010	yamt	sync with head
1.1.4.1	07-Jan-2010	yamt	file if_casreg.h was added on branch yamt-nfs-mp on 2010-03-11 15:03:45 +0000
1.2.2.1	29-Feb-2020	ad	Sync with head.
1.6.10.1	02-Aug-2025	perseant	Sync with HEAD
1.9	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.8	29-Jan-2020	msaitoh	u_int{32,64}_t -> uint{32,64}_t. No functional change.
1.7	26-Dec-2019	msaitoh	branches: 1.7.2; Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants.
1.6	05-Mar-2019	msaitoh	branches: 1.6.4; Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.5	14-Apr-2015	riastradh	branches: 1.5.18; Include <sys/rndsource.h> where it is actually used. I had removed <sys/rnd.h> from files that didn't mention anything of the rnd(9) API. But they included other files which assumed <sys/rnd.h> had already been included.
1.4	02-Feb-2012	tls	branches: 1.4.6; 1.4.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.3	19-Nov-2011	tls	branches: 1.3.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.2	18-Jan-2010	jdc	branches: 1.2.4; 1.2.14; Add definitions for CAS_INTR_REG and CAS_INTR_PCI.
1.1	07-Jan-2010	jdc	Driver for Sun Cassini/Cassini+ (GigaSwift) Ethernet (also known as National Semiconductor Saturn). Based on the OpenBSD driver written by Mark Kettenis: detach support added MAC address lookup modified (only tested on sparc64) Tested on single GigaSwift (UTP and MMF) and quad GigaSwift PCI cards.
1.2.14.1	17-Apr-2012	yamt	sync with head
1.2.4.2	11-Mar-2010	yamt	sync with head
1.2.4.1	18-Jan-2010	yamt	file if_casvar.h was added on branch yamt-nfs-mp on 2010-03-11 15:03:45 +0000
1.3.2.1	18-Feb-2012	mrg	merge to -current.
1.4.24.1	06-Jun-2015	skrll	Sync with HEAD
1.4.6.1	03-Dec-2017	jdolecek	update from HEAD
1.5.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.5.18.1	10-Jun-2019	christos	Sync with HEAD
1.6.4.1	21-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/arch/amd64/conf/GENERIC: revision 1.552 sys/dev/pci/if_casreg.h: revision 1.2 sys/dev/pci/if_cas.c: revision 1.37 share/man/man4/cas.4: revision 1.4 sys/dev/pci/if_casvar.h: revision 1.7 Update supported device list with part number. From FreeBSD. Fix a bug that the driver sometimes incorrectly attach gentbi(4) instead of gphyter(4) on non-fiber device. Tested with Sun Quad GigaSwift Ethernet UTP (QGE) (part no. 501-6522) on amd64. On this environment, the problem frequently had occurred. I also tested with other 4 cas(4) variants. Add cas(4).
1.7.2.1	29-Feb-2020	ad	Sync with head.
1.166	19-Jan-2020	thorpej	Remove the de(4) driver, which has long since been supplanted by the tlp(4) driver, which supports more chips and more board variants.
1.165	06-Jan-2020	msaitoh	branches: 1.165.2; Protect ec_multicnt.
1.164	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.163	29-May-2019	msaitoh	branches: 1.163.2; Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.162	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.161	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.160	21-May-2019	msaitoh	- I think it's not required to test "#if defined(IFM_ETHER)" now. - Fix compile error. It seems no one use this file. Is it OK to remove this file?
1.159	24-Apr-2019	msaitoh	Remove extra test of SIOCSIFMEDIA.
1.158	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.157	08-Jan-2019	msaitoh	KNF. No functional change.
1.156	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.155	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.154	26-Jun-2018	msaitoh	branches: 1.154.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.153	22-Jun-2018	msaitoh	Remove bpf_tap() call in rx. It's done in if_percpuq_softint().
1.152	22-Jun-2018	msaitoh	It's not required to include net/bpfdesc.h. Remove it.
1.151	19-Mar-2018	msaitoh	s/been been/been/ in comment.
1.150	15-Dec-2016	ozaki-r	branches: 1.150.8; 1.150.14; Move bpf_mtap in Tx hardware intrrupt to if_start The intention of the change is to prevent bpf_mtap from running in hardware interrupt context. if_start is a usual place to do bpf_mtap on Tx. Proposed on tech-kern and tech-net
1.149	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.148	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.147	10-Jun-2016	ozaki-r	branches: 1.147.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.146	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.145	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.144	10-Aug-2014	tls	branches: 1.144.4; Merge tls-earlyentropy branch into HEAD.
1.143	18-May-2014	christos	remove stray line accidentally committed.
1.142	29-Mar-2014	christos	branches: 1.142.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.141	30-Mar-2013	christos	branches: 1.141.4; remove trailing whitespace
1.140	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.139	02-Feb-2012	tls	branches: 1.139.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.138	13-Nov-2010	uebayasi	branches: 1.138.8; 1.138.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.137	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.136	02-Feb-2010	wiz	branches: 1.136.2; 1.136.4; Add missing parentheses, found by cppcheck and reported by Henning Petersen in PR 42718.
1.135	19-Jan-2010	pooka	Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.134	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.133	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.132	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.131	29-Mar-2009	dholland	Merge patch from PR 12605, which tidies up allocation of transmit DMA maps. Generalize it to also tidy up allocation of receive DMA maps. And change a few of the symbol names involved to (1) make sure all uses have been fixed and (2) make it clearer what's actually going on. Previously the driver was using DMA maps off the free list without fully allocating them, apparently in order to save two or three lines releasing them on error paths. According to the submitter of the PR (H.Saito) this was causing it to reuse a map already in use when under load, resulting in panics. I'm not sure if that ought to have been possible or if it reflected an interrupt handling bug somewhere else, but the change is an improvement regardless, so we'll go with it. Compile-tested only, but I've crosschecked the diffs and all that and it's a pretty noninvasive change. (Is anyone actually using this driver rather than tlp?)
1.130	18-Mar-2009	cegger	bzero -> memset
1.129	07-Nov-2008	dyoung	branches: 1.129.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.128	10-Apr-2008	cegger	branches: 1.128.4; 1.128.10; 1.128.12; use aprint_*_dev and device_xname
1.127	07-Feb-2008	dyoung	branches: 1.127.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.126	19-Oct-2007	ad	branches: 1.126.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.125	01-Sep-2007	dyoung	branches: 1.125.4; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.124	09-Jul-2007	ad	branches: 1.124.2; 1.124.6; 1.124.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.123	04-Mar-2007	christos	branches: 1.123.2; 1.123.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.122	24-Oct-2006	drochner	branches: 1.122.4; add zero initializers to make that compile again
1.121	07-Sep-2006	dogcow	branches: 1.121.2; 1.121.4; remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
1.120	15-May-2006	christos	Move the lmc match test in the right place (not in the bsdi match). From David Boggs.
1.119	20-Apr-2006	christos	Put back the code that avoids matching the lmc card. As pointed out by David Boggs, it does not matter if the lmc driver routine returns 100 for the match, because if the lmc driver is loaded via an LKM after boot, the other drivers would have claimed the card already.
1.118	24-Dec-2005	perry	branches: 1.118.4; 1.118.6; 1.118.8; 1.118.10; 1.118.12; __inline__ -> inline
1.117	06-Dec-2005	christos	Remove the "don't match lmc cards" test, because the lmc driver returns a match of 100 which will supersede the de and tlp match if present. If not present, then these two drivers will fail to match an mii. Thanks to thorpej for the explanation.
1.116	06-Dec-2005	thorpej	Don't use magic constants in previous. (Not that we should be doing this here in the first place...)
1.115	06-Dec-2005	christos	Don't match lmc cards by accident.
1.114	01-Jun-2005	drochner	branches: 1.114.2; more cast-qual and shadow fallout
1.113	27-Feb-2005	perry	nuke trailing whitespace
1.112	30-Oct-2004	thorpej	branches: 1.112.4; 1.112.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.111	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.110	03-May-2003	wiz	branches: 1.110.2; 1.110.6; DMA, not dma nor Dma.
1.109	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.108	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.107	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.106	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.105	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.104	08-Jun-2002	yamt	make an auto const variable static.
1.103	13-Nov-2001	lukem	branches: 1.103.8; add RCSID
1.102	07-Jul-2001	thorpej	branches: 1.102.2; bcmp -> memcmp
1.101	07-Jul-2001	thorpej	bzero -> memset
1.100	07-Jul-2001	thorpej	bcopy -> memcpy, strcpy
1.99	13-Jun-2001	wiz	withough -> without
1.98	16-May-2001	lukem	delint: zero sized array initalisation is a gccism
1.97	04-Mar-2001	kim	branches: 1.97.2; Help messages appear each on their own line when discovering additional ports on a multi-port card.
1.96	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.95	15-Nov-2000	thorpej	Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
1.94	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.93	28-Sep-2000	tsutsui	6 -> ETHER_ADDR_LEN
1.92	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.91	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.90	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.89	23-Feb-2000	fair	Fix a missing closing parenthesis in an error diagnostic
1.88	15-Jan-2000	matt	Fixes for 21143 and MII in full-duplex mode.
1.87	04-Dec-1999	ragge	CL* discarding.
1.86	01-Jun-1999	thorpej	branches: 1.86.2; 1.86.8; Make sure bus mastering is enabled; some broken BIOSes don't set this up properly.
1.85	18-May-1999	thorpej	Rework layer 2 protocol input routines. Instead of calling e.g. ether_input() directly, call the function pointer (if_input)(ifp, m). The input routine expects the packet header to be at the head of the packet, and will adjust as necessary. Privatize the layer 2 input and output routines, allowing _ifattach() to set them up as appropriate.
1.84	12-Apr-1999	cjs	PR 6852 from Chris Jones <cjones@hideo.nervana.montana.edu>. This handles the interrupt structure of Compex boards.
1.83	01-Apr-1999	tsubai	branches: 1.83.2; Fix big-endian support. TULIP_BUSMODE_BIGENDIAN does bswap packet buffers also, so we should use TULIP_BUSMODE_DESC_BIGENDIAN on big-endian machines. (PR 7027) XXX 21040 doesn't have this bit, but supporting only 21041+ is better than nothing.
1.82	28-Feb-1999	explorer	Update to slightly altered rnd_attach_source() api
1.81	09-Nov-1998	matt	branches: 1.81.2; Cleanup media set on Alpha. ring txpoll after setting proper tx interrupt mask.
1.80	25-Sep-1998	matt	Fix transmit underflow messages. (\0128 != \0 128)
1.79	22-Sep-1998	matt	Fix the accton support. Only set the board type it isn't an ISV.
1.78	15-Sep-1998	matt	Fix a problem where txput could "lose" dmamaps if it couldn't actually place the packet into the ring for some reason.
1.77	15-Sep-1998	matt	This commit fixes two bugs in the de driver. The first is that the descriptor count gets miscounted on txprobes. The second (and more important) is that transmit stalls should now be fixed. The problem was not due to lack of ring resources but dmamaps. When Jason changed the driver to use pre-allocated maps (instead of the dynamic ones I used), and when there were no more maps, the driver just gave up instead of calling tx_intr to free any transmitted but unreclaimed dma maps. Since there was nothing being transmitted, no transmit interrupts would fire to restore things (and OACTIVE prevented other transmits from happenning). So it stayed starved until another interrupt cause (like a received packet) "woke" it up.
1.76	28-Aug-1998	drochner	some fixes: -use of uninitialized variable in attach (mostly harmless) -removal of dead code in attach() -fix of interpretation of SIA blocks on >=21142 (serious, caused strange things with 10Mbit)
1.75	24-Aug-1998	matt	Fix bug in transmit logic that would transmits to stop. Boolean dyslexia stricks again.
1.74	18-Aug-1998	thorpej	Add some braces to make egcs happy.
1.73	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.72	05-Jul-1998	jonathan	branches: 1.72.2; defopt NS, NSIP.
1.71	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.70	21-Jun-1998	matt	Workaround problem with stalling output. Fix later after I've moved.
1.69	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.68	25-May-1998	mark	Don't make the inclusion of machine/intr.h alpha specific as other ports require this as well.
1.67	22-May-1998	matt	Change 21040 modes slightly. Prevent infinite autosense loop of 21143 that is not connected. Make TULIP_STS_LINFAIL only valid for chips with that status bit defined.
1.66	02-Apr-1998	mycroft	Make sure to unload and recycle the DMA map for a bad received packet. This was (indirectly) the reason for machines `hanging' (actually, looping in tulip_rx_intr()) with de cards.
1.65	29-Mar-1998	mycroft	Oops; make the table const.
1.64	29-Mar-1998	mycroft	Use a 4-bit table to speed up the CRC even further, without increasing the code size too much.
1.63	29-Mar-1998	mycroft	Optimize the address CRC routines a bit.
1.62	11-Feb-1998	thorpej	Don't try to dynamically create and destroy DMA maps. Instead, keep it simple, and allocate one for each transmit and receive descriptor. In addition to being simpler and faster, this fixes a serious memory leak in the transmit path.
1.61	07-Feb-1998	thorpej	Properly handle error conditions from bus_dmamap_load_mbuf().
1.60	07-Feb-1998	thorpej	Use BUS_DMA_NOWAIT where appropriate. Don't create DMA maps with BUS_DMA_ALLOCNOW.
1.59	06-Feb-1998	matt	A version of the de driver which does bus_dma. Note that it has not been tested on any other platform other than i386. Use of bus_dma(9) can be turned by defining TULIP_BUS_DMA_NOTX and TULIP_BUS_DMA_NORX. These allow one to determine if the problem is in the transmit or receive path. A problem reported by mycroft is also fixed.
1.58	12-Jan-1998	thorpej	Update for config changes.
1.57	27-Oct-1997	thorpej	Print interrupt information like all of the other PCI drivers.
1.56	20-Oct-1997	matt	branches: 1.56.2; 21142 / 21143 chips always use ISVSROM format. (print out message is SROM checksum is invalid).
1.55	16-Oct-1997	matt	Fix pr's port-i386/4127 kern/4190 kern/3700.
1.54	16-Oct-1997	matt	Fix a thinko in last night fix. We must set the media if we have reset the chip but if we haven't the the media isn't changing, don't do anything.
1.53	15-Oct-1997	matt	Fix media/console interaction problem on Alpha. Fix 21142/21143 problems with 10Mb/s links.
1.52	13-Oct-1997	explorer	o Make usage of /dev/random dependant on pseudo-device rnd # /dev/random and in-kernel generator in config files. o Add declaration to all architectures. o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include that this code is derived in part from Ted Tyso's linux code.
1.51	10-Oct-1997	explorer	Add hooks to insert timing info into the random system
1.50	24-Sep-1997	thorpej	Mimmick the logic used by other drivers to determine if we should use memory-mapped or i/o-mapped access to the device registers, and always choose memory-mapped if it is enabled. In particular, do _not_ explictly disable the space we decide to not use, as this confuses some versions of Alpha console software (which are arguably buggy because of this problem). Also, fix a logic error pointed out by Ross Harvey <ross@teraflop.com> that would cause memory-mapped access to never be enabled.
1.49	12-Sep-1997	thorpej	Add missing "break;" statements in the code that determines the media setup by the Alpha SRM console software. From Ross Harvey <ross@teraflop.com>.
1.48	25-Aug-1997	thorpej	- Add RCS tags that got lost. - Fix a couple of typos in PCI space configuration.
1.47	25-Aug-1997	thorpej	Merge with 970703 snapshot of Matt Thomas's de driver.
1.46	19-Jul-1997	cgd	branches: 1.46.2; don't expect/provide pci_decompose_tag to be a MI, public function. It wasn't intended to be to begin with, and uses of it (e.g. the one in the 'de' driver) are quite likely to be incorrect.
1.45	09-Jun-1997	thorpej	Fix a slight bogon in merge of this driver; make the device map all of its space, and pick the space used by default based on what is appropriate for the CPU we're running on.
1.44	08-Jun-1997	thorpej	Correct a couple of typos so that this compiles with TULIP_DEBUG, per PR 3723, submitted by enami tsugutomo <enami@ba2.so-net.or.jp>.
1.43	08-Jun-1997	thorpej	Put back Matt Thomas's revision ID, per PR 3723, submitted by enami tsugutomo <enami@ba2.so-net.or.jp>.
1.42	06-Jun-1997	mellon	Clean up conflicts during merge (thomas de-970603)
1.41	07-May-1997	thorpej	Merge 970327 "de" update.
1.40	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.39	13-Apr-1997	cgd	use pci_map_register(). map both mem and I/O spaces, preferring mem space by default if it's usable, but falling back to I/O space if mem isn't usable. If TULIP_IOMAPPED is defined (default on the x86), prefer I/O space then fall back to mem.
1.38	26-Mar-1997	thorpej	Resolve conflicts from merge.
1.37	23-Mar-1997	veego	Resolve conflicts from merge.
1.36	19-Mar-1997	thorpej	Resolve conflicts from merge.
1.35	17-Mar-1997	thorpej	Resolve conflicts from merge.
1.34	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.33	06-Mar-1997	thorpej	Patch from Matt Thomas to deal with the fact that Pass 2.[012] of the 21140A-A[CDE] have a buggy receiver which may hand and/or corrupt data on receive overflows.
1.32	13-Jan-1997	thorpej	branches: 1.32.4; Fix a couple of minor nits so that this compiles on the alpha. Noticed by Matt Jacob (mjacob@feral.com).
1.31	11-Jan-1997	thorpej	Merge into mainline, tidy up some debug code.
1.30	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.29	25-Oct-1996	cgd	update alpha_XXX_dmamap() invocations to match prototype.
1.28	23-Oct-1996	cgd	update drivers to use the new NetBSD/alpha DMA mapping hack, since the old one is not workable with the new bus.h.
1.27	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.26	13-Oct-1996	christos	backout kprintf changes
1.25	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.24	15-Jul-1996	cgd	before #defining the vtophys() hack for the alpha, #undef vtophys since it may be a macro.
1.23	03-Jun-1996	cgd	convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make the code more correct.
1.22	13-May-1996	mycroft	branches: 1.22.4; Use intr.h.
1.21	11-May-1996	mycroft	There was no need to change ac_ipaddr here.
1.20	07-May-1996	christos	#ifdef FreeBSD out unit which is not used by NetBSD.
1.19	07-May-1996	thorpej	Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.18	03-May-1996	christos	Remove unused printf arguments
1.17	01-Apr-1996	cgd	fix inverted sense of tests in DE500 media select function. (Would auto-detect as 100Mbps if 10Mbps, 10Mbps if 100Mbps.)
1.16	30-Mar-1996	cgd	make sure that chipset tags are used properly (because the alpha cares), and include <machine/intr.h> on the alpha, for interrupt information. (i386 will probably switch to this eventually.)
1.15	27-Mar-1996	cgd	modify these to use the new PCI interfaces: (1) use pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register, and bus_{io,mem}_map() to map it. (2) use pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (5) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.14	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.13	14-Mar-1996	cgd	include pcireg before pcivar.h
1.12	11-Mar-1996	cgd	mostly convert to use interfaces provided by <machine/bus.h>. (For now, assumes that pci_map_mem/pci_map_io provide interfaces which utilitize bus_{io,mem}_handle_t's, or types which are compatible. This works on the i386, and will change eventually anyway.)
1.11	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.10	23-Dec-1995	tls	merge latest Thomas version with our minimal changes
1.9	19-Aug-1995	mycroft	Fix oversights in previous change. Remove ugly casts and correct spelling.
1.8	19-Aug-1995	cgd	use I/O ports on the i386
1.7	17-Aug-1995	cgd	clean up 7/26 de snapshot import
1.6	24-Jul-1995	cgd	merge updated version from Matt Thomas, which includes dc21041 support. There are a couple of questionable changes in here, that i'm going to hammer out of him in the next week or so...
1.5	28-Jun-1995	cgd	header file got renamed
1.4	06-Jun-1995	cgd	update from matt thomas
1.3	05-Jun-1995	cgd	set a shutdown hook
1.2	05-Jun-1995	cgd	kill old Id, add NetBSD Id
1.1	05-Jun-1995	cgd	branches: 1.1.1; Initial revision
1.1.1.12	25-Aug-1997	thorpej	Import the 970703 snapshot of Matt Thomas's "de" driver.
1.1.1.11	06-Jun-1997	mellon	Matt Thomas's if_de driver, 1997/6/3 snapshot
1.1.1.10	07-May-1997	thorpej	Update from Matt Thomas (oof, this has been on my todo list for a while!) This adds support for the Cogent 4-port cards.
1.1.1.9	26-Mar-1997	thorpej	Update from Matt Thomas <matt@3am-software.com>. Fixes a couple of media-related bugs, and now properly groks Asante versions of Tulip cards.
1.1.1.8	23-Mar-1997	veego	New DEC Tulip driver from Matt Thomas <matt@3am-software.com>. - Fix the mbuf leak while probing. - Stop probing after saying "autosense problem" if the interface is down.
1.1.1.7	19-Mar-1997	thorpej	Update from Matt Thomas. Fixes several media-selection and full-duplex related bugs. Also, fix a bug that was causing hangs and machine-checks on Alpha systems in some cases.
1.1.1.6	17-Mar-1997	thorpej	New DEC Tulip driver from Matt Thomas <matt@3am-software.com>. Highligts: - Supports media selection via if_media - Support for Asante and better support for ZNYX boards - Support for 21143 chips (such as found on the DE500-BA) - Should be able to pick the BNC on the on-board 21040-based interface in DEC Multias. - Squish a bug here and there.
1.1.1.5	11-Jan-1997	thorpej	Update from Matt Thomas <matt@3am-software.com>. Supports more Tulip-based cards, including new SMC and Kingston cards.
1.1.1.4	14-Dec-1995	tls	95/12/12 snapshot of Matt Thomas's 'de' driver
1.1.1.3	17-Aug-1995	cgd	Matt Thomas's dc21040 'de' driver, July 26, '95 snapshot.
1.1.1.2	22-Jul-1995	cgd	latest if_de, from Matt Thomas
1.1.1.1	05-Jun-1995	cgd	dc21040 driver, from Matt Thomas
1.22.4.1	03-Jun-1996	cgd	pull up from trunk: >convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to >get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make >the code more correct.
1.32.4.3	12-Mar-1997	is	Merge in changes from The Trunk
1.32.4.2	09-Mar-1997	is	netinet/if_ether.h -> netinet/if_inarp.h
1.32.4.1	27-Feb-1997	is	This is a maze of twisty little #ifdef's., all different. I hope I didn't break anything when converting it.
1.46.2.4	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.46.2.3	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.46.2.2	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.46.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.56.2.1	27-Oct-1997	thorpej	Pull up from trunk: Print interrupt information like all of the other PCI drivers.
1.72.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.81.2.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.83.2.2	21-Jun-1999	perry	pullup 1.85->1.86 (thorpej): Make sure bus mastering is enabled
1.83.2.1	28-Apr-1999	perry	branches: 1.83.2.1.2; 1.83.2.1.4; pullup 1.83->1.84 (cjs): recognise/support Comtex boards.
1.83.2.1.4.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.83.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.86.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.86.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.86.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.86.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.86.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.97.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.97.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.97.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.97.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.97.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.102.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.102.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.102.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.103.8.1	20-Jun-2002	gehenna	catch up with -current.
1.110.6.1	24-Jan-2005	he	Pull up revision 1.112 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.110.2.7	11-Dec-2005	christos	Sync with head.
1.110.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.110.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.110.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.110.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.110.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.110.2.1	03-Aug-2004	skrll	Sync with HEAD
1.112.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.112.4.1	29-Apr-2005	kent	sync with -current
1.114.2.5	11-Feb-2008	yamt	sync with head.
1.114.2.4	27-Oct-2007	yamt	sync with head.
1.114.2.3	03-Sep-2007	yamt	sync with head.
1.114.2.2	30-Dec-2006	yamt	sync with head.
1.114.2.1	21-Jun-2006	yamt	sync with head.
1.118.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.118.10.1	11-May-2006	elad	sync with head
1.118.8.2	14-Sep-2006	yamt	sync with head.
1.118.8.1	24-May-2006	yamt	sync with head.
1.118.6.2	01-Jun-2006	kardel	Sync with head.
1.118.6.1	22-Apr-2006	simonb	Sync with head.
1.118.4.1	09-Sep-2006	rpaulo	sync with head
1.121.4.1	10-Dec-2006	yamt	sync with head.
1.121.2.1	18-Nov-2006	ad	Sync with head.
1.122.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.123.4.1	11-Jul-2007	mjf	Sync with head.
1.123.2.3	23-Oct-2007	ad	Sync with head.
1.123.2.2	09-Oct-2007	ad	Sync with head.
1.123.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.124.8.2	23-Mar-2008	matt	sync with HEAD
1.124.8.1	06-Nov-2007	matt	sync with HEAD
1.124.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.124.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.124.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.125.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.126.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.127.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.127.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.128.12.2	28-Apr-2009	skrll	Sync with HEAD.
1.128.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.128.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.128.4.4	11-Aug-2010	yamt	sync with head.
1.128.4.3	11-Mar-2010	yamt	sync with head
1.128.4.2	16-May-2009	yamt	sync with head
1.128.4.1	04-May-2009	yamt	sync with head.
1.129.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.136.4.2	05-Mar-2011	rmind	sync with head
1.136.4.1	30-May-2010	rmind	sync with head
1.136.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.138.12.1	18-Feb-2012	mrg	merge to -current.
1.138.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.138.8.2	30-Oct-2012	yamt	sync with head
1.138.8.1	17-Apr-2012	yamt	sync with head
1.139.6.4	03-Dec-2017	jdolecek	update from HEAD
1.139.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.139.6.2	23-Jun-2013	tls	resync from head
1.139.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.141.4.1	18-May-2014	rmind	sync with head
1.142.2.2	10-Aug-2014	tls	Rebase.
1.142.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.144.4.4	05-Feb-2017	skrll	Sync with HEAD
1.144.4.3	09-Jul-2016	skrll	Sync with HEAD
1.144.4.2	19-Mar-2016	skrll	Sync with HEAD
1.144.4.1	06-Jun-2015	skrll	Sync with HEAD
1.147.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.150.14.6	18-Jan-2019	pgoyette	Synch with HEAD
1.150.14.5	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.150.14.4	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.150.14.3	28-Jul-2018	pgoyette	Sync with HEAD
1.150.14.2	25-Jun-2018	pgoyette	Sync with HEAD
1.150.14.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.150.8.2	31-Jul-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #947): sys/dev/pci/if_de.c: revision 1.153 Remove bpf_tap() call in rx. It's done in if_percpuq_softint().
1.150.8.1	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.154.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.154.2.1	10-Jun-2019	christos	Sync with HEAD
1.163.2.1	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #662): sys/dev/pcmcia/if_xi.c: revision 1.93 sys/arch/x86/pci/if_vmx.c: revision 1.54 sys/dev/pci/if_de.c: revision 1.165 sys/arch/arm/ti/if_cpsw.c: revision 1.10 sys/arch/arm/omap/if_cpsw.c: revision 1.26 sys/dev/isa/if_iy.c: revision 1.112 sys/dev/pcmcia/if_ray.c: revision 1.96 Add ETHER_LOCK() and ETHER_UNLOCK() to protect ec_multiaddrs. XXX These drivers don't check whether enm_addrlo and enm_addrhi are the same or not, so it won't work correctly if an multicast address entry has a range. Protect ec_multicnt.
1.165.2.1	25-Jan-2020	ad	Sync with head.
1.64	19-Jan-2020	thorpej	Remove the de(4) driver, which has long since been supplanted by the tlp(4) driver, which supports more chips and more board variants.
1.63	27-Dec-2019	msaitoh	branches: 1.63.2; s/recevive/receive/ in comment.
1.62	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.61	08-Jan-2019	msaitoh	KNF. No functional change.
1.60	06-Sep-2015	dholland	branches: 1.60.16; 1.60.18; More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.59	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.58	27-Oct-2012	chs	branches: 1.58.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.57	02-Feb-2012	tls	branches: 1.57.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.56	19-Nov-2011	tls	branches: 1.56.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.55	05-Apr-2010	joerg	branches: 1.55.8; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.54	19-Jan-2010	pooka	branches: 1.54.2; 1.54.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.53	18-Jan-2010	pooka	Remove unused macro definition (the NetBSD version would just not go through a compiler).
1.52	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.51	29-Mar-2009	dholland	Merge patch from PR 12605, which tidies up allocation of transmit DMA maps. Generalize it to also tidy up allocation of receive DMA maps. And change a few of the symbol names involved to (1) make sure all uses have been fixed and (2) make it clearer what's actually going on. Previously the driver was using DMA maps off the free list without fully allocating them, apparently in order to save two or three lines releasing them on error paths. According to the submitter of the PR (H.Saito) this was causing it to reuse a map already in use when under load, resulting in panics. I'm not sure if that ought to have been possible or if it reflected an interrupt handling bug somewhere else, but the change is an improvement regardless, so we'll go with it. Compile-tested only, but I've crosschecked the diffs and all that and it's a pretty noninvasive change. (Is anyone actually using this driver rather than tlp?)
1.50	12-Jun-2008	cegger	branches: 1.50.4; 1.50.10; use device_lookup_private to get softc
1.49	22-May-2008	dyoung	branches: 1.49.2; Delete non-illuminating #ifdef-OTHER_OPERATING_SYSTEM code that confuses ctags(1).
1.48	04-Mar-2007	chris	branches: 1.48.36; 1.48.38; 1.48.40; 1.48.42; Fix caddr_t fallout for cats kernel builds.
1.47	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.46	22-Nov-2006	uebayasi	branches: 1.46.4; Use C99 initializers where needed.
1.45	29-May-2006	drochner	branches: 1.45.6; 1.45.8; gcc4 found that it can't store #define TULIP_TXTIMER 4 in u_int32_t tulip_txtimer : 2; /* transmission timer */ gcc4 appears to be right. "tulip_txtimer" is purely used by software, so give it the full integer. (Appearently the txtimer could never work.)
1.44	16-Feb-2006	perry	branches: 1.44.2; 1.44.8; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.43	24-Dec-2005	perry	branches: 1.43.2; 1.43.4; 1.43.6; __inline__ -> inline
1.42	24-Dec-2005	perry	__asm__ -> __asm
1.41	11-Dec-2005	christos	merge ktrace-lwp.
1.40	27-Feb-2005	perry	branches: 1.40.4; nuke trailing whitespace
1.39	11-Feb-2005	simonb	White space nit- don't put a space before/after increment/decrement operators.
1.38	01-Oct-2002	thorpej	branches: 1.38.6; 1.38.14; 1.38.16; Remove uneecessary extern decl of de_ca.
1.37	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.36	13-Jun-2001	wiz	branches: 1.36.2; withough -> without
1.35	03-May-2000	thorpej	branches: 1.35.6; `typedef void ifnet_ret_t;' isn't valid C -- use #define, instead.
1.34	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.33	15-Jan-2000	matt	Fixes for 21143 and MII in full-duplex mode.
1.32	01-Apr-1999	tsubai	branches: 1.32.8; Fix typo. (From PR 7027)
1.31	29-Sep-1998	matt	branches: 1.31.4; Invert the RXCOPY test. Note that most new platforms will not have relaxed alignment so the de driver should just work on them.
1.30	15-Sep-1998	matt	Fix a problem where txput could "lose" dmamaps if it couldn't actually place the packet into the ring for some reason.
1.29	15-Sep-1998	matt	This commit fixes two bugs in the de driver. The first is that the descriptor count gets miscounted on txprobes. The second (and more important) is that transmit stalls should now be fixed. The problem was not due to lack of ring resources but dmamaps. When Jason changed the driver to use pre-allocated maps (instead of the dynamic ones I used), and when there were no more maps, the driver just gave up instead of calling tx_intr to free any transmitted but unreclaimed dma maps. Since there was nothing being transmitted, no transmit interrupts would fire to restore things (and OACTIVE prevented other transmits from happenning). So it stayed starved until another interrupt cause (like a received packet) "woke" it up.
1.28	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.27	25-May-1998	mark	branches: 1.27.2; Define TULIP_COPY_RXDATA for arm32 as well as alpha.
1.26	11-Feb-1998	thorpej	Don't try to dynamically create and destroy DMA maps. Instead, keep it simple, and allocate one for each transmit and receive descriptor. In addition to being simpler and faster, this fixes a serious memory leak in the transmit path.
1.25	07-Feb-1998	thorpej	Create RX DMA maps with ALLOCNOW. These maps are essentially always loaded, and recovering from a resource shortage error is much harder in the receive case.
1.24	07-Feb-1998	thorpej	Use BUS_DMA_NOWAIT where appropriate. Don't create DMA maps with BUS_DMA_ALLOCNOW.
1.23	06-Feb-1998	matt	A version of the de driver which does bus_dma. Note that it has not been tested on any other platform other than i386. Use of bus_dma(9) can be turned by defining TULIP_BUS_DMA_NOTX and TULIP_BUS_DMA_NORX. These allow one to determine if the problem is in the transmit or receive path. A problem reported by mycroft is also fixed.
1.22	20-Oct-1997	matt	21142 / 21143 chips always use ISVSROM format. (print out message is SROM checksum is invalid).
1.21	16-Oct-1997	matt	Fix pr's port-i386/4127 kern/4190 kern/3700.
1.20	15-Oct-1997	matt	Fix media/console interaction problem on Alpha. Fix 21142/21143 problems with 10Mb/s links.
1.19	15-Oct-1997	explorer	Add and/or update rnd_*() hooks
1.18	13-Oct-1997	explorer	o Make usage of /dev/random dependant on pseudo-device rnd # /dev/random and in-kernel generator in config files. o Add declaration to all architectures. o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include that this code is derived in part from Ted Tyso's linux code.
1.17	10-Oct-1997	explorer	Add hooks to insert timing info into the random system
1.16	25-Aug-1997	thorpej	Add RCS tags that got lost.
1.15	25-Aug-1997	thorpej	Merge with 970703 snapshot of Matt Thomas's de driver.
1.14	19-Jul-1997	cgd	branches: 1.14.2; don't expect/provide pci_decompose_tag to be a MI, public function. It wasn't intended to be to begin with, and uses of it (e.g. the one in the 'de' driver) are quite likely to be incorrect.
1.13	08-Jun-1997	thorpej	Put back Matt Thomas's revision ID, per PR 3723, submitted by enami tsugutomo <enami@ba2.so-net.or.jp>.
1.12	06-Jun-1997	mellon	Clean up conflicts during merge (thomas de-970603)
1.11	07-May-1997	thorpej	Merge 970327 "de" update.
1.10	13-Apr-1997	cgd	clean up space/tag comments
1.9	26-Mar-1997	thorpej	Resolve conflicts from merge.
1.8	23-Mar-1997	veego	Resolve conflicts from merge.
1.7	19-Mar-1997	thorpej	Resolve conflicts from merge.
1.6	17-Mar-1997	thorpej	Resolve conflicts from merge.
1.5	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.4	06-Mar-1997	thorpej	Patch from Matt Thomas to deal with the fact that Pass 2.[012] of the 21140A-A[CDE] have a buggy receiver which may hand and/or corrupt data on receive overflows.
1.3	13-Jan-1997	thorpej	branches: 1.3.4; Fix a couple of minor nits so that this compiles on the alpha. Noticed by Matt Jacob (mjacob@feral.com).
1.2	11-Jan-1997	thorpej	Merge into mainline, tidy up some debug code.
1.1	11-Jan-1997	thorpej	branches: 1.1.1; Initial revision
1.1.1.8	25-Aug-1997	thorpej	Import the 970703 snapshot of Matt Thomas's "de" driver.
1.1.1.7	06-Jun-1997	mellon	Matt Thomas's if_de driver, 1997/6/3 snapshot
1.1.1.6	07-May-1997	thorpej	Update from Matt Thomas (oof, this has been on my todo list for a while!) This adds support for the Cogent 4-port cards.
1.1.1.5	26-Mar-1997	thorpej	Update from Matt Thomas <matt@3am-software.com>. Fixes a couple of media-related bugs, and now properly groks Asante versions of Tulip cards.
1.1.1.4	23-Mar-1997	veego	New DEC Tulip driver from Matt Thomas <matt@3am-software.com>. - Fix the mbuf leak while probing. - Stop probing after saying "autosense problem" if the interface is down.
1.1.1.3	19-Mar-1997	thorpej	Update from Matt Thomas. Fixes several media-selection and full-duplex related bugs. Also, fix a bug that was causing hangs and machine-checks on Alpha systems in some cases.
1.1.1.2	17-Mar-1997	thorpej	New DEC Tulip driver from Matt Thomas <matt@3am-software.com>. Highligts: - Supports media selection via if_media - Support for Asante and better support for ZNYX boards - Support for 21143 chips (such as found on the DE500-BA) - Should be able to pick the BNC on the on-board 21040-based interface in DEC Multias. - Squish a bug here and there.
1.1.1.1	11-Jan-1997	thorpej	Update from Matt Thomas <matt@3am-software.com>. Supports more Tulip-based cards, including new SMC and Kingston cards.
1.3.4.2	12-Mar-1997	is	Merge in changes from The Trunk
1.3.4.1	27-Feb-1997	is	This is a maze of twisty little #ifdef's., all different. I hope I didn't break anything when converting it.
1.14.2.2	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.27.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.31.4.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.32.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.35.6.2	18-Oct-2002	nathanw	Catch up to -current.
1.35.6.1	21-Jun-2001	nathanw	Catch up to -current.
1.36.2.1	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.38.16.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.38.16.1	12-Feb-2005	yamt	sync with head.
1.38.14.1	29-Apr-2005	kent	sync with -current
1.38.6.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.38.6.1	15-Feb-2005	skrll	Sync with HEAD.
1.40.4.3	03-Sep-2007	yamt	sync with head.
1.40.4.2	30-Dec-2006	yamt	sync with head.
1.40.4.1	21-Jun-2006	yamt	sync with head.
1.43.6.2	01-Jun-2006	kardel	Sync with head.
1.43.6.1	22-Apr-2006	simonb	Sync with head.
1.43.4.1	09-Sep-2006	rpaulo	sync with head
1.43.2.1	18-Feb-2006	yamt	sync with head.
1.44.8.1	19-Jun-2006	chap	Sync with head.
1.44.2.1	26-Jun-2006	yamt	sync with head.
1.45.8.1	10-Dec-2006	yamt	sync with head.
1.45.6.1	12-Jan-2007	ad	Sync with head.
1.46.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.48.42.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.48.40.4	11-Aug-2010	yamt	sync with head.
1.48.40.3	11-Mar-2010	yamt	sync with head
1.48.40.2	16-May-2009	yamt	sync with head
1.48.40.1	04-May-2009	yamt	sync with head.
1.48.38.2	17-Jun-2008	yamt	sync with head.
1.48.38.1	04-Jun-2008	yamt	sync with head
1.48.36.2	29-Jun-2008	mjf	Sync with HEAD.
1.48.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.49.2.1	18-Jun-2008	simonb	Sync with head.
1.50.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.4.1	28-Apr-2009	skrll	Sync with HEAD.
1.54.4.1	30-May-2010	rmind	sync with head
1.54.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.55.8.2	30-Oct-2012	yamt	sync with head
1.55.8.1	17-Apr-2012	yamt	sync with head
1.56.2.1	18-Feb-2012	mrg	merge to -current.
1.57.6.2	03-Dec-2017	jdolecek	update from HEAD
1.57.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.58.14.2	22-Sep-2015	skrll	Sync with HEAD
1.58.14.1	06-Jun-2015	skrll	Sync with HEAD
1.60.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.60.18.1	10-Jun-2019	christos	Sync with HEAD
1.60.16.1	18-Jan-2019	pgoyette	Synch with HEAD
1.63.2.1	25-Jan-2020	ad	Sync with head.
1.65	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.64	09-Feb-2024	andvar	branches: 1.64.2; fix spelling mistakes, mainly in comments and log messages.
1.63	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.62	05-Dec-2021	msaitoh	s/checkum/checksum/ in comment.
1.61	10-Nov-2021	msaitoh	s/threshhold/threshold/ in comment.
1.60	09-Aug-2021	andvar	s/aligment/alignment/ + one more typo fix in comments.
1.59	08-May-2021	thorpej	Use pci_compatible_match().
1.58	01-Mar-2020	thorpej	branches: 1.58.10; Support 64-bit DMA addressing, and use the 64-bit PCI DMA tag if available.
1.57	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.56	10-Nov-2019	chs	branches: 1.56.2; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.55	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.54	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.53	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.52	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.51	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.50	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.49	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.48	26-Jun-2018	msaitoh	branches: 1.48.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.47	15-Dec-2016	ozaki-r	branches: 1.47.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.46	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.45	07-Jul-2016	msaitoh	branches: 1.45.2; KNF. Remove extra spaces. No functional change.
1.44	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.43	01-Jun-2016	pgoyette	Fix typo that somehow snuck in.
1.42	01-Jun-2016	pgoyette	Add support for Intel 82597EX_SR - from PR kern/47750
1.41	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.40	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.39	11-Jan-2015	chs	fix device/softc split botch.
1.38	10-Aug-2014	tls	branches: 1.38.2; 1.38.4; Merge tls-earlyentropy branch into HEAD.
1.37	29-Mar-2014	christos	branches: 1.37.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.36	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.35	27-Oct-2012	chs	branches: 1.35.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.34	02-Feb-2012	tls	branches: 1.34.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.33	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.32	19-Nov-2011	tls	branches: 1.32.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.31	13-Nov-2010	uebayasi	branches: 1.31.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.30	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.29	19-Jan-2010	pooka	branches: 1.29.2; 1.29.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.28	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.27	27-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Compile test only.
1.26	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.25	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.24	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.23	13-Feb-2009	abs	avoid long/unsigned long printf warning on some platforms
1.22	07-Nov-2008	dyoung	branches: 1.22.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.21	10-Apr-2008	cegger	branches: 1.21.4; 1.21.10; 1.21.12; 1.21.14; 1.21.18; use aprint_*_dev and device_xname
1.20	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.19	07-Feb-2008	dyoung	branches: 1.19.2; 1.19.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.18	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.17	19-Oct-2007	ad	branches: 1.17.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	26-Aug-2007	dyoung	branches: 1.16.2; 1.16.6; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.15	04-Mar-2007	christos	branches: 1.15.2; 1.15.10; 1.15.14; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.14	16-Nov-2006	christos	branches: 1.14.4; __unused removal on arguments; approved by core.
1.13	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.12	11-Dec-2005	christos	branches: 1.12.20; 1.12.22; merge ktrace-lwp.
1.11	02-May-2005	yamt	branches: 1.11.2; split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.10	27-Feb-2005	perry	nuke trailing whitespace
1.9	21-Feb-2005	thorpej	Define M_CSUM_DATA_IPv4_IPHL() and M_CSUM_DATA_IPv4_OFFSET() macros to extract data from csum_data, rather than just open-coding it.
1.8	18-Feb-2005	heas	Adjust csum_data usage; mask off the high 16 bits for L4 header stuff offset and make use of the L3 header length from the high 16 bits such that an m_pullup is no londer necessary.
1.7	30-Oct-2004	thorpej	branches: 1.7.4; 1.7.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.6	21-Sep-2004	yamt	introduce M_EXT_RW to allow mbuf external storage R/W. discussed on tech-net@.
1.5	10-May-2004	thorpej	branches: 1.5.2; Fix a couple of whitespace nits.
1.4	22-Apr-2004	itojun	sprintf -> snprintf
1.3	14-Apr-2004	ragge	The 82597 chip may write back some receive descriptors off-by-one when the PCIX bus start to get heavily loaded. This only happen when using chained descriptors, therefore work-around it by providing a private pool of receive buffers able to contain a full-sized packet.
1.2	13-Apr-2004	ragge	Make it possible to use 16k MTU.
1.1	12-Mar-2004	ragge	branches: 1.1.2; Add driver for Intel PRO/10GbE ethernet adapter. Now NetBSD has no problem with intercontinental TCP connections of over 4Gbit/s, which is where my test hardware runs out of bus bandwidth. Stuff that is on the TODO list: * HW VLAN support. * Large jumbo buffers (16k). * TCP Segmentation Offload * RAIDC (receive interrupt delay adaptation) * Understand how to use memory above 4GB.
1.1.2.2	16-Apr-2004	tron	branches: 1.1.2.2.2; Pull up revision 1.3 (requested by ragge in ticket #139): The 82597 chip may write back some receive descriptors off-by-one when the PCIX bus start to get heavily loaded. This only happen when using chained descriptors, therefore work-around it by providing a private pool of receive buffers able to contain a full-sized packet.
1.1.2.1	16-Apr-2004	tron	Pull up revision 1.2 (requested by ragge in ticket #139): Make it possible to use 16k MTU.
1.1.2.2.2.1	24-Jan-2005	he	Pull up revision 1.7 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.5.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.5.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.5.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.2	03-Aug-2004	skrll	Sync with HEAD
1.5.2.1	10-May-2004	skrll	file if_dge.c was added on branch ktrace-lwp on 2004-08-03 10:49:08 +0000
1.7.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.4.1	29-Apr-2005	kent	sync with -current
1.11.2.5	17-Mar-2008	yamt	sync with head.
1.11.2.4	11-Feb-2008	yamt	sync with head.
1.11.2.3	15-Nov-2007	yamt	sync with head.
1.11.2.2	27-Oct-2007	yamt	sync with head.
1.11.2.1	03-Sep-2007	yamt	sync with head.
1.12.22.2	10-Dec-2006	yamt	sync with head.
1.12.22.1	22-Oct-2006	yamt	sync with head
1.12.20.1	18-Nov-2006	ad	Sync with head.
1.14.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.15.14.3	11-Nov-2007	joerg	Sync with HEAD.
1.15.14.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.15.14.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.15.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.15.2.3	23-Oct-2007	ad	Sync with head.
1.15.2.2	09-Oct-2007	ad	Sync with head.
1.15.2.1	01-Sep-2007	ad	Update for pool_cache API changes.
1.16.6.2	13-Nov-2007	bouyer	Sync with HEAD
1.16.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.2.3	23-Mar-2008	matt	sync with HEAD
1.16.2.2	08-Nov-2007	matt	sync with -HEAD
1.16.2.1	06-Nov-2007	matt	sync with HEAD
1.17.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.17.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.19.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.19.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.19.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.19.2.1	24-Mar-2008	keiichi	sync with head.
1.21.18.1	21-Apr-2010	matt	sync to netbsd-5
1.21.14.1	30-Sep-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #1040): sys/dev/pci/if_dge.c: revision 1.23 avoid long/unsigned long printf warning on some platforms
1.21.12.2	03-Mar-2009	skrll	Sync with HEAD.
1.21.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.21.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.21.4.4	11-Aug-2010	yamt	sync with head.
1.21.4.3	11-Mar-2010	yamt	sync with head
1.21.4.2	16-May-2009	yamt	sync with head
1.21.4.1	04-May-2009	yamt	sync with head.
1.22.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.29.4.2	05-Mar-2011	rmind	sync with head
1.29.4.1	30-May-2010	rmind	sync with head
1.29.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.31.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.31.8.2	30-Oct-2012	yamt	sync with head
1.31.8.1	17-Apr-2012	yamt	sync with head
1.32.2.1	18-Feb-2012	mrg	merge to -current.
1.34.6.3	03-Dec-2017	jdolecek	update from HEAD
1.34.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.34.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.35.2.1	18-May-2014	rmind	sync with head
1.37.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.38.4.5	05-Feb-2017	skrll	Sync with HEAD
1.38.4.4	09-Jul-2016	skrll	Sync with HEAD
1.38.4.3	19-Mar-2016	skrll	Sync with HEAD
1.38.4.2	06-Jun-2015	skrll	Sync with HEAD
1.38.4.1	06-Apr-2015	skrll	Sync with HEAD
1.38.2.1	11-Jan-2015	martin	Pull up following revision(s) (requested by chs in ticket #406): sys/dev/pci/if_dge.c: revision 1.39 fix device/softc split botch.
1.45.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.47.14.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.47.14.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.47.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.48.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.48.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.48.2.1	10-Jun-2019	christos	Sync with HEAD
1.56.2.1	29-Feb-2020	ad	Sync with head.
1.58.10.1	13-May-2021	thorpej	Sync with HEAD.
1.64.2.1	02-Aug-2025	perseant	Sync with HEAD
1.3	25-Dec-2007	perry	Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.2	11-Dec-2005	christos	branches: 1.2.46; 1.2.52; 1.2.56; 1.2.60; merge ktrace-lwp.
1.1	12-Mar-2004	ragge	branches: 1.1.4; 1.1.18; Add driver for Intel PRO/10GbE ethernet adapter. Now NetBSD has no problem with intercontinental TCP connections of over 4Gbit/s, which is where my test hardware runs out of bus bandwidth. Stuff that is on the TODO list: * HW VLAN support. * Large jumbo buffers (16k). * TCP Segmentation Offload * RAIDC (receive interrupt delay adaptation) * Understand how to use memory above 4GB.
1.1.18.1	21-Jan-2008	yamt	sync with head
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	12-Mar-2004	skrll	file if_dgereg.h was added on branch ktrace-lwp on 2004-08-03 10:49:08 +0000
1.2.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.2.56.1	26-Dec-2007	ad	Sync with head.
1.2.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.46.1	09-Jan-2008	matt	sync with HEAD
1.38	06-Sep-2018	maxv	Retire the 'midway' driver. Discussed on tech-net@ recently and also three years ago, part of removing the network ATM code.
1.37	29-Mar-2014	christos	branches: 1.37.28; 1.37.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.36	26-Jul-2011	dyoung	branches: 1.36.2; 1.36.12; 1.36.16; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.35	18-Jul-2011	mrg	convert the remaining dev/sbus drivers to CFATTACH_DECL_NEW, cfdata_t and device_t, including the pci frontend and backend for en(4).
1.34	01-Feb-2011	chuck	udpate license clauses on my code to match the new-style BSD licenses. remove no-longer-valid wustl email address for me. based on diff that rmind@ sent me. no functional change with this commit.
1.33	26-Nov-2009	njoly	branches: 1.33.4; 1.33.6; 1.33.8; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.32	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.31	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.30	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.29	15-Mar-2009	cegger	ansify function definitions
1.28	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.27	10-Apr-2008	cegger	branches: 1.27.4; 1.27.12; 1.27.18; use aprint_*_dev and device_xname
1.26	16-Nov-2006	christos	branches: 1.26.48; __unused removal on arguments; approved by core.
1.25	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.24	11-May-2006	mrg	branches: 1.24.8; 1.24.10; don't try to #define bus_space_read_1() if defined(sparc).
1.23	11-Dec-2005	christos	branches: 1.23.4; 1.23.6; 1.23.8; 1.23.10; 1.23.12; merge ktrace-lwp.
1.22	27-Feb-2005	perry	branches: 1.22.4; nuke trailing whitespace
1.21	04-Feb-2005	perry	de-__P
1.20	30-Oct-2003	simonb	branches: 1.20.8; 1.20.10; Remove some assigned-to but otherwise unused variables.
1.19	31-Jan-2003	thorpej	branches: 1.19.2; Use aprint_*().
1.18	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.17	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.16	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.15	13-Nov-2001	lukem	add RCSID
1.14	28-Dec-2000	sommerfeld	branches: 1.14.2; 1.14.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.13	01-Jul-1999	itojun	branches: 1.13.2; IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628. (Sorry for a big commit, I can't separate this into several pieces...) Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details. - sys/kern: do not assume single mbuf, accept chained mbuf on passing data from userland to kernel (or other way round). - "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ package (ftp://ftp.csl.sony.co.jp/pub/kjc/). - sys/netinet/tcp: IPv4/v6 dual stack tcp support. - sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those file to be there so we patch it up. - sys/netinet: IPsec additions are here and there. - sys/netinet6/: most of IPv6 code sits here. - sys/netkey: IPsec key management code - dev/pci/pcidevs: regen In my understanding no code here is subject to export control so it should be safe.
1.12	24-Mar-1999	cgd	branches: 1.12.4; 1.12.6; in general, it's a good idea to make sure you at least have all of the variables you reference, before you commit something...
1.11	24-Mar-1999	thorpej	Make sure bus mastering is enabled.
1.10	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.9	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.8	13-Apr-1997	cgd	use pci_map_register().
1.7	20-Mar-1997	chuck	add support for adaptec PCI atm cards (see midway.c log for details)
1.6	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.5	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.4	13-Oct-1996	christos	backout kprintf changes
1.3	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.2	27-Aug-1996	cgd	remove include of <sys/types.h> (it's implied by including <sys/param.h>, and historically has been), and add include of <sys/systm.h> for prototypes.
1.1	22-Jun-1996	chuck	pci specific code for the eni155p card
1.12.6.2	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.12.6.1	28-Jun-1999	itojun	KAME/NetBSD 1.4 SNAP kit, dated 19990628. NOTE: this branch (kame) is used just for refernce. this may not compile due to multiple reasons.
1.12.4.1	01-Jul-1999	thorpej	Sync w/ -current.
1.13.2.1	05-Jan-2001	bouyer	Sync with HEAD
1.14.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.14.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.14.2.2	18-Oct-2002	nathanw	Catch up to -current.
1.14.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.19.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.19.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.19.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.19.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.10.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.20.10.1	12-Feb-2005	yamt	sync with head.
1.20.8.1	29-Apr-2005	kent	sync with -current
1.22.4.2	30-Dec-2006	yamt	sync with head.
1.22.4.1	21-Jun-2006	yamt	sync with head.
1.23.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.23.10.1	11-May-2006	elad	sync with head
1.23.8.1	24-May-2006	yamt	sync with head.
1.23.6.1	01-Jun-2006	kardel	Sync with head.
1.23.4.1	09-Sep-2006	rpaulo	sync with head
1.24.10.2	10-Dec-2006	yamt	sync with head.
1.24.10.1	22-Oct-2006	yamt	sync with head
1.24.8.1	18-Nov-2006	ad	Sync with head.
1.26.48.1	02-Jun-2008	mjf	Sync with HEAD.
1.27.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.27.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.27.4.3	11-Mar-2010	yamt	sync with head
1.27.4.2	16-May-2009	yamt	sync with head
1.27.4.1	04-May-2009	yamt	sync with head.
1.33.8.1	08-Feb-2011	bouyer	Sync with HEAD
1.33.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.33.4.1	05-Mar-2011	rmind	sync with head
1.36.16.1	18-May-2014	rmind	sync with head
1.36.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.36.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.30.1	10-Jun-2019	christos	Sync with HEAD
1.37.28.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.43	26-Feb-2025	andvar	Fix typos in comments, mainly s/calcurate/calculate/.
1.42	29-Jun-2024	riastradh	branches: 1.42.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.41	20-Dec-2023	skrll	Remove unnecssary #include
1.40	05-Nov-2023	jdolecek	ena(4): stop management first when detaching Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.39	05-Nov-2023	jdolecek	ena(4): establish interrupt after setting up resources Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.38	05-Nov-2023	jdolecek	ena(4): prevent AENQ handler from use-after-free Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.37	05-Nov-2023	jdolecek	ena(4) is MP-ready, always use MPSAFE Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.36	05-Nov-2023	jdolecek	ena(4): support RSS and delete FreeBSD-specified code Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.35	05-Nov-2023	jdolecek	ena(4): replace malloc(9) to kmem(9) Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.34	21-Sep-2023	msaitoh	s/ for for / for / in comment.
1.33	23-May-2022	rin	branches: 1.33.4; Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.32	23-Sep-2021	jmcneill	ena: fix packet reordering issue A reorder occures when - memory allocation fails in bus_dmamap_load_mbuf() or - submission queue is full This patch makes ena(4) to - allocate memory in advance (BUS_DMA_ALLOCNOW flag in bus_dmamap_create()) - check if the queue is vacant before pcq_get() Patch from KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.31	16-Sep-2021	andvar	fix typos in word "successfully", mainly s/succesfully/successfully/.
1.30	24-Jul-2021	andvar	Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889. Also fixed some additional typos in comments, found on review of same files or typos.
1.29	19-Jul-2021	jmcneill	Various ena(4) bug fixes, from KUSABA Takeshi <t-kusaba@iij.ad.jp>: [PATCH 01/32] include device_xname for evcnt group. [PATCH 02/32] fix improper NULL check. [PATCH 03/32] add tx drop counter [PATCH 04/32] remove unimplemented m_getjcl [PATCH 05/32] make ENA_MEM_{ALLOC,FREE}_COHERENT symmetric. [PATCH 06/32] disestablish the correct interrupt. [PATCH 07/32] fix null check target. [PATCH 08/32] use if_initialize() and if_register() instead of [PATCH 09/32] free all pci-related resource. [PATCH 10/32] no need to call if_free(), but to call if_detach [PATCH 11/32] add some locking assertions. [PATCH 12/32] use bus_size_t for bus_dma instead of uint32_t. [PATCH 13/32] no need to pass interlock, ena_timer_service does not [PATCH 14/32] enable ena(4) to down I/F [PATCH 15/32] destroy I/O queues before disestablishing msix [PATCH 16/32] count input/output packet stats. [PATCH 17/32] fix missing #ifdef LRO, NetBSD does not support LRO [PATCH 18/32] MP-ify TX, allocate mbuf queue to each TX ring. [PATCH 19/32] down the interface first when to detach, to prevent [PATCH 20/32] default link speed should be unknown. [PATCH 21/32] protect ena_adapter members only by "global_mtx". [PATCH 22/32] lower global_mtx interrupt level [PATCH 23/32] reorder function declaration [PATCH 24/32] add locking notes and some marking. [PATCH 25/32] process RX in workqueue context, as same as FreeBSD [PATCH 26/32] count rx_drops correctly. [PATCH 27/32] no need to lock when attach/detach, down/up [PATCH 28/32] fix memory leak. [PATCH 29/32] add "stopping" flag to ena_ring. [PATCH 30/32] make the flags atomic. [PATCH 31/32] do not schedule timer when device is down. [PATCH 32/32] no need to start timer if I/F is down.
1.28	01-Jul-2021	jmcneill	port-evbarm/56274: no network on ec2 arm64 9.99.85 Remove custom SIOCSIFFLAGS handling and rely on ether_ioctl to DTRT
1.27	23-Jan-2021	jmcneill	branches: 1.27.4; kern/55942: destroyed ena(4) evcnts cause panic Fix event counter teardown code. From KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.26	18-Jan-2021	skrll	Trailing whitespace
1.25	03-Aug-2020	jmcneill	branches: 1.25.2; port-arm/55532: kernel panic with ena on AWS a1.2xlarge Do not mark callout and workqueues as mpsafe unless the NET_MPSAFE option is present.
1.24	03-Mar-2020	jdolecek	replace #if condition for irq_slot with __diagused on the cleanup path fixes PR kern/55044 by Uwe Toenjes
1.23	24-Feb-2020	rin	0x%p --> %p for non-external codes.
1.22	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.21	04-Feb-2020	thorpej	Use ifmedia_fini().
1.20	01-Feb-2020	riastradh	Use atomic_load/store_* in ena(4), not questionable membar_sync.
1.19	02-Dec-2019	msaitoh	branches: 1.19.2; Use PCI_MSIX_"TBL"BIR_MASK instead of PCI_MSIX_"PBA"BIR_MASK for MSI-X table. This is not a real bug because both macros have the same value.
1.18	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.17	18-Oct-2019	msaitoh	s/initalize/initialize/ in comment or printf message.
1.16	01-Aug-2019	msaitoh	If the memory space's BAR and the MSI-X table is on the same BAR, reserve the space. Almost the same as nvme_pci.c and xhci_pci.c. Tested by ryoon@.
1.15	29-May-2019	msaitoh	branches: 1.15.2; Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.14	26-Apr-2019	msaitoh	Remove duplicated inclusion of sys/bus.h.
1.13	23-Jan-2019	ryoon	Fix no options DEBUG nor DIAGNOSTIC build
1.12	22-Dec-2018	roy	remove needless parens, makes clang happy
1.11	30-Nov-2018	jmcneill	Get this driver into a functional state.
1.10	30-Nov-2018	jmcneill	workqueue and callout fixes
1.9	28-Nov-2018	jmcneill	Add calls to if_attach and if_deferred_start_init
1.8	28-Nov-2018	jmcneill	Our softc is the private data of self, not parent. Spotted by mlelstv@
1.7	28-Nov-2018	jdolecek	whitespace fix, NFC
1.6	28-Nov-2018	bad	Adapt ena_rx_checksum() to NetBSD. It wasn't ported to the NetBSD conventions of indicating hardware checkum status. Compile tested only.
1.5	26-Jun-2018	msaitoh	branches: 1.5.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.4	16-Jun-2018	jdolecek	make ena(4) compile, with best efford exercised to convert code to NetBSD equivalents; where possible left the original code intact to make it easier to compare against FreeBSD original made non-working stubs for some functions which don't have direct equivalent yet - this includes all of <sys/buf_ring.h>, m_getjcl(), m_append, and m_collapse(); these need to be adressed left XXX/TODO in if_enavar.h, colocated with the non-working stubs
1.3	01-Jun-2018	maxv	Fix M_PKTHDR use in if_alc, if_age and if_ena. if_alc and if_age always put in _rxhead a M_PKTHDR-flagged mbuf, so the flag must always be present. Instead of manually adding the flag, add a KASSERT to ensure it is already there. If it weren't, there would be memory corruptions. Same in if_ena, but this one does not compile so we don't really care. Also, use m_remove_pkthdr to remove the flag, instead of doing it manually. This ensures the tags get freed (even though these drivers don't seem to be using mtags).
1.2	19-May-2018	jdolecek	branches: 1.2.2; Intermediate changes to make it further into actually at least compile, not nearly yet complete. Committed so that others can possibly pick it off, as discussed on tech-userlevel@ with Martin
1.1	19-May-2018	jdolecek	branches: 1.1.1; Initial revision
1.1.1.1	19-May-2018	jdolecek	Import source for FreeBSD Amazon Elastic Network Adapter (ENA) NIC driver for reference. Needs a lot of work to port over. Remapped filenames from FreeBSD to NetBSD structure: sys/dev/ena/ena.c -> sys/dev/pci/if_ena.c sys/dev/ena/ena.h -> sys/dev/pci/if_enavar.h ena_sysctl.* not imported, if needed later will be merged into if_ena.c
1.2.2.6	26-Jan-2019	pgoyette	Sync with HEAD
1.2.2.5	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.2.2.4	28-Jul-2018	pgoyette	Sync with HEAD
1.2.2.3	25-Jun-2018	pgoyette	Sync with HEAD
1.2.2.2	21-May-2018	pgoyette	Sync with HEAD
1.2.2.1	19-May-2018	pgoyette	file if_ena.c was added on branch pgoyette-compat on 2018-05-21 04:36:06 +0000
1.5.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.5.2.1	10-Jun-2019	christos	Sync with HEAD
1.15.2.4	23-Jan-2021	martin	Pull up following revision(s) (requested by jmcneill in ticket #1186): sys/dev/pci/if_ena.c: revision 1.27 PR kern/55942: destroyed ena(4) evcnts cause panic Fix event counter teardown code. From KUSABA Takeshi
1.15.2.3	05-Aug-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #1038): sys/dev/pci/if_ena.c: revision 1.25 PR port-arm/55532: kernel panic with ena on AWS a1.2xlarge Do not mark callout and workqueues as mpsafe unless the NET_MPSAFE option is present.
1.15.2.2	08-Mar-2020	martin	Pull up following revision(s) (requested by jdolecek in ticket #759): sys/dev/pci/if_ena.c: revision 1.24 replace #if condition for irq_slot with __diagused on the cleanup path fixes PR kern/55044 by Uwe Toenjes
1.15.2.1	09-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #37): doc/3RDPARTY: revision 1.1641 sys/dev/pci/if_ena.c: revision 1.16 sys/arch/amd64/conf/GENERIC: revision 1.532 If the memory space's BAR and the MSI-X table is on the same BAR, reserve the space. Almost the same as nvme_pci.c and xhci_pci.c. Tested by ryoon@. Add ena(4).
1.19.2.1	29-Feb-2020	ad	Sync with head.
1.25.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.27.4.1	01-Aug-2021	thorpej	Sync with HEAD.
1.33.4.1	26-Nov-2023	bouyer	Pull up following revision(s) (requested by jdolecek in ticket #460): sys/dev/pci/if_ena.c: revision 1.35 sys/dev/pci/if_ena.c: revision 1.36 sys/dev/pci/if_ena.c: revision 1.37 sys/dev/pci/if_ena.c: revision 1.38 sys/dev/pci/if_ena.c: revision 1.39 sys/external/bsd/ena-com/ena_plat.h: revision 1.10 sys/dev/pci/if_enavar.h: revision 1.9 sys/external/bsd/ena-com/ena_com.c: revision 1.2 sys/external/bsd/ena-com/ena_com.c: revision 1.3 sys/external/bsd/ena-com/ena_com.c: revision 1.4 sys/dev/pci/if_ena.c: revision 1.40 sys/external/bsd/ena-com/ena_com.h: revision 1.2 ena(4): replace malloc(9) to kmem(9) Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): prevent AENQ handler from use-after-free Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): destroy all wait_event Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): support RSS and delete FreeBSD-specified code Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4) is MP-ready, always use MPSAFE Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): establish interrupt after setting up resources Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): stop management first when detaching Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.42.2.1	02-Aug-2025	perseant	Sync with HEAD
1.10	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.9	05-Nov-2023	jdolecek	ena(4): support RSS and delete FreeBSD-specified code Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.8	19-Jul-2021	jmcneill	branches: 1.8.10; Various ena(4) bug fixes, from KUSABA Takeshi <t-kusaba@iij.ad.jp>: [PATCH 01/32] include device_xname for evcnt group. [PATCH 02/32] fix improper NULL check. [PATCH 03/32] add tx drop counter [PATCH 04/32] remove unimplemented m_getjcl [PATCH 05/32] make ENA_MEM_{ALLOC,FREE}_COHERENT symmetric. [PATCH 06/32] disestablish the correct interrupt. [PATCH 07/32] fix null check target. [PATCH 08/32] use if_initialize() and if_register() instead of [PATCH 09/32] free all pci-related resource. [PATCH 10/32] no need to call if_free(), but to call if_detach [PATCH 11/32] add some locking assertions. [PATCH 12/32] use bus_size_t for bus_dma instead of uint32_t. [PATCH 13/32] no need to pass interlock, ena_timer_service does not [PATCH 14/32] enable ena(4) to down I/F [PATCH 15/32] destroy I/O queues before disestablishing msix [PATCH 16/32] count input/output packet stats. [PATCH 17/32] fix missing #ifdef LRO, NetBSD does not support LRO [PATCH 18/32] MP-ify TX, allocate mbuf queue to each TX ring. [PATCH 19/32] down the interface first when to detach, to prevent [PATCH 20/32] default link speed should be unknown. [PATCH 21/32] protect ena_adapter members only by "global_mtx". [PATCH 22/32] lower global_mtx interrupt level [PATCH 23/32] reorder function declaration [PATCH 24/32] add locking notes and some marking. [PATCH 25/32] process RX in workqueue context, as same as FreeBSD [PATCH 26/32] count rx_drops correctly. [PATCH 27/32] no need to lock when attach/detach, down/up [PATCH 28/32] fix memory leak. [PATCH 29/32] add "stopping" flag to ena_ring. [PATCH 30/32] make the flags atomic. [PATCH 31/32] do not schedule timer when device is down. [PATCH 32/32] no need to start timer if I/F is down.
1.7	23-Dec-2018	jmcneill	branches: 1.7.16; Use ___STRING instead of __STRING to expand driver version numbers properly
1.6	30-Nov-2018	jmcneill	Get this driver into a functional state.
1.5	28-Nov-2018	jmcneill	Use correct PCI BAR offsets
1.4	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.3	16-Jun-2018	jdolecek	branches: 1.3.2; make ena(4) compile, with best efford exercised to convert code to NetBSD equivalents; where possible left the original code intact to make it easier to compare against FreeBSD original made non-working stubs for some functions which don't have direct equivalent yet - this includes all of <sys/buf_ring.h>, m_getjcl(), m_append, and m_collapse(); these need to be adressed left XXX/TODO in if_enavar.h, colocated with the non-working stubs
1.2	19-May-2018	jdolecek	branches: 1.2.2; Intermediate changes to make it further into actually at least compile, not nearly yet complete. Committed so that others can possibly pick it off, as discussed on tech-userlevel@ with Martin
1.1	19-May-2018	jdolecek	branches: 1.1.1; Initial revision
1.1.1.1	19-May-2018	jdolecek	Import source for FreeBSD Amazon Elastic Network Adapter (ENA) NIC driver for reference. Needs a lot of work to port over. Remapped filenames from FreeBSD to NetBSD structure: sys/dev/ena/ena.c -> sys/dev/pci/if_ena.c sys/dev/ena/ena.h -> sys/dev/pci/if_enavar.h ena_sysctl.* not imported, if needed later will be merged into if_ena.c
1.2.2.5	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.2.2.4	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.2.2.3	25-Jun-2018	pgoyette	Sync with HEAD
1.2.2.2	21-May-2018	pgoyette	Sync with HEAD
1.2.2.1	19-May-2018	pgoyette	file if_enavar.h was added on branch pgoyette-compat on 2018-05-21 04:36:06 +0000
1.3.2.1	10-Jun-2019	christos	Sync with HEAD
1.7.16.1	01-Aug-2021	thorpej	Sync with HEAD.
1.8.10.1	26-Nov-2023	bouyer	Pull up following revision(s) (requested by jdolecek in ticket #460): sys/dev/pci/if_ena.c: revision 1.35 sys/dev/pci/if_ena.c: revision 1.36 sys/dev/pci/if_ena.c: revision 1.37 sys/dev/pci/if_ena.c: revision 1.38 sys/dev/pci/if_ena.c: revision 1.39 sys/external/bsd/ena-com/ena_plat.h: revision 1.10 sys/dev/pci/if_enavar.h: revision 1.9 sys/external/bsd/ena-com/ena_com.c: revision 1.2 sys/external/bsd/ena-com/ena_com.c: revision 1.3 sys/external/bsd/ena-com/ena_com.c: revision 1.4 sys/dev/pci/if_ena.c: revision 1.40 sys/external/bsd/ena-com/ena_com.h: revision 1.2 ena(4): replace malloc(9) to kmem(9) Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): prevent AENQ handler from use-after-free Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): destroy all wait_event Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): support RSS and delete FreeBSD-specified code Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4) is MP-ready, always use MPSAFE Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): establish interrupt after setting up resources Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp> ena(4): stop management first when detaching Code contributed by KUSABA Takeshi <t-kusaba@iij.ad.jp>
1.55	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.54	23-Jun-2018	maxv	branches: 1.54.2; constify
1.53	29-Mar-2014	christos	branches: 1.53.28; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.52	26-Jul-2011	dyoung	branches: 1.52.2; 1.52.12; 1.52.16; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.51	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.50	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.49	27-Aug-2008	christos	branches: 1.49.8; device_t changes, pmf, minor knf.
1.48	28-Apr-2008	martin	branches: 1.48.2; 1.48.6; Remove clause 3 and 4 from TNF licenses
1.47	10-Apr-2008	cegger	branches: 1.47.2; 1.47.4; use aprint_*_dev and device_xname
1.46	19-Oct-2007	ad	branches: 1.46.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.45	16-Nov-2006	christos	branches: 1.45.8; 1.45.22; 1.45.24; 1.45.28; __unused removal on arguments; approved by core.
1.44	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.43	11-Dec-2005	christos	branches: 1.43.20; 1.43.22; merge ktrace-lwp.
1.42	27-Feb-2005	perry	branches: 1.42.4; nuke trailing whitespace
1.41	21-Aug-2004	thorpej	branches: 1.41.4; 1.41.6; Use ANSI function decls and make use of static.
1.40	31-Jan-2003	thorpej	branches: 1.40.2; Use aprint_*().
1.39	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.38	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.37	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.36	13-Nov-2001	lukem	add RCSID
1.35	08-Jul-2001	thorpej	branches: 1.35.2; Remove unnecessary include files. From Onno van der Linden.
1.34	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.33	28-Dec-2000	sommerfeld	branches: 1.33.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.32	19-Feb-1999	thorpej	branches: 1.32.8; Improve card descriptions.
1.31	09-Nov-1998	thorpej	Const poisoning.
1.30	07-Nov-1998	fvdl	Rely on priority match to resolve conflict with the ex driver, not on a define.
1.29	04-Nov-1998	fvdl	Add the 'ex' driver, a DMA driver for 3Com 90x and 90xB cards. Rename constants from EP_ to ELINK_ since they're now used in the ex driver as well.
1.28	12-Aug-1998	thorpej	Completely rewrite media selection. Add support for the MII/PHY on the 3Com 3c905-TX and 3c905-T4. Fix handling of the MII on the 3c595-MII and 3c597-MII (can't talk to the PHYs on these cards; just use "manual" for the external MII port). Fixes kern/4782 (Chris Demetriou).
1.27	05-Jul-1998	jonathan	defopt NS, NSIP.
1.26	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.25	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.24	14-Oct-1997	thorpej	Pull down from marc-pcmcia branch: Minor tweaks to deal with PCMCIA differences, and support for power management.
1.23	13-Apr-1997	cgd	branches: 1.23.4; rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.22	13-Apr-1997	cgd	use pci_map_register().
1.21	30-Mar-1997	jonathan	Add ifmedia hooks: include <ifmedia.h> and add sc_media to softc.
1.20	30-Mar-1997	jonathan	Match missing variants: 3c900 twisted-pair only, 3c595 mii, 3c595 100bT4, 3c905 100bT4. Is there an MII version of the 3c905?
1.19	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.18	18-Feb-1997	jonathan	Move if_ep media-sense code from front ends to inside back end, to prepare for if_media, and to support EISA 3c59x cards: * change epconfig() to take a short that encodes the chipset type (3c509, or Demon/Vortex/Boomerang). * add distinct 3c509 and Demon/Vortex/Boomerang media-sense functions to back end. * Add EISA match/attach support for the 3c592 and 3c597 (Demon) cards as well as the 3c509-comatibl EISA 3c509/3c579. * Assume that ISA and ISA PnP cards are 3c509-style, until proof to the contrary (e.g. large packet support). Tested on 3c595, 3c590, and 3c509b. Not tested on 3c515 or Demon.
1.17	31-Dec-1996	jonathan	branches: 1.17.4; Use the same symbolic cnames for 100Mbit media as epconfig().
1.16	30-Dec-1996	jonathan	* make sure 100mbit media-present bits don't collied with 3c509 CONFIG_CNTRL bits by making the "conn" argument to epconfig() a u_int, not a u_int16_t, and by defining 100mbit media bits in the upper 16 bits. A redesigned interface that fixes this properly is in the pipeline.
1.15	29-Dec-1996	jonathan	* Accept 100baseT4, MII variants of 3c595 as 3c595-model devices, rather than unknown. * Look for 100mbit interfaces, and for any present, set corresponding bits in `conn' argument to epconfig(). epconfig() will now print 100Mbit media by name.
1.14	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.13	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.12	15-Oct-1996	christos	Use the official 3com part names; from Curt Sampson
1.11	13-Oct-1996	christos	Apply PR/2840: Recognize more 3com models...
1.10	13-Oct-1996	christos	backout kprintf changes
1.9	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.8	29-Sep-1996	christos	Don't epstop() at the and of the attach
1.7	13-May-1996	mycroft	branches: 1.7.4; Use intr.h.
1.6	12-May-1996	thorpej	Value returned by pci_conf_read() is a pcireg_t, so don't try to store it in a u_short.
1.5	12-May-1996	thorpej	Don't use home-grown macros to extract PCI vendor and product IDs. Also, be truthful; We're actually looking for Configuration Base IO Address, not Configuration Base Memory Address.
1.4	10-May-1996	thorpej	Update Herb's e-mail address; beer.org feels a lot more friendly.
1.3	03-May-1996	christos	- include <sys/systm.h> - remove unused variables
1.2	30-Apr-1996	thorpej	Convert the Etherlink III (ep) driver to use <machine/bus.h>.
1.1	25-Apr-1996	thorpej	Break up the ep driver into isa, eisa, and pci front-ends and a bus-independent core driver. Tested on all three bus types, including an isa 3c509 masquerading as an eisa device (use ep* at eisa? slot ? in your kernel config file to catch this one). XXX Driver still needs to be converted to <machine/bus.h>
1.7.4.2	04-Mar-1997	mycroft	Recognize 3c900 and 3c905 cards.
1.7.4.1	04-Mar-1997	mycroft	Update for new pcidevs.
1.17.4.3	09-Mar-1997	is	netinet/if_ether.h -> netinet/if_inarp.h
1.17.4.2	20-Feb-1997	is	Merged in recent changes from Trunk.
1.17.4.1	07-Feb-1997	is	Snapshot of new ARP code. Our old ARP code was hardwired for 6-byte length medium addresses, while the protocol is designed for any size. This snapshot contains a first hack at getting rid of Ethernet specific data structures. The ep driver is updated (and tested on the PCI bus), the iy and fpa drivers have been updated, but not real life tested yet. If you want to test this with other drivers, you have to update them first yourself, and probably tag the relevant directories. Better contact me if you want to do this.
1.23.4.2	27-Sep-1997	marc	initialize enable/disable hooks to NULL so the bus independent code dtrt.
1.23.4.1	30-Jul-1997	marc	added pcmcia infrastructure and a few devices
1.32.8.1	05-Jan-2001	bouyer	Sync with HEAD
1.33.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.33.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.33.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.33.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.35.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.35.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.40.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.40.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.40.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.40.2.1	25-Aug-2004	skrll	Sync with HEAD.
1.41.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.41.4.1	29-Apr-2005	kent	sync with -current
1.42.4.2	27-Oct-2007	yamt	sync with head.
1.42.4.1	30-Dec-2006	yamt	sync with head.
1.43.22.2	10-Dec-2006	yamt	sync with head.
1.43.22.1	22-Oct-2006	yamt	sync with head
1.43.20.1	18-Nov-2006	ad	Sync with head.
1.45.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.45.24.1	06-Nov-2007	matt	sync with HEAD
1.45.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.45.8.1	23-Oct-2007	ad	Sync with head.
1.46.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.46.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.47.4.4	11-Mar-2010	yamt	sync with head
1.47.4.3	16-May-2009	yamt	sync with head
1.47.4.2	04-May-2009	yamt	sync with head.
1.47.4.1	16-May-2008	yamt	sync with head.
1.47.2.1	18-May-2008	yamt	sync with head.
1.48.6.1	19-Oct-2008	haad	Sync with HEAD.
1.48.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.49.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.52.16.1	18-May-2014	rmind	sync with head
1.52.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.52.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.28.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.53.28.1	25-Jun-2018	pgoyette	Sync with HEAD
1.54.2.1	10-Jun-2019	christos	Sync with HEAD
1.43	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.42	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.41	29-Mar-2014	christos	branches: 1.41.28; 1.41.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.40	30-Jan-2012	drochner	branches: 1.40.6; 1.40.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.39	26-Jul-2011	dyoung	branches: 1.39.2; 1.39.6; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.38	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.37	06-Jul-2008	tsutsui	Split device_t/softc with misc cosmetic changes.
1.36	28-Apr-2008	martin	branches: 1.36.2; 1.36.4; Remove clause 3 and 4 from TNF licenses
1.35	21-Mar-2008	dyoung	branches: 1.35.2; 1.35.4; Use aprint_*_dev() instead of accessing struct device member dv_xname directly.
1.34	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.33	19-Oct-2007	ad	branches: 1.33.12; 1.33.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.32	16-Nov-2006	christos	branches: 1.32.8; 1.32.22; 1.32.24; 1.32.28; __unused removal on arguments; approved by core.
1.31	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.30	17-Jun-2006	christos	branches: 1.30.4; 1.30.6; re-factor the pci powestate api. reviewed by gimpy
1.29	11-Dec-2005	christos	branches: 1.29.4; 1.29.8; 1.29.14; 1.29.16; merge ktrace-lwp.
1.28	27-Feb-2005	perry	branches: 1.28.4; nuke trailing whitespace
1.27	21-Aug-2004	thorpej	branches: 1.27.4; 1.27.6; Use ANSI function decls and make use of static.
1.26	31-Jan-2003	thorpej	branches: 1.26.2; Use aprint_*().
1.25	23-Dec-2002	tsutsui	Always print the device name first in epic_pci_attach().
1.24	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.23	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.22	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.21	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.20	15-Jul-2002	drochner	recognize a SMC9432BTX from PCI subsys id and tell the bus independant backend that it bas a BNC port (just like the SMC9432BTX1)
1.19	13-Nov-2001	lukem	branches: 1.19.8; add RCSID
1.18	08-Jul-2001	thorpej	branches: 1.18.2; Remove unnecessary include files. From Onno van der Linden.
1.17	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.16	17-May-2001	drochner	check PCI subsystem ID for a SMC9432BTX1 or SMC9432FTX and tell the common driver about their media support (BNC and 100FX, respectively)
1.15	28-Dec-2000	sommerfeld	branches: 1.15.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.14	17-Jul-2000	tron	Don't make assumptions of the actual value of PCI_PMCSR_STATE_D0. Problem noted by John Hawkinson.
1.13	17-Jul-2000	tron	Don't clobber bits from 2 to 31 when trying to change the power state. Problem noted by John Hawkinson.
1.12	16-Jul-2000	tron	If card is in power state D3 put it into power state D0 so that it will at least work after the next reboot.
1.11	15-Jul-2000	tron	Avoid printing duplicate colon when card is waked up.
1.10	15-Jul-2000	tron	Un-__P'ify kernel prototypes.
1.9	15-Jul-2000	tron	Improve kernel message when card is in power state D3. Instead of... epic0 at pci0 dev 11 function 0 epic0: unable to wake up from ... ... print this: epic0 at pci0 dev 11 function 0: unable to wake up from power state D3
1.8	14-Jul-2000	tron	Check if card was put into sleep mode using PCI/ACPI power management and try to restart it. This patch based on hints by Jason Thorpe.
1.7	27-Jul-1999	thorpej	branches: 1.7.2; 1.7.12; Add support for the SMC 83c175. Also, print out the chip rev when we attach.
1.6	24-Mar-1999	thorpej	branches: 1.6.4; Make sure bus mastering is enabled.
1.5	11-Aug-1998	thorpej	Adapt to new MII code.
1.4	05-Jul-1998	jonathan	defopt NS, NSIP.
1.3	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.2	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.1	02-Jun-1998	thorpej	Device driver for the SMC 83c170 Ethernet PCI Integrated Controller (EPIC/100) used in the SMC EtherPower II. Media control isn't yet supported, due to some MII infrastructure problems which I hope to address soon. This isn't a huge deal, since the PHY defaults to auto-negotiate mode. Also, the device just programs the multicast hash table to accept all multicast, to avoid a hardware bug that causes the multicast address filter to lose in 10Mb/s mode. This bug will be fixed in a more sane way once the media control issues are dealt with.
1.6.4.1	02-Aug-1999	thorpej	Update from trunk.
1.7.12.4	16-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.12 syssrc/sys/dev/pci/if_tlp_pci.c, 1.43 If card is in power state D3 put it into power state D0 so that it will at least work after the next reboot.
1.7.12.3	15-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.11 syssrc/sys/dev/pci/if_tlp_pci.c, 1.42 Avoid printing duplicate colon when card is waked up.
1.7.12.2	15-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.9 syssrc/sys/dev/pci/if_tlp_pci.c, 1.41 Improve kernel message when card is in power state D3. Instead of... epic0 at pci0 dev 11 function 0 epic0: unable to wake up from ... ... print this: epic0 at pci0 dev 11 function 0: unable to wake up from power state D3
1.7.12.1	14-Jul-2000	tron	Pull up revision 1.8 (approved by thorpej): Check if card was put into sleep mode using PCI/ACPI power management and try to restart it. This patch based on hints by Jason Thorpe.
1.7.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.7.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.15.2.6	29-Dec-2002	thorpej	Sync with HEAD.
1.15.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.15.2.4	01-Aug-2002	nathanw	Catch up to -current.
1.15.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.15.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.15.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.18.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.18.2.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.18.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.8.1	20-Jul-2002	gehenna	catch up with -current.
1.26.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.26.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.26.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.26.2.1	25-Aug-2004	skrll	Sync with HEAD.
1.27.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.27.4.1	29-Apr-2005	kent	sync with -current
1.28.4.4	24-Mar-2008	yamt	sync with head.
1.28.4.3	27-Oct-2007	yamt	sync with head.
1.28.4.2	30-Dec-2006	yamt	sync with head.
1.28.4.1	21-Jun-2006	yamt	sync with head.
1.29.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.29.14.1	19-Jun-2006	chap	Sync with head.
1.29.8.1	26-Jun-2006	yamt	sync with head.
1.29.4.1	09-Sep-2006	rpaulo	sync with head
1.30.6.2	10-Dec-2006	yamt	sync with head.
1.30.6.1	22-Oct-2006	yamt	sync with head
1.30.4.1	18-Nov-2006	ad	Sync with head.
1.32.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.32.24.2	23-Mar-2008	matt	sync with HEAD
1.32.24.1	06-Nov-2007	matt	sync with HEAD
1.32.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.32.8.1	23-Oct-2007	ad	Sync with head.
1.33.16.3	28-Sep-2008	mjf	Sync with HEAD.
1.33.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.33.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.33.12.1	24-Mar-2008	keiichi	sync with head.
1.35.4.3	11-Mar-2010	yamt	sync with head
1.35.4.2	04-May-2009	yamt	sync with head.
1.35.4.1	16-May-2008	yamt	sync with head.
1.35.2.1	18-May-2008	yamt	sync with head.
1.36.4.1	18-Jul-2008	simonb	Sync with head.
1.36.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.39.6.1	18-Feb-2012	mrg	merge to -current.
1.39.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.39.2.1	17-Apr-2012	yamt	sync with head
1.40.10.1	18-May-2014	rmind	sync with head
1.40.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.41.30.1	10-Jun-2019	christos	Sync with HEAD
1.41.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.4	19-Dec-2023	skrll	Remove unnecessary #include
1.3	31-Oct-2023	msaitoh	branches: 1.3.2; eqos_pci: Limit to 32bit DMA only for PSE devices.
1.2	26-Oct-2023	msaitoh	eqos(4): Set TX/RX DMA burst length to improve performance.
1.1	20-Oct-2023	msaitoh	eqos(4): Add initial support for Intel Elkhart Lake internal Ethernet devices. - Only tested on PSE SGMII 1G Ethernet MAC with MaxLinear GPY115. - I don't know why dmat64 doesn't work. eqos_attach() have a special code if EQOS_HW_FEATURE_ADDR64_32BIT(sc) is true, but it seems it doesn't work. - TODO: Multiqueue support. Detach support.
1.3.2.2	03-Nov-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #446): sys/dev/pci/if_eqos_pci.c: revision 1.3 sys/arch/i386/conf/GENERIC: revision 1.1251 sys/arch/i386/conf/GENERIC: revision 1.1252 sys/arch/amd64/conf/GENERIC: revision 1.607 sys/arch/amd64/conf/GENERIC: revision 1.608 sys/dev/ic/dwc_eqos.c: revision 1.20 sys/dev/ic/dwc_eqos.c: revision 1.21 share/man/man4/eqos.4: revision 1.2 sys/dev/ic/dwc_eqos.c: revision 1.22 sys/dev/ic/dwc_eqos_reg.h: revision 1.7 sys/dev/ic/dwc_eqos.c: revision 1.23 sys/dev/ic/dwc_eqos_reg.h: revision 1.8 sys/dev/ic/dwc_eqos.c: revision 1.24 sys/dev/ic/dwc_eqos.c: revision 1.25 sys/dev/ic/dwc_eqos.c: revision 1.26 sys/dev/ic/dwc_eqos.c: revision 1.27 sys/dev/ic/dwc_eqos_var.h: revision 1.5 sys/dev/ic/dwc_eqos.c: revision 1.28 sys/dev/ic/dwc_eqos_var.h: revision 1.6 sys/dev/ic/dwc_eqos.c: revision 1.29 sys/dev/ic/dwc_eqos.c: revision 1.18 sys/dev/ic/dwc_eqos.c: revision 1.19 sys/dev/pci/files.pci: revision 1.448 sys/dev/pci/if_eqos_pci.c: revision 1.1 sys/dev/pci/if_eqos_pci.c: revision 1.2 eqos(4): Fix definition of GMAC_MAC_HW_FEATURE1_RXFIFOSIZE. eqos(4): Fix a bug that the MAC address is swapped. Don't swap the MAC address in eqos_get_eaddr(). Other OSes except FreeBSD (which was based on NetBSD's) don't swap it. With this change, my own OnLogic Helix 330's MAC address becomes correct. The OUI is 84:8b:cd:4d. It's owned by Logic Supply and they were acquired by OnLogic. On Quartz64 with UEFI, the MAC address is wrongly set and the multicast bit might be set. To do workaround, clear the bit if it's set. eqos(4): Add missing clock range. eqos(4): Accept if snpsver == 0x52. Tested with Intel Elkhart Lake. TODO: Multiqueue support. Add watchdog timer. Add detach function. eqos(4): Add initial support for Intel Elkhart Lake internal Ethernet devices. - Only tested on PSE SGMII 1G Ethernet MAC with MaxLinear GPY115. - I don't know why dmat64 doesn't work. eqos_attach() have a special code if EQOS_HW_FEATURE_ADDR64_32BIT(sc) is true, but it seems it doesn't work. - TODO: Multiqueue support. Detach support. eqos(4): Fix compile error for arch that sizeof(bus_size_t) == 4 (i386). Trailing whitespace eqos(4): Disable eqos(4) by default because it's not stable on x86. eqos(4): KNF. No functional change. eqos(4): Add and modify some DPRINTF()s. eqos(4): Add sysctls for debugging. eqos(4): Use EQOS_TXLOCK() more to be stable. Fix a bug that sc_tx.{cur,next,queued} become inconsitent. Use txlock when accessing TX data. eqos(4): Set TX/RX DMA burst length to improve performance. eqos(4): Set flow control correctly. eqos_pci: Limit to 32bit DMA only for PSE devices. eqos(4): Fix typo in comment.
1.3.2.1	31-Oct-2023	martin	file if_eqos_pci.c was added on branch netbsd-10 on 2023-11-03 08:56:36 +0000
1.34	19-Jan-2020	thorpej	Remove HIPPI support and the esh(4) driver that uses it. There have not been any users of HIPPI for some time, and it is unlikely to be resurrected.
1.33	09-Dec-2018	jdolecek	branches: 1.33.6; use pci_intr_establish_xname() everywhere
1.32	07-Jul-2016	msaitoh	branches: 1.32.16; 1.32.18; KNF. Remove extra spaces. No functional change.
1.31	29-Mar-2014	christos	branches: 1.31.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.30	27-Oct-2012	chs	branches: 1.30.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.29	26-Jul-2011	dyoung	branches: 1.29.2; 1.29.12; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.28	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.27	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.26	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.25	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.24	28-Apr-2008	martin	branches: 1.24.14; Remove clause 3 and 4 from TNF licenses
1.23	10-Apr-2008	cegger	branches: 1.23.2; 1.23.4; use aprint_*_dev and device_xname
1.22	19-Oct-2007	ad	branches: 1.22.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.21	16-Nov-2006	christos	branches: 1.21.8; 1.21.22; 1.21.24; 1.21.28; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	11-Dec-2005	christos	branches: 1.19.20; 1.19.22; merge ktrace-lwp.
1.18	30-May-2005	christos	branches: 1.18.2; - const poisoning - avoid variable shadowing.
1.17	27-Feb-2005	perry	nuke trailing whitespace
1.16	28-Oct-2004	yamt	branches: 1.16.4; 1.16.6; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.15	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.14	31-Jan-2003	thorpej	branches: 1.14.2; Use aprint_*().
1.13	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	13-Nov-2001	lukem	add RCSID
1.9	08-Jul-2001	thorpej	branches: 1.9.2; Remove unnecessary include files. From Onno van der Linden.
1.8	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.7	28-Dec-2000	sommerfeld	branches: 1.7.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.6	21-Jan-2000	thorpej	Update for sys/buf.h/disksort_*() changes.
1.5	05-Jul-1998	jonathan	branches: 1.5.14; defopt NS, NSIP.
1.4	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.3	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.2	17-May-1998	kml	Correct copyright date.
1.1	14-May-1998	kml	Driver for Essential Communications' RoadRunner HIPPI (800 Mb/sec network) card. With some modification, this could probably also work for their Gigabit Ethernet card based on the same chipset...
1.5.14.2	05-Jan-2001	bouyer	Sync with HEAD
1.5.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.7.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.9.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.9.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.14.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.14.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.1	25-Aug-2004	skrll	Sync with HEAD.
1.16.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.16.4.1	29-Apr-2005	kent	sync with -current
1.18.2.2	27-Oct-2007	yamt	sync with head.
1.18.2.1	30-Dec-2006	yamt	sync with head.
1.19.22.2	10-Dec-2006	yamt	sync with head.
1.19.22.1	22-Oct-2006	yamt	sync with head
1.19.20.1	18-Nov-2006	ad	Sync with head.
1.21.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.21.24.1	06-Nov-2007	matt	sync with HEAD
1.21.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.21.8.1	23-Oct-2007	ad	Sync with head.
1.22.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.23.4.3	11-Mar-2010	yamt	sync with head
1.23.4.2	16-May-2009	yamt	sync with head
1.23.4.1	16-May-2008	yamt	sync with head.
1.23.2.1	18-May-2008	yamt	sync with head.
1.24.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.29.12.3	03-Dec-2017	jdolecek	update from HEAD
1.29.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.29.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.29.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.29.2.1	30-Oct-2012	yamt	sync with head
1.30.2.1	18-May-2014	rmind	sync with head
1.31.6.1	09-Jul-2016	skrll	Sync with HEAD
1.32.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.32.18.1	10-Jun-2019	christos	Sync with HEAD
1.32.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.33.6.1	25-Jan-2020	ad	Sync with head.
1.36	01-Aug-2023	andvar	s/diabled/disabled/ in comments.
1.35	10-Sep-2021	jakllsch	if_et.c: nothing in this driver needs netinet includes now, so drop #include "opt_inet.h" and the includes of netinet headers conditioned on it.
1.34	10-Sep-2021	jakllsch	No need to #include "vlan.h", as NVLAN is not used in these drivers now.
1.33	08-May-2021	thorpej	Use pci_compatible_match().
1.32	01-Mar-2020	thorpej	branches: 1.32.10; Use the 64-bit PCI DMA tag if available.
1.31	28-Feb-2020	msaitoh	Move PHY_RESET() in the media change path from if_et.c to etphy.c. I don't know if the reset is required or not. I tested some media transitions without the reset and it worked. It might be OK to remove but Linux does it only in et1011c_config_aneg(). So move the reset to etphy_service(,,MII_MEDIACHG)'s autonego case. This change is also required for future locking changes.
1.30	04-Feb-2020	thorpej	Use ifmedia_fini().
1.29	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.28	28-Nov-2019	maxv	branches: 1.28.2; localify
1.27	12-Oct-2019	msaitoh	Fix a bug that multicast address filter doesn't work correctly. XXX pullup-[789].
1.26	01-Aug-2019	msaitoh	Make et(4) work: - Enabling TX/RX in et_init() will always fail when cable is not plugged in, if this happens, we delay TX/RX enablement until link is up. From FreeBSD. - Modify flow control stuff a little (from FrerBSD). It still doesn't work. - KNF. Part of OpenBSD 1.12.
1.25	31-Jul-2019	msaitoh	Add missing ifioctl_common() for SIOCSIFFLAGS to make if_flags controllable.
1.24	28-May-2019	msaitoh	branches: 1.24.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.23	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.22	22-Apr-2019	msaitoh	On drivers which use MII(4) and have hook SIOC[GS]IFMEDIA which just pass to ifmedia_ioctl(), the hook is not required because ether_ioctl has it (if_ethersubr.c rev. 1.160). These drivers don't return ENETRESET in ifmedia_ioctl(), so no functional change.
1.21	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.20	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.19	22-Dec-2018	maxv	Replace: M_COPY_PKTHDR -> m_copy_pkthdr. No functional change, since the former is a macro to the latter.
1.18	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.17	26-Jun-2018	msaitoh	branches: 1.17.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.16	08-Feb-2018	mrg	branches: 1.16.2; update for GCC 6: - fix an array bounds violation and pass the right address to ether_crc32_be(). (i assume this actually makes et(4) multicast work.)
1.15	29-Jul-2017	riastradh	Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.14	15-Dec-2016	ozaki-r	branches: 1.14.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.13	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.12	10-Jun-2016	ozaki-r	branches: 1.12.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.11	17-Feb-2016	christos	PR/50819: David Binderman: Delete dead code.
1.10	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.9	29-Jun-2015	maxv	Use-after-free. ok christos@ Found by Brainy.
1.8	29-Mar-2014	christos	branches: 1.8.4; 1.8.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.7	30-Mar-2013	christos	branches: 1.7.4; remove trailing whitespace
1.6	08-Nov-2012	dyoung	Use native NetBSD macros instead of compatibility macros. This change compiles, but I don't have an et(4) with which to test.
1.5	29-Oct-2012	msaitoh	Don't use old NBPFILTER macro and use new bpf_mtap() API. It fixes a bug that et(4) can't use bpf.
1.4	22-Jul-2012	matt	branches: 1.4.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.3	30-Jan-2012	drochner	branches: 1.3.2; 1.3.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.2	18-May-2011	dyoung	branches: 1.2.4; 1.2.8; #include <sys/bus.h>, not <machine/bus.h>.
1.1	13-Nov-2010	jnemeth	branches: 1.1.2; 1.1.6; PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) network driver The et(4) driver supports PCI Express Ethernet adapters based on the Agere/LSI ET1310/ET1301 integrated MAC/PHY. The et(4) driver was written by Sepherosa Ziehau for DragonFlyBSD, ported to OpenBSD by Jonathan Gray and subsequently ported to NetBSD by Kaspar Brand.
1.1.6.3	31-May-2011	rmind	sync with head
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	13-Nov-2010	rmind	file if_et.c was added on branch rmind-uvmplock on 2011-03-05 20:53:40 +0000
1.1.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.2.8.1	18-Feb-2012	mrg	merge to -current.
1.2.4.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.3	16-Jan-2013	yamt	sync with (a bit old) head
1.2.4.2	30-Oct-2012	yamt	sync with head
1.2.4.1	17-Apr-2012	yamt	sync with head
1.3.6.1	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1474): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.3.2.2	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1474): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.3.2.1	19-Nov-2012	riz	branches: 1.3.2.1.2; Pull up following revision(s) (requested by msaitoh in ticket #655): sys/dev/pci/if_et.c: revision 1.5 Don't use old NBPFILTER macro and use new bpf_mtap() API. It fixes a bug that et(4) can't use bpf.
1.3.2.1.2.1	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1474): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.4.2.4	03-Dec-2017	jdolecek	update from HEAD
1.4.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.4.2.2	23-Jun-2013	tls	resync from head
1.4.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.7.4.1	18-May-2014	rmind	sync with head
1.8.6.5	28-Aug-2017	skrll	Sync with HEAD
1.8.6.4	05-Feb-2017	skrll	Sync with HEAD
1.8.6.3	09-Jul-2016	skrll	Sync with HEAD
1.8.6.2	19-Mar-2016	skrll	Sync with HEAD
1.8.6.1	22-Sep-2015	skrll	Sync with HEAD
1.8.4.2	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1470): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.8.4.1	04-Aug-2015	snj	branches: 1.8.4.1.2; 1.8.4.1.6; Pull up following revision(s) (requested by maxv in ticket #914): sys/dev/pci/if_et.c: revision 1.9 Use-after-free. ok christos@ Found by Brainy.
1.8.4.1.6.1	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1470): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.8.4.1.2.1	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1470): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.12.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.14.8.4	17-Oct-2019	martin	Pull up the following revision, requested by msaitoh in ticket #1407: sys/dev/pci/if_et.c 1.27 via patch Fix a bug that multicast address filter doesn't work correctly.
1.14.8.3	06-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1329: sys/dev/pci/if_etreg.h 1.2-1.3 sys/dev/pci/if_et.c 1.23,1.25 and 1.26 via patch Make et(4) work: - Add missing ifioctl_common() for SIOCSIFFLAGS to make if_flags controllable. - Enabling TX/RX in et_init() will always fail when cable is not plugged in. If this happens, we delay TX/RX enablement until link is up. - Modify flow control stuff a little. It still doesn't work. - Avoid undefined behavior when reset the chip. - Code cleanup.
1.14.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.14.8.1	05-Aug-2017	snj	Pull up following revision(s) (requested by spz in ticket #183): sys/dev/pci/if_et.c: revision 1.15 Check for MCLGET failure in et_newbuf. From Ilja Van Sprundel.
1.16.2.3	26-Jan-2019	pgoyette	Sync with HEAD
1.16.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.16.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.17.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.17.2.1	10-Jun-2019	christos	Sync with HEAD
1.24.2.2	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #346): sys/dev/pci/if_et.c: revision 1.27 Fix a bug that multicast address filter doesn't work correctly. XXX pullup-[789].
1.24.2.1	09-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #36): sys/dev/pci/if_et.c: revision 1.25 sys/dev/pci/if_et.c: revision 1.26 sys/dev/pci/if_etreg.h: revision 1.2 sys/dev/pci/if_etreg.h: revision 1.3 Avoid undefined behavior when reset the chip. found by kUBSan. Add missing ifioctl_common() for SIOCSIFFLAGS to make if_flags controllable. Make et(4) work: - Enabling TX/RX in et_init() will always fail when cable is not plugged in, if this happens, we delay TX/RX enablement until link is up. From FreeBSD. - Modify flow control stuff a little (from FrerBSD). It still doesn't work. - KNF. Part of OpenBSD 1.12.
1.28.2.1	29-Feb-2020	ad	Sync with head.
1.32.10.1	13-May-2021	thorpej	Sync with HEAD.
1.5	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.4	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.3	01-Aug-2019	msaitoh	Make et(4) work: - Enabling TX/RX in et_init() will always fail when cable is not plugged in, if this happens, we delay TX/RX enablement until link is up. From FreeBSD. - Modify flow control stuff a little (from FrerBSD). It still doesn't work. - KNF. Part of OpenBSD 1.12.
1.2	30-Jul-2019	msaitoh	Avoid undefined behavior when reset the chip. found by kUBSan.
1.1	13-Nov-2010	jnemeth	branches: 1.1.6; 1.1.54; 1.1.62; 1.1.68; PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) network driver The et(4) driver supports PCI Express Ethernet adapters based on the Agere/LSI ET1310/ET1301 integrated MAC/PHY. The et(4) driver was written by Sepherosa Ziehau for DragonFlyBSD, ported to OpenBSD by Jonathan Gray and subsequently ported to NetBSD by Kaspar Brand.
1.1.68.1	09-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #36): sys/dev/pci/if_et.c: revision 1.25 sys/dev/pci/if_et.c: revision 1.26 sys/dev/pci/if_etreg.h: revision 1.2 sys/dev/pci/if_etreg.h: revision 1.3 Avoid undefined behavior when reset the chip. found by kUBSan. Add missing ifioctl_common() for SIOCSIFFLAGS to make if_flags controllable. Make et(4) work: - Enabling TX/RX in et_init() will always fail when cable is not plugged in, if this happens, we delay TX/RX enablement until link is up. From FreeBSD. - Modify flow control stuff a little (from FrerBSD). It still doesn't work. - KNF. Part of OpenBSD 1.12.
1.1.62.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.54.1	06-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1329: sys/dev/pci/if_etreg.h 1.2-1.3 sys/dev/pci/if_et.c 1.23,1.25 and 1.26 via patch Make et(4) work: - Add missing ifioctl_common() for SIOCSIFFLAGS to make if_flags controllable. - Enabling TX/RX in et_init() will always fail when cable is not plugged in. If this happens, we delay TX/RX enablement until link is up. - Modify flow control stuff a little. It still doesn't work. - Avoid undefined behavior when reset the chip. - Code cleanup.
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	13-Nov-2010	rmind	file if_etreg.h was added on branch rmind-uvmplock on 2011-03-05 20:53:40 +0000
1.58	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.57	29-Mar-2014	christos	branches: 1.57.28; 1.57.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	13-Sep-2013	martin	Remove unused variable
1.55	30-Mar-2013	christos	branches: 1.55.4; remove trailing whitespace
1.54	26-Jul-2011	dyoung	branches: 1.54.2; 1.54.12; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.53	22-Mar-2010	dyoung	ex_softc.ex_bustype is written but never read. Just get rid of it.
1.52	22-Mar-2010	dyoung	psc_pwrmgmt_csr_reg was never initialized to the actual offset of the PCI PMCSR! Just delete psc_pwrmgmt_csr_reg and all uses of it. In this way, ex_pci_disable() becomes a no-op, so delete it, too.
1.51	26-Feb-2010	dyoung	branches: 1.51.2; Move the definitions for PCI_BAR0, PCI_BAR1, PCI_BAR2, PCI_BAR3, PCI_BAR4, and PCI_BAR5 to pcireg.h for re-use.
1.50	05-May-2009	cegger	branches: 1.50.2; print autoconf error message with aprint_error
1.49	28-Apr-2008	martin	branches: 1.49.14; Remove clause 3 and 4 from TNF licenses
1.48	14-Apr-2008	cegger	branches: 1.48.2; 1.48.4; - ansify - use POSIX integers
1.47	14-Apr-2008	spz	pretty up if_ex: pmf'ify, device split, and improve printing a bit too. Reviewed and function-tested on if_ex_cardbus by cube (thanks). Function-tested on if_ex_pci by myself.
1.46	10-Apr-2008	cegger	use aprint_*_dev and device_xname
1.45	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.44	19-Oct-2007	ad	branches: 1.44.12; 1.44.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.43	16-Nov-2006	christos	branches: 1.43.8; 1.43.22; 1.43.24; 1.43.28; __unused removal on arguments; approved by core.
1.42	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.41	17-Jun-2006	christos	branches: 1.41.4; 1.41.6; re-factor the pci powestate api. reviewed by gimpy
1.40	11-Dec-2005	christos	branches: 1.40.4; 1.40.8; 1.40.14; 1.40.16; merge ktrace-lwp.
1.39	27-Feb-2005	perry	branches: 1.39.4; nuke trailing whitespace
1.38	24-Feb-2005	martin	Fix the size of psc_regs (0x3c >> 2 is the biggest index used to access it now, so pick 0x40 >> 2). Fixes "Bug 6", reported by Ted Unangst on tech-kern.
1.37	21-Aug-2004	thorpej	branches: 1.37.4; 1.37.6; Use ANSI function decls and make use of static.
1.36	15-Jul-2004	junyoung	Add support for 3Com 3c920B-EMB-WNM Integrated Fast Ethernet, as found on Asus P4R800-VM motherboards. From BL on Korea BSD User Forum (sorry but I don't know his real name).
1.35	05-Jun-2003	dogcow	branches: 1.35.2; 1.35.4; Some models of the 556B don't turn on the transmitter without magic values getting written. Fix from FreeBSD; cf sys/pci/if_xl.c, rev 1.110
1.34	19-Apr-2003	christos	Put back bits of pr 9286 that got lost. Thanks to Onno van der Linden for pointing this.
1.33	12-Apr-2003	christos	PR/9286: Steven Grunza: Win98 warm boot puts 3c905B in state D3 and causes ex driver failure
1.32	31-Jan-2003	thorpej	Use aprint_*().
1.31	16-Jan-2003	gendalia	Add support for 3com 3c905CX-TX found on nForce2 MCP-T.
1.30	23-Dec-2002	tsutsui	Set PCI_PMCSR more properly.
1.29	23-Dec-2002	tsutsui	Save PCI_PMCSR value correctly.
1.28	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.27	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.26	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.25	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.24	01-Jul-2002	thorpej	Also save/restore the PCI_INTERRUPT_REG.
1.23	20-Jun-2002	itojun	more proper fix to ex_disable.
1.22	20-Jun-2002	itojun	make it compile (ex_disable). not sure if it is 100% right.
1.21	20-Jun-2002	thorpej	If the chip supports ACPI power management: * Go to D3 if IFF_UP == 0. * Go to D0 if IFF_UP == 1. Saves battery life on my T21, mmm.
1.20	07-Feb-2002	christos	branches: 1.20.8; Turn on mii on 3C980CTXM
1.19	13-Nov-2001	lukem	add RCSID
1.18	18-Aug-2001	kanaoka	Improve intr_ack code to support for 3c556,3c556B. Some cards don't have an interrupt acknowledge register in PCI config space, so we map and write it, such as CardBus driver does.
1.17	08-Jul-2001	thorpej	branches: 1.17.2; Remove unnecessary include files. From Onno van der Linden.
1.16	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.15	28-Dec-2000	sommerfeld	branches: 1.15.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.14	19-Sep-2000	fvdl	Add support for the 3c555, 556 and 556B MiniPCI cards. Based on the cardbus code and the FreeBSD xl driver.
1.13	07-Aug-2000	billc	Added PCI IDs for 3Com's Home Connect (3c450) and Office Connect (3cSOHO100). They both are register compat with later versions of the 3c905 (tornado/ vortex) series.
1.12	12-May-2000	thorpej	branches: 1.12.4; Treat the 3c980C-TXM like the 3c980 Server Adapter. From Paul J. Lavoie, kern/10037.
1.11	23-Mar-2000	mycroft	Rather than guessing at the location of the PMCSR, use the pointer returned by pci_get_capability(). (This is, after all, a standardized interface...)
1.10	13-Jan-2000	mycroft	Remove EX_CONF_MII for the 3c980. At least 2 people need this for the card to work.
1.9	30-Nov-1999	drochner	-support 3c905B-COMBO - somewhat preliminary: the chip has both an internal MII PHY capable of NWAY (according to the Linux driver) and plain BNC/ transceiver connectors. The shared code can't handle this yet, so leave out MII/NWAY for now. -print out revision number from PCI config header, this has been useful to identify buggy chips
1.8	01-Sep-1999	fvdl	branches: 1.8.2; 1.8.8; Don't assume that an -FX card has MII (because it doesn't).
1.7	15-Jul-1999	ross	Recognize 3c905C.
1.6	29-May-1999	fvdl	Recognize 900B-TPC. From Robert Elz.
1.5	20-Feb-1999	ross	branches: 1.5.2; 1.5.4; 1.5.6; Fix typo.
1.4	19-Feb-1999	thorpej	Improve product descriptions, and add 3c905-FX and 3c980 Server Adapter products.
1.3	09-Nov-1998	thorpej	Make the device lookup table-driven, like if_ep_pci.c.
1.2	07-Nov-1998	drochner	use common code for extended capability (power management) let "match" return 2 (> ep_pci)
1.1	04-Nov-1998	fvdl	Add the 'ex' driver, a DMA driver for 3Com 90x and 90xB cards. Rename constants from EP_ to ELINK_ since they're now used in the ex driver as well.
1.5.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.5.4.2	02-Aug-1999	thorpej	Update from trunk.
1.5.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.5.2.4	11-Sep-2000	he	Pull up revision 1.13 (requested by billc): Recognize 3Com's Home Connect (3c450) and Office Connect (3cSOHO100) Ethernet cards.
1.5.2.3	15-Jan-2000	he	Pull up revision 1.10 (requested by mycroft): Make the 3c980 actually work.
1.5.2.2	29-Aug-1999	he	Pull up revision 1.7: Recognize 3c905C-TX card. (bouyer)
1.5.2.1	22-Jun-1999	perry	pullup 1.5->1.6 (fvdl): recognize another 390xB card
1.8.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.8.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.8.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.12.4.3	01-Jan-2002	he	Pull up revision 1.18 (requested by kanaoka): Improve intr_ack code to support 3c556 and 3c556B.
1.12.4.2	20-Mar-2001	he	Pull up revision 1.14 (requested by chuck): Add support for the 3Com 3c555, 3c556 and 3c556B MiniPCI Ethernet cards.
1.12.4.1	19-Oct-2000	he	Pull up revision 1.13 (requested by billc): Add PCI IDs for 3Com Home Connect (3c450) and Office Connect (3cSOHO100), register compatible with later versions of 3c905.
1.15.2.8	17-Jan-2003	thorpej	Sync with HEAD.
1.15.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.15.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.15.2.5	01-Aug-2002	nathanw	Catch up to -current.
1.15.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.15.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.15.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.15.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.17.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.17.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.17.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.17.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.17.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.20.8.1	15-Jul-2002	gehenna	catch up with -current.
1.35.4.1	22-Aug-2004	tron	Pull up revision 1.36 (requested by junyoung in ticket #781): Add support for 3Com 3c920B-EMB-WNM Integrated Fast Ethernet, as found on Asus P4R800-VM motherboards. From BL on Korea BSD User Forum (sorry but I don't know his real name).
1.35.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.35.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.35.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.35.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.35.2.1	03-Aug-2004	skrll	Sync with HEAD
1.37.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.37.4.1	29-Apr-2005	kent	sync with -current
1.39.4.4	24-Mar-2008	yamt	sync with head.
1.39.4.3	27-Oct-2007	yamt	sync with head.
1.39.4.2	30-Dec-2006	yamt	sync with head.
1.39.4.1	21-Jun-2006	yamt	sync with head.
1.40.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.40.14.1	19-Jun-2006	chap	Sync with head.
1.40.8.1	26-Jun-2006	yamt	sync with head.
1.40.4.1	09-Sep-2006	rpaulo	sync with head
1.41.6.2	10-Dec-2006	yamt	sync with head.
1.41.6.1	22-Oct-2006	yamt	sync with head
1.41.4.1	18-Nov-2006	ad	Sync with head.
1.43.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.43.24.2	23-Mar-2008	matt	sync with HEAD
1.43.24.1	06-Nov-2007	matt	sync with HEAD
1.43.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.43.8.1	23-Oct-2007	ad	Sync with head.
1.44.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.44.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.44.12.1	24-Mar-2008	keiichi	sync with head.
1.48.4.4	11-Aug-2010	yamt	sync with head.
1.48.4.3	11-Mar-2010	yamt	sync with head
1.48.4.2	16-May-2009	yamt	sync with head
1.48.4.1	16-May-2008	yamt	sync with head.
1.48.2.1	18-May-2008	yamt	sync with head.
1.49.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.51.2.1	30-May-2010	rmind	sync with head
1.54.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.54.12.1	23-Jun-2013	tls	resync from head
1.54.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.55.4.1	18-May-2014	rmind	sync with head
1.57.30.1	10-Jun-2019	christos	Sync with HEAD
1.57.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.64	20-Jan-2020	thorpej	Remove FDDI support.
1.63	10-Nov-2019	chs	branches: 1.63.2; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.62	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.61	22-Jun-2018	msaitoh	branches: 1.61.2; It's not required to include net/bpfdesc.h. Remove it.
1.60	25-Jan-2015	martin	branches: 1.60.10; 1.60.16; Fix obviously missing "sc->sc_dev = self" initialization. Might help PR port-i386/49602.
1.59	29-Mar-2014	christos	branches: 1.59.4; 1.59.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.58	27-Oct-2012	chs	branches: 1.58.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.57	19-Jan-2010	pooka	branches: 1.57.12; 1.57.22; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.56	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.55	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.54	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.53	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.52	18-Apr-2009	tsutsui	Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.51	18-Mar-2009	cegger	bcopy -> memcpy
1.50	12-Jun-2008	cegger	branches: 1.50.4; 1.50.10; use device_lookup_private to get softc
1.49	10-Apr-2008	cegger	branches: 1.49.2; 1.49.4; 1.49.6; 1.49.8; use aprint_*_dev and device_xname
1.48	04-Mar-2007	christos	branches: 1.48.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.47	16-Nov-2006	christos	branches: 1.47.4; __unused removal on arguments; approved by core.
1.46	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.45	11-Dec-2005	christos	branches: 1.45.20; 1.45.22; merge ktrace-lwp.
1.44	27-Feb-2005	perry	branches: 1.44.4; nuke trailing whitespace
1.43	31-Jan-2003	thorpej	branches: 1.43.2; 1.43.10; 1.43.12; Use aprint_*().
1.42	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.41	02-Oct-2002	thorpej	Remove unnecessary extern decl of fpa_ca.
1.40	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.39	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.38	14-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of bzero() after malloc(). XXX It seems these attachments need to be rewritten for modern config(9)..
1.37	13-Nov-2001	lukem	add RCSID
1.36	07-Jul-2001	thorpej	branches: 1.36.2; bcopy -> strcpy
1.35	13-Jun-2001	wiz	withough -> without
1.34	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.33	28-Dec-2000	sommerfeld	branches: 1.33.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.32	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.31	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.30	30-Mar-2000	augustss	Remove register declarations.
1.29	01-Jun-1999	thorpej	branches: 1.29.2; Make sure bus mastering is enabled; some broken BIOSes don't set this up properly.
1.28	13-Aug-1998	eeh	branches: 1.28.4; 1.28.6; 1.28.8; 1.28.10; Merge paddr_t changes into the main branch.
1.27	05-Jul-1998	jonathan	branches: 1.27.2; defopt INET, NETATALK.
1.26	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.25	21-May-1998	matt	Add bus_dma(9) support for the DEFxA FDDI driver
1.24	12-Jan-1998	thorpej	Update for config changes.
1.23	09-Jun-1997	thorpej	Duplicate register mapping logic from Tulip driver.
1.22	08-Jun-1997	thorpej	Merge pdq-970604 update.
1.21	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.20	13-Apr-1997	cgd	use pci_map_register(). map both mem and I/O spaces, preferring mem space if it's usable, but falling back to I/O space if mem isn't usable.
1.19	11-Apr-1997	cgd	add a missing newline
1.18	24-Mar-1997	thorpej	Resolve conflicts from merge.
1.17	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.16	05-Dec-1996	cgd	branches: 1.16.4; update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.15	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.14	13-Oct-1996	christos	backout kprintf changes
1.13	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.12	17-Jun-1996	cgd	FPA probe would always return zero, because of incorrect return statement. Patch provided by Peter Galbavy in PR 2555.
1.11	20-May-1996	thorpej	branches: 1.11.2; Remove some extra printfs, now that Matt has lessened my confusion.
1.10	20-May-1996	thorpej	Fix conflicts from import, clean up a few printfs, and use the generic PCI vendor/product definitions, rather than home-grown versions.
1.9	13-May-1996	mycroft	Use intr.h.
1.8	07-May-1996	thorpej	Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.7	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.6	14-Mar-1996	cgd	include pcireg before pcivar.h
1.5	11-Mar-1996	thorpej	Resolve conflicts from merge.
1.4	09-Mar-1996	thorpej	Fill in the missing pieces and make the NetBSD version of Matt Thomas's DEFPA FDDI driver actually compile and link.
1.3	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.2	19-Aug-1995	cgd	local adaptations, and fix a could of compilation errors
1.1	19-Aug-1995	cgd	branches: 1.1.1; Initial revision
1.1.1.5	08-Jun-1997	thorpej	Update of DEC PDQ FDDI driver, from Matt Thomas <matt@3am-software.com>.
1.1.1.4	24-Mar-1997	thorpej	Update from Matt Thomas <matt@3am-software.com>. Notable highlight: the DEFTA (TurboChannel) card now works on the Alpha!
1.1.1.3	20-May-1996	thorpej	Update to the Digital Equipment PDQ-based FDDI controllers, from Matt Thomas. Now includes support for EISA (DEFEA) and first-cut of support for TurboChannel (DEFTA) cards.
1.1.1.2	11-Mar-1996	thorpej	New snapshot of DEC PDQ-based FDDI driver, from Matt Thomas.
1.1.1.1	19-Aug-1995	cgd	Generic FDDI support by Matt Thomas. Support for DEC "PDQ" FDDI chipset and for the PCI attachment of said chipset ("if_fpa"), also from Matt Thomas. Arguably, pdq* doesn't belong in sys/dev/ic, but it's going to be shared by various bus attachment devices at some point in the future, and there's no other place that seems to fit as well.
1.11.2.1	17-Jun-1996	cgd	pull up from trunk: >FPA probe would always return zero, because of incorrect return statement. >Patch provided by Peter Galbavy in PR 2555.
1.16.4.4	09-Mar-1997	is	netinet/if_ether.h -> netinet/if_inarp.h
1.16.4.3	27-Feb-1997	is	Removed #include <route.h> and #include <netisr.h>. They're not needed, and the latter will go away soon.
1.16.4.2	20-Feb-1997	is	Give fddi_ifattach() a 2nd parameter, like ether_ifattach(): a pointer to the link level address. xxx_ifattach() copy it to the sockaddr_dl structure associated with the interface. Change pdq to pass that parameter. Change frontend to not copy the l.l.a. itself.
1.16.4.1	07-Feb-1997	is	Snapshot of new ARP code. Our old ARP code was hardwired for 6-byte length medium addresses, while the protocol is designed for any size. This snapshot contains a first hack at getting rid of Ethernet specific data structures. The ep driver is updated (and tested on the PCI bus), the iy and fpa drivers have been updated, but not real life tested yet. If you want to test this with other drivers, you have to update them first yourself, and probably tag the relevant directories. Better contact me if you want to do this.
1.27.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.28.10.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.28.8.1	21-Jun-1999	thorpej	Sync w/ -current.
1.28.6.1	21-Jun-1999	perry	pullup 1.28->1.29 (thorpej): Make sure bus mastering is enabled
1.28.4.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.29.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.29.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.33.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.33.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.33.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.33.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.33.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.36.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.36.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.36.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.43.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.43.10.1	29-Apr-2005	kent	sync with -current
1.43.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.44.4.1	03-Sep-2007	yamt	sync with head.
1.45.22.2	10-Dec-2006	yamt	sync with head.
1.45.22.1	22-Oct-2006	yamt	sync with head
1.45.20.1	18-Nov-2006	ad	Sync with head.
1.47.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.48.36.2	29-Jun-2008	mjf	Sync with HEAD.
1.48.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.49.8.1	18-Jun-2008	simonb	Sync with head.
1.49.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.49.4.3	11-Mar-2010	yamt	sync with head
1.49.4.2	16-May-2009	yamt	sync with head
1.49.4.1	04-May-2009	yamt	sync with head.
1.49.2.1	17-Jun-2008	yamt	sync with head.
1.50.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.4.1	28-Apr-2009	skrll	Sync with HEAD.
1.57.22.3	03-Dec-2017	jdolecek	update from HEAD
1.57.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.57.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.57.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.57.12.1	30-Oct-2012	yamt	sync with head
1.58.2.1	18-May-2014	rmind	sync with head
1.59.6.1	06-Apr-2015	skrll	Sync with HEAD
1.59.4.1	31-Jan-2015	snj	Pull up following revision(s) (requested by martin in ticket #469): sys/dev/pci/if_fpa.c: revision 1.60 Fix obviously missing "sc->sc_dev = self" initialization. Might help PR port-i386/49602.
1.60.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.60.16.1	25-Jun-2018	pgoyette	Sync with HEAD
1.60.10.1	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.61.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.61.2.1	10-Jun-2019	christos	Sync with HEAD
1.63.2.1	25-Jan-2020	ad	Sync with head.
1.35	20-Jun-1999	thorpej	Intel i82557 driver is now split into bus and chip bits.
1.34	18-May-1999	thorpej	Rework layer 2 protocol input routines. Instead of calling e.g. ether_input() directly, call the function pointer (if_input)(ifp, m). The input routine expects the packet header to be at the head of the packet, and will adjust as necessary. Privatize the layer 2 input and output routines, allowing _ifattach() to set them up as appropriate.
1.33	23-Mar-1999	thorpej	branches: 1.33.2; 1.33.4; 1.33.6; Make sure bus mastering is enabled.
1.32	28-Feb-1999	explorer	Update to slightly altered rnd_attach_source() api
1.31	18-Feb-1999	thorpej	Fix omission in previous.
1.30	18-Feb-1999	thorpej	Don't reset the interface when the media changes.
1.29	18-Feb-1999	thorpej	Back out to the previous stable revision of the fxp driver. The latest one had too many stability problems.
1.28	13-Feb-1999	thorpej	Restructure the interrupt handler and 1-second timer a bit, and tidy up the lazy-transmit-interrupt logic, fixing a few minor logic problems. Now unable to reproduce the lockup problem described in PR #6767. Changing PR's state to "feedback".
1.27	19-Dec-1998	thorpej	Changes from FreeBSD: Implemented a more sophisticated mechanism for handling transmitter interrupts which now defers them until the transmit queue if filled up with completed buffers. This has two advantages: first, it reduces the number of transmitter interrupts to just 1/120th of the rate that they occured previously, and two, running down many buffers at once has much improved cache effects. Defer rundown (m_freem) of completed transmit buffers for no longer than 1 second. This brings us up-to-date with the most recent "fxp" driver in FreeBSD.
1.26	17-Dec-1998	explorer	teach fxp about rnd_*()
1.25	25-Nov-1998	bouyer	branches: 1.25.2; Use bus_dma(9) for the DMA ops in fxp_init() too. Without this the driver was stuck in the while (!(cbp->cb_status & FXP_CB_STATUS_C)); loop, I think because the relevent CPU cache entry was never updated. This was on a compaq deskpro system.
1.24	03-Nov-1998	thorpej	Add support for the Seeq 80c24 AutoDUPLEX media interface which was nuked when this driver was changed to use the MAC-independent MII code. This "PHY" does not have a programming interface, but simply senses the duplex mode from the link partner. Since it is not possible to read the 80c24 to determine the link type or status, we consider this to be "manual".
1.23	03-Nov-1998	thorpej	Make sure that a delay loop actually has some delay in it, and use IFF_OACTIVE as appropriate.
1.22	19-Oct-1998	thorpej	Fix a memory leak in fxp_stop(), from the FreeBSD driver.
1.21	25-Aug-1998	thorpej	Change the 32-bit fields of the RFA to an arrays of bytes, so that egcs won't attempt to optimize inline copies of them into a single load/store (the RFA is not 4-byte aligned in the DMA area). Just use memcpy() to copy those fields, now, not our own home-grown.
1.20	11-Aug-1998	thorpej	Adapt to new MII code.
1.19	08-Aug-1998	mycroft	Use splnet, not splimp.
1.18	05-Jul-1998	jonathan	defopt NS, NSIP.
1.17	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.16	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.15	18-May-1998	cgd	check that memory access is enabled before bothering to do the pci_mapreg_info() call. pci_mapreg_map() implies this check, but code which calls pci_mapreg_info() has to check it explicitly. Otherwise, if memory space is disabled, the driver does the wrong thing, and tries to use memory space anyway, potentially resulting incorrect driver operation and no useful error message.
1.14	04-Feb-1998	thorpej	Add necessary bus_dmamap_sync() calls for memory mapped with the COHERENT hint, per bus_dma(9).
1.13	04-Feb-1998	thorpej	Add offset and length parameters to bus_dmamap_sync(), used for specifiying partial syncs of a DMA mapping.
1.12	04-Feb-1998	thorpej	Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC to BUS_DMA_COHERENT.
1.11	28-Jan-1998	thorpej	Add support for the 82553[AC] PHY by treating it like an 82555. From Soren S. Jorvang <soren@t.dk>, PR #4840.
1.10	28-Jan-1998	thorpej	Convert this driver to use the bus_dma interfaces.
1.9	22-Jan-1998	thorpej	Sync w/ FreeBSD-current-980122: - Rewrite fxp_start() for better clarity and efficiency. - Remove unused #includes.
1.8	22-Jan-1998	thorpej	Fix two register mapping problems: - Fall back on i/o space if mem space isn't available. - Card reports mem space as prefetchable, but mapping the card into dense space fails in nasty ways on the Alpha. Force mapping into sparse space by clearing the BUS_SPACE_MAP_CACHEABLE bit (XXX!).
1.7	12-Jan-1998	thorpej	Update for config changes.
1.6	11-Nov-1997	thorpej	Type/alignment cleanup in EEPROM code.
1.5	20-Oct-1997	thorpej	branches: 1.5.2; Port the latest FreeBSD "fxp" driver to NetBSD-current. Changes include: - Work around a bug in the 82557 that causes the receiver to lock up in certain conditions by kicking the multicast address filter if we haven't heard anything come down the wire for some period of time. - Fix a bug that could cause TxCB descriptor chains to cross page boundaries on the Alpha. - Remove some unneeded register masking. - Fix a bug where too much data was copied from the config template, causing memory corruption. - Fix handing of if_timer (it was be cleared too early in some cases). - Attempt to reduce the chances of receiver overrun by doubling the number of receive DMA segments, and processing receive interrupts before transmit interrupts. - Remove a gratuitous assignment. - Fix a bug where incoming packets were counted twice.
1.4	25-Aug-1997	thorpej	Use if_media to change media type, not IFF_LINK*.
1.3	23-Aug-1997	thorpej	Merge the FreeBSD-current-970823 version of the "fxp" driver, adds support for more PHYs. Closes PR #3953.
1.2	05-Jun-1997	thorpej	branches: 1.2.4; Changes to support NetBSD, written by me, tested by Matthias Drochner: - Add NetBSD autoconfiguration support. - Rearrange code slightly to minimize the number of #ifdefs. - Don't use a structure to access CSRs. Use macros that DTRT for the NetBSD and FreeBSD cases. - Deal with alignment contraint on Alpha - add 2-byte padding at the beginning of the RFA, so that the data will be 4-byte aligned, after the 14-byte Ethernet header. Thanks to Matthias Drochner for the testing, and David Greenman for the feedback on the changes.
1.1	05-Jun-1997	thorpej	branches: 1.1.1; Initial revision
1.1.1.4	22-Jan-1998	thorpej	Import FreeBSD's latest "fxp" driver, as of Jan 22, 1998.
1.1.1.3	20-Oct-1997	thorpej	Import the latest FreeBSD "fxp" driver, as of 971019.
1.1.1.2	23-Aug-1997	thorpej	Import the latest "fxp" driver from FreeBSD-current, adds support for additional PHYs.
1.1.1.1	05-Jun-1997	thorpej	Import FreeBSD's Intel EtherExpress PRO 10/100B driver, written by David Greenman.
1.2.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.5.2.3	07-Feb-1998	mellon	Pull up 1.11 (thorpej)
1.5.2.2	07-Feb-1998	mellon	Pull up 1.8 (thorpej)
1.5.2.1	11-Nov-1997	thorpej	Sync w/ trunk: type/alignment cleanup.
1.25.2.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.33.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.33.4.2	01-Jul-1999	thorpej	Sync w/ -current.
1.33.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.33.2.2	01-Jun-2000	he	Apply patch (requested by jhawk): Calibrate the timeouts by using DELAY(1) so their values are less machine-dependant. Fixes PR#10141, where spurious timeouts were being seen. Also report the line numbers of the timeouts so it can be determined which is being seen.
1.33.2.1	13-May-2000	he	Apply patch (requested by jhawk): Fix the fxp driver so that it works after a warm-boot from windows and across an APM suspend/resume or hibernate/resume cycle. Fixes PR#9370, PR#9548, PR#9571 and PR#9573.
1.87	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.86	08-May-2021	thorpej	Use pci_compatible_match().
1.85	23-Jan-2019	msaitoh	branches: 1.85.16; Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them expilicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.84	23-Jan-2019	msaitoh	KNF. No functional change.
1.83	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.82	13-Apr-2015	riastradh	branches: 1.82.10; 1.82.16; 1.82.18; Convert sys/dev to use <sys/rndsource.h>.
1.81	29-Mar-2014	christos	branches: 1.81.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.80	30-Mar-2013	christos	branches: 1.80.4; remove trailing whitespace
1.79	02-Feb-2012	tls	branches: 1.79.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.78	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.77	05-Sep-2011	msaitoh	branches: 1.77.2; 1.77.6; Change 0x1229's product name from 82557 to 8255x. 82550, 82551, 82558 and 82559 share the same product id. This change makes "pcictl dump" happy.
1.76	02-Sep-2011	msaitoh	Add support for some fxp devices from FreeBSD and OpenBSD. {Free,Open}BSD say that we have to do some work to make fxp stable.
1.75	17-May-2011	dyoung	PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.74	22-Mar-2010	dyoung	branches: 1.74.2; Add a detach routine. Enable detachment at shutdown. psc_pwrmgmt_csr_reg was never initialized to the actual offset of the PCI PMCSR! Just delete psc_pwrmgmt_csr_reg and all uses of it. In this way, fxp_pci_disable() becomes a no-op, so delete it, too.
1.73	24-Feb-2010	dyoung	branches: 1.73.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.72	08-Jan-2010	dyoung	branches: 1.72.2; Expand PMF_FN_* macros.
1.71	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.70	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.69	17-Apr-2009	tsutsui	Check revisions and print names of i82551/i82551ER, per Intel manual.
1.68	16-Apr-2009	tsutsui	There is a report that 82559ER doesn't have RX hardware checksum support (even though it has a newer revision number than 82559_A0), so handle it separately per its own PCI device ID.
1.67	19-Mar-2009	tsutsui	Add braces missed on RX hardware cksum support. Fixes RX failure on 82557/82558.
1.66	11-Mar-2009	tsutsui	u_intNN_t -> uintNN_t
1.65	07-Mar-2009	tsutsui	Add TCPv4/UDPv4 RX hardware checksum support for i82559 and later chips which don't have EXT_RFA and IPCB support. From hme(4) driver and FreeBSD's fxp(4). Tested on i82559. XXX: Probably we should have a common function to parse RX packet headers XXX: to handle a raw checksum value and share it among hme(4) and gem(4) etc.
1.64	06-Mar-2009	tsutsui	The Intel manual says that ICH and 82562 chips are compatible with 82559, not 82550, so don't set FXPF_EXT_RFA (available only on 82550/82551) for those variants. Tested on Intel PRO/100 VM with 82562EM/EX PHY by ISHII Kimio on current-users, and should also fix PR kern/40431.
1.63	03-Mar-2009	mrg	enable FXPF_EXT_RFA the same way the cardbus frontend does.
1.62	20-Feb-2009	mrg	- remove FXPF_IPCB flag. it should always/only be used with the code conditional on FXPF_EXT_TXCB, so, replace all uses with that - for the pci frontend, reestablish some flags lost the the prior changes and simplify one of the cases this fixes PR 40677 and may fix PR 40431.
1.61	18-Jan-2009	mrg	branches: 1.61.2; The PCI revision numbers are unique to a PCI vendor/product ID pair. Misuse of the revision numbers was causing some of the chip features to be disabled on some integrated Intel chips. So, move the determination of the features into the bus frontend, where the vendor/product ID is known. (Note: sc_rev should be removed. The microcode patch stuff is also busted and needs to be fixed.) Also, poll the actual flow control status in inphy, rather than making assumptions. contributed anonymously.
1.60	09-Jul-2008	joerg	branches: 1.60.2; - device/softc split for fxp(4)
1.59	28-Apr-2008	martin	branches: 1.59.2; 1.59.4; 1.59.6; Remove clause 3 and 4 from TNF licenses
1.58	10-Apr-2008	cegger	branches: 1.58.2; 1.58.4; use aprint_*_dev and device_xname
1.57	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.56	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.55	24-Dec-2007	hamajima	branches: 1.55.2; 1.55.6; Match on board ethernet controller on VGN-TX91PS.
1.54	10-Dec-2007	degroote	branches: 1.54.2; Register fxp @ pci to the pmf framework Kill reference to old powerhook framework.
1.53	19-Oct-2007	ad	branches: 1.53.4; 1.53.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.52	26-Nov-2006	enami	branches: 1.52.8; 1.52.22; 1.52.24; 1.52.28; Match on board ethernet controller on VGN-TX72PZ.
1.51	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.50	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.49	27-Sep-2006	cube	Make fxp(4) support the PRO/100 VE variant just added to pcidevs. Reported by rix on #NetBSD-code on freenode.
1.48	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.47	30-Jul-2006	oster	branches: 1.47.4; 1.47.6; Recognize another Intel PRO/100 VM Network Controller.
1.46	19-Jun-2006	jmcneill	Restore PCI configuration space on resume.
1.45	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.44	10-Jan-2006	christos	branches: 1.44.2; 1.44.6; 1.44.12; 1.44.14; PR/32494: Dennis Chernoivanov: Intel PRO/100 VE (LOM) not detected on Sony Vaio
1.43	11-Dec-2005	christos	branches: 1.43.2; merge ktrace-lwp.
1.42	29-Jul-2005	cube	New Dell Day, continued: add support for the fxp(4) variant in the 802801G* family of SB.
1.41	18-May-2005	riz	branches: 1.41.2; Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.40	27-Feb-2005	perry	branches: 1.40.2; nuke trailing whitespace
1.39	21-Aug-2004	thorpej	branches: 1.39.4; 1.39.6; Use ANSI function decls and make use of static.
1.38	28-Apr-2004	briggs	Add a default case to the product ID attachment to print the chip description and revision information if a chip is matched and this switch statement isn't updated to match. Addresses PR kern/18019.
1.37	12-Mar-2004	drochner	branches: 1.37.2; deal with 82801EB/ER (ICH5) Network Controller
1.36	13-Sep-2003	jdolecek	fix description of Intel PRO/100 VM Network Controller with 82562ET/EZ PHY don't activate any quirks for VM_6 variant problem reported by Lars Heidieker on current-users also don't activate any quirks for VM_3 - VM_5 variants, neither VE_2 - VE_4, in anticipation that the more modern variants don't need the quirks
1.35	14-Aug-2003	nonaka	Add support for another Intel Pro/100 variant.
1.34	28-Jun-2003	bouyer	branches: 1.34.2; Add support for another Intel Pro/100 variant. From Quentin Garnier in private mail.
1.33	14-Mar-2003	jdolecek	apparently this driver works fine for Intel PRO/100 VM with 82562EM/EX PHY, so make driver match it while here, also make the driver match 'Intel PRO/100 VM with 82562EM/EX (CNR) PHY', with anticipation it would work fine too activate the resume and dynamic standby bug workaround for these; might not be necessary (not confirmed to be needed), but better safe than sorry Adresses PR kern/20689
1.32	18-Feb-2003	grant	add Intel PRO/100M. from Quentin Garnier in PR kern/20400.
1.31	31-Jan-2003	thorpej	Use aprint_*().
1.30	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.29	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.28	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.27	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.26	27-Sep-2002	itojun	don't go into D3 state if revision <= 82558_B0. PR 18011
1.25	18-Sep-2002	abs	Recognise the Intel PRO/100 VM Network Controller as seen in the HP omnibook 6100.
1.24	22-Jul-2002	msaitoh	add support intel C-ICH's embeded Ether MACs. Tested on RadiSys Endura CH815C (http://www.radisys.com/)
1.23	22-Jun-2002	cjs	Match PCI IDs of new PRO/100 chips.
1.22	04-Apr-2002	thorpej	branches: 1.22.2; 1.22.4; Address Intel 82801BA/82801BAM Specification Update, Errata #30: The ICH on-board Ethernet and some i82559 chips have a bug which will cause a PCI protocol violation if the chip receives a CU_RESUME command as it is entering the IDLE state by deasserting #CLKRUN. (This is the so-called "resume bug" that we previously had an incomplete work-around for on ICH chipsets.) The work-around is to disable Dynamic Standby Mode, such that the chip will never deasert #CLKRUN. Dynamic Standby Mode is disabled by clearing a bit in the EEPROM and updating the EEPROM (and EEPROM checksum). Unfortunately, the chip will only consult the EEPROM setting after a PCI bus reset, so a system reboot is required once the EEPROM has been updated (the EEPROM update only needs to happen once, and the driver usses a warning instructing the user to reboot the system once the work-around has been applied). Issue pointed out by David Brownlee, and code more-or-less lifted from FreeBSD.
1.21	13-Nov-2001	lukem	add RCSID
1.20	02-Nov-2001	itojun	recognize fxp on IBM thinkpad T23. from akimichi@wide.ad.jp
1.19	05-Aug-2001	thorpej	branches: 1.19.4; Add enable/disable hooks on devices that have ACPI power management capability, using D3 and D0 power states. This saves me quite a bit of battery when not using the built-in Ethernet on my IBM T-20 during long IETF meetings.
1.18	08-Jul-2001	thorpej	branches: 1.18.2; Remove unnecessary include files. From Onno van der Linden.
1.17	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.16	22-May-2001	thorpej	Enable the PCI MWI command for memory writes on the i82558 and higher.
1.15	22-May-2001	thorpej	Report 82558, 82559, etc. based on chip revision. Also, remember which revsion we have for feature enables later.
1.14	21-May-2001	thorpej	The ICH3 also has the Resume Bug.
1.13	21-May-2001	thorpej	The 82801BA built-in Ethernet has a bug which requires us to issue a NOP before a CU_RESUME when in 10baseT mode. Handle this.
1.12	28-Dec-2000	sommerfeld	branches: 1.12.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.11	27-Sep-2000	ad	Match the i82562.
1.10	15-Jul-2000	jhawk	Fix uninitialized variable (reg) in fxp_pci_confreg_restore(); this could result in garbage being written to the PCI status register, which is unlikely to have had a serious effect. This was with us from an #if 0 added in rev 1.6.
1.9	09-Jul-2000	mycroft	Recognize the i82559ER.
1.8	12-May-2000	jhawk	branches: 1.8.4; sc_ih (interrupt cookie) is not pci-specific data and needs to remain in fxp_softc rather than fxp_pci_softc; otherwise we break cardbus.
1.7	12-May-2000	jhawk	Move struct fxp_pci_softc declaration to if_fxp_pci.c removing an unnecessary header file (if_fxp_pcivar.h); oh, how quickly the nascent inodes are subject to infant mortality.
1.6	12-May-2000	jhawk	Fix bugs seen on Sony Vaios. Address kern/9370, kern/9548, kern/9571, and kern/9573. Save and restore critical PCI configuration registers because they are sometimes bashed. Add a powerhook to restore said PCI registers after APM events. The fxp now works across a warm boot from windows (vaio). The fxp now works across an apm suspend/resume or hibernate/resume cycle.
1.5	16-Mar-2000	thorpej	Recognize the Intel InBusiness.
1.4	25-Jan-2000	drochner	use BUS_SPACE_MAP_PREFETCHABLE instead BUS_SPACE_MAP_CACHEABLE where the PCI BAR bit is referred to (XXX is the hack still necessary?)
1.3	12-Dec-1999	thorpej	Take a stab at making this work on big-endian systems.
1.2	30-Oct-1999	sommerfeld	Add enable/disable support and EEPROM size selection, prerequisites for getting the cardbus fxp attachment to work. Tested on the built-in 82559 in a VAIO Z505S. From Johan Danielsson / PR8631; changed slightly to KNF.
1.1	20-Jun-1999	thorpej	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Intel i82557 driver is now split into bus and chip bits.
1.1.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.6.1	15-Nov-1999	fvdl	Sync with -current
1.1.4.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.4.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	20-Jun-1999	thorpej	file if_fxp_pci.c was added on branch chs-ubc2 on 1999-07-01 23:34:06 +0000
1.8.4.5	30-Sep-2002	itojun	sys/dev/pci/if_fxp_pci.c 1.26 Don't go into D3 state if revision <= 82558_B0. Fixes PR#18011. (itojun)
1.8.4.4	06-Jun-2002	he	Pull up revisions 1.13-1.18,1.20,1.22 (requested by mycroft): Work around a bug which can cause some machines to randomly hang. Drag along performance improvements and improved hardware support.
1.8.4.3	16-Aug-2001	tv	Pullup [he]: sys/dev/pci/if_fxp_pci.c 1.19 Add enable/disable hooks on devices that have ACPI power management capability, using D3 and D0 power states.
1.8.4.2	22-Mar-2001	he	Pull up revision 1.11 (requested by hubertf): Make the Intel Pro/100 as e.g. found on Toshiba Satellite Pro 4600s work.
1.8.4.1	16-Jul-2000	jhawk	Pullup rev 1.9 (approved by jhawk), rev 1.10 (approved by thorpej): revision 1.9: Recognize the i82559ER. revision 1.10: Fix uninitialized variable (reg) in fxp_pci_confreg_restore(); this could result in garbage being written to the PCI status register, which is unlikely to have had a serious effect. This was with us from an #if 0 added in rev 1.6.
1.12.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.12.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.12.2.5	01-Aug-2002	nathanw	Catch up to -current.
1.12.2.4	17-Apr-2002	nathanw	Catch up to -current.
1.12.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.12.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.12.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.18.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.18.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.18.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.18.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.19.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.22.4.8	26-Oct-2005	jmc	Pullup (via patch) requested by riz in ticket #5773) Add support for 82801FB 10/100 ethernet
1.22.4.7	06-Apr-2004	jmc	Pullup patch (requested by briggs in ticket #1648) Pull in changes to support the Intel 82801EB/ER (ICH5) Controller. PR#25002
1.22.4.6	15-Aug-2003	tron	Pull up revision 1.34 via patch (requested by bouyer in ticket #1368): Add support for another Intel Pro/100 variant. From Quentin Garnier in private mail.
1.22.4.5	30-Jun-2003	grant	Apply patch (requested by drochner in ticket #1343): add Intel PRO/100M. apparently this driver works fine for Intel PRO/100 VM with 82562EM/EX PHY, so make driver match it while here, also make the driver match 'Intel PRO/100 VM with 82562EM/EX (CNR) PHY', with anticipation it would work fine too activate the resume and dynamic standby bug workaround for these; might not be necessary (not confirmed to be needed), but better safe than sorry
1.22.4.4	27-Jan-2003	jmc	Pullup revisions 1.23-1.24 (requested by masanobu in ticket #1110) Add support intel C-ICH's embeded Ether MACs.
1.22.4.3	13-Oct-2002	lukem	Pull up revision 1.25 (requested by abs in ticket #844): Recognise the Intel PRO/100 VM Network Controller as seen in the HP omnibook 6100.
1.22.4.2	02-Oct-2002	lukem	Pull up revision 1.26 (requested by itojun in ticket #870): don't go into D3 state if revision <= 82558_B0. PR 18011
1.22.4.1	25-Jun-2002	lukem	Pull up revision 1.23 (requested by cjs in ticket #362): Add new Intel Ethernet cards
1.22.2.2	29-Aug-2002	gehenna	catch up with -current.
1.22.2.1	15-Jul-2002	gehenna	catch up with -current.
1.34.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.34.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.34.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.34.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.34.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.34.2.1	03-Aug-2004	skrll	Sync with HEAD
1.37.2.1	28-Apr-2004	tron	branches: 1.37.2.1.2; Pull up revision 1.38 (requested by briggs in ticket #220): Add a default case to the product ID attachment to print the chip description and revision information if a chip is matched and this switch statement isn't updated to match. Addresses PR kern/18019.
1.37.2.1.2.1	19-May-2005	tron	Pull up revision 1.41 (requested by riz in ticket #1532): Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.39.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.39.4.1	29-Apr-2005	kent	sync with -current
1.40.2.3	27-Aug-2006	ghen	Pull up following revision(s) (requested by oster in ticket #1479): sys/dev/pci/pcidevs: revision 1.810 sys/dev/pci/if_fxp_pci.c: revision 1.44 sys/dev/pci/if_fxp_pci.c: revision 1.47 sys/dev/pci/pcidevs: revision 1.756 PR/32494: Add ID 0x1068 for Intel PRO/100 VE (LOM) found on Sony Vaio S4HRP laptops. PR/32494: Dennis Chernoivanov: Intel PRO/100 VE (LOM) not detected on Sony Vaio Recognize another Intel PRO/100 VM Network Controller.
1.40.2.2	21-Nov-2005	tron	Pull up following revision(s) (requested by cube in ticket #957): sys/dev/pci/if_fxp_pci.c: revision 1.42 New Dell Day, continued: add support for the fxp(4) variant in the 802801G* family of SB.
1.40.2.1	28-May-2005	snj	Pull up revision 1.41 (requested by riz in ticket #323): Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.41.2.6	24-Mar-2008	yamt	sync with head.
1.41.2.5	17-Mar-2008	yamt	sync with head.
1.41.2.4	21-Jan-2008	yamt	sync with head
1.41.2.3	27-Oct-2007	yamt	sync with head.
1.41.2.2	30-Dec-2006	yamt	sync with head.
1.41.2.1	21-Jun-2006	yamt	sync with head.
1.43.2.1	15-Jan-2006	yamt	sync with head.
1.44.14.1	13-Jul-2006	gdamore	Merge from HEAD.
1.44.12.2	22-Jun-2006	chap	Complete a sync sys/ with head.
1.44.12.1	19-Jun-2006	chap	Sync with head.
1.44.6.2	11-Aug-2006	yamt	sync with head
1.44.6.1	26-Jun-2006	yamt	sync with head.
1.44.2.1	09-Sep-2006	rpaulo	sync with head
1.47.6.2	10-Dec-2006	yamt	sync with head.
1.47.6.1	22-Oct-2006	yamt	sync with head
1.47.4.2	12-Jan-2007	ad	Sync with head.
1.47.4.1	18-Nov-2006	ad	Sync with head.
1.52.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.52.24.3	23-Mar-2008	matt	sync with HEAD
1.52.24.2	09-Jan-2008	matt	sync with HEAD
1.52.24.1	06-Nov-2007	matt	sync with HEAD
1.52.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.52.8.1	23-Oct-2007	ad	Sync with head.
1.53.6.1	11-Dec-2007	yamt	sync with head.
1.53.4.1	26-Dec-2007	ad	Sync with head.
1.54.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.55.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.55.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.55.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.55.2.1	24-Mar-2008	keiichi	sync with head.
1.58.4.5	11-Aug-2010	yamt	sync with head.
1.58.4.4	11-Mar-2010	yamt	sync with head
1.58.4.3	16-Sep-2009	yamt	sync with head
1.58.4.2	04-May-2009	yamt	sync with head.
1.58.4.1	16-May-2008	yamt	sync with head.
1.58.2.1	18-May-2008	yamt	sync with head.
1.59.6.1	19-Oct-2008	haad	Sync with HEAD.
1.59.4.1	18-Jul-2008	simonb	Sync with head.
1.59.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.60.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.60.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.60.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.61.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.72.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.73.2.2	31-May-2011	rmind	sync with head
1.73.2.1	30-May-2010	rmind	sync with head
1.74.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.77.6.1	18-Feb-2012	mrg	merge to -current.
1.77.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.77.2.1	17-Apr-2012	yamt	sync with head
1.79.6.3	03-Dec-2017	jdolecek	update from HEAD
1.79.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.79.6.1	23-Jun-2013	tls	resync from head
1.80.4.1	18-May-2014	rmind	sync with head
1.81.6.1	06-Jun-2015	skrll	Sync with HEAD
1.82.18.1	10-Jun-2019	christos	Sync with HEAD
1.82.16.2	26-Jan-2019	pgoyette	Sync with HEAD
1.82.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.82.10.1	27-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1172): sys/dev/pci/nvme_pci.c: revision 1.26 sys/dev/pci/xhci_pci.c: revision 1.21 sys/dev/pci/ixgbe/ixv.c: revision 1.108 sys/dev/pci/ixgbe/ixgbe.c: revision 1.171 sys/dev/pci/if_fxp_pci.c: revision 1.84 sys/dev/pci/if_fxp_pci.c: revision 1.85 sys/dev/pci/xhci_pci.c: revision 1.16 remove #ifdef DEBUG printf, it seems to have outlived it's usefulness - KNF. No functional change. - Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them explicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.85.16.1	13-May-2021	thorpej	Sync with HEAD.
1.2	12-May-2000	jhawk	Move struct fxp_pci_softc declaration to if_fxp_pci.c removing an unnecessary header file (if_fxp_pcivar.h); oh, how quickly the nascent inodes are subject to infant mortality.
1.1	12-May-2000	jhawk	Move pci-specific data to fxp_pci_softc from fxp_pci_softc. Add some more pci-specific data to fxp_pci_softc.
1.10	20-Jun-1999	thorpej	Intel i82557 driver is now split into bus and chip bits.
1.9	25-Aug-1998	thorpej	branches: 1.9.8; Change the 32-bit fields of the RFA to an arrays of bytes, so that egcs won't attempt to optimize inline copies of them into a single load/store (the RFA is not 4-byte aligned in the DMA area). Just use memcpy() to copy those fields, now, not our own home-grown.
1.8	11-Aug-1998	thorpej	G/c the 83840 register definitions.
1.7	05-Feb-1998	thorpej	Fix the FXP_TXDESCOFF() macro.
1.6	04-Feb-1998	thorpej	Add necessary bus_dmamap_sync() calls for memory mapped with the COHERENT hint, per bus_dma(9).
1.5	28-Jan-1998	thorpej	Convert this driver to use the bus_dma interfaces.
1.4	20-Oct-1997	thorpej	Port the latest FreeBSD "fxp" driver to NetBSD-current. Changes include: - Work around a bug in the 82557 that causes the receiver to lock up in certain conditions by kicking the multicast address filter if we haven't heard anything come down the wire for some period of time. - Fix a bug that could cause TxCB descriptor chains to cross page boundaries on the Alpha. - Remove some unneeded register masking. - Fix a bug where too much data was copied from the config template, causing memory corruption. - Fix handing of if_timer (it was be cleared too early in some cases). - Attempt to reduce the chances of receiver overrun by doubling the number of receive DMA segments, and processing receive interrupts before transmit interrupts. - Remove a gratuitous assignment. - Fix a bug where incoming packets were counted twice.
1.3	23-Aug-1997	thorpej	Merge the FreeBSD-current-970823 version of the "fxp" driver, adds support for more PHYs. Closes PR #3953.
1.2	05-Jun-1997	thorpej	branches: 1.2.4; Changes to support NetBSD, written by me, tested by Matthias Drochner: - Add NetBSD autoconfiguration support. - Rearrange code slightly to minimize the number of #ifdefs. - Don't use a structure to access CSRs. Use macros that DTRT for the NetBSD and FreeBSD cases. - Deal with alignment contraint on Alpha - add 2-byte padding at the beginning of the RFA, so that the data will be 4-byte aligned, after the 14-byte Ethernet header. Thanks to Matthias Drochner for the testing, and David Greenman for the feedback on the changes.
1.1	05-Jun-1997	thorpej	branches: 1.1.1; Initial revision
1.1.1.3	20-Oct-1997	thorpej	Import the latest FreeBSD "fxp" driver, as of 971019.
1.1.1.2	23-Aug-1997	thorpej	Import the latest "fxp" driver from FreeBSD-current, adds support for additional PHYs.
1.1.1.1	05-Jun-1997	thorpej	Import FreeBSD's Intel EtherExpress PRO 10/100B driver, written by David Greenman.
1.2.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.9.8.1	01-Jul-1999	thorpej	Sync w/ -current.
1.10	20-Jun-1999	thorpej	Intel i82557 driver is now split into bus and chip bits.
1.9	18-Feb-1999	thorpej	branches: 1.9.2; 1.9.4; Back out to the previous stable revision of the fxp driver. The latest one had too many stability problems.
1.8	19-Dec-1998	thorpej	Changes from FreeBSD: Implemented a more sophisticated mechanism for handling transmitter interrupts which now defers them until the transmit queue if filled up with completed buffers. This has two advantages: first, it reduces the number of transmitter interrupts to just 1/120th of the rate that they occured previously, and two, running down many buffers at once has much improved cache effects. Defer rundown (m_freem) of completed transmit buffers for no longer than 1 second. This brings us up-to-date with the most recent "fxp" driver in FreeBSD.
1.7	17-Dec-1998	explorer	teach fxp about rnd_*()
1.6	11-Aug-1998	thorpej	Adapt to new MII code.
1.5	28-Jan-1998	thorpej	Convert this driver to use the bus_dma interfaces.
1.4	22-Jan-1998	thorpej	Sync w/ FreeBSD-current-980122: - Rearrange softc for better cache line behavior.
1.3	20-Oct-1997	thorpej	Port the latest FreeBSD "fxp" driver to NetBSD-current. Changes include: - Work around a bug in the 82557 that causes the receiver to lock up in certain conditions by kicking the multicast address filter if we haven't heard anything come down the wire for some period of time. - Fix a bug that could cause TxCB descriptor chains to cross page boundaries on the Alpha. - Remove some unneeded register masking. - Fix a bug where too much data was copied from the config template, causing memory corruption. - Fix handing of if_timer (it was be cleared too early in some cases). - Attempt to reduce the chances of receiver overrun by doubling the number of receive DMA segments, and processing receive interrupts before transmit interrupts. - Remove a gratuitous assignment. - Fix a bug where incoming packets were counted twice.
1.2	25-Aug-1997	thorpej	Use if_media to change media type, not IFF_LINK*.
1.1	05-Jun-1997	thorpej	branches: 1.1.1; 1.1.4; Changes to support NetBSD, written by me, tested by Matthias Drochner: - Add NetBSD autoconfiguration support. - Rearrange code slightly to minimize the number of #ifdefs. - Don't use a structure to access CSRs. Use macros that DTRT for the NetBSD and FreeBSD cases. - Deal with alignment contraint on Alpha - add 2-byte padding at the beginning of the RFA, so that the data will be 4-byte aligned, after the 14-byte Ethernet header. Thanks to Matthias Drochner for the testing, and David Greenman for the feedback on the changes.
1.1.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.1.2	22-Jan-1998	thorpej	Import FreeBSD's latest "fxp" driver, as of Jan 22, 1998.
1.1.1.1	20-Oct-1997	thorpej	Import the latest FreeBSD "fxp" driver, as of 971019.
1.9.4.1	01-Jul-1999	thorpej	Sync w/ -current.
1.9.2.1	13-May-2000	he	Apply patch (requested by jhawk): Fix the fxp driver so that it works after a warm-boot from windows and across an APM suspend/resume or hibernate/resume cycle. Fixes PR#9370, PR#9548, PR#9571 and PR#9573.
1.57	04-Oct-2025	thorpej	Use pci_compatible_{match,lookup}().
1.56	04-Oct-2025	thorpej	Add a shared function to query the common properties used for configuring an Ethernet address.
1.55	15-Feb-2025	hannken	Use "int" format for constant "GEM_TMP_BUFSIZE - 1". Kernel ALL/amd64 builds again.
1.54	13-Feb-2025	joe	use the right buffer size for shared pin check and debug avoid the use of sizeof(pointer) and use the Temp BUF limit
1.53	13-Feb-2025	joe	fix: wrong size of buffer limit used when reading through the pci bus use the GEM_TMP_BUFSIZE rather than the sizeof pointer
1.52	02-Jul-2020	msaitoh	branches: 1.52.20; 1.52.26; prop_data_data_nocopy -> prop_data_value
1.51	30-May-2020	thorpej	gem_pci_attach(): avoid allocating a 2K buffer on the stack.
1.50	02-Mar-2020	msaitoh	KNF. No functional change.
1.49	01-Mar-2020	thorpej	Use the 64-bit PCI DMA tag if available.
1.48	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.47	29-Mar-2014	christos	branches: 1.47.28; 1.47.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.46	30-Mar-2013	christos	branches: 1.46.4; remove trailing whitespace
1.45	23-Sep-2012	chs	fix a debug printf.
1.44	30-Jan-2012	drochner	branches: 1.44.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.43	13-Nov-2010	uebayasi	branches: 1.43.8; 1.43.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.42	24-Feb-2010	dyoung	branches: 1.42.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.41	21-Jan-2010	macallan	branches: 1.41.2; get rid of architecture-specific firmware calls to determine the MAC address on sparc(64) and macppc - use device properties instead tested by myself on macppc and martin on sparc64
1.40	19-Jan-2010	pooka	Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.39	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.38	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.37	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.36	27-Jul-2009	dyoung	Do a complete device_t/softc split for the PCI attachment. I have not finished the device_t/softc split for the SBus attachment because I don't have an SBus gem(4) to test with. Convert from legacy shutdownhooks to a PMF shutdown hook. Add PMF suspend/resume handlers. Factor a detachment hook out of gem_attach(). During device attachment, track which resources are reserved in sc_attach_state, and release only those resources during detachment. Tested on gem0 and gem1 at pci1 on a Sun Fire V120.
1.35	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.34	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.31	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.30	25-May-2008	chs	branches: 1.30.6; 1.30.12; recognize the variant found in the last model of powermac G5.
1.29	10-Apr-2008	cegger	branches: 1.29.2; 1.29.4; 1.29.6; use aprint_*_dev and device_xname
1.28	04-Apr-2008	martin	Make this compile w/o GEM_DEBUG on non-sparc systems
1.27	01-Feb-2008	jdc	branches: 1.27.6; Set the chip revision in the softc.
1.26	19-Jan-2008	aymeric	recognize (again) the Intrepid 2 gem adapter in the Powerbook5,8
1.25	05-Jan-2008	jdc	Add support for Sun PCI X1141A (SX fibre) cards. Add support for GEM cards on non-OFW machines.
1.24	19-Oct-2007	ad	branches: 1.24.2; 1.24.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.23	25-Jun-2007	aymeric	branches: 1.23.6; 1.23.8; 1.23.12; Recognize the Intrepid2 gem. Needed at least on the last PowerBook G4.
1.22	24-Nov-2006	martin	branches: 1.22.2; 1.22.8; 1.22.10; 1.22.16; In preparation for gem @ sbus: split the GEM registers into two halfs and provide a bus_space_handle_t for each. While there, fix a typo in gem_bitwait().
1.21	05-Aug-2006	sanjayl	branches: 1.21.4; 1.21.6; 1st cut of Powermac G5 support (uses bridge mode).
1.20	11-Dec-2005	christos	branches: 1.20.4; 1.20.8; merge ktrace-lwp.
1.19	27-Feb-2005	perry	branches: 1.19.4; nuke trailing whitespace
1.18	04-Feb-2005	perry	de-__P
1.17	17-Oct-2004	heas	branches: 1.17.4; 1.17.6; Adjust/add a few comments and add defines for PCS (aka GEM_MII) FSM and interrupt registers. drop some trailing WS. no functional changes.
1.16	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.15	17-Mar-2004	martin	For sparc*: myetheraddr is no more
1.14	31-Jan-2003	thorpej	branches: 1.14.2; Use aprint_*().
1.13	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	09-Jul-2002	matt	Match Apple GMAC id #3.
1.9	21-May-2002	matt	branches: 1.9.2; Add code to distinguish between apple and sun gem cores.
1.8	13-Nov-2001	lukem	branches: 1.8.2; 1.8.10; add RCSID
1.7	18-Oct-2001	thorpej	Remove more cruft left over from copying bits of other drivers.
1.6	18-Oct-2001	thorpej	Remove some more stuff left over from copying bits of the HME driver.
1.5	18-Oct-2001	thorpej	Remove some unneeded stuff from the gem_pci_softc.
1.4	18-Oct-2001	thorpej	Add goo to fetch the MAC address from OpenFirmware on the macppc port. Copied from the macppc "gmac" driver, but otherwise untested.
1.3	18-Oct-2001	thorpej	- The GEM registers are always memory type registers. Don't use a broken hueristic to determine memory vs. i/o (one should never make an assumption that the bus_space_tag_t is a pointer, as this code did). - Fix the "can't map registers" error message. - Garbage-collect some code that is not relevant to the GEM (which was already #if 0'd out). - Cluster all the SPARC-specific code into one place (will be replaced with Properties once that is fleshed out).
1.2	24-Sep-2001	eeh	branches: 1.2.2; Print out device details.
1.1	16-Sep-2001	eeh	branches: 1.1.2; Driver for Sun GEM gigabit ethernet, Sun ERI 10/100, and Apple GMAC.
1.1.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.7	01-Aug-2002	nathanw	Catch up to -current.
1.1.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.5	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.4	22-Oct-2001	nathanw	Catch up to -current.
1.1.2.3	26-Sep-2001	nathanw	Catch up to -current. Again.
1.1.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.2.1	16-Sep-2001	nathanw	file if_gem_pci.c was added on branch nathanw_sa on 2001-09-21 22:35:55 +0000
1.2.2.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.2.1	24-Sep-2001	fvdl	file if_gem_pci.c was added on branch thorpej-devvp on 2001-10-01 12:45:54 +0000
1.8.10.2	15-Jul-2002	gehenna	catch up with -current.
1.8.10.1	30-May-2002	gehenna	Catch up with -current.
1.8.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.8.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.8.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.1	13-Nov-2001	thorpej	file if_gem_pci.c was added on branch kqueue on 2002-01-10 19:56:40 +0000
1.9.2.1	12-Jul-2002	thorpej	pullup-1-6 ticket #491 (matt). Original log message: Match Apple GMAC id #3.
1.14.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.14.2.4	19-Oct-2004	skrll	Sync with HEAD
1.14.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.1	03-Aug-2004	skrll	Sync with HEAD
1.17.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.17.6.1	12-Feb-2005	yamt	sync with head.
1.17.4.1	29-Apr-2005	kent	sync with -current
1.19.4.5	04-Feb-2008	yamt	sync with head.
1.19.4.4	21-Jan-2008	yamt	sync with head
1.19.4.3	27-Oct-2007	yamt	sync with head.
1.19.4.2	03-Sep-2007	yamt	sync with head.
1.19.4.1	30-Dec-2006	yamt	sync with head.
1.20.8.1	11-Aug-2006	yamt	sync with head
1.20.4.1	09-Sep-2006	rpaulo	sync with head
1.21.6.1	10-Dec-2006	yamt	sync with head.
1.21.4.1	12-Jan-2007	ad	Sync with head.
1.22.16.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.22.10.1	11-Jul-2007	mjf	Sync with head.
1.22.8.2	23-Oct-2007	ad	Sync with head.
1.22.8.1	15-Jul-2007	ad	Sync with head.
1.22.2.1	17-May-2008	bouyer	Pull up following revision(s) (requested by jdc in ticket #1145): src/share/man/man4/gem.4 1.5,1.6 via patch src/sys/arch/sparc64/conf/GENERIC 1.97 via patch src/sys/dev/ic/gem.c 1.59,1.61-1.68,1.70-1.73 via patch src/sys/dev/ic/gemreg.h 1.11 via patch src/sys/dev/ic/gemvar.h 1.16,1.18 via patch src/sys/dev/mii/bmtphy.c 1.24 via patch src/sys/dev/pci/if_gem_pci.c 1.25,1.27 via patch src/sys/dev/pci/pcidevs 1.12 via patch src/sys/dev/pci/pcidevs.h regenerate src/sys/dev/pci/pcidevs_data.h regenerate Add support for Sun PCI X1141A (SX fibre) cards. Add improvements to the driver. Disable HW UDP4 and RX TCP4 checksums because they are buggy (PR#34799) Add improvements to bmtphy.
1.23.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.23.8.3	23-Mar-2008	matt	sync with HEAD
1.23.8.2	09-Jan-2008	matt	sync with HEAD
1.23.8.1	06-Nov-2007	matt	sync with HEAD
1.23.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.24.8.2	20-Jan-2008	bouyer	Sync with HEAD
1.24.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.24.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.27.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.29.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.29.4.5	11-Mar-2010	yamt	sync with head
1.29.4.4	16-Sep-2009	yamt	sync with head
1.29.4.3	19-Aug-2009	yamt	sync with head.
1.29.4.2	16-May-2009	yamt	sync with head
1.29.4.1	04-May-2009	yamt	sync with head.
1.29.2.1	04-Jun-2008	yamt	sync with head
1.30.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.30.6.1	28-Apr-2009	skrll	Sync with HEAD.
1.41.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.42.2.1	05-Mar-2011	rmind	sync with head
1.43.12.1	18-Feb-2012	mrg	merge to -current.
1.43.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.43.8.2	30-Oct-2012	yamt	sync with head
1.43.8.1	17-Apr-2012	yamt	sync with head
1.44.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.44.6.2	23-Jun-2013	tls	resync from head
1.44.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.46.4.1	18-May-2014	rmind	sync with head
1.47.30.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.47.30.1	10-Jun-2019	christos	Sync with HEAD
1.47.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.52.26.1	02-Aug-2025	perseant	Sync with HEAD
1.52.20.2	22-Feb-2025	martin	Pull up following revision(s) (requested by joe in ticket #1054): sys/dev/pci/if_gem_pci.c: revision 1.54 sys/dev/pci/if_gem_pci.c: revision 1.55 use the right buffer size for shared pin check and debug avoid the use of sizeof(pointer) and use the Temp BUF limit. Use "int" format for constant "GEM_TMP_BUFSIZE - 1". Kernel ALL/amd64 builds again.
1.52.20.1	22-Feb-2025	martin	Pull up following revision(s) (requested by joe in ticket #1053): sys/dev/pci/if_gem_pci.c: revision 1.53 fix: wrong size of buffer limit used when reading through the pci bus use the GEM_TMP_BUFSIZE rather than the sizeof pointer
1.3	16-Dec-2007	dyoung	Collect garbage. All of this code is in pci/if_sip.c, now.
1.2	13-Nov-2001	lukem	branches: 1.2.32; 1.2.90; 1.2.96; 1.2.100; 1.2.104; add RCSID
1.1	18-May-2001	thorpej	branches: 1.1.2; 1.1.4; Add a driver for the National Semiconductor DP83820 Gigabit Ethernet chip. This is found on the NetGear GA-622 and Asante FriendlyNet GigaNIX.
1.1.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.1	18-May-2001	nathanw	file if_gsip.c was added on branch nathanw_sa on 2001-06-21 20:04:43 +0000
1.2.104.1	02-Jan-2008	bouyer	Sync with HEAD
1.2.100.1	26-Dec-2007	ad	Sync with head.
1.2.96.1	27-Dec-2007	mjf	Sync with HEAD.
1.2.90.1	23-Mar-2008	matt	sync with HEAD
1.2.32.1	21-Jan-2008	yamt	sync with head
1.41	04-Oct-2025	thorpej	Add a shared function to query the common properties used for configuring an Ethernet address.
1.40	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.39	02-Jul-2020	msaitoh	prop_data_data_nocopy -> prop_data_value
1.38	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.37	29-Mar-2014	christos	branches: 1.37.28; 1.37.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.36	10-May-2011	dyoung	branches: 1.36.4; 1.36.14; 1.36.18; Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.35	23-Mar-2010	dyoung	branches: 1.35.2; In the #definition of PCI_HME_BASEADDR, use PCI_BAR(0) instead of the anonymous constant 0x10.
1.34	23-Mar-2010	dyoung	Put back PCI_HME_BASEADDR.
1.33	23-Mar-2010	dyoung	Use the PCI BAR type to decide which I/O enables to set. TBD: Put back the old BAR #definition, s/PCI_BAR(0)/PCI_HME_BASEADDR/.
1.32	11-Mar-2010	mrg	branches: 1.32.2; various aprint_* fixes.
1.31	20-Jan-2010	martin	branches: 1.31.2; Remove sparc specific code and HME_USE_LOCAL_MAC_ADDRESS hack, instead use the "mac-address" device property if present.
1.30	17-May-2009	tsutsui	u_intNN_t -> uintNN_t
1.29	17-May-2009	tsutsui	Split device_t/softc. Tested PCI HME on i386. Sbus HME is compile test only.
1.28	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.27	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.26	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.25	29-May-2008	mrg	branches: 1.25.12; remove clause #3 from my license where there are no other copyright holders involved.
1.24	10-Apr-2008	cegger	branches: 1.24.2; 1.24.4; 1.24.6; use aprint_*_dev and device_xname
1.23	19-Oct-2007	ad	branches: 1.23.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.22	16-Nov-2006	christos	branches: 1.22.8; 1.22.22; 1.22.24; 1.22.28; __unused removal on arguments; approved by core.
1.21	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.20	11-Dec-2005	christos	branches: 1.20.20; 1.20.22; merge ktrace-lwp.
1.19	26-Apr-2005	atatat	branches: 1.19.2; Reverse order of args in hmevpdoff()
1.18	21-Apr-2005	christos	simplify the mac address retrieval code.
1.17	21-Apr-2005	christos	PR/30018: Peter Eisch: Sun QFE support in non-sparc hardware
1.16	27-Feb-2005	perry	branches: 1.16.2; nuke trailing whitespace
1.15	04-Feb-2005	perry	de-__P
1.14	17-Mar-2004	martin	branches: 1.14.6; 1.14.8; 1.14.10; Fix hme for pci based sparcs (krups).
1.13	26-Dec-2002	itohy	branches: 1.13.2; Lookup and use the Ethernet address in the FCode PROM, and avoid using MD myetheraddr() function. This makes the driver MI, and closes PR kern/13797. The PCI HME is a PCIO chip, which is composed of two functions: function 0: PCI-EBus2 bridge, and function 1: HappyMeal Ethernet controller. The Ethernet address is (expected to be) in the PCI FCode PROM connected to the EBus bridge (function 0) of the device. Since the HME is on function 1, some magic is used to access to the PROM. We don't have MI EBus driver since no EBus device exists (besides the FCode PROM) on add-on HME boards. The ``not configured'' message for function 0 is what is expected. The SPARC case is currently unchanged. It needs interaction with OpenBoot.
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	13-Nov-2001	lukem	add RCSID
1.8	18-Oct-2001	thorpej	Add a missing return statement in an error case, and g/c some cruft that's not relevant in the PCI world.
1.7	05-Oct-2001	thorpej	Minor tidying up of the HME driver attach: - Print nicer-looking message. - Use bus_space_subregion() when appropriate, rather than arithmetic on a bus_space_handle_t.
1.6	26-Sep-2001	eeh	getprop* -> PROM_getprop*
1.5	27-Aug-2001	augustss	branches: 1.5.2; Use #include <machine/intr.h> instead of #include <machine/cpu.h> to improve portability.
1.4	27-Aug-2001	augustss	Remove an unneeded #include. This makes the file compilable on more platforms.
1.3	28-Dec-2000	sommerfeld	branches: 1.3.2; 1.3.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.2	25-Jun-2000	eeh	branches: 1.2.2; Get PCI HME working.
1.1	05-Apr-2000	mrg	branches: 1.1.4; add "hme at pci" support.
1.1.4.1	31-Jul-2000	mrg	pull up 1.2 (approved by thorpej): >Get PCI HME working.
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.1	25-Jun-2000	bouyer	file if_hme_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:22 +0000
1.3.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.3.2.6	29-Dec-2002	thorpej	Sync with HEAD.
1.3.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.3	22-Oct-2001	nathanw	Catch up to -current.
1.3.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.3.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.5.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.5.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.13.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.13.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.13.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.2.1	03-Aug-2004	skrll	Sync with HEAD
1.14.10.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.10.1	12-Feb-2005	yamt	sync with head.
1.14.8.1	29-Apr-2005	kent	sync with -current
1.14.6.1	12-May-2005	riz	Pull up revisions 1.15-1.19 (requested by christos in ticket #1444): revision 1.19 date: 2005/04/26 13:02:06; author: atatat; state: Exp; lines: +3 -3 Reverse order of args in hmevpdoff()
1.18	21-Apr-2005	christos	simplify the mac address retrieval code.
1.17	21-Apr-2005	christos	PR/30018: Peter Eisch: Sun QFE support in non-sparc hardware
1.16	27-Feb-2005	perry	nuke trailing whitespace
1.15	04-Feb-2005	perry	de-__P
1.16.2.3	26-Apr-2005	tron	Pull up revision 1.19 (requested by atatat in ticket #208): Reverse order of args in hmevpdoff()
1.16.2.2	25-Apr-2005	tron	Pull up revision 1.18 (requested by christos in ticket #191): simplify the mac address retrieval code.
1.16.2.1	25-Apr-2005	tron	Pull up revision 1.17 (requested by christos in ticket #191): PR/30018: Peter Eisch: Sun QFE support in non-sparc hardware
1.19.2.2	27-Oct-2007	yamt	sync with head.
1.19.2.1	30-Dec-2006	yamt	sync with head.
1.20.22.2	10-Dec-2006	yamt	sync with head.
1.20.22.1	22-Oct-2006	yamt	sync with head
1.20.20.1	18-Nov-2006	ad	Sync with head.
1.22.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.22.24.1	06-Nov-2007	matt	sync with HEAD
1.22.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.22.8.1	23-Oct-2007	ad	Sync with head.
1.23.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.24.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.24.4.5	11-Aug-2010	yamt	sync with head.
1.24.4.4	11-Mar-2010	yamt	sync with head
1.24.4.3	20-Jun-2009	yamt	sync with head
1.24.4.2	16-May-2009	yamt	sync with head
1.24.4.1	04-May-2009	yamt	sync with head.
1.24.2.1	04-Jun-2008	yamt	sync with head
1.25.12.2	23-Jul-2009	jym	Sync with HEAD.
1.25.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.31.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.32.2.2	31-May-2011	rmind	sync with head
1.32.2.1	30-May-2010	rmind	sync with head
1.35.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.36.18.1	18-May-2014	rmind	sync with head
1.36.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.36.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.30.1	10-Jun-2019	christos	Sync with HEAD
1.37.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.20	23-Mar-2025	joe	avoid NULL pointer dereference on failed memory allocation for sc_ihs
1.19	25-Feb-2025	joe	initialize post to 0 this prevents the use of unitialized variable when we hit an error branch on the first attempt/iteration that will cause it to break out of the do while loop
1.18	29-Jun-2024	riastradh	branches: 1.18.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.17	02-Jun-2024	andvar	s/interuppts/interrupts/ and s/intruppts/interrupts/ in comment and log message.
1.16	17-Jun-2022	yamaguchi	ixl(4), iavf(4): fix endian bug in vlan tag
1.15	06-Nov-2021	andvar	s/invlaid/invalid/
1.14	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.13	05-Mar-2021	yamaguchi	branches: 1.13.4; Fix missing IFM_FDX
1.12	09-Feb-2021	jakllsch	Zero interrupt handles upon allocation. Might prevent detach-time bugs.
1.11	10-Dec-2020	yamaguchi	Add a string literal for IAVF_VC_OP_CONFIG_RSS_LUT
1.10	10-Dec-2020	yamaguchi	remove unnecessary lock acquire and release
1.9	10-Dec-2020	yamaguchi	Fix a wrong value about HENA configuration
1.8	08-Dec-2020	yamaguchi	Change the default interrupt-affinity of iavf(4)
1.7	01-Dec-2020	yamaguchi	Dequeue aqb from sc_atq_live even when the last command is failed iavf(4) didn't dequeue aqb from sc_atq_live that is a list for buffer in use when a command is failed by ETIMEDOUT. This causes a panic in the following sequence: 1. enqueue an aqb to sc_atq_live at iavf_aqb_post() 2. the last command is failed by ETIMEDOUT 3. enqueue the aqb used in the failed command to sc_atq_idle at an error handling in iavf_attach() 4. dequeue the same aqb from sc_atq_live and enqueue sc_atq_idle again at iavf_cleanup_admin_queue() - sc_atq_idle is broken at that time 5. free the aqb in sc_atq_idle more than once Fix PR/55822 reviewed by knakahara@n.o.
1.6	17-Sep-2020	yamaguchi	branches: 1.6.2; Fix a panic caused while detaching iavf(4) The sequence of the panic is follow: 1. failed to disable queues in iavf_stop() 2. release ec->ec_lock in ether_ifdetach() 3. do reset scheduled at iavf_stop() 4. acquire ec->ec_lock in iavf_reset_finish() - panic reviewed by knakahara@n.o, thanks.
1.5	10-Sep-2020	yamaguchi	Fix a possible deadlock in iavf(4)
1.4	09-Sep-2020	yamaguchi	Added the explain about locking order in iavf(4)
1.3	09-Sep-2020	jakllsch	fix bad conditional in iavf_parse_modprop() found by clang
1.2	08-Sep-2020	jakllsch	iaq_datalen is 16-bit, always use htole16 fixes build failure on big endian
1.1	08-Sep-2020	yamaguchi	Added iavf(4) that is based on OpenBSD's iavf(4) implementation reviewed by msaitoh@n.o and knakahara@n.o
1.6.2.2	03-Apr-2021	thorpej	Sync with HEAD.
1.6.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.13.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.18.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1	08-Sep-2020	yamaguchi	Added iavf(4) that is based on OpenBSD's iavf(4) implementation reviewed by msaitoh@n.o and knakahara@n.o
1.77	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.76	22-Aug-2022	thorpej	ipw_start(): Replace "IF_DEQUEUE() -> IF_PREPEND() on failure" with "IF_POLL() -> IF_DEQUEUE() on success".
1.75	22-Aug-2022	thorpej	We come into our if_start routine via ether_output(), therefore there is no need to check-and-pullup to sizeof(struct ether_header). Instead, we can simply assert it.
1.74	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.73	30-Jan-2020	thorpej	branches: 1.73.10; Adopt <net/if_stats.h>.
1.72	03-Feb-2019	mrg	branches: 1.72.6; - add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.71	22-Dec-2018	maxv	Replace: M_COPY_PKTHDR -> m_copy_pkthdr. No functional change, since the former is a macro to the latter.
1.70	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.69	26-Jun-2018	msaitoh	branches: 1.69.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.68	01-May-2018	maya	GC private 802.11 rateset declarations, use the standard ones. Build tested only.
1.67	16-Jan-2018	maxv	branches: 1.67.2; Fix overflow.
1.66	23-Oct-2017	msaitoh	If if_initialize() failed in the attach function, free resources and return.
1.65	29-Jul-2017	riastradh	Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.64	23-May-2017	ozaki-r	branches: 1.64.2; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.63	02-Feb-2017	nonaka	wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.62	01-Feb-2017	nonaka	sync only mapped size.
1.61	08-Dec-2016	ozaki-r	branches: 1.61.2; Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.60	10-Jun-2016	ozaki-r	branches: 1.60.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.59	20-Nov-2015	christos	use copyout instead of suword.
1.58	07-Jan-2015	ozaki-r	Pass a correct firmware size (instead of 0) to firmware_free firmware_free now uses kmem_free(9) instead of free(9), so we need to pass a correct size to it.
1.57	29-Mar-2014	christos	branches: 1.57.4; 1.57.6; 1.57.8; 1.57.12; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.55	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.54	27-Oct-2012	chs	branches: 1.54.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.53	30-Jan-2012	drochner	branches: 1.53.2; 1.53.6; 1.53.8; 1.53.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.52	15-Nov-2010	uebayasi	branches: 1.52.8; 1.52.12; tsleep needs sys/proc.h.
1.51	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.50	19-Jan-2010	pooka	branches: 1.50.2; 1.50.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.49	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.48	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.47	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.46	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.45	09-Jan-2009	jmcneill	branches: 1.45.2; Make the "EULA not accepted" message a bit more informative, now reads: EULA not accepted; please see the <driver>(4) man page.
1.44	07-Nov-2008	joerg	Match image name with upstream distfiles.
1.43	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.42	30-Oct-2008	joerg	branches: 1.42.2; Add glue to ship Intel firmware images. For ipw and iwi require user to accept the EULA via sysctl as discussed with core@.
1.41	27-May-2008	dyoung	branches: 1.41.4; 1.41.6; Initialize the channel info in the tx/rx radiotap headers in ipw_newstate(), instead of reinitializing every time a packet is sent or received. Drop the unused radiotap 'flags' field.
1.40	25-May-2008	jmcneill	PR#34477: Don't dereference ni_chan if it points to the special ``any channel'' token IEEE80211_CHAN_ANYC
1.39	10-Apr-2008	cegger	branches: 1.39.2; 1.39.4; 1.39.6; use aprint_*_dev and device_xname
1.38	11-Mar-2008	dyoung	pci_disable_retry() is now a no-op, so don't call it any more.
1.37	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.36	20-Dec-2007	dyoung	branches: 1.36.2; 1.36.6; Constify struct ifnet->if_sadl and every use throughout the tree. Add if_set_sadl() that both sets the link-layer address length and replaces the current link-layer address with a new one, and use it throughout the tree.
1.35	09-Dec-2007	jmcneill	branches: 1.35.2; Merge jmcneill-pm branch.
1.34	19-Oct-2007	ad	branches: 1.34.4; 1.34.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.33	01-Sep-2007	dyoung	branches: 1.33.4; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.32	04-Mar-2007	christos	branches: 1.32.2; 1.32.10; 1.32.14; 1.32.16; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.31	16-Nov-2006	christos	branches: 1.31.2; 1.31.4; 1.31.8; __unused removal on arguments; approved by core.
1.30	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.29	29-Sep-2006	christos	Coverity CID 4151: Don't forget to close the firmware file (from Arnaud Lacombe)
1.28	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.27	02-Aug-2006	rpaulo	branches: 1.27.4; 1.27.6; Don't read the firmware twice. Noticed by Nick.
1.26	28-May-2006	blymn	Clean up bogus whitespace
1.25	01-May-2006	rpaulo	branches: 1.25.2; aprint_debug -> aprint_error.
1.24	29-Apr-2006	rpaulo	Describe the capabilities supported.
1.23	29-Apr-2006	rpaulo	There's no power management support under ipw, so don't announce it to the net80211 stack.
1.22	28-Apr-2006	rpaulo	bzero -> memset bcopy -> memcpy
1.21	28-Apr-2006	rpaulo	Print the Ethernet address on attachement.
1.20	18-Apr-2006	rpaulo	Use strlcpy instead of strncpy and sizeof. Requested by Luke.
1.19	17-Apr-2006	rpaulo	Remove SIOCKILLFW.
1.18	17-Apr-2006	rpaulo	Oops, 16 is not enough to store the firmware filenames, use 24.
1.17	17-Apr-2006	rpaulo	Use the firmload API to retrive the firmware from the filesystem.
1.16	24-Dec-2005	perry	branches: 1.16.4; 1.16.6; 1.16.8; 1.16.10; 1.16.12; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.15	20-Dec-2005	skrll	Merge the FreeBSD code and adapt to NetBSD: - add power management. - update ioctl handling. Fix PR 27439 as the FreeBSD code has joerg's fix for when the RX intr routine can't allocate a new mbuf. Fix PR 30806. Thanks to cube@ for testing and some struct ipw_soft_hdr handling code. XXX The state machine doesn't sync with the net80211 layer, but that's XXX not a surprise.
1.14	11-Dec-2005	christos	merge ktrace-lwp.
1.13	06-Jul-2005	dyoung	Avoid an unnecessary API difference between NetBSD and FreeBSD: back out my change to ieee80211_crypto_encap that made it free its mbuf argument on error. I had thought it was a bug. It was not. It's the drivers that are broken. Make an(4), atw(4), ipw(4), iwi(4), ral(4), rtw(4), ural(4), and wi(4) free the mbuf when ieee80211_crypto_encap returns NULL. Also, return ath(4) to the way it was---i.e., free the mbuf. Thanks to Sam Leffler to pointing out my mistake.
1.12	22-Jun-2005	dyoung	branches: 1.12.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.11	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.10	19-Jan-2005	dyoung	Keep a count of free transmit/command descriptors, sc->txfree. Make the transmit section reserve one descriptor for issuing a command at all times. If either transmit descriptors run out, or header/buffer software descriptors run out, set IFF_OACTIVE and get out of ipw_start, rather than re-using a descriptor or trying to use a NULL descriptor that comes off the front of an empty descriptor tailqueue. This ought to fix port-i386/27439 and kern/28683.
1.9	18-Jan-2005	dyoung	Use IFQ_DEQUEUE instead of IF_DEQUEUE on if->if_snd to be consistent with IFQ_SET_READY(&if->if_snd) elsewhere, and also to enable ALTQ(9) on ipw(4).
1.8	14-Sep-2004	lukem	branches: 1.8.4; Whitespace tweaks. Remove vendor RCSID, since it's out of sync already (due to manual patches).
1.7	14-Sep-2004	lukem	Use ieee80211_find_rxnode() function instead of ieee80211_find_node() and ieee80211_ref_node(). From Damien Bergamini.
1.6	14-Sep-2004	lukem	Correctly synchronize DMA mapping for transmission buffer descriptors. From Damien Bergamini.
1.5	14-Sep-2004	lukem	Improve monitor mode support. From Damien Bergamini
1.4	14-Sep-2004	lukem	Add IEEE802.11 radiotap support. From Damien Bergamini.
1.3	27-Aug-2004	lukem	Minor cosmetic consistency tweaks, from Damien.
1.2	23-Aug-2004	lukem	branches: 1.2.2; NetBSD RCSID
1.1	23-Aug-2004	lukem	branches: 1.1.1; Initial revision
1.1.1.2	20-Dec-2005	skrll	Import of the ipw(4) FreeBSD sources from 14-Nov-2005 (just after the last net80211 merge) as a step towards a working driver and to make future merging easier.
1.1.1.1	23-Aug-2004	lukem	ipw(4) - Intel PRO/Wireless 2100 driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/
1.2.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.6	24-Jan-2005	skrll	Sync with HEAD.
1.2.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.3	03-Sep-2004	skrll	Sync with HEAD
1.2.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.2.2.1	23-Aug-2004	skrll	file if_ipw.c was added on branch ktrace-lwp on 2004-08-25 06:58:05 +0000
1.8.4.1	29-Apr-2005	kent	sync with -current
1.12.2.6	17-Mar-2008	yamt	sync with head.
1.12.2.5	21-Jan-2008	yamt	sync with head
1.12.2.4	27-Oct-2007	yamt	sync with head.
1.12.2.3	03-Sep-2007	yamt	sync with head.
1.12.2.2	30-Dec-2006	yamt	sync with head.
1.12.2.1	21-Jun-2006	yamt	sync with head.
1.16.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.16.10.4	11-May-2006	elad	sync with head
1.16.10.3	19-Apr-2006	elad	sync with head.
1.16.10.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.16.10.1	08-Mar-2006	elad	Adapt to kernel authorization KPI.
1.16.8.3	11-Aug-2006	yamt	sync with head
1.16.8.2	26-Jun-2006	yamt	sync with head.
1.16.8.1	24-May-2006	yamt	sync with head.
1.16.6.2	01-Jun-2006	kardel	Sync with head.
1.16.6.1	22-Apr-2006	simonb	Sync with head.
1.16.4.1	09-Sep-2006	rpaulo	sync with head
1.25.2.1	19-Jun-2006	chap	Sync with head.
1.27.6.2	10-Dec-2006	yamt	sync with head.
1.27.6.1	22-Oct-2006	yamt	sync with head
1.27.4.1	18-Nov-2006	ad	Sync with head.
1.31.8.1	04-Sep-2008	skrll	Sync with netbsd-4.
1.31.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.31.2.1	08-Jun-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1156): sys/dev/pci/if_ipw.c: revision 1.40 PR#34477: Don't dereference ni_chan if it points to the special ``any channel'' token IEEE80211_CHAN_ANYC
1.32.16.3	23-Mar-2008	matt	sync with HEAD
1.32.16.2	09-Jan-2008	matt	sync with HEAD
1.32.16.1	06-Nov-2007	matt	sync with HEAD
1.32.14.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.32.14.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.32.14.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.32.14.2	05-Oct-2007	joerg	Convert ipw(4) to PNP power management and use pci_disable_retry.
1.32.14.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.32.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.32.2.2	23-Oct-2007	ad	Sync with head.
1.32.2.1	09-Oct-2007	ad	Sync with head.
1.33.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.34.6.1	11-Dec-2007	yamt	sync with head.
1.34.4.1	26-Dec-2007	ad	Sync with head.
1.35.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.36.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.36.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.36.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.36.2.1	24-Mar-2008	keiichi	sync with head.
1.39.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.39.4.5	11-Aug-2010	yamt	sync with head.
1.39.4.4	11-Mar-2010	yamt	sync with head
1.39.4.3	16-Sep-2009	yamt	sync with head
1.39.4.2	16-May-2009	yamt	sync with head
1.39.4.1	04-May-2009	yamt	sync with head.
1.39.2.1	04-Jun-2008	yamt	sync with head
1.41.6.1	19-Jan-2009	skrll	Sync with HEAD.
1.41.4.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.42.2.2	14-Jan-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #250): sys/dev/pci/if_iwi.c: revision 1.78 sys/dev/pci/if_ipw.c: revision 1.45 Make the "EULA not accepted" message a bit more informative, now reads: EULA not accepted; please see the <driver>(4) man page.
1.42.2.1	09-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #25): sys/dev/pci/if_ipw.c: revision 1.44 sys/dev/pci/if_ipwvar.h: revision 1.14 sys/dev/pci/if_iwi.c: revision 1.76 sys/dev/pci/if_iwn.c: revision 1.26 Match image name with upstream distfiles.
1.45.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.4.2	05-Mar-2011	rmind	sync with head
1.50.4.1	30-May-2010	rmind	sync with head
1.50.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.52.12.1	18-Feb-2012	mrg	merge to -current.
1.52.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.52.8.2	30-Oct-2012	yamt	sync with head
1.52.8.1	17-Apr-2012	yamt	sync with head
1.53.10.1	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1473): sys/dev/pci/if_ipw.c: revision 1.65 via patch Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.53.8.1	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1473): sys/dev/pci/if_ipw.c: revision 1.65 via patch Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.53.6.3	03-Dec-2017	jdolecek	update from HEAD
1.53.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.53.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.53.2.1	18-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1473): sys/dev/pci/if_ipw.c: revision 1.65 via patch Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.54.2.1	18-May-2014	rmind	sync with head
1.57.12.1	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1469): sys/dev/pci/if_ipw.c: revision 1.65 Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.57.8.1	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1469): sys/dev/pci/if_ipw.c: revision 1.65 Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.57.6.5	28-Aug-2017	skrll	Sync with HEAD
1.57.6.4	05-Feb-2017	skrll	Sync with HEAD
1.57.6.3	09-Jul-2016	skrll	Sync with HEAD
1.57.6.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.57.6.1	06-Apr-2015	skrll	Sync with HEAD
1.57.4.1	12-Aug-2017	snj	Pull up following revision(s) (requested by mrg in ticket #1469): sys/dev/pci/if_ipw.c: revision 1.65 Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.60.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.60.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.61.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.64.2.2	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.64.2.1	05-Aug-2017	snj	Pull up following revision(s) (requested by spz in ticket #182): sys/dev/pci/if_ipw.c: revision 1.65 Null out sbuf->m on failure to avoid double-free later. From Ilja Van Sprundel. Also null out sbuf->map out of paranoia.
1.67.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.67.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.67.2.1	02-May-2018	pgoyette	Synch with HEAD
1.69.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.69.2.1	10-Jun-2019	christos	Sync with HEAD
1.72.6.1	29-Feb-2020	ad	Sync with head.
1.73.10.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.6	25-Dec-2007	perry	Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.5	20-Dec-2005	skrll	branches: 1.5.46; 1.5.52; 1.5.56; 1.5.60; Merge the FreeBSD code and adapt to NetBSD: - add power management. - update ioctl handling. Fix PR 27439 as the FreeBSD code has joerg's fix for when the RX intr routine can't allocate a new mbuf. Fix PR 30806. Thanks to cube@ for testing and some struct ipw_soft_hdr handling code. XXX The state machine doesn't sync with the net80211 layer, but that's XXX not a surprise.
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	14-Sep-2004	lukem	branches: 1.3.12; Whitespace tweaks from Damien Bergamini. Whitespace tweaks by me. Remove vendor RCSID, since it's out of sync already (due to manual patchs).
1.2	23-Aug-2004	lukem	branches: 1.2.2; NetBSD RCSID
1.1	23-Aug-2004	lukem	branches: 1.1.1; Initial revision
1.1.1.2	20-Dec-2005	skrll	Import of the ipw(4) FreeBSD sources from 14-Nov-2005 (just after the last net80211 merge) as a step towards a working driver and to make future merging easier.
1.1.1.1	23-Aug-2004	lukem	ipw(4) - Intel PRO/Wireless 2100 driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/
1.2.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.2.2.1	23-Aug-2004	skrll	file if_ipwreg.h was added on branch ktrace-lwp on 2004-08-25 06:58:05 +0000
1.3.12.2	21-Jan-2008	yamt	sync with head
1.3.12.1	21-Jun-2006	yamt	sync with head.
1.5.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.5.56.1	26-Dec-2007	ad	Sync with head.
1.5.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.5.46.1	09-Jan-2008	matt	sync with HEAD
1.19	05-Oct-2019	mrg	remove __packed attribute from struct ieee80211_radiotap_header and all the structures that include it. this should not change anything while avoiding packed vs alignment warnings from GCC 8, and potentially pessimised code generation due to the packed marker (there are no misaligned members, just that the per-device parts may end unaligned.) all consumers of these members are done from the properly aligned packet members directly, or, as a union with a 64 byte member, also properly aligned. codegen didn't appear to change, except for the definition of sizeof(struct driver_[rt]x_radiotap_header) in debug info, which is not directly used anywhere.
1.18	02-Feb-2017	nonaka	branches: 1.18.14; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.17	06-Sep-2015	dholland	branches: 1.17.2; 1.17.4; More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.16	27-Oct-2012	chs	branches: 1.16.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.15	19-Jan-2010	pooka	branches: 1.15.12; 1.15.22; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.14	07-Nov-2008	joerg	Match image name with upstream distfiles.
1.13	27-May-2008	dyoung	branches: 1.13.4; 1.13.6; 1.13.8; Straggler from last commit: drop unused radiotap flags field, set the __packed attribute on the radiotap headers.
1.12	09-Dec-2007	jmcneill	branches: 1.12.10; 1.12.12; 1.12.14; 1.12.16; Merge jmcneill-pm branch.
1.11	17-Apr-2006	rpaulo	branches: 1.11.32; 1.11.34; 1.11.44; 1.11.46; Remove SIOCKILLFW.
1.10	17-Apr-2006	rpaulo	Oops, 16 is not enough to store the firmware filenames, use 24.
1.9	17-Apr-2006	rpaulo	Use the firmload API to retrive the firmware from the filesystem.
1.8	20-Dec-2005	skrll	branches: 1.8.4; 1.8.6; 1.8.8; 1.8.10; 1.8.12; Merge the FreeBSD code and adapt to NetBSD: - add power management. - update ioctl handling. Fix PR 27439 as the FreeBSD code has joerg's fix for when the RX intr routine can't allocate a new mbuf. Fix PR 30806. Thanks to cube@ for testing and some struct ipw_soft_hdr handling code. XXX The state machine doesn't sync with the net80211 layer, but that's XXX not a surprise.
1.7	11-Dec-2005	christos	merge ktrace-lwp.
1.6	22-Jun-2005	dyoung	branches: 1.6.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.5	19-Jan-2005	dyoung	Keep a count of free transmit/command descriptors, sc->txfree. Make the transmit section reserve one descriptor for issuing a command at all times. If either transmit descriptors run out, or header/buffer software descriptors run out, set IFF_OACTIVE and get out of ipw_start, rather than re-using a descriptor or trying to use a NULL descriptor that comes off the front of an empty descriptor tailqueue. This ought to fix port-i386/27439 and kern/28683.
1.4	14-Sep-2004	lukem	branches: 1.4.4; Add IEEE802.11 radiotap support. From Damien Bergamini.
1.3	14-Sep-2004	lukem	Whitespace tweaks from Damien Bergamini. Whitespace tweaks by me. Remove vendor RCSID, since it's out of sync already (due to manual patchs).
1.2	23-Aug-2004	lukem	branches: 1.2.2; NetBSD RCSID
1.1	23-Aug-2004	lukem	branches: 1.1.1; Initial revision
1.1.1.2	20-Dec-2005	skrll	Import of the ipw(4) FreeBSD sources from 14-Nov-2005 (just after the last net80211 merge) as a step towards a working driver and to make future merging easier.
1.1.1.1	23-Aug-2004	lukem	ipw(4) - Intel PRO/Wireless 2100 driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/
1.2.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.5	24-Jan-2005	skrll	Sync with HEAD.
1.2.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.2.2.1	23-Aug-2004	skrll	file if_ipwvar.h was added on branch ktrace-lwp on 2004-08-25 06:58:05 +0000
1.4.4.1	29-Apr-2005	kent	sync with -current
1.6.2.2	21-Jan-2008	yamt	sync with head
1.6.2.1	21-Jun-2006	yamt	sync with head.
1.8.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.8.10.1	19-Apr-2006	elad	sync with head.
1.8.8.1	24-May-2006	yamt	sync with head.
1.8.6.1	22-Apr-2006	simonb	Sync with head.
1.8.4.1	09-Sep-2006	rpaulo	sync with head
1.11.46.1	11-Dec-2007	yamt	sync with head.
1.11.44.1	26-Dec-2007	ad	Sync with head.
1.11.34.1	09-Jan-2008	matt	sync with HEAD
1.11.32.1	05-Oct-2007	joerg	Convert ipw(4) to PNP power management and use pci_disable_retry.
1.12.16.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.12.14.2	11-Mar-2010	yamt	sync with head
1.12.14.1	04-May-2009	yamt	sync with head.
1.12.12.1	04-Jun-2008	yamt	sync with head
1.12.10.2	17-Jan-2009	mjf	Sync with HEAD.
1.12.10.1	02-Jun-2008	mjf	Sync with HEAD.
1.13.8.1	09-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #25): sys/dev/pci/if_ipw.c: revision 1.44 sys/dev/pci/if_ipwvar.h: revision 1.14 sys/dev/pci/if_iwi.c: revision 1.76 sys/dev/pci/if_iwn.c: revision 1.26 Match image name with upstream distfiles.
1.13.6.1	19-Jan-2009	skrll	Sync with HEAD.
1.13.4.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.15.22.2	03-Dec-2017	jdolecek	update from HEAD
1.15.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.15.12.1	30-Oct-2012	yamt	sync with head
1.16.14.2	05-Feb-2017	skrll	Sync with HEAD
1.16.14.1	22-Sep-2015	skrll	Sync with HEAD
1.17.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.17.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.18.14.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.121	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.120	22-Aug-2022	thorpej	branches: 1.120.10; iwi_start(): Replace "IFQ_DEQUEUE() -> IF_PREPEND() on failure" with "IFQ_POLL() -> IFQ_DEQUEUE() on success (or fatal error)".
1.119	22-Aug-2022	thorpej	We come into our if_start routine via ether_output(), therefore there is no need to check-and-pullup to sizeof(struct ether_header). Instead, we can simply assert it.
1.118	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.117	09-Sep-2021	riastradh	sys/dev: Memset zero before copyout. Just in case of uninitialized padding which would lead to kernel stack disclosure. If the compiler can prove the memset redundant then it can optimize it away; otherwise better safe than sorry. I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks; the raidframe(4) change probably doesn't (but doesn't hurt).
1.116	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.115	08-May-2021	thorpej	Use pci_compatible_match().
1.114	20-Mar-2020	thorpej	branches: 1.114.8; This driver sleeps during iwi_media_change(), and thus requires an adaptive mutex for the media lock.
1.113	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.112	10-Nov-2019	chs	branches: 1.112.2; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.111	03-Feb-2019	mrg	branches: 1.111.4; - add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.110	22-Dec-2018	maxv	Replace: M_COPY_PKTHDR -> m_copy_pkthdr. No functional change, since the former is a macro to the latter.
1.109	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.108	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.107	26-Jun-2018	msaitoh	branches: 1.107.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.106	01-May-2018	maya	GC private 802.11 rateset declarations, use the standard ones. Build tested only.
1.105	16-Jan-2018	maxv	branches: 1.105.2; Fix overflow.
1.104	23-Oct-2017	msaitoh	- If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc.
1.103	23-May-2017	ozaki-r	branches: 1.103.2; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.102	02-Feb-2017	nonaka	wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.101	08-Dec-2016	ozaki-r	branches: 1.101.2; Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.100	03-Aug-2016	mlelstv	be quiet about regular intermediate authentication states.
1.99	10-Jun-2016	ozaki-r	branches: 1.99.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.98	07-Jan-2015	ozaki-r	Pass a correct firmware size (instead of 0) to firmware_free firmware_free now uses kmem_free(9) instead of free(9), so we need to pass a correct size to it.
1.97	29-Mar-2014	christos	branches: 1.97.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.96	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.95	26-Nov-2013	roy	iwi_newstate should work along with ieee80211_new_state, not always override it.
1.94	21-Nov-2013	riz	Wrap debug code in {} because "a label can only be part of a statement and a declaration is not a statement." I.e., make it a statement by making it part of a block.
1.93	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.92	30-Mar-2013	christos	branches: 1.92.4; remove trailing whitespace
1.91	02-Jun-2012	dsl	branches: 1.91.2; Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.90	21-Mar-2012	nisimura	Unbreak the endian issue in firmware header decoding. Comfirmed good and running by a powerpc machine.
1.89	30-Jan-2012	drochner	branches: 1.89.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.88	19-Nov-2011	tls	branches: 1.88.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.87	23-May-2011	drochner	branches: 1.87.4; -remove references to crypto/arc4/arc4.* -- the code isn't used anywhere afaics (The confusion comes probably from use of arc4random() at various places, but this lives in libkern and doesn't share code with the former.) -g/c non-implementation of arc4 encryption in swcrypto(4) -remove special casing of ARC4 in crypto(4) -- the point is that it doesn't use an IV, and this fact is made explicit by the new "ivsize" property of xforms
1.86	31-Jan-2011	christos	clear register 0x41 as FreeBSD and OpenBSD do. Update copyright to the latest.
1.85	15-Nov-2010	uebayasi	branches: 1.85.2; 1.85.4; tsleep needs sys/proc.h.
1.84	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.83	19-Jan-2010	pooka	branches: 1.83.2; 1.83.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.82	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.81	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.80	13-Mar-2009	jym	Fixes in the iwi(4) driver to better handle errors: - bus_dmamap fixes, with checks against NULL - move the rings' allocation code later in iwi_attach() to avoid NULL pointer dereference if allocation fails - avoid double free() in case of failure during attach. If an allocation fails, do not free the ring directly, as it is handled by iwi_detach() - only set ring->count for RX/TX rings when ring allocation is successful, or else the for loop during detach will fail with a NULL dereference. - call pci_intr_disestablish() if iwi_reset() fails during attach; driver is in an incoherent state, interrupt handler should not be used. Proposed by me on tech-net@, approved by christos@. Compiled and tested with GENERIC and XEN3_DOM0 kernels.
1.79	13-Feb-2009	bouyer	make iwi_alloc_tx_ring() prototype match function declaration (shows up when bus_addr_t != bus_size_t). Fix cast for 64bit paddr_t on i386.
1.78	09-Jan-2009	jmcneill	branches: 1.78.2; Make the "EULA not accepted" message a bit more informative, now reads: EULA not accepted; please see the <driver>(4) man page.
1.77	10-Nov-2008	joerg	Fix typo in fw name.
1.76	07-Nov-2008	joerg	Match image name with upstream distfiles.
1.75	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.74	30-Oct-2008	joerg	branches: 1.74.2; Add glue to ship Intel firmware images. For ipw and iwi require user to accept the EULA via sysctl as discussed with core@.
1.73	16-Jun-2008	mlelstv	branches: 1.73.2; 1.73.4; Compute the correct number of pages spanned by the firmware. The old code just uses the firmware size and rounds up to full pages. However, if the firmware isn't loaded aligned to a page boundary (which it isn't), then one more page (and thus DMA segment) might be necessary.
1.72	21-Mar-2008	dyoung	branches: 1.72.2; 1.72.4; 1.72.6; 1.72.8; pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.71	11-Mar-2008	dyoung	Finish the job: pci_disable_retry() is now a no-op, so don't call it any more.
1.70	11-Mar-2008	dyoung	pci_disable_retry() is now a no-op, so don't call it any more.
1.69	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.68	09-Dec-2007	jmcneill	branches: 1.68.6; 1.68.10; Merge jmcneill-pm branch.
1.67	22-Oct-2007	joerg	branches: 1.67.4; 1.67.6; Also convert detach function.
1.66	22-Oct-2007	joerg	Convert to device_t/softc split and aprint usage.
1.65	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.64	01-Sep-2007	dyoung	branches: 1.64.4; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.63	26-Aug-2007	dyoung	branches: 1.63.2; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.62	04-Mar-2007	sketch	branches: 1.62.2; 1.62.10; 1.62.14; Put unformatted firmware version output behind IWI_DEBUG
1.61	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.60	23-Jan-2007	skrll	branches: 1.60.2; Pass the right pointer to firmware_free. Fixes the DIAGNOSTIC panic reported by Sverre Froyen on current-users.
1.59	20-Dec-2006	skrll	- Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.58	16-Nov-2006	christos	branches: 1.58.2; __unused removal on arguments; approved by core.
1.57	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.56	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.55	30-Aug-2006	christos	branches: 1.55.2; 1.55.4; fix initializers.
1.54	19-Aug-2006	skrll	Add some sysctl knobs to bring us closer to the FreeBSD driver. Also sysctl enable the debug message printing (if IWI_DEBUG) is defined.
1.53	09-Aug-2006	skrll	s/printf/aprint_error/ for one message.
1.52	09-Aug-2006	skrll	Adapt iwi(4) to use firmload(9) and remove firmware loading from iwictl(8).
1.51	21-Jul-2006	ad	- Use the LWP cached credentials where sane. - Minor cosmetic changes.
1.50	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.49	28-May-2006	blymn	branches: 1.49.2; Clean up bogus whitespace
1.48	14-May-2006	elad	branches: 1.48.2; integrate kauth.
1.47	11-Apr-2006	rpaulo	Coverity ID 2442: check if one can dereference ifp before calling iwi_stop() in iwi_detach().
1.46	09-Mar-2006	jmcneill	branches: 1.46.2; Save and restore PCI configuration state on suspend/resume.
1.45	21-Feb-2006	skrll	branches: 1.45.2; 1.45.4; Fix from Sam Leffler "correct check for whether wpa is enabled". wpa_supplicant and static WEP doesn't crash the firmware now. YAY!
1.44	24-Dec-2005	perry	branches: 1.44.2; 1.44.4; 1.44.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.43	05-Dec-2005	skrll	Finally (I hope) fix iwi_ioctl: - handle SIOCADDMULTI/SIOCDELMULTI properly, i.e. no need to reset anything as we don't do multicast filters (yet) - restructure some code and use an IS_RUNNING macro Fix iwi_init to set if_flags before the net80211 state machine is kicked and init to IEEE80211_S_INIT.
1.42	29-Nov-2005	rpaulo	Add support for software controlled LEDs as found on some boards. Currently, only the associated LED is being used because that's the only LED my card has. The other two LEDs (OFDM and activity) can later be set by someone, easily, who owns some board with that LEDs. Reviewed by Nick Hudson.
1.41	26-Nov-2005	skrll	Fixes from Steve Woodford: - add some missing bus_dmamap_sync operations. - don't process other interrupts if we get an error/radio off interrupt. - improve command handling - sleep against the descriptor instead of the descriptor set.
1.40	23-Nov-2005	skrll	Fix the handling of ENETRESET in iwi_ioctl. This fixes PR 32143
1.39	23-Nov-2005	skrll	Correct a bpf tap.
1.38	18-Nov-2005	skrll	Resolve conflicts. Changelog to follow.
1.37	14-Nov-2005	skrll	branches: 1.37.2; Don't use DPRINTF for previous. Doh!
1.36	14-Nov-2005	skrll	Dump the error log if we get an IWI_INTR_FATAL_ERROR interrupt.
1.35	29-Oct-2005	scw	Expunge an extraneous bus_dmamap_sync() call in iwi_frame_intr(). The entire frame has already been sync'd by iwi_rx_intr().
1.34	29-Oct-2005	scw	- Use the DMA map size when calling bus_dmamap_sync(), instead of blindly assuming MCLBYTES will DTRT. - Use bus_dmamap_load_mbuf() instead of bus_dmamap_load() where appropriate. - If we have to coalesce a Tx mbuf chain comprised of more than IWI_MAX_NSEG segments, allocate a cluster iff the payload won't fit in the header.
1.33	29-Oct-2005	skrll	Don't byte swap unnecessarily in iwi_read_prom_word. Fix the user instead.
1.32	29-Oct-2005	scw	Support big-endian hosts: - Always byte-swap data read from EEPROM. - Byte-swap firmware image when host is big-endian. In iwi_tx_start(): - Fix descriptor initialisation bogon (only problematic on big-endian hosts). - Fix bus_dmamap_sync() 'len' parameter..
1.31	19-Oct-2005	joerg	Fix a kernel segfault when a RX interrupt can't allocate a new mbuf. The change adopts the idea of fxp to drop the incoming packet and panic if the old mbuf cannot be reloaded. Since the bus_dmamap is allocated during attach, this is not supposed to happen. Since a lot of code moves anyway, factor out the allocation of RX ring elements, which is shared between the init path and the RX interrupt path. XXX A better fix might be to borrow the mbuf from the logic end of the XXX ring buffer, but that needs more involved driver changes. Reviewed by dyoung@ and nick@
1.30	08-Oct-2005	skrll	branches: 1.30.2; Use the new scan command for monitor mode as well.
1.29	29-Sep-2005	skrll	Improve interrupt handling: - don't disable/enable as we're already at splnet() - ack the interrupts early Fixes my "lost interrupt" problem. Thanks to dyoung and scw for the suggestions.
1.28	25-Sep-2005	skrll	Whitespace and change a comment slightly.
1.27	25-Sep-2005	skrll	Don't try to bus_dmamem_alloc everytime we want to load the firmware as it can fail all too easily. Instead bus_dmamap_load the cached copy and create the command blocks for the device to load it accordingly. Thanks to scw and mrg for reviewing this. Closes PR 29892 (I hope).
1.26	25-Sep-2005	skrll	Support hidden ESSID APs. Use the newer scan command as this one doesn't crash the firmware when scanning 802.11a channels. Thanks to scw and blymn for testing. Closes PR 31295.
1.25	25-Sep-2005	skrll	Resolve conflicts.
1.24	17-Sep-2005	skrll	KNF. That is change u_intN_t to uintN_t.
1.23	15-Sep-2005	skrll	Correct a debug message.
1.22	12-Sep-2005	skrll	Fix spello.
1.21	12-Sep-2005	skrll	Use the correct length parameter when capturing rx packets.
1.20	04-Sep-2005	skrll	Re-organise a bit to reduce diff to FreeBSD.
1.19	30-Aug-2005	skrll	Let the net80211 layer handle the transition to INIT as well. Testing by salo and me suggests that PR 31010 is fixed.
1.18	30-Aug-2005	skrll	Disestablish the power and shutdown hooks when detaching.
1.17	30-Aug-2005	skrll	KNF (and reduced diff to FreeBSD)
1.16	27-Aug-2005	skrll	Be explicit about which cards to setup .11a rates for. hi matt.
1.15	19-Aug-2005	skrll	Add shutdown and power hooks.
1.14	19-Aug-2005	skrll	More sync with the FreeBSD driver.
1.13	01-Aug-2005	skrll	- Match more product ids. - Don't initialise ic_phytype twice and do initialise ic_state. - announce available rates. - mark interface down if firmware crashes for the radio transmitter gets turned off.
1.12	30-Jul-2005	christos	Partially fix. Now it comes up, associates with wep, but something is wrong with tcp since it crashes in wep encryption when I cvs update.
1.11	07-Jul-2005	dyoung	Cosmetic: use the idiom &ic->ic_nw_keys[0] instead of ic->ic_nw_keys, just be a little more parallel to &ic->ic_nw_keys[IEEE80211_WEP_NKID] on the same line.
1.10	25-Jun-2005	dyoung	branches: 1.10.2; Let ieee80211_input do WEP-decapsulation. Make iwi(4) provide a key-allocator that returns valid indices for the four global keys. Now net80211 should let the h/w do WEP encryption/decryption.
1.9	22-Jun-2005	dyoung	Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.8	20-Jun-2005	sekiya	Implement monitor mode. From OpenBSD.
1.7	07-Jun-2005	skrll	Sync with pcidevs changes.
1.6	13-Apr-2005	mrg	fix an error message: could -> could not
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	15-Jan-2005	skrll	branches: 1.4.2; 1.4.4; 1.4.6; Remove local PCI_PRODUCT defines
1.3	11-Jan-2005	skrll	Disable the debug stuff
1.2	11-Jan-2005	skrll	NetBSD RCSID
1.1	11-Jan-2005	skrll	branches: 1.1.1; Initial revision
1.1.1.3	18-Nov-2005	skrll	Import FreeBSD's iwi(4) of 1-nov-2005
1.1.1.2	25-Sep-2005	skrll	Import of the FreeBSD sources from just after the net80211 merge. There are no major changes, but this will make future merging easier.
1.1.1.1	11-Jan-2005	skrll	iwi(4) - Intel PRO/Wireless 2200AG/2915ABG driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/ Ad-hoc mode and roaming are not fully implemented yet.
1.4.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.4.5	11-Dec-2005	christos	Sync with head.
1.4.4.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.4.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.4.4.1	15-Jan-2005	skrll	file if_iwi.c was added on branch ktrace-lwp on 2005-01-17 19:31:24 +0000
1.4.2.1	29-Apr-2005	kent	sync with -current
1.10.2.8	24-Mar-2008	yamt	sync with head.
1.10.2.7	17-Mar-2008	yamt	sync with head.
1.10.2.6	21-Jan-2008	yamt	sync with head
1.10.2.5	27-Oct-2007	yamt	sync with head.
1.10.2.4	03-Sep-2007	yamt	sync with head.
1.10.2.3	26-Feb-2007	yamt	sync with head.
1.10.2.2	30-Dec-2006	yamt	sync with head.
1.10.2.1	21-Jun-2006	yamt	sync with head.
1.30.2.2	02-Nov-2005	yamt	sync with head.
1.30.2.1	26-Oct-2005	yamt	sync with head
1.37.2.3	29-Nov-2005	yamt	sync with head.
1.37.2.2	22-Nov-2005	yamt	sync with head.
1.37.2.1	14-Nov-2005	yamt	file if_iwi.c was added on branch yamt-readahead on 2005-11-22 16:08:11 +0000
1.44.6.2	01-Jun-2006	kardel	Sync with head.
1.44.6.1	22-Apr-2006	simonb	Sync with head.
1.44.4.1	09-Sep-2006	rpaulo	sync with head
1.44.2.1	01-Mar-2006	yamt	sync with head.
1.45.4.4	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.45.4.3	19-Apr-2006	elad	sync with head.
1.45.4.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.45.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI.
1.45.2.5	03-Sep-2006	yamt	sync with head.
1.45.2.4	11-Aug-2006	yamt	sync with head
1.45.2.3	26-Jun-2006	yamt	sync with head.
1.45.2.2	24-May-2006	yamt	sync with head.
1.45.2.1	13-Mar-2006	yamt	sync with head.
1.46.2.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.48.2.1	19-Jun-2006	chap	Sync with head.
1.49.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.55.4.3	21-Dec-2006	yamt	sync with head.
1.55.4.2	10-Dec-2006	yamt	sync with head.
1.55.4.1	22-Oct-2006	yamt	sync with head
1.55.2.3	01-Feb-2007	ad	Sync with head.
1.55.2.2	12-Jan-2007	ad	Sync with head.
1.55.2.1	18-Nov-2006	ad	Sync with head.
1.58.2.2	28-Jan-2007	tron	Pull up following revision(s) (requested by skrll in ticket #385): sys/dev/pci/if_iwi.c: revision 1.60 Pass the right pointer to firmware_free. Fixes the DIAGNOSTIC panic reported by Sverre Froyen on current-users.
1.58.2.1	29-Dec-2006	riz	Pull up following revision(s) (requested by skrll in ticket #296): sys/dev/pci/if_iwireg.h: revision 1.17 sys/dev/pci/if_iwi.c: revision 1.59 sys/dev/pci/if_iwivar.h: revision 1.13 - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things. - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.60.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.62.14.7	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.62.14.6	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.62.14.5	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.62.14.4	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.62.14.3	03-Sep-2007	jmcneill	Sync with HEAD.
1.62.14.2	23-Aug-2007	joerg	Introduce pci_net_generic_power, which should be enough for most network drivers and be a good foundation for C&P for the rest. For iwi(4), don't reset the PCI retry register again, pci_conf_restore should take care of that already. For bge(4), add a NetBSD style if_stop.
1.62.14.1	21-Aug-2007	joerg	Convert iwi(4) to pnp_register/pnp_deregister and drop explicit shutdown hooks.
1.62.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.62.2.2	23-Oct-2007	ad	Sync with head.
1.62.2.1	09-Oct-2007	ad	Sync with head.
1.63.2.3	23-Mar-2008	matt	sync with HEAD
1.63.2.2	09-Jan-2008	matt	sync with HEAD
1.63.2.1	06-Nov-2007	matt	sync with HEAD
1.64.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.67.6.1	11-Dec-2007	yamt	sync with head.
1.67.4.1	26-Dec-2007	ad	Sync with head.
1.68.10.3	17-Jan-2009	mjf	Sync with HEAD.
1.68.10.2	29-Jun-2008	mjf	Sync with HEAD.
1.68.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.68.6.1	24-Mar-2008	keiichi	sync with head.
1.72.8.1	18-Jun-2008	simonb	Sync with head.
1.72.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.72.4.5	11-Aug-2010	yamt	sync with head.
1.72.4.4	11-Mar-2010	yamt	sync with head
1.72.4.3	16-Sep-2009	yamt	sync with head
1.72.4.2	16-May-2009	yamt	sync with head
1.72.4.1	04-May-2009	yamt	sync with head.
1.72.2.1	17-Jun-2008	yamt	sync with head.
1.73.4.3	28-Apr-2009	skrll	Sync with HEAD.
1.73.4.2	03-Mar-2009	skrll	Sync with HEAD.
1.73.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.73.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.74.2.4	29-Sep-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #1040): sys/dev/pci/if_iwi.c: revision 1.79 make iwi_alloc_tx_ring() prototype match function declaration (shows up when bus_addr_t != bus_size_t). Fix cast for 64bit paddr_t on i386.
1.74.2.3	14-Jan-2009	snj	branches: 1.74.2.3.4; Pull up following revision(s) (requested by jmcneill in ticket #250): sys/dev/pci/if_iwi.c: revision 1.78 sys/dev/pci/if_ipw.c: revision 1.45 Make the "EULA not accepted" message a bit more informative, now reads: EULA not accepted; please see the <driver>(4) man page.
1.74.2.2	10-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #43): sys/dev/pci/if_iwi.c: revision 1.77 Fix typo in fw name.
1.74.2.1	09-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #25): sys/dev/pci/if_ipw.c: revision 1.44 sys/dev/pci/if_ipwvar.h: revision 1.14 sys/dev/pci/if_iwi.c: revision 1.76 sys/dev/pci/if_iwn.c: revision 1.26 Match image name with upstream distfiles.
1.74.2.3.4.1	21-Apr-2010	matt	sync to netbsd-5
1.78.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.83.4.3	31-May-2011	rmind	sync with head
1.83.4.2	05-Mar-2011	rmind	sync with head
1.83.4.1	30-May-2010	rmind	sync with head
1.83.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.85.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.85.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.87.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.87.4.2	30-Oct-2012	yamt	sync with head
1.87.4.1	17-Apr-2012	yamt	sync with head
1.88.2.2	05-Apr-2012	mrg	sync to latest -current.
1.88.2.1	18-Feb-2012	mrg	merge to -current.
1.89.2.1	22-Mar-2012	riz	Pull up following revision(s) (requested by nisimura in ticket #134): sys/dev/pci/if_iwi.c: revision 1.90 Unbreak the endian issue in firmware header decoding. Comfirmed good and running by a powerpc machine.
1.91.2.3	03-Dec-2017	jdolecek	update from HEAD
1.91.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.91.2.1	23-Jun-2013	tls	resync from head
1.92.4.1	18-May-2014	rmind	sync with head
1.97.6.5	28-Aug-2017	skrll	Sync with HEAD
1.97.6.4	05-Feb-2017	skrll	Sync with HEAD
1.97.6.3	05-Oct-2016	skrll	Sync with HEAD
1.97.6.2	09-Jul-2016	skrll	Sync with HEAD
1.97.6.1	06-Apr-2015	skrll	Sync with HEAD
1.99.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.99.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.99.2.1	06-Aug-2016	pgoyette	Sync with HEAD
1.101.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.103.2.2	21-Jun-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #1835): sys/dev/pci/if_iwi.c: revision 1.117 sys/dev/raidframe/rf_netbsdkintf.c: revision 1.401 sys/dev/scsipi/ses.c: revision 1.52 sys/dev/isa/mcd.c: revision 1.121 (all via patch) sys/dev: Memset zero before copyout. Just in case of uninitialized padding which would lead to kernel stack disclosure. If the compiler can prove the memset redundant then it can optimize it away; otherwise better safe than sorry. I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks; the raidframe(4) change probably doesn't (but doesn't hurt).
1.103.2.1	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.105.2.4	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.105.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.105.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.105.2.1	02-May-2018	pgoyette	Synch with HEAD
1.107.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.107.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.107.2.1	10-Jun-2019	christos	Sync with HEAD
1.111.4.1	03-Aug-2022	martin	Pull up following revision(s) (requested by riastradh in ticket #1485): sys/dev/pci/if_iwi.c: revision 1.117 sys/dev/raidframe/rf_netbsdkintf.c: revision 1.401 sys/dev/scsipi/ses.c: revision 1.52 sys/dev/isa/mcd.c: revision 1.121 sys/dev: Memset zero before copyout. Just in case of uninitialized padding which would lead to kernel stack disclosure. If the compiler can prove the memset redundant then it can optimize it away; otherwise better safe than sorry. I think the iwi(4), mcd(4), and ses(4) changes actually plug leaks; the raidframe(4) change probably doesn't (but doesn't hurt).
1.112.2.1	29-Feb-2020	ad	Sync with head.
1.114.8.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.114.8.1	13-May-2021	thorpej	Sync with HEAD.
1.120.10.1	02-Aug-2025	perseant	Sync with HEAD
1.19	08-Sep-2008	gmcgarry	Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.18	25-Dec-2007	perry	branches: 1.18.6; 1.18.10; 1.18.12; 1.18.16; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.17	20-Dec-2006	skrll	branches: 1.17.20; 1.17.26; 1.17.28; 1.17.32; - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.16	09-Aug-2006	skrll	branches: 1.16.2; 1.16.4; 1.16.6; Adapt iwi(4) to use firmload(9) and remove firmware loading from iwictl(8).
1.15	29-Nov-2005	rpaulo	branches: 1.15.4; 1.15.8; Add support for software controlled LEDs as found on some boards. Currently, only the associated LED is being used because that's the only LED my card has. The other two LEDs (OFDM and activity) can later be set by someone, easily, who owns some board with that LEDs. Reviewed by Nick Hudson.
1.14	18-Nov-2005	skrll	Resolve conflicts. Changelog to follow.
1.13	14-Nov-2005	skrll	branches: 1.13.2; Dump the error log if we get an IWI_INTR_FATAL_ERROR interrupt.
1.12	29-Oct-2005	scw	Re-arrange a bitfield according to host byte-order.
1.11	08-Oct-2005	skrll	branches: 1.11.2; Use the new scan command for monitor mode as well.
1.10	29-Sep-2005	skrll	Fix the endianness of the scan command. All the channels get the right scan type/dwelltime now.
1.9	25-Sep-2005	skrll	Support hidden ESSID APs. Use the newer scan command as this one doesn't crash the firmware when scanning 802.11a channels. Thanks to scw and blymn for testing. Closes PR 31295.
1.8	25-Sep-2005	skrll	Resolve conflicts.
1.7	17-Sep-2005	skrll	KNF. That is change u_intN_t to uintN_t.
1.6	12-Sep-2005	skrll	Fix spello.
1.5	19-Aug-2005	skrll	More sync with the FreeBSD driver.
1.4	30-Jul-2005	christos	Partially fix. Now it comes up, associates with wep, but something is wrong with tcp since it crashes in wep encryption when I cvs update.
1.3	20-Jun-2005	sekiya	branches: 1.3.2; Implement monitor mode. From OpenBSD.
1.2	11-Jan-2005	skrll	branches: 1.2.4; NetBSD RCSID
1.1	11-Jan-2005	skrll	branches: 1.1.1; Initial revision
1.1.1.3	18-Nov-2005	skrll	Import FreeBSD's iwi(4) of 1-nov-2005
1.1.1.2	25-Sep-2005	skrll	Import of the FreeBSD sources from just after the net80211 merge. There are no major changes, but this will make future merging easier.
1.1.1.1	11-Jan-2005	skrll	iwi(4) - Intel PRO/Wireless 2200AG/2915ABG driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/ Ad-hoc mode and roaming are not fully implemented yet.
1.2.4.4	11-Dec-2005	christos	Sync with head.
1.2.4.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.2.4.1	11-Jan-2005	skrll	file if_iwireg.h was added on branch ktrace-lwp on 2005-01-17 19:31:24 +0000
1.3.2.3	21-Jan-2008	yamt	sync with head
1.3.2.2	30-Dec-2006	yamt	sync with head.
1.3.2.1	21-Jun-2006	yamt	sync with head.
1.11.2.1	02-Nov-2005	yamt	sync with head.
1.13.2.3	29-Nov-2005	yamt	sync with head.
1.13.2.2	22-Nov-2005	yamt	sync with head.
1.13.2.1	14-Nov-2005	yamt	file if_iwireg.h was added on branch yamt-readahead on 2005-11-22 16:08:11 +0000
1.15.8.1	11-Aug-2006	yamt	sync with head
1.15.4.1	09-Sep-2006	rpaulo	sync with head
1.16.6.1	29-Dec-2006	riz	Pull up following revision(s) (requested by skrll in ticket #296): sys/dev/pci/if_iwireg.h: revision 1.17 sys/dev/pci/if_iwi.c: revision 1.59 sys/dev/pci/if_iwivar.h: revision 1.13 - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things. - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.16.4.1	21-Dec-2006	yamt	sync with head.
1.16.2.1	12-Jan-2007	ad	Sync with head.
1.17.32.1	02-Jan-2008	bouyer	Sync with HEAD
1.17.28.1	26-Dec-2007	ad	Sync with head.
1.17.26.1	18-Feb-2008	mjf	Sync with HEAD.
1.17.20.1	09-Jan-2008	matt	sync with HEAD
1.18.16.1	19-Oct-2008	haad	Sync with HEAD.
1.18.12.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.18.10.1	04-May-2009	yamt	sync with head.
1.18.6.1	28-Sep-2008	mjf	Sync with HEAD.
1.20	20-Mar-2020	thorpej	This driver sleeps during iwi_media_change(), and thus requires an adaptive mutex for the media lock.
1.19	02-Feb-2017	nonaka	branches: 1.19.14; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.18	06-Sep-2015	dholland	branches: 1.18.2; 1.18.4; More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.17	07-Jan-2015	ozaki-r	Pass a correct firmware size (instead of 0) to firmware_free firmware_free now uses kmem_free(9) instead of free(9), so we need to pass a correct size to it.
1.16	19-Jan-2010	pooka	branches: 1.16.22; 1.16.40; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.15	09-Dec-2007	jmcneill	branches: 1.15.14; Merge jmcneill-pm branch.
1.14	22-Oct-2007	joerg	branches: 1.14.4; 1.14.6; Convert to device_t/softc split and aprint usage.
1.13	20-Dec-2006	skrll	branches: 1.13.18; 1.13.20; 1.13.24; - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.12	19-Aug-2006	skrll	branches: 1.12.2; 1.12.4; 1.12.6; Add some sysctl knobs to bring us closer to the FreeBSD driver. Also sysctl enable the debug message printing (if IWI_DEBUG) is defined.
1.11	09-Aug-2006	skrll	Adapt iwi(4) to use firmload(9) and remove firmware loading from iwictl(8).
1.10	09-Mar-2006	jmcneill	Save and restore PCI configuration state on suspend/resume.
1.9	29-Nov-2005	rpaulo	branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; Add support for software controlled LEDs as found on some boards. Currently, only the associated LED is being used because that's the only LED my card has. The other two LEDs (OFDM and activity) can later be set by someone, easily, who owns some board with that LEDs. Reviewed by Nick Hudson.
1.8	18-Nov-2005	skrll	Resolve conflicts. Changelog to follow.
1.7	17-Sep-2005	skrll	branches: 1.7.6; KNF. That is change u_intN_t to uintN_t.
1.6	19-Aug-2005	skrll	Add shutdown and power hooks.
1.5	19-Aug-2005	skrll	More sync with the FreeBSD driver.
1.4	30-Jul-2005	christos	Partially fix. Now it comes up, associates with wep, but something is wrong with tcp since it crashes in wep encryption when I cvs update.
1.3	22-Jun-2005	dyoung	branches: 1.3.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.2	11-Jan-2005	skrll	branches: 1.2.4; NetBSD RCSID
1.1	11-Jan-2005	skrll	branches: 1.1.1; Initial revision
1.1.1.3	18-Nov-2005	skrll	Import FreeBSD's iwi(4) of 1-nov-2005
1.1.1.2	25-Sep-2005	skrll	Import of the FreeBSD sources from just after the net80211 merge. There are no major changes, but this will make future merging easier.
1.1.1.1	11-Jan-2005	skrll	iwi(4) - Intel PRO/Wireless 2200AG/2915ABG driver. Written by Damien Bergamini, from http://damien.bergamini.free.fr/ipw/ Ad-hoc mode and roaming are not fully implemented yet.
1.2.4.4	11-Dec-2005	christos	Sync with head.
1.2.4.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.2.4.1	11-Jan-2005	skrll	file if_iwivar.h was added on branch ktrace-lwp on 2005-01-17 19:31:24 +0000
1.3.2.4	21-Jan-2008	yamt	sync with head
1.3.2.3	27-Oct-2007	yamt	sync with head.
1.3.2.2	30-Dec-2006	yamt	sync with head.
1.3.2.1	21-Jun-2006	yamt	sync with head.
1.7.6.2	29-Nov-2005	yamt	sync with head.
1.7.6.1	22-Nov-2005	yamt	sync with head.
1.9.10.1	19-Apr-2006	elad	sync with head.
1.9.8.3	03-Sep-2006	yamt	sync with head.
1.9.8.2	11-Aug-2006	yamt	sync with head
1.9.8.1	13-Mar-2006	yamt	sync with head.
1.9.6.1	22-Apr-2006	simonb	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.12.6.1	29-Dec-2006	riz	Pull up following revision(s) (requested by skrll in ticket #296): sys/dev/pci/if_iwireg.h: revision 1.17 sys/dev/pci/if_iwi.c: revision 1.59 sys/dev/pci/if_iwivar.h: revision 1.13 - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things. - Upgrade to the latest firmware and as a result remove the error log dump. - Mostly sync with OpenBSD - Serialise sending commands to the firmware - Remove redundant calls to bpfdetach. - use bus_size_t where appropriate and not fetch iobase as it's not used. - improve 802.11 radiotap support (correct rx rate) - add short preamble flag - add short slot time support - ignore parity errors (as per the Linux driver) - Set Tx power for all channels. - disable bluetooth co-existance - Check that ni->ni_rates.rs_nrates is not greater than the maximum handled by the firmware. - Begin syncing with the FreeBSD driver by renaming a few things.
1.12.4.1	21-Dec-2006	yamt	sync with head.
1.12.2.1	12-Jan-2007	ad	Sync with head.
1.13.24.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.20.2	09-Jan-2008	matt	sync with HEAD
1.13.20.1	06-Nov-2007	matt	sync with HEAD
1.13.18.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.18.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.13.18.1	21-Aug-2007	joerg	Convert iwi(4) to pnp_register/pnp_deregister and drop explicit shutdown hooks.
1.14.6.1	11-Dec-2007	yamt	sync with head.
1.14.4.1	26-Dec-2007	ad	Sync with head.
1.15.14.1	11-Mar-2010	yamt	sync with head
1.16.40.3	05-Feb-2017	skrll	Sync with HEAD
1.16.40.2	22-Sep-2015	skrll	Sync with HEAD
1.16.40.1	06-Apr-2015	skrll	Sync with HEAD
1.16.22.1	03-Dec-2017	jdolecek	update from HEAD
1.18.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.18.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.19.14.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.90	10-Nov-2024	mlelstv	Add MBUFTRACE
1.89	09-Feb-2024	mlelstv	branches: 1.89.2; Fix DMA sync sizes. Don't panic for inconsistent queue counter, just print an error to console.
1.88	21-Sep-2023	msaitoh	s/ for for / for / in comment.
1.87	24-Jun-2021	riastradh	iwm(4): Disentangle attach. Don't attach a half-baked interface and then detach it and then reattach it after mountroot when we can read firmware; just defer attaching the interface altogether until mountroot. Likely fixes some panics I've seen every now and then at boot with iwm(4).
1.86	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.85	22-May-2020	thorpej	branches: 1.85.6; This driver sleeps during iwm_media_change(), and thus requires an adaptive mutex for the media lock.
1.84	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.83	26-Jun-2018	msaitoh	branches: 1.83.2; 1.83.10; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.82	08-Jun-2018	knakahara	iwm(4) workaround as ifp->if_hwdl can be set only once. Advised nonaka@n.o and ozaki-r@n.o, thanks. To avoid for iwm(4) to set if_hwdl before loading firmware, use local address as dummy. See sys/net/if.c::if_set_sadl().
1.81	08-Jun-2018	knakahara	Fix "ifconfig iwm0" failures reported by ryoon@n.o. Advised nonaka@n.o and ozaki-r@n.o, thanks.
1.80	06-Jun-2018	maya	Remove duplicate ;
1.79	05-Jun-2018	knakahara	Fix panic on boot with iwm(4). Advised by nonaka@n.o, thanks. XXX pullup-8
1.78	21-Jan-2018	christos	branches: 1.78.2; CID-1427736: Appease coverity using KASSERT
1.77	10-Jan-2018	mlelstv	Match Intel Dual Band Wireless AC 3168. The code already supports it.
1.76	23-Oct-2017	msaitoh	branches: 1.76.2; - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly.
1.75	23-Jul-2017	para	branches: 1.75.2; kmem_intr_free kmem_intr_[z]alloced memory the underlying pools are the same but api-wise those should match
1.74	09-Jun-2017	nonaka	fix typo. reported by ryo@.
1.73	01-Jun-2017	chs	branches: 1.73.2; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.72	19-May-2017	nonaka	iwm(4): match Wireless 8265 device.
1.71	14-Mar-2017	nonaka	iwm(4): update firmware to maintained core release. firmware from https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=4b9559fb0565f582a21a4070a81a3b6a7d77610b
1.70	02-Feb-2017	nonaka	wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.69	21-Jan-2017	nonaka	PR/51797: initialize hardware completely when receiving a deauth frame from AP. avoid "iwm0: fatal firmware error".
1.68	21-Jan-2017	nonaka	more block network interrupt.
1.67	21-Jan-2017	nonaka	increase command response timeout value.
1.66	21-Jan-2017	nonaka	include <sys/intr.h> for softint(9).
1.65	19-Jan-2017	nonaka	newstate should be blocked network interrupt, because it called iwm_start.
1.64	17-Jan-2017	nonaka	show NVM channel flags when AB_DEBUG is set.
1.63	17-Jan-2017	nonaka	more acknowledge interrupts from iwn(4).
1.62	17-Jan-2017	nonaka	only bus_dmamap_sync(9) once for ICT region.
1.61	13-Jan-2017	nonaka	branches: 1.61.2; iwm(4): 3165 works fine with iwlwifi-7265D-17.ucode firmware. firmware from http://git.kernel.org/cgit/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=be8a6fcba2e48e00674d1f35e6def56e2268039c
1.60	10-Jan-2017	nonaka	check LAR support in NVM for 8260/4165.
1.59	10-Jan-2017	nonaka	Add apmg_wake_up_wa support from DragonFly BSD.
1.58	10-Jan-2017	nonaka	iwm(4): 8260 works now. 4165 may works too.
1.57	10-Jan-2017	nonaka	set NULL after free.
1.56	09-Jan-2017	nonaka	Apply OpenBSD if_iwm.c rev.1.148. > While setting up the basic rate bitmask for iwm's firmware, if the AP does > not specify basic rates for either the CCK or OFDM set, add just the most > basic rate to that set (1 Mbit/s in case of CCK, 6 Mbit/s in case of OFDM). > This behaviour matches what code comments seem to imply. > The previous code would add all possible basic rates in such cases. > So if all basic rates were CCK only, the code would add all possible OFDM basic > rates on top. Then the firmware would send some frames at too high rates, e.g. > RTS frames would be sent at 24Mbit/s which is a bit risky on noisy channels.
1.55	09-Jan-2017	khorben	Do not use the "flags" field uninitialized A backport if iwm(4) (to netbsd-7) does not build without this change. LGTM nonaka@
1.54	09-Jan-2017	nonaka	Added enum iwm_tx_pm_timeouts from FreeBSD.
1.53	09-Jan-2017	nonaka	cosmetic.
1.52	09-Jan-2017	nonaka	use correct header size.
1.51	08-Jan-2017	nonaka	fix indent
1.50	08-Jan-2017	nonaka	iwm(4): make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html
1.49	08-Jan-2017	nonaka	use pci_intr_establish_xname(9).
1.48	08-Jan-2017	nonaka	twiddle bpf related.
1.47	08-Jan-2017	nonaka	cosmetic
1.46	04-Jan-2017	nonaka	Export some 802.11 IE manipulate functions.
1.45	18-Dec-2016	nonaka	iwm(4): Sync with OpenBSD. 7265 works fine. 7260 and 3160 maybe work. 7265D, 3165 and 8260 not work yet. if_iwm.c: r1.147. if_iwmreg.h: r1.19 without r1.16. if_iwmvar.h: r1.24. firmware is taken from https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi - iwlwifi-7260-ucode-16.242414.0.tgz - iwlwifi-3160-ucode-16.242414.0.tgz - iwlwifi-7265-ucode-16.242414.0.tgz - iwlwifi-8000-ucode-16.242414.0.tgz
1.44	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.43	23-Sep-2016	maya	Fix inverted logic in iwm(4). The PROT_REQUIRE flag in should be set for data frames above a certain length, but we were setting it for !data frames above a certain length, which makes no sense at all. - taken from all other BSDs (OpenBSD, FreeBSD, DragonflyBSD...)
1.42	10-Jun-2016	ozaki-r	branches: 1.42.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.41	26-May-2016	ozaki-r	Use M_GETCTX No functional change.
1.40	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.39	06-Nov-2015	nonaka	Always use pci_intr_alloc(9)/pci_intr_release(9).
1.38	14-Oct-2015	ozaki-r	Remove leading whitespaces and kind
1.37	24-Jul-2015	nonaka	Use pci_intr_alloc().
1.36	22-Jul-2015	nonaka	iwm(4) can read the firmware image from disk when interface up. This change is useful for full-disk encryption environment.
1.35	22-May-2015	nonaka	Limit the number of dma segments used for transmitting packets to IWM_NUM_OF_TBS - 2. We have IWM_NUM_OF_TBS slots, but use two of those for sending commands to the firmware. From OpenBSD rev. 1.41.
1.34	16-May-2015	nonaka	should be able to use 5GHz wireless network.
1.33	15-May-2015	knakahara	if_iwm use unified establish API.
1.32	29-Apr-2015	nonaka	Added control debugging output via sysctl.
1.31	28-Apr-2015	nonaka	Use MSI instead of legacy INTx, if available. should fix PR/49663.
1.30	15-Apr-2015	nonaka	Register iwm(4) with pmf(9). Patch from Pierre Pronchery.
1.29	26-Mar-2015	nonaka	branches: 1.29.2; Not all iwm hardware supports the 5GHz band. Make setting 11a rates and scanning conditional on the 5GHz support bit in the nvm. From OpenBSD rev. 1.39. Arranged by me.
1.28	07-Mar-2015	nonaka	Fix radiotap and fixed rates. Radiotap was reporting the wrong rate for data frames and fixed rates weren't really fixed and were converted into the wrong hardware rate. From OpenBSD rev.1.36.
1.27	07-Mar-2015	khorben	Avoid kmem_free(NULL) in iwm_read_firmware() This code path can be hit if the firmware failed to load, for instance if the file is not present on the filesystem. In this case firmware_open() fails, and fw->fw_rawdata never gets allocated in the first place.
1.26	04-Mar-2015	nonaka	stop the calibration timeout when stopping the interface. From OpenBSD rev.1.35.
1.25	04-Mar-2015	nonaka	Set/clear the IWM_MAC_FILTER_IN_BEACON flag under conditions that match what Linux does. From OpenBSD rev.1.34.
1.24	03-Mar-2015	nonaka	Fix 11a support. From OpenBSD rev.1.33.
1.23	03-Mar-2015	nonaka	Make setting the MAC context bit for 11g protection when transmitting OFDM frames conditional on the node via IEEE80211_F_USEPROT. From OpenBSD rev.1.31.
1.22	03-Mar-2015	nonaka	Don't set CTS to self. Linux stopped setting this back in dc271ee0d04d12d6bfabacbec803289a7072fbd9 as it is known to cause problems. From OpenBSD rev.1.30.
1.21	03-Mar-2015	nonaka	sc_fixed_ridx is initialised to 0 but tested as it it were initialised to -1. The result of this is tx frames were always sent out at fixed rate 0 instead of ni_txrate. Match the iwn behaviour and test ic_fixed_rate for -1 instead. From OpenBSD rev.1.29.
1.20	03-Mar-2015	nonaka	Correctly set lowest_present_ofdm in iwm_mvm_ack_rates(). From OpenBSD rev.1.28.
1.19	03-Mar-2015	nonaka	init error to 0 in iwm_send_cmd(). From OpenBSD rev.1.27.
1.18	03-Mar-2015	nonaka	match iwn and init txrate at lowest available bit-rate. From OpenBSD rev.1.26.
1.17	03-Mar-2015	nonaka	workaround for interrupt coalescing bug not needed on 7265. From OpenBSD if_iwm.c rev.1.25, if_iwmvar.h rev.1.7.
1.16	02-Mar-2015	nonaka	Correctly set lowest_present_ofdm in iwm_mvm_ack_rates(). From OpenBSD rev.1.23.
1.15	25-Feb-2015	nonaka	Display code number of unhandled packet.
1.14	24-Feb-2015	christos	Work around Coverity CIDs 127102{0,1}: Confused by the comparison nrates > __arraycount(lq->rs_table), coverity erroneously binds nrates to the range [0..16), whereas it is really bound to [0..15) by ni->ni_rates.rs_nrates (IEEE80211_RATE_MAXSIZE). We add an explicit comparison to unconfuse it that will always succeed.
1.13	24-Feb-2015	nonaka	Revert previous change.
1.12	23-Feb-2015	nonaka	CID 1271021: Overrunning array "in->in_ridx" of 15 bytes at byte offset 15 using index "i" (which evaluates to 15). http://mail-index.netbsd.org/coverity-updates/2015/02/21/msg000115.html
1.11	20-Feb-2015	nonaka	Make the iwm(4) hardware receive filter pass multicast traffic from OpenBSD. Fixes inet6 autoconf, for example.
1.10	20-Feb-2015	nonaka	Don't return success when hw is not ready.
1.9	17-Feb-2015	nonaka	fix compile failure with IWM_DEBUG on i386.
1.8	16-Feb-2015	nonaka	whitespace
1.7	16-Feb-2015	nonaka	Fix compile failure with clang. Pointed out by Herbert J. Skuhra.
1.6	14-Feb-2015	nonaka	print error message when fail to load firmware.
1.5	13-Feb-2015	nonaka	iwm(4) works now, but 2.4GHz only. > iwm0 at pci2 dev 0 function 0: Intel Dual Band Wireless AC 3160 (rev. 0x83) > iwm0: interrupting at ioapic0 pin 18 > iwm0: hw rev: 0x160, fw ver 25.228 (API ver 9), address XX:XX:XX:XX:XX:XX > iwm0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps > iwm0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
1.4	13-Feb-2015	nonaka	Make some functions and variables static.
1.3	13-Feb-2015	nonaka	Convert printf() to aprint_*().
1.2	13-Feb-2015	nonaka	Sync with OpenBSD if_iwm.c rev.1.18.
1.1	07-Feb-2015	pooka	Add a driver for Intel Centrino 7260 and similar wireless cards. Supported devices should more or less match ones supported by the Linux iwlwifi mvm driver. Sponsored by genua mbh for OpenBSD. This is probably the world's first Canadian cross device driver: it was created for OpenBSD by writing and porting a NetBSD driver which was developed in a rump kernel in Linux userspace. Note: I don't have access to the hardware anymore, so this version is not tested and not enabled by default. While I tried to be careful in adding the NetBSD bits back, it's probable that there's a snafu or two. Feel free to send private email in case you have the hardware and there are issues.
1.29.2.11	28-Aug-2017	skrll	Sync with HEAD
1.29.2.10	05-Feb-2017	skrll	Sync with HEAD
1.29.2.9	05-Oct-2016	skrll	Sync with HEAD
1.29.2.8	09-Jul-2016	skrll	Sync with HEAD
1.29.2.7	29-May-2016	skrll	Sync with HEAD
1.29.2.6	19-Mar-2016	skrll	Sync with HEAD
1.29.2.5	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.29.2.4	22-Sep-2015	skrll	Sync with HEAD
1.29.2.3	06-Jun-2015	skrll	Sync with HEAD
1.29.2.2	06-Apr-2015	skrll	Sync with HEAD
1.29.2.1	26-Mar-2015	skrll	file if_iwm.c was added on branch nick-nhusb on 2015-04-06 15:18:10 +0000
1.42.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.42.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.42.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.61.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.73.2.3	02-Jan-2018	snj	Pull up following revision(s) (requested by ozaki-r in ticket #456): sys/arch/arm/sunxi/sunxi_emac.c: 1.9 sys/dev/ic/dwc_gmac.c: 1.43-1.44 sys/dev/pci/if_iwm.c: 1.75 sys/dev/pci/if_wm.c: 1.543 sys/dev/pci/ixgbe/ixgbe.c: 1.112 sys/dev/pci/ixgbe/ixv.c: 1.74 sys/kern/sys_socket.c: 1.75 sys/net/agr/if_agr.c: 1.43 sys/net/bpf.c: 1.219 sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416 sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257 sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146 sys/net/if_etherip.c: 1.40 sys/net/if_ethersubr.c: 1.243, 1.246 sys/net/if_faith.c: 1.57 sys/net/if_gif.c: 1.132 sys/net/if_l2tp.c: 1.15, 1.17 sys/net/if_loop.c: 1.98-1.101 sys/net/if_media.c: 1.35 sys/net/if_pppoe.c: 1.131-1.132 sys/net/if_spppsubr.c: 1.176-1.177 sys/net/if_tun.c: 1.142 sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121 sys/net/npf/npf_ifaddr.c: 1.3 sys/net/npf/npf_os.c: 1.8-1.9 sys/net/rtsock.c: 1.230 sys/netcan/if_canloop.c: 1.3-1.5 sys/netinet/if_arp.c: 1.255 sys/netinet/igmp.c: 1.65 sys/netinet/in.c: 1.210-1.211 sys/netinet/in_pcb.c: 1.180 sys/netinet/ip_carp.c: 1.92, 1.94 sys/netinet/ip_flow.c: 1.81 sys/netinet/ip_input.c: 1.362 sys/netinet/ip_mroute.c: 1.147 sys/netinet/ip_output.c: 1.283, 1.285, 1.287 sys/netinet6/frag6.c: 1.61 sys/netinet6/in6.c: 1.251, 1.255 sys/netinet6/in6_pcb.c: 1.162 sys/netinet6/ip6_flow.c: 1.35 sys/netinet6/ip6_input.c: 1.183 sys/netinet6/ip6_output.c: 1.196 sys/netinet6/mld6.c: 1.90 sys/netinet6/nd6.c: 1.239-1.240 sys/netinet6/nd6_nbr.c: 1.139 sys/netinet6/nd6_rtr.c: 1.136 sys/netipsec/ipsec_output.c: 1.65 sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10 kmem_intr_free kmem_intr_[z]alloced memory the underlying pools are the same but api-wise those should match Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@ Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..." scattered all over the source code and makes it easy to identify remaining KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE. No functional change Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE If IFEF_MPSAFE is set, hold the lock and otherwise don't hold. This change requires additions of KERNEL_LOCK to subsequence functions from if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe components. Proposed on tech-kern@ and tech-net@ Ensure to hold if_ioctl_lock when calling if_flags_set Fix locking against myself on ifpromisc vlan_unconfig_locked could be called with holding if_ioctl_lock. Ensure to not turn on IFF_RUNNING of an interface until its initialization completes And ensure to turn off it before destruction as per IFF_RUNNING's description "resource allocated". (The description is a bit doubtful though, I believe the change is still proper.) Ensure to hold if_ioctl_lock on if_up and if_down One exception for if_down is if_detach; in the case the lock isn't needed because it's guaranteed that no other one can access ifp at that point. Make if_link_queue MP-safe if IFEF_MPSAFE if_link_queue is a queue to store events of link state changes, which is used to pass events from (typically) an interrupt handler to if_link_state_change softint. The queue was protected by KERNEL_LOCK so far, but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it by a spin mutex. Additionally with this change KERNEL_LOCK of if_link_state_change softint is omitted if NET_MPSAFE is enabled. Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of if_timer (see the comment). Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH At that point no other one modifies the list so IFADDR_READER_FOREACH is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though, if we try to detect contract violations of pserialize, using it violates the contract. So avoid using it makes life easy. Ensure to call if_addr_init with holding if_ioctl_lock Get rid of outdated comments Fix build of kernels without ether By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that created a unnecessary dependency from if.c to if_ethersubr.c. PR kern/52790 Rename IFNET_LOCK to IFNET_GLOBAL_LOCK IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then). Wrap if_ioctl_lock with IFNET_* macros (NFC) Also if_ioctl_lock perhaps needs to be renamed to something because it's now not just for ioctl... Reorder some destruction routines in if_detach - Destroy if_ioctl_lock at the end of the if_detach because it's used in various destruction routines - Move psref_target_destroy after pr_purgeif because we want to use psref in pr_purgeif (otherwise destruction procedures can be tricky) Ensure to call if_mcast_op with holding IFNET_LOCK Note that CARP doesn't deal with IFNET_LOCK yet. Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held Describe which lock is used to protect each member variable of struct ifnet Requested by skrll@ Write a guideline for converting an interface to IFEF_MPSAFE Requested by skrll@ Note that IFNET_LOCK must not be held in softint Don't set IFEF_MPSAFE unless NET_MPSAFE at this point Because recent investigations show that interfaces with IFEF_MPSAFE need to follow additional restrictions to work with the flag safely. We should enable it on an interface by default only if the interface surely satisfies the restrictions, which are described in if.h. Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because the network stack is still serialized by the big kernel locks by default.
1.73.2.2	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.73.2.1	10-Jun-2017	snj	Pull up following revision(s) (requested by nonaka in ticket #21): sys/dev/pci/if_iwm.c: revision 1.74 fix typo. reported by ryo@.
1.75.2.3	26-Jul-2017	snj	Pull up following revision(s) (requested by nonaka in ticket #1459): sys/dev/pci/if_iwm.c: revision 1.72 iwm(4): match Wireless 8265 device.
1.75.2.2	25-Jul-2017	snj	Apply patch (requested by nonaka in ticket #1413): iwm(4): Add a driver for Intel Wireless 726x, 316x, 826x and 416x series. Ported from OpenBSD.
1.75.2.1	23-Jul-2017	snj	file if_iwm.c was added on branch netbsd-7 on 2017-07-25 19:43:03 +0000
1.76.2.2	03-Dec-2017	jdolecek	update from HEAD
1.76.2.1	23-Oct-2017	jdolecek	file if_iwm.c was added on branch tls-maxphys on 2017-12-03 11:37:07 +0000
1.78.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.78.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.83.10.1	29-Feb-2020	ad	Sync with head.
1.83.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.85.6.2	01-Aug-2021	thorpej	Sync with HEAD.
1.85.6.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.89.2.1	02-Aug-2025	perseant	Sync with HEAD
1.14	01-Sep-2023	andvar	fix typos in comments, mainly s/innner/inner/.
1.13	24-Jun-2023	msaitoh	Fix typo in comment.
1.12	05-Dec-2021	msaitoh	s/coresponding/corresponding/ in comment.
1.11	05-Dec-2021	msaitoh	availabe -> available in comment.
1.10	10-Nov-2021	msaitoh	s/unoticable/unnoticeable/ in comment.
1.9	17-Aug-2021	andvar	fix multiplei repetitive typos in comments, messages and documentation. mainly because copy paste code big amount of files are affected.
1.8	02-Jun-2021	nia	Can't left shift a signed int by 31. Found by kubsan.
1.7	19-Apr-2018	christos	branches: 1.7.8; 1.7.20; s/static inline/static __inline/g for consistency.
1.6	14-Mar-2017	nonaka	branches: 1.6.6; 1.6.10; 1.6.14; 1.6.16; iwm(4): update firmware to maintained core release. firmware from https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=4b9559fb0565f582a21a4070a81a3b6a7d77610b
1.5	13-Jan-2017	nonaka	branches: 1.5.2; iwm(4): 3165 works fine with iwlwifi-7265D-17.ucode firmware. firmware from http://git.kernel.org/cgit/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=be8a6fcba2e48e00674d1f35e6def56e2268039c
1.4	09-Jan-2017	nonaka	Added enum iwm_tx_pm_timeouts from FreeBSD.
1.3	18-Dec-2016	nonaka	iwm(4): Sync with OpenBSD. 7265 works fine. 7260 and 3160 maybe work. 7265D, 3165 and 8260 not work yet. if_iwm.c: r1.147. if_iwmreg.h: r1.19 without r1.16. if_iwmvar.h: r1.24. firmware is taken from https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi - iwlwifi-7260-ucode-16.242414.0.tgz - iwlwifi-3160-ucode-16.242414.0.tgz - iwlwifi-7265-ucode-16.242414.0.tgz - iwlwifi-8000-ucode-16.242414.0.tgz
1.2	16-Feb-2015	nonaka	branches: 1.2.2; 1.2.4; whitespace
1.1	07-Feb-2015	pooka	Add a driver for Intel Centrino 7260 and similar wireless cards. Supported devices should more or less match ones supported by the Linux iwlwifi mvm driver. Sponsored by genua mbh for OpenBSD. This is probably the world's first Canadian cross device driver: it was created for OpenBSD by writing and porting a NetBSD driver which was developed in a rump kernel in Linux userspace. Note: I don't have access to the hardware anymore, so this version is not tested and not enabled by default. While I tried to be careful in adding the NetBSD bits back, it's probable that there's a snafu or two. Feel free to send private email in case you have the hardware and there are issues.
1.2.4.2	20-Mar-2017	pgoyette	Sync with HEAD
1.2.4.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.2.2.4	28-Aug-2017	skrll	Sync with HEAD
1.2.2.3	05-Feb-2017	skrll	Sync with HEAD
1.2.2.2	06-Apr-2015	skrll	Sync with HEAD
1.2.2.1	16-Feb-2015	skrll	file if_iwmreg.h was added on branch nick-nhusb on 2015-04-06 15:18:10 +0000
1.5.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.6.16.1	22-Apr-2018	pgoyette	Sync with HEAD
1.6.14.2	03-Dec-2017	jdolecek	update from HEAD
1.6.14.1	14-Mar-2017	jdolecek	file if_iwmreg.h was added on branch tls-maxphys on 2017-12-03 11:37:07 +0000
1.6.10.2	25-Jul-2017	snj	Apply patch (requested by nonaka in ticket #1413): iwm(4): Add a driver for Intel Wireless 726x, 316x, 826x and 416x series. Ported from OpenBSD.
1.6.10.1	14-Mar-2017	snj	file if_iwmreg.h was added on branch netbsd-7 on 2017-07-25 19:43:03 +0000
1.6.6.1	05-Jun-2021	martin	Pull up following revision(s) (requested by nia in ticket #1682): sys/dev/pci/if_iwmreg.h: revision 1.8 Can't left shift a signed int by 31. Found by kubsan.
1.7.20.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.7.8.1	05-Jun-2021	martin	Pull up following revision(s) (requested by nia in ticket #1282): sys/dev/pci/if_iwmreg.h: revision 1.8 Can't left shift a signed int by 31. Found by kubsan.
1.20	22-May-2020	thorpej	This driver sleeps during iwm_media_change(), and thus requires an adaptive mutex for the media lock.
1.19	05-Oct-2019	mrg	remove __packed attribute from struct ieee80211_radiotap_header and all the structures that include it. this should not change anything while avoiding packed vs alignment warnings from GCC 8, and potentially pessimised code generation due to the packed marker (there are no misaligned members, just that the per-device parts may end unaligned.) all consumers of these members are done from the properly aligned packet members directly, or, as a union with a 64 byte member, also properly aligned. codegen didn't appear to change, except for the definition of sizeof(struct driver_[rt]x_radiotap_header) in debug info, which is not directly used anywhere.
1.18	14-Mar-2017	nonaka	branches: 1.18.10; 1.18.14; 1.18.18; iwm(4): update firmware to maintained core release. firmware from https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=4b9559fb0565f582a21a4070a81a3b6a7d77610b
1.17	02-Feb-2017	nonaka	wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.16	21-Jan-2017	nonaka	Remove unused iwm_softc member.
1.15	13-Jan-2017	nonaka	branches: 1.15.2; iwm(4): 3165 works fine with iwlwifi-7265D-17.ucode firmware. firmware from http://git.kernel.org/cgit/linux/kernel/git/iwlwifi/linux-firmware.git/tree/?id=be8a6fcba2e48e00674d1f35e6def56e2268039c
1.14	10-Jan-2017	nonaka	check LAR support in NVM for 8260/4165.
1.13	10-Jan-2017	nonaka	Add apmg_wake_up_wa support from DragonFly BSD.
1.12	10-Jan-2017	nonaka	iwm(4): 8260 works now. 4165 may works too.
1.11	08-Jan-2017	nonaka	iwm(4): make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html
1.10	18-Dec-2016	nonaka	iwm(4): Sync with OpenBSD. 7265 works fine. 7260 and 3160 maybe work. 7265D, 3165 and 8260 not work yet. if_iwm.c: r1.147. if_iwmreg.h: r1.19 without r1.16. if_iwmvar.h: r1.24. firmware is taken from https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi - iwlwifi-7260-ucode-16.242414.0.tgz - iwlwifi-3160-ucode-16.242414.0.tgz - iwlwifi-7265-ucode-16.242414.0.tgz - iwlwifi-8000-ucode-16.242414.0.tgz
1.9	06-Nov-2015	nonaka	branches: 1.9.2; Always use pci_intr_alloc(9)/pci_intr_release(9).
1.8	22-Jul-2015	nonaka	iwm(4) can read the firmware image from disk when interface up. This change is useful for full-disk encryption environment.
1.7	15-May-2015	knakahara	if_iwm use unified establish API.
1.6	28-Apr-2015	nonaka	Use MSI instead of legacy INTx, if available. should fix PR/49663.
1.5	03-Mar-2015	nonaka	branches: 1.5.2; workaround for interrupt coalescing bug not needed on 7265. From OpenBSD if_iwm.c rev.1.25, if_iwmvar.h rev.1.7.
1.4	16-Feb-2015	nonaka	whitespace
1.3	13-Feb-2015	nonaka	iwm(4) works now, but 2.4GHz only. > iwm0 at pci2 dev 0 function 0: Intel Dual Band Wireless AC 3160 (rev. 0x83) > iwm0: interrupting at ioapic0 pin 18 > iwm0: hw rev: 0x160, fw ver 25.228 (API ver 9), address XX:XX:XX:XX:XX:XX > iwm0: 11b rates: 1Mbps 2Mbps 5.5Mbps 11Mbps > iwm0: 11g rates: 1Mbps 2Mbps 5.5Mbps 11Mbps 6Mbps 9Mbps 12Mbps 18Mbps 24Mbps 36Mbps 48Mbps 54Mbps
1.2	13-Feb-2015	nonaka	Sync with OpenBSD if_iwm.c rev.1.18.
1.1	07-Feb-2015	pooka	Add a driver for Intel Centrino 7260 and similar wireless cards. Supported devices should more or less match ones supported by the Linux iwlwifi mvm driver. Sponsored by genua mbh for OpenBSD. This is probably the world's first Canadian cross device driver: it was created for OpenBSD by writing and porting a NetBSD driver which was developed in a rump kernel in Linux userspace. Note: I don't have access to the hardware anymore, so this version is not tested and not enabled by default. While I tried to be careful in adding the NetBSD bits back, it's probable that there's a snafu or two. Feel free to send private email in case you have the hardware and there are issues.
1.5.2.7	28-Aug-2017	skrll	Sync with HEAD
1.5.2.6	05-Feb-2017	skrll	Sync with HEAD
1.5.2.5	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.5.2.4	22-Sep-2015	skrll	Sync with HEAD
1.5.2.3	06-Jun-2015	skrll	Sync with HEAD
1.5.2.2	06-Apr-2015	skrll	Sync with HEAD
1.5.2.1	03-Mar-2015	skrll	file if_iwmvar.h was added on branch nick-nhusb on 2015-04-06 15:18:10 +0000
1.9.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.9.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.15.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.18.18.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.18.14.2	03-Dec-2017	jdolecek	update from HEAD
1.18.14.1	14-Mar-2017	jdolecek	file if_iwmvar.h was added on branch tls-maxphys on 2017-12-03 11:37:08 +0000
1.18.10.2	25-Jul-2017	snj	Apply patch (requested by nonaka in ticket #1413): iwm(4): Add a driver for Intel Wireless 726x, 316x, 826x and 416x series. Ported from OpenBSD.
1.18.10.1	14-Mar-2017	snj	file if_iwmvar.h was added on branch netbsd-7 on 2017-07-25 19:43:03 +0000
1.101	10-Nov-2024	mlelstv	Add MBUFTRACE
1.100	03-Apr-2024	gutteridge	branches: 1.100.2; if_iwn.c: pick up proper firmware for Centrino Wireless-N 130 Both variants should use iwlwifi-6000g2b-6 not iwlwifi-6000g2a-6. (It seems only two specific product variants use iwlwifi-6000g2a-6. We could simplify by reversing the sense of the test, as OpenBSD does, but it doesn't seem to matter much, as what we now match seems to be the full gamut possible, so the simpler diff was chosen here.) Addresses PR kern/58105 from wandrien.dev@gmail.com, with the PCI_PRODUCT_INTEL_WIFI_LINK_130_1 match tested by the reporter. Code inspection of the FreeBSD driver indicates we should safely be able to match PCI_PRODUCT_INTEL_WIFI_LINK_130_2 also.
1.99	25-Apr-2022	gutteridge	branches: 1.99.4; Update firmware for Intel Centrino Advanced-N 6205 WiFi cards Use the final version of firmware provided for these cards. This has been tested with both variants of impacted iwn(4) cards (6005_2X2_1 and 6005_2X2_2, as found in a Dell E6230 and a Lenovo T420), improvements noted.
1.98	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.97	19-Sep-2021	andvar	fix few more typos in comments, messages and documentation.
1.96	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.95	08-May-2021	thorpej	Use pci_compatible_match().
1.94	20-Mar-2020	sevan	branches: 1.94.8; Apply the same change as for if_iwi.c r1.114 here, as part of kern/55090. This driver sleeps during iwn_media_change(), and thus requires an adaptive mutex for the media lock.
1.93	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.92	10-Oct-2019	bad	branches: 1.92.2; magic number police. use ETHER_ADDR_LEN.
1.91	19-Apr-2019	gutteridge	branches: 1.91.4; if_iwn.c: add a comment noting which PCI IDs are covered by an else statement that loads a specific firmware variant. While here, also amend the formatting of two other comments.
1.90	26-Jun-2018	msaitoh	branches: 1.90.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.89	01-May-2018	maya	GC private 802.11 rateset declarations, use the standard ones. Build tested only.
1.88	28-Jan-2018	christos	branches: 1.88.2; add an annotation.
1.87	16-Jan-2018	maxv	Fix overflow.
1.86	23-Oct-2017	msaitoh	If if_initialize() failed in the attach function, free resources and return.
1.85	19-Jul-2017	mlelstv	Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.84	02-Feb-2017	nonaka	branches: 1.84.6; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.83	02-Feb-2017	nonaka	iwn(4): use MSI if available.
1.82	04-Jan-2017	nonaka	branches: 1.82.2; Export some 802.11 IE manipulate functions.
1.81	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.80	24-Nov-2016	hkenken	bus_dmamap_sync() is required for ICT table read/write. Tested by arm platform.
1.79	03-Aug-2016	mlelstv	be less noisy with concurrent scan requests.
1.78	10-Jun-2016	ozaki-r	branches: 1.78.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.77	26-May-2016	ozaki-r	Use M_GETCTX No functional change.
1.76	22-Sep-2015	nonaka	PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.75	24-Aug-2015	pooka	Remove "#ifdef INET" code. Por que? Because opt_inet.h was not included and "#ifdef INET" was just a fancy way of saying "#ifndef NetBSD".
1.74	09-Nov-2014	nonaka	branches: 1.74.2; Use correct IWN_HW_REV_TYPE_MASK.
1.73	09-Nov-2014	nonaka	fix error messages.
1.72	30-Oct-2014	nonaka	Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.71	29-Mar-2014	christos	branches: 1.71.4; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.70	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.69	14-Sep-2013	joerg	Conditionalize iwn_memwrite_2 and friends under !IEEE80211_NO_HT.
1.68	23-Aug-2013	christos	PR/48150: Noriyuki Koizumi: Add support for Intel Centrino Advanced-N 6235 Wi-Fi controller
1.67	28-Jul-2013	prlw1	Add support for Centrino Wireless-N 1030 and Centrino Advanced-N 6230 http://mail-index.netbsd.org/tech-net/2013/01/04/msg003802.html
1.66	15-Jun-2013	christos	branches: 1.66.2; reflect the correct OpenBSD versions we are synced with.
1.65	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.64	30-Mar-2013	christos	replace function with macro
1.63	30-Mar-2013	christos	remove trailing whitespace
1.62	30-Jan-2012	drochner	branches: 1.62.2; 1.62.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.61	08-Oct-2011	elric	branches: 1.61.2; 1.61.6; Use aprint_error_dev().
1.60	08-Oct-2011	mbalmer	Make this compile again: Use the device_xname() macro to get the devic name.
1.59	07-Oct-2011	elric	Update from OpenBSD merging in 1.111 and 1.112 with log messages: 1.112 Differential gain calibration makes the 6005 firmware crap out, so skip it for now until we figure out why. This probably means the device won't function optimally, but that's better than not functioning at all. Makes my "Intel Centrinto Advanced-N 6205" work quite well. 1.111 The 6005 and 6050-based parts need DC calibration turned on, otherwise the firmware will crap out, at least on the 6005. First step to getting my "Intel Centrino Advanced-N 6205" to work. Both are commited by kettenis@openbsd.org.
1.58	28-Aug-2011	elric	Revert prior inadvertent checkin.
1.57	28-Aug-2011	elric	Remove crypto/dist/heimdal and its associated build infrastructure as it has been replaced by crypto/external/bsd/heimdal.
1.56	21-May-2011	msaitoh	Add newer Wireless WiFi Link 4965 devices, Centrino Wireless-N 1030 and Centrino Advanced-N 6230
1.55	20-May-2011	msaitoh	Add support for 6050(6250 and 6150) and 6005(6205). TODO: update iwn.4 manpage.
1.54	18-May-2011	dyoung	#include <sys/bus.h>, not <machine/bus.h>.
1.53	15-May-2011	christos	update from OpenBSD by msaitoh. Tested on amd64.
1.52	30-Dec-2010	jruoho	branches: 1.52.2; Use "temperature" rather than "TEMP" as the sensor name. No functional change.
1.51	30-Dec-2010	jruoho	Fill out struct ifnet::if_stop, which is required for network class devices in pmf(9). Fixes the panic reported in PR # kern/44296.
1.50	29-Aug-2010	christos	Don't try to hold the mutex if we are disabling the card because we can be called from an interrupt context. While this is not strictly correct, we are SoL anyway so this does not matter.
1.49	10-Jul-2010	christos	V() the mutex when firmware is initialized, duh!
1.48	03-Jul-2010	christos	Don't use spinlocks.
1.47	02-Jul-2010	christos	- Fix the recursive iwn_init() call problem properly. - Don't print the scan request while scanning all the time. Perhaps I will remove this completely.
1.46	18-Jun-2010	christos	Patch from Sverre Froyen to avoid panic when an ioctl happens while the driver is initializing.
1.45	12-May-2010	christos	Use device names from the OpenBSD iwn driver, adding and modifying device names as needed. From Sverre Froyen
1.44	02-May-2010	christos	Replace iwn_tx from rev. 1.39 with a port of the current OpenBSD version. Remove superfluous call to bpf_detach. Add comments regarding porting issues and add a couple of cosmetic changes that reduce the diffs to the OpenBSD version. From Sverre Froyen
1.43	28-Apr-2010	christos	kill stray free. from Sverre Froyen.
1.42	23-Apr-2010	christos	Align usage of free and alloc and clean up the firmware read code. This fixes two panics that occur for kernels compiled with DIAGNOSTIC and DEBUG. From Sverre Froyen
1.41	17-Apr-2010	christos	add support for the 6000. From sverre froyen
1.40	16-Apr-2010	christos	Update with newer version from Sverre Froyen
1.39	11-Apr-2010	mrg	avoid rbuf starvation. when grabbing a recieve buffer, if a local 4KB buffer isn't available, try to MEXTMALLOC() one. only if both of these fail to we abort the recieve operation. based on a patch from Sverre Froyen <sverre@viewmark.com>, thanks! tested by myself, sverre and jeremy reed.
1.38	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.37	24-Feb-2010	dyoung	branches: 1.37.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.36	19-Jan-2010	pooka	branches: 1.36.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.35	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.34	21-Nov-2009	njoly	Cleanup boot messages, use aprint functions.
1.33	11-Sep-2009	christos	Add support for the 5100 and newer firmware. Tested also with the 4900. Thanks to everyone for helping and testing.
1.32	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.31	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.30	07-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.29	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.28	22-Dec-2008	blymn	branches: 1.28.2; Pay more attention to the state of the RF kill switch, such as don't bring the interface up when the RF is off.
1.27	13-Nov-2008	blymn	Actually declare the structures used in iwn_scan_hdr instead of using pointer arithmetic to create the elements on the fly. Other changes to scan routine taken from FreeBSD driver.
1.26	07-Nov-2008	joerg	Match image name with upstream distfiles.
1.25	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.24	06-Nov-2008	blymn	Move init of essid to the attach code.
1.23	05-Nov-2008	blymn	* Disable aborting the scan for the moment - it causes a firmware error if a scan was never initiated which breaks WEP and open wireless connections * Make sure the ESSID is cleared.
1.22	20-Oct-2008	rtr	branches: 1.22.2; 1.22.4; - fix crazy condition that always evaluates to false without this iwn_fix_channel() never gets called. - remove hardcode to ic->ic_channels[11] fixes a problem in associating to access points
1.21	18-Oct-2008	blymn	Make scan abort command async as newstate is called from an interrupt context so we cannot ltsleep.
1.20	13-Oct-2008	blymn	DMA fixes thanks to rtr & skrll Don't set tsf in auth, this stops firmware errors on WPA renegotiation Various other tweaks that the linux driver seems to have that we did not
1.19	20-Sep-2008	freza	iwn_node_alloc(): Allocate 'iwn_node' in M_80211_NODE (which is what ieee80211_node:node_free() expects), not M_DEVBUF. Fixes DIAGNOSTIC crashes due to suspected double-free.
1.18	14-Sep-2008	freza	* iwn_rx_intr(): free rbuf is available if 'nb_free_entries > 0'. This used to have '>=' resulting in kernel crash accessing NULL rbuf under non-trivial Rx load.
1.17	18-Aug-2008	cube	- Make RX queue free list code look like in if_wpi.c, including: - Protect RX queue free list with a mutex, as it was done in so many network drivers now that it calls for common code, as dyoung@ points out. However, for now it should improve a bit iwn(4)'s stability.
1.16	29-Jul-2008	christos	fix typo.
1.15	28-Jul-2008	christos	just whitespace fixes.
1.14	28-Jul-2008	christos	fix type punned warnings.
1.13	28-Jul-2008	christos	lint wins again :-) From Anon Ymous: - fix a mis-placed parenthesis. - fix assignment to the wrong variable.
1.12	26-Jul-2008	christos	make sure we write little endian. From Anon Ymous
1.11	24-Jul-2008	blymn	* Add the BSS node during the auth phase so negotiation can occur * Make the adding of a node into a function instead of duplicating code
1.10	10-May-2008	degroote	branches: 1.10.2; 1.10.4; Fix error handling after m_pullup
1.9	29-Mar-2008	blymn	branches: 1.9.2; 1.9.4; 1.9.6; Remove erroneous byteswap of flags - the flags are byteswapped when assigned to the tx struct.
1.8	29-Mar-2008	blymn	Clean up trailing whitespace.
1.7	13-Mar-2008	taca	branches: 1.7.2; Reflect change of PMF_FN_ARGS.
1.6	11-Mar-2008	dyoung	pci_disable_retry() is now a no-op, so don't call it any more.
1.5	16-Feb-2008	ober	branches: 1.5.2; 1.5.4; 1.5.8; More cleanup. Removal of debug comments.
1.4	10-Feb-2008	skrll	branches: 1.4.2; Use PRIu64 in a DPRINTF.
1.3	09-Feb-2008	skrll	More whitespace and RCSIds.
1.2	09-Feb-2008	ober	Clean up white space and KNF everything.
1.1	09-Feb-2008	ober	Intel PRO/Wireless LAN 4965AGN Mini-PCI Adapter Driver ported from OpenBSD by deroote@. OK nick@ moof@
1.4.2.4	17-Mar-2008	yamt	sync with head.
1.4.2.3	27-Feb-2008	yamt	sync with head.
1.4.2.2	11-Feb-2008	yamt	sync with head.
1.4.2.1	10-Feb-2008	yamt	file if_iwn.c was added on branch yamt-lazymbuf on 2008-02-11 14:59:38 +0000
1.5.8.4	17-Jan-2009	mjf	Sync with HEAD.
1.5.8.3	28-Sep-2008	mjf	Sync with HEAD.
1.5.8.2	02-Jun-2008	mjf	Sync with HEAD.
1.5.8.1	03-Apr-2008	mjf	Sync with HEAD.
1.5.4.1	24-Mar-2008	keiichi	sync with head.
1.5.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.5.2.1	16-Feb-2008	mjf	file if_iwn.c was added on branch mjf-devfs on 2008-02-18 21:05:57 +0000
1.7.2.2	23-Mar-2008	matt	sync with HEAD
1.7.2.1	13-Mar-2008	matt	file if_iwn.c was added on branch matt-armv6 on 2008-03-23 02:04:47 +0000
1.9.6.4	10-Oct-2008	skrll	Sync with HEAD.
1.9.6.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.9.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.9.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.9.4.7	09-Oct-2010	yamt	sync with head
1.9.4.6	11-Aug-2010	yamt	sync with head.
1.9.4.5	11-Mar-2010	yamt	sync with head
1.9.4.4	16-Sep-2009	yamt	sync with head
1.9.4.3	16-May-2009	yamt	sync with head
1.9.4.2	04-May-2009	yamt	sync with head.
1.9.4.1	16-May-2008	yamt	sync with head.
1.9.2.1	18-May-2008	yamt	sync with head.
1.10.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.10.4.1	19-Oct-2008	haad	Sync with HEAD.
1.10.2.2	31-Jul-2008	simonb	Sync with head.
1.10.2.1	28-Jul-2008	simonb	Sync with head.
1.22.4.3	15-Nov-2008	snj	Pull up following revision(s) (requested by blymn in ticket #58): sys/dev/pci/if_iwn.c: revision 1.24 Move init of essid to the attach code.
1.22.4.2	15-Nov-2008	snj	Pull up following revision(s) (requested by blymn in ticket #58): sys/dev/pci/if_iwn.c: revision 1.23 * Disable aborting the scan for the moment - it causes a firmware error if a scan was never initiated which breaks WEP and open wireless connections * Make sure the ESSID is cleared.
1.22.4.1	09-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #25): sys/dev/pci/if_ipw.c: revision 1.44 sys/dev/pci/if_ipwvar.h: revision 1.14 sys/dev/pci/if_iwi.c: revision 1.76 sys/dev/pci/if_iwn.c: revision 1.26 Match image name with upstream distfiles.
1.22.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.28.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.36.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.36.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.36.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.37.2.4	31-May-2011	rmind	sync with head
1.37.2.3	05-Mar-2011	rmind	sync with head
1.37.2.2	03-Jul-2010	rmind	sync with head
1.37.2.1	30-May-2010	rmind	sync with head
1.52.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.61.6.1	18-Feb-2012	mrg	merge to -current.
1.61.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.61.2.1	17-Apr-2012	yamt	sync with head
1.62.6.3	03-Dec-2017	jdolecek	update from HEAD
1.62.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.62.6.1	23-Jun-2013	tls	resync from head
1.62.2.1	15-Nov-2015	bouyer	Pull up following revision(s) (requested by nonaka in ticket #1330): sys/dev/pci/if_iwnvar.h: revision 1.17 sys/dev/pci/if_iwn.c: revision 1.76 PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.66.2.2	18-May-2014	rmind	sync with head
1.66.2.1	28-Aug-2013	rmind	sync with head
1.71.4.2	06-Nov-2015	riz	Pull up following revision(s) (requested by nonaka in ticket #987): sys/dev/pci/if_iwnvar.h: revision 1.17 sys/dev/pci/if_iwn.c: revision 1.76 PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.71.4.1	28-Jan-2015	martin	Pull up following revision(s) (requested by nonaka in ticket #467): external/intel-fw-public/iwl2030/dist/iwlwifi-2030-6.ucode: revision 1.1 distrib/sets/lists/base/mi: revision 1.1091 share/man/man4/iwn.4: revision 1.13 external/intel-fw-public/iwl100/Makefile: revision 1.1 external/intel-fw-public/iwl2030/Makefile: revision 1.1 external/intel-fw-public/iwl135/Makefile: revision 1.1 sys/dev/pci/pcidevs: revision 1.1205 external/intel-fw-public/Makefile: revision 1.8 external/intel-fw-public/iwl135/dist/iwlwifi-135-6.ucode: revision 1.1 external/intel-fw-public/iwl2030/dist/LICENSE.iwlwifi-2030-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/LICENSE.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/LICENSE.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/iwlwifi-100-5.ucode: revision 1.1 external/intel-fw-public/iwl135/dist/LICENSE.iwlwifi-135-ucode: revision 1.1 external/intel-fw-public/iwl135/dist/README.iwlwifi-135-ucode: revision 1.1 sys/dev/pci/if_iwnvar.h: revision 1.16 external/intel-fw-public/iwl2000/dist/iwlwifi-2000-6.ucode: revision 1.1 external/intel-fw-public/iwl100/dist/README.iwlwifi-100-ucode: revision 1.1 sys/dev/pci/if_iwn.c: revision 1.72 sys/dev/pci/if_iwnreg.h: revision 1.14 external/intel-fw-public/iwl105/dist/README.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/LICENSE.iwlwifi-100-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/README.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/iwlwifi-105-6.ucode: revision 1.1 external/intel-fw-public/iwl2000/Makefile: revision 1.1 external/intel-fw-public/iwl105/Makefile: revision 1.1 external/intel-fw-public/iwl2030/dist/README.iwlwifi-2030-ucode: revision 1.1 Add some Intel Wi-Fi devices. Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.74.2.8	28-Aug-2017	skrll	Sync with HEAD
1.74.2.7	05-Feb-2017	skrll	Sync with HEAD
1.74.2.6	05-Dec-2016	skrll	Sync with HEAD
1.74.2.5	05-Oct-2016	skrll	Sync with HEAD
1.74.2.4	09-Jul-2016	skrll	Sync with HEAD
1.74.2.3	29-May-2016	skrll	Sync with HEAD
1.74.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.74.2.1	22-Sep-2015	skrll	Sync with HEAD
1.78.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.78.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.78.2.1	06-Aug-2016	pgoyette	Sync with HEAD
1.82.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.84.6.3	17-Apr-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #1953): sys/dev/pci/if_iwn.c: revision 1.100 if_iwn.c: pick up proper firmware for Centrino Wireless-N 130 Both variants should use iwlwifi-6000g2b-6 not iwlwifi-6000g2a-6. (It seems only two specific product variants use iwlwifi-6000g2a-6. We could simplify by reversing the sense of the test, as OpenBSD does, but it doesn't seem to matter much, as what we now match seems to be the full gamut possible, so the simpler diff was chosen here.) Addresses PR kern/58105 from wandrien.dev%gmail.com@localhost, with the PCI_PRODUCT_INTEL_WIFI_LINK_130_1 match tested by the reporter. Code inspection of the FreeBSD driver indicates we should safely be able to match PCI_PRODUCT_INTEL_WIFI_LINK_130_2 also.
1.84.6.2	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.84.6.1	25-Aug-2017	snj	Pull up following revision(s) (requested by mlelstv in ticket #218): sys/dev/pci/if_iwn.c: revision 1.85 sys/dev/pci/if_iwnreg.h: revision 1.17 sys/dev/pci/if_iwnvar.h: revision 1.20 Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.88.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.88.2.1	02-May-2018	pgoyette	Synch with HEAD
1.90.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.90.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.90.2.1	10-Jun-2019	christos	Sync with HEAD
1.91.4.1	17-Apr-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #1825): sys/dev/pci/if_iwn.c: revision 1.100 if_iwn.c: pick up proper firmware for Centrino Wireless-N 130 Both variants should use iwlwifi-6000g2b-6 not iwlwifi-6000g2a-6. (It seems only two specific product variants use iwlwifi-6000g2a-6. We could simplify by reversing the sense of the test, as OpenBSD does, but it doesn't seem to matter much, as what we now match seems to be the full gamut possible, so the simpler diff was chosen here.) Addresses PR kern/58105 from wandrien.dev%gmail.com@localhost, with the PCI_PRODUCT_INTEL_WIFI_LINK_130_1 match tested by the reporter. Code inspection of the FreeBSD driver indicates we should safely be able to match PCI_PRODUCT_INTEL_WIFI_LINK_130_2 also.
1.92.2.1	29-Feb-2020	ad	Sync with head.
1.94.8.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.94.8.1	13-May-2021	thorpej	Sync with HEAD.
1.99.4.1	17-Apr-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #653): sys/dev/pci/if_iwn.c: revision 1.100 if_iwn.c: pick up proper firmware for Centrino Wireless-N 130 Both variants should use iwlwifi-6000g2b-6 not iwlwifi-6000g2a-6. (It seems only two specific product variants use iwlwifi-6000g2a-6. We could simplify by reversing the sense of the test, as OpenBSD does, but it doesn't seem to matter much, as what we now match seems to be the full gamut possible, so the simpler diff was chosen here.) Addresses PR kern/58105 from wandrien.dev%gmail.com@localhost, with the PCI_PRODUCT_INTEL_WIFI_LINK_130_1 match tested by the reporter. Code inspection of the FreeBSD driver indicates we should safely be able to match PCI_PRODUCT_INTEL_WIFI_LINK_130_2 also.
1.100.2.1	02-Aug-2025	perseant	Sync with HEAD
1.18	17-Jun-2021	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.17	19-Jul-2017	mlelstv	branches: 1.17.22; Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.16	11-Mar-2017	maya	branches: 1.16.6; Add missing comma, else the strings are concatenated (and possibly the wrong message is used). noted by coverity. Linux iwlwifi strings seem similar.
1.15	09-Nov-2014	nonaka	branches: 1.15.2; 1.15.4; 1.15.6; Use correct IWN_HW_REV_TYPE_MASK.
1.14	30-Oct-2014	nonaka	Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.13	28-Jul-2013	prlw1	branches: 1.13.4; Add support for Centrino Wireless-N 1030 and Centrino Advanced-N 6230 http://mail-index.netbsd.org/tech-net/2013/01/04/msg003802.html
1.12	15-Jun-2013	christos	branches: 1.12.2; reflect the correct OpenBSD versions we are synced with.
1.11	30-Mar-2013	christos	Use only one set of EDCA variables
1.10	08-Oct-2011	elric	branches: 1.10.2; 1.10.12; Update from OpenBSD merging in 1.111 and 1.112 with log messages: 1.112 Differential gain calibration makes the 6005 firmware crap out, so skip it for now until we figure out why. This probably means the device won't function optimally, but that's better than not functioning at all. Makes my "Intel Centrinto Advanced-N 6205" work quite well. 1.111 The 6005 and 6050-based parts need DC calibration turned on, otherwise the firmware will crap out, at least on the 6005. First step to getting my "Intel Centrino Advanced-N 6205" to work. Both are commited by kettenis@openbsd.org.
1.9	15-May-2011	christos	update from OpenBSD by msaitoh. Tested on amd64.
1.8	18-Jun-2010	christos	branches: 1.8.2; Patch from Sverre Froyen to avoid panic when an ioctl happens while the driver is initializing.
1.7	16-Apr-2010	christos	Update with newer version from Sverre Froyen
1.6	11-Sep-2009	christos	branches: 1.6.2; 1.6.4; Add support for the 5100 and newer firmware. Tested also with the 4900. Thanks to everyone for helping and testing.
1.5	13-Nov-2008	blymn	Actually declare the structures used in iwn_scan_hdr instead of using pointer arithmetic to create the elements on the fly. Other changes to scan routine taken from FreeBSD driver.
1.4	13-Oct-2008	blymn	branches: 1.4.2; Correct beacon structure Add scan abort command Add TX status codes from FreeBSD driver
1.3	09-Feb-2008	skrll	branches: 1.3.2; 1.3.4; 1.3.10; 1.3.12; 1.3.16; 1.3.22; More whitespace and RCSIds.
1.2	09-Feb-2008	ober	Clean up white space and KNF everything.
1.1	09-Feb-2008	ober	Intel PRO/Wireless LAN 4965AGN Mini-PCI Adapter Driver ported from OpenBSD by deroote@. OK nick@ moof@
1.3.22.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.3.22.1	19-Oct-2008	haad	Sync with HEAD.
1.3.16.3	11-Aug-2010	yamt	sync with head.
1.3.16.2	16-Sep-2009	yamt	sync with head
1.3.16.1	04-May-2009	yamt	sync with head.
1.3.12.2	23-Mar-2008	matt	sync with HEAD
1.3.12.1	09-Feb-2008	matt	file if_iwnreg.h was added on branch matt-armv6 on 2008-03-23 02:04:47 +0000
1.3.10.1	17-Jan-2009	mjf	Sync with HEAD.
1.3.4.2	18-Feb-2008	mjf	Sync with HEAD.
1.3.4.1	09-Feb-2008	mjf	file if_iwnreg.h was added on branch mjf-devfs on 2008-02-18 21:05:57 +0000
1.3.2.2	11-Feb-2008	yamt	sync with head.
1.3.2.1	09-Feb-2008	yamt	file if_iwnreg.h was added on branch yamt-lazymbuf on 2008-02-11 14:59:38 +0000
1.4.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.6.4.3	31-May-2011	rmind	sync with head
1.6.4.2	03-Jul-2010	rmind	sync with head
1.6.4.1	30-May-2010	rmind	sync with head
1.6.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.6.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.8.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.10.12.3	03-Dec-2017	jdolecek	update from HEAD
1.10.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.10.12.1	23-Jun-2013	tls	resync from head
1.10.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.2.1	28-Aug-2013	rmind	sync with head
1.13.4.1	28-Jan-2015	martin	Pull up following revision(s) (requested by nonaka in ticket #467): external/intel-fw-public/iwl2030/dist/iwlwifi-2030-6.ucode: revision 1.1 distrib/sets/lists/base/mi: revision 1.1091 share/man/man4/iwn.4: revision 1.13 external/intel-fw-public/iwl100/Makefile: revision 1.1 external/intel-fw-public/iwl2030/Makefile: revision 1.1 external/intel-fw-public/iwl135/Makefile: revision 1.1 sys/dev/pci/pcidevs: revision 1.1205 external/intel-fw-public/Makefile: revision 1.8 external/intel-fw-public/iwl135/dist/iwlwifi-135-6.ucode: revision 1.1 external/intel-fw-public/iwl2030/dist/LICENSE.iwlwifi-2030-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/LICENSE.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/LICENSE.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/iwlwifi-100-5.ucode: revision 1.1 external/intel-fw-public/iwl135/dist/LICENSE.iwlwifi-135-ucode: revision 1.1 external/intel-fw-public/iwl135/dist/README.iwlwifi-135-ucode: revision 1.1 sys/dev/pci/if_iwnvar.h: revision 1.16 external/intel-fw-public/iwl2000/dist/iwlwifi-2000-6.ucode: revision 1.1 external/intel-fw-public/iwl100/dist/README.iwlwifi-100-ucode: revision 1.1 sys/dev/pci/if_iwn.c: revision 1.72 sys/dev/pci/if_iwnreg.h: revision 1.14 external/intel-fw-public/iwl105/dist/README.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/LICENSE.iwlwifi-100-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/README.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/iwlwifi-105-6.ucode: revision 1.1 external/intel-fw-public/iwl2000/Makefile: revision 1.1 external/intel-fw-public/iwl105/Makefile: revision 1.1 external/intel-fw-public/iwl2030/dist/README.iwlwifi-2030-ucode: revision 1.1 Add some Intel Wi-Fi devices. Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.15.6.1	21-Apr-2017	bouyer	Sync with HEAD
1.15.4.1	20-Mar-2017	pgoyette	Sync with HEAD
1.15.2.1	28-Aug-2017	skrll	Sync with HEAD
1.16.6.1	25-Aug-2017	snj	Pull up following revision(s) (requested by mlelstv in ticket #218): sys/dev/pci/if_iwn.c: revision 1.85 sys/dev/pci/if_iwnreg.h: revision 1.17 sys/dev/pci/if_iwnvar.h: revision 1.20 Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.17.22.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.22	20-Mar-2020	sevan	Apply the same change as for if_iwi.c r1.114 here, as part of kern/55090. This driver sleeps during iwn_media_change(), and thus requires an adaptive mutex for the media lock.
1.21	05-Oct-2019	mrg	remove __packed attribute from struct ieee80211_radiotap_header and all the structures that include it. this should not change anything while avoiding packed vs alignment warnings from GCC 8, and potentially pessimised code generation due to the packed marker (there are no misaligned members, just that the per-device parts may end unaligned.) all consumers of these members are done from the properly aligned packet members directly, or, as a union with a 64 byte member, also properly aligned. codegen didn't appear to change, except for the definition of sizeof(struct driver_[rt]x_radiotap_header) in debug info, which is not directly used anywhere.
1.20	19-Jul-2017	mlelstv	branches: 1.20.4; Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.19	02-Feb-2017	nonaka	branches: 1.19.6; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.18	02-Feb-2017	nonaka	iwn(4): use MSI if available.
1.17	22-Sep-2015	nonaka	branches: 1.17.2; 1.17.4; PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.16	30-Oct-2014	nonaka	branches: 1.16.2; Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.15	28-Jul-2013	prlw1	branches: 1.15.4; Add support for Centrino Wireless-N 1030 and Centrino Advanced-N 6230 http://mail-index.netbsd.org/tech-net/2013/01/04/msg003802.html
1.14	15-Jun-2013	christos	branches: 1.14.2; reflect the correct OpenBSD versions we are synced with.
1.13	15-May-2011	christos	branches: 1.13.4; 1.13.10; 1.13.14; update from OpenBSD by msaitoh. Tested on amd64.
1.12	02-Jul-2010	christos	branches: 1.12.2; - Fix the recursive iwn_init() call problem properly. - Don't print the scan request while scanning all the time. Perhaps I will remove this completely.
1.11	18-Jun-2010	christos	Patch from Sverre Froyen to avoid panic when an ioctl happens while the driver is initializing.
1.10	23-Apr-2010	christos	Align usage of free and alloc and clean up the firmware read code. This fixes two panics that occur for kernels compiled with DIAGNOSTIC and DEBUG. From Sverre Froyen
1.9	16-Apr-2010	christos	Update with newer version from Sverre Froyen
1.8	19-Jan-2010	pooka	branches: 1.8.2; 1.8.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.7	17-Jan-2010	pooka	Forward declare struct bpf_if and use that as the type for bpf_if instead of "void *". Buys us oo times the type-safety for 0 times the price. (no functional change)
1.6	11-Sep-2009	christos	Add support for the 5100 and newer firmware. Tested also with the 4900. Thanks to everyone for helping and testing.
1.5	22-Dec-2008	blymn	Pay more attention to the state of the RF kill switch, such as don't bring the interface up when the RF is off.
1.4	18-Aug-2008	cube	branches: 1.4.2; - Make RX queue free list code look like in if_wpi.c, including: - Protect RX queue free list with a mutex, as it was done in so many network drivers now that it calls for common code, as dyoung@ points out. However, for now it should improve a bit iwn(4)'s stability.
1.3	09-Feb-2008	skrll	branches: 1.3.2; 1.3.4; 1.3.10; 1.3.12; 1.3.16; 1.3.18; 1.3.22; More whitespace and RCSIds.
1.2	09-Feb-2008	ober	Clean up white space and KNF everything.
1.1	09-Feb-2008	ober	Intel PRO/Wireless LAN 4965AGN Mini-PCI Adapter Driver ported from OpenBSD by deroote@. OK nick@ moof@
1.3.22.1	19-Oct-2008	haad	Sync with HEAD.
1.3.18.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.3.16.4	11-Aug-2010	yamt	sync with head.
1.3.16.3	11-Mar-2010	yamt	sync with head
1.3.16.2	16-Sep-2009	yamt	sync with head
1.3.16.1	04-May-2009	yamt	sync with head.
1.3.12.2	23-Mar-2008	matt	sync with HEAD
1.3.12.1	09-Feb-2008	matt	file if_iwnvar.h was added on branch matt-armv6 on 2008-03-23 02:04:47 +0000
1.3.10.2	17-Jan-2009	mjf	Sync with HEAD.
1.3.10.1	28-Sep-2008	mjf	Sync with HEAD.
1.3.4.2	18-Feb-2008	mjf	Sync with HEAD.
1.3.4.1	09-Feb-2008	mjf	file if_iwnvar.h was added on branch mjf-devfs on 2008-02-18 21:05:57 +0000
1.3.2.2	11-Feb-2008	yamt	sync with head.
1.3.2.1	09-Feb-2008	yamt	file if_iwnvar.h was added on branch yamt-lazymbuf on 2008-02-11 14:59:38 +0000
1.4.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.8.4.3	31-May-2011	rmind	sync with head
1.8.4.2	03-Jul-2010	rmind	sync with head
1.8.4.1	30-May-2010	rmind	sync with head
1.8.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.8.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.12.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.13.14.3	03-Dec-2017	jdolecek	update from HEAD
1.13.14.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.13.14.1	23-Jun-2013	tls	resync from head
1.13.10.1	15-Nov-2015	bouyer	Pull up following revision(s) (requested by nonaka in ticket #1330): sys/dev/pci/if_iwnvar.h: revision 1.17 sys/dev/pci/if_iwn.c: revision 1.76 PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.13.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.14.2.1	28-Aug-2013	rmind	sync with head
1.15.4.2	06-Nov-2015	riz	Pull up following revision(s) (requested by nonaka in ticket #987): sys/dev/pci/if_iwnvar.h: revision 1.17 sys/dev/pci/if_iwn.c: revision 1.76 PR/50187: Don't use DS parameter set when 5GHz channel is scanning. XXX: pullup-7
1.15.4.1	28-Jan-2015	martin	Pull up following revision(s) (requested by nonaka in ticket #467): external/intel-fw-public/iwl2030/dist/iwlwifi-2030-6.ucode: revision 1.1 distrib/sets/lists/base/mi: revision 1.1091 share/man/man4/iwn.4: revision 1.13 external/intel-fw-public/iwl100/Makefile: revision 1.1 external/intel-fw-public/iwl2030/Makefile: revision 1.1 external/intel-fw-public/iwl135/Makefile: revision 1.1 sys/dev/pci/pcidevs: revision 1.1205 external/intel-fw-public/Makefile: revision 1.8 external/intel-fw-public/iwl135/dist/iwlwifi-135-6.ucode: revision 1.1 external/intel-fw-public/iwl2030/dist/LICENSE.iwlwifi-2030-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/LICENSE.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/LICENSE.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/iwlwifi-100-5.ucode: revision 1.1 external/intel-fw-public/iwl135/dist/LICENSE.iwlwifi-135-ucode: revision 1.1 external/intel-fw-public/iwl135/dist/README.iwlwifi-135-ucode: revision 1.1 sys/dev/pci/if_iwnvar.h: revision 1.16 external/intel-fw-public/iwl2000/dist/iwlwifi-2000-6.ucode: revision 1.1 external/intel-fw-public/iwl100/dist/README.iwlwifi-100-ucode: revision 1.1 sys/dev/pci/if_iwn.c: revision 1.72 sys/dev/pci/if_iwnreg.h: revision 1.14 external/intel-fw-public/iwl105/dist/README.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/LICENSE.iwlwifi-100-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/README.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/iwlwifi-105-6.ucode: revision 1.1 external/intel-fw-public/iwl2000/Makefile: revision 1.1 external/intel-fw-public/iwl105/Makefile: revision 1.1 external/intel-fw-public/iwl2030/dist/README.iwlwifi-2030-ucode: revision 1.1 Add some Intel Wi-Fi devices. Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.16.2.3	28-Aug-2017	skrll	Sync with HEAD
1.16.2.2	05-Feb-2017	skrll	Sync with HEAD
1.16.2.1	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.17.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.17.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.19.6.1	25-Aug-2017	snj	Pull up following revision(s) (requested by mlelstv in ticket #218): sys/dev/pci/if_iwn.c: revision 1.85 sys/dev/pci/if_iwnreg.h: revision 1.17 sys/dev/pci/if_iwnvar.h: revision 1.20 Pull in some hardware support from FreeBSD. Also fix 5GHz mode by adapting the method described in: https://forums.freebsd.org/threads/53574/.
1.20.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.20.4.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.100	25-Mar-2025	christos	Don't return random value from interrupt. (coverity)
1.99	29-Jun-2024	riastradh	branches: 1.99.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.98	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.97	05-Feb-2024	andvar	fix various typos in comments.
1.96	20-Dec-2023	skrll	Remove unnecssary #include
1.95	13-Oct-2023	yamaguchi	ixl(4): limit receive frame size upto 9600 even if ETHERCAP_VLAN_MTU is enabled
1.94	13-Oct-2023	yamaguchi	ixl(4): added ETHERCAP_VLAN_MTU support PR kern/57652
1.93	11-Oct-2023	rin	ixl(4): Print device info on attach Without PCIVERBOSE option: ixl0 at pci1 dev 0 function 0: vendor 8086 product 1572 (rev. 0x01) With PCIVERBOSE option: ixl0 at pci1 dev 0 function 0: Intel XL710 SFP+ Ethernet (SFI) (rev. 0x01) OK yamaguchi@
1.92	11-Oct-2023	yamaguchi	ixl(4): update link status in workqueue
1.91	11-Oct-2023	yamaguchi	ixl(4): use cv_broadcast to wakeup ioctl and workqueue context. Commands may be issued at the same time from the both context.
1.90	11-Oct-2023	yamaguchi	ixl(4): skip getting link status if the last command is running
1.89	26-Mar-2023	andvar	fix various typos in documentation, comments and sysctl device description. mainly aion -> ation and inlude -> include.
1.88	16-Sep-2022	knakahara	branches: 1.88.4; Add ALWAYS_TXDEFER option to ixl(4), too.
1.87	28-Aug-2022	skrll	Empty if_watchdog functions are useless and now waste resources. Let's remove them.
1.86	25-Aug-2022	knakahara	Tune number of Rx descriptors for ixl(4), ok'ed by yamaguchi@n.o. In some tunneling measurement cases, that improve performance. In other cases, that keep up performance.
1.85	17-Jun-2022	yamaguchi	ixl(4): fix typo
1.84	17-Jun-2022	yamaguchi	ixl(4), iavf(4): fix endian bug in vlan tag
1.83	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.82	31-Mar-2022	yamaguchi	call detach functions before stopping the interrupt for atq
1.81	31-Mar-2022	yamaguchi	Use xc_barrier to wait for finish of the interrupt handler poited out by riastradh@n.o, thanks.
1.80	31-Mar-2022	yamaguchi	ixl(4): remove unnecessary lock acquirement Running interrupt handlers is finished while accessing to I40E_Q[TR]X_ENA register. Therefore, it is not need to acquire locks to wait for the finish. pointed out by riastradh@n.o, thanks.
1.79	24-Mar-2022	yamaguchi	ixl(4): Added Intel V710 for 5000BaseT support
1.78	16-Mar-2022	yamaguchi	ixl: Added support for Intel ethernet X710-T*L
1.77	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.76	09-Feb-2021	jakllsch	branches: 1.76.4; Zero interrupt handles upon allocation. Might prevent detach-time bugs.
1.75	08-Sep-2020	yamaguchi	branches: 1.75.2; Added iavf(4) that is based on OpenBSD's iavf(4) implementation reviewed by msaitoh@n.o and knakahara@n.o
1.74	19-Aug-2020	yamaguchi	Make descriptor_num in ixl(4) readonly because ixl(4) does not support reallocating related resources pointed out by knakahara@, thanks.
1.73	19-Aug-2020	yamaguchi	Adjust the default descriptor sizes for ixl(4) These decreasements has no impact for throughput while forwarding 64-1518 byte packets.
1.72	19-Aug-2020	yamaguchi	whitespace fix From msaitoh@, thanks.
1.71	31-Jul-2020	yamaguchi	Use if_initialize() and if_register()
1.70	31-Jul-2020	yamaguchi	Added sysctl nodes for the number of descriptors in ixl(4)
1.69	31-Jul-2020	yamaguchi	Added sysctl nodes for ixl(4) Interrupt throttling(ITR)
1.68	16-Jul-2020	yamaguchi	Set PCI_COMMAND_MASTER_ENABLE and PCI_COMMAND_MEM_ENABLE to activate the pci devices This configuration is needed when BIOS or UEFI do not make them set.
1.67	11-Jun-2020	thorpej	Update for proplib(3) API changes.
1.66	14-May-2020	msaitoh	Remove extra semicolon.
1.65	19-Mar-2020	yamaguchi	branches: 1.65.2; Fix panic due to aquire an adaptive mutex with spin mutex held The panic is caused by the spin mutex held in ifmedia before calling callbacks. ixl(4) registers the adaptive mutex used in the callbacks at initialization to prevent the panic.
1.64	13-Mar-2020	yamaguchi	Use I40E_PFINT_ITRN registers to enable ITR for MSI-X interrupts
1.63	13-Mar-2020	yamaguchi	Fix locking against myself at ixl_link_state_update() called by ixl_get_link_status() Though the timing to call the function is change, there is no order to get link status at initialization.
1.62	12-Mar-2020	yamaguchi	hold the lock for struct ixl_softc during read or write to sc->sc_media_* OKed by knakahara@n.o
1.61	12-Mar-2020	yamaguchi	remove unnecessary code because IFM_ACTIVE is setted in ixl_set_link_status() OKed by knakahara@n.o
1.60	03-Mar-2020	yamaguchi	Send packets even if link state is not up ixl(4) has not the limitation that it must not send packet when link is down. And most of the other NIC drivers do not refer to it when sending packets. reviewed by knakahara@n.o, thanks.
1.59	03-Mar-2020	yamaguchi	Use ixl_atq_exec for waiting the Get link status response of ixl(4) not to send the same command at once reviewed by knakahara@n.o., thanks.
1.58	03-Mar-2020	yamaguchi	Added KASSERT
1.57	03-Mar-2020	yamaguchi	Fix the crash when ixl(4) sends Get link status command and receives Get link status response in parallel sc->sc_link_state_atq.iatq_fn is used in ixl_atq_done_locked() with sc->sc_atq_lock, but it was cleared in ixl_get_link_status() without the lock.
1.56	25-Feb-2020	yamaguchi	Add a comment about RSS and remove a wrong comment pointed out and reviewed by knakahara@n.o., thanks
1.55	25-Feb-2020	yamaguchi	Use IFM_25G_ACC pointed out and reviewed by msaitoh@n.o., thanks
1.54	25-Feb-2020	yamaguchi	Use kmem_alloc instead of malloc pointed out and reviewed by knakahar@n.o., thanks
1.53	25-Feb-2020	yamaguchi	Change the number of retry to be the same as FreeBSD and Linux pointed out and reviewed by knakahara@n.o., thanks.
1.52	25-Feb-2020	yamaguchi	Disable all queues before waiting for the completion on each queue to reduce the number of delays pointed out and reviewed by knakahara@n.o., thanks
1.51	25-Feb-2020	yamaguchi	Use callout_halt() instead of callout_stop for safety pointed out and reviewed by knakahara@n.o., thanks
1.50	25-Feb-2020	yamaguchi	Stop callout for statistics while the interface is not running
1.49	25-Feb-2020	yamaguchi	Change position of the cpuid::ixl_hmc_txq to be same with ixl_hmc_pack_txq[] This has no functionality impact because the position on memory is defined in ixl_hmc_pack_txq[]. pointed out and reviewed by knakahara@n.o., thanks
1.48	25-Feb-2020	yamaguchi	Hold per-queue locks when clearing config of queues for safety
1.47	25-Feb-2020	yamaguchi	Add defines for the max number of queue pointed out and reviewed by knakahara@n.o., thanks
1.46	25-Feb-2020	yamaguchi	Set 0 to txr_{prod,cons} and rxr_{prod,cons} when HMC objects are cleared Those values should be the same as head and tail in an HMC object
1.45	25-Feb-2020	yamaguchi	Use workqueue API directly, without the wrapper in ixl(4) to improve performace by removing atomic_ops(3) pointed out and reviewed by knakahara@n.o., thanks
1.44	25-Feb-2020	yamaguchi	Remove WQ_PERCPU flag for the workqueue that does configurations such as link up and down And added kpreempt_disable and kpreempt_enable around workqueue_enqueue to call it in non-WQ_PERCPU thread context. pointed out and reviewed by knakahara@n.o., thanks.
1.43	25-Feb-2020	yamaguchi	Added __KERNEL_RCSID
1.42	12-Feb-2020	yamaguchi	Remove wrong KASSERTs because mutex_owned for a spin lock doesn't check that a lock is held in the current context.
1.41	12-Feb-2020	yamaguchi	Use Set RSS key command and Set LUT command to enable RSS on X722
1.40	12-Feb-2020	yamaguchi	xl(4) should use the aq command to access RX control registers if the device is X722 or others whose API version is upper than 1.5
1.39	12-Feb-2020	yamaguchi	Fix a wrong comment in ixl(4)
1.38	12-Feb-2020	yamaguchi	Change the command sequence and timing to prevent EIO in "Get PHY Abilities" command The error has been appeared on X722 attaching an SFP module without link up.
1.37	07-Feb-2020	yamaguchi	Use if_link_state_change() even if ixl(4) is attaching
1.36	04-Feb-2020	thorpej	Use ifmedia_fini().
1.35	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.34	31-Jan-2020	yamaguchi	Revert r1.32 because some documents say that the device supports SFP+ SX/LX I referred the document #332191-024 "Intel Ethernet Controller X710/XXV710/XL710 Feature Support Matrix" pointed out by msaitoh@, thanks.
1.33	31-Jan-2020	yamaguchi	Fix missing kpreempt_disabe()
1.32	31-Jan-2020	yamaguchi	Remove unsupported media on ixl(4)
1.31	31-Jan-2020	yamaguchi	Add media and flow changes support for ixl(4)
1.30	31-Jan-2020	yamaguchi	Fix a wrong return code from ixl_get_phy_types()
1.29	30-Jan-2020	yamaguchi	Add SIOCSIFMTU support for ixl(4)
1.28	27-Jan-2020	yamaguchi	Add TX checksum offload supports for ixl(4)
1.27	27-Jan-2020	yamaguchi	Fix missing PCI_INTR_MPSAFE settings in ixl(4)
1.26	17-Jan-2020	yamaguchi	Print nvm version of ixl(4) when attaching reviewed by msaitoh
1.25	17-Jan-2020	yamaguchi	Use flags instead of bool that changes behavior related to adminq command of ixl(4)
1.24	17-Jan-2020	yamaguchi	Fix the wrong expression in ixl(4) Closes PR/54860.
1.23	16-Jan-2020	yamaguchi	Fix wrong usage of PCI_PRODUCT(), ixl(4)
1.22	16-Jan-2020	yamaguchi	Fix ixl(4) not to assign interrupts to cpu #0 if ncpu > queue pairs
1.21	16-Jan-2020	yamaguchi	Use def{flag,param} for parameters in ixl(4)
1.20	15-Jan-2020	yamaguchi	Fix typo (s/txr_opackets/txr_obytes/) pointed out by ryo@n.o, thanks.
1.19	09-Jan-2020	yamaguchi	Make some parametes in ixl(4) be able to change when loading for debugging - nomsix(boolean) - disable msix support - stats_interval(signed integer) - change interval for collecting statistic counters - nqps_limit(signed integer) - limitation for the number of queue pairs - {tx,rx}_ndescs(unsigned integer) - the number of discriptors in txqueue or rxqueue
1.18	09-Jan-2020	yamaguchi	ixl(4) supports in-chip statistic counters per VSI
1.17	09-Jan-2020	yamaguchi	ixl(4) supports in-chip statistic counters per port
1.16	08-Jan-2020	yamaguchi	branches: 1.16.2; Not stop kpreempt if unnecessary, ixl(4)
1.15	26-Dec-2019	yamaguchi	Fix lack of of sysctl_teardown() in ixl(4)
1.14	26-Dec-2019	yamaguchi	ixl(4) supports IFCAP_CSUM_*_Rx OKed by msaitoh@n.o.
1.13	26-Dec-2019	yamaguchi	Improve DDPRINTF for ixl(4)
1.12	20-Dec-2019	yamaguchi	ixl(4) supports ETHERCAP_VLAN_HWFILTER the feature is disable by default. reviewed by msaitoh and knakahara
1.11	20-Dec-2019	yamaguchi	ixl(4) supports ETHERCAP_VLAN_HWTAGGING It is enabled by defualt The features is realized by the following operations: - internal switch - use "0b00: Show VLAN, DEI and UP in descriptor" mode - TX - set VLAN tag and IL2TAG1 flag to each descriptor - RX - use 32byte descriptor to use SHOWIV and L2SEL flag included in RX queue configuration - VLAN tags are not stored in descriptor without these config. - get VLAN tags from L2TAG1 field included in RX descriptor and set them to mbuf reviewed by msaitoh and knakahara
1.10	20-Dec-2019	yamaguchi	enable other interrupt even if ixl(4) is down
1.9	20-Dec-2019	yamaguchi	Do IXL_AQ_OP_SET_VSI_PROMISC in ixl_ifflags_cb() for applying IFF_PROMISC
1.8	20-Dec-2019	yamaguchi	Fix ixl(4) to set ENETRESET when IFF_ALLMULTI is changed
1.7	20-Dec-2019	yamaguchi	Fix ixl_{add\|remove}_macvlan() to return errno instead of command status
1.6	20-Dec-2019	yamaguchi	Fix softint leak in ixl(4) when detaching
1.5	20-Dec-2019	yamaguchi	Use name of variable instead of that of data structure
1.4	11-Dec-2019	yamaguchi	Fix to detect link state down Link status event that is used to update link status sometimes is not notified when link down. So, use Get Link Status response that has the same information with the event. The handling of the event is needed to detect link state up because the response sometimes does not notify link up.
1.3	11-Dec-2019	yamaguchi	Fix build error related to ixl(4) on i386
1.2	11-Dec-2019	yamaguchi	Not use _LP64 but sizeof(bus_addr_t) for getting size of address It makes ixl(4) be able to work on ILP32 pointed out by thorpej@. thanks.
1.1	10-Dec-2019	yamaguchi	Ported driver for Intel Ethernet 700 series reviewed by msaitoh and knakahara
1.16.2.2	29-Feb-2020	ad	Sync with head.
1.16.2.1	17-Jan-2020	ad	Sync with head.
1.65.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.65.2.1	19-Mar-2020	martin	file if_ixl.c was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.75.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.76.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.88.4.2	18-Oct-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #417): sys/dev/pci/if_ixl.c: revision 1.94 sys/dev/pci/if_ixl.c: revision 1.95 ixl(4): added ETHERCAP_VLAN_MTU support PR kern/57652 ixl(4): limit receive frame size upto 9600 even if ETHERCAP_VLAN_MTU is enabled
1.88.4.1	14-Oct-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #409): sys/dev/pci/if_ixl.c: revision 1.90 sys/dev/pci/if_ixl.c: revision 1.91 sys/dev/pci/if_ixl.c: revision 1.92 sys/dev/pci/if_ixl.c: revision 1.93 sys/arch/evbarm/conf/GENERIC64: revision 1.214 ixl(4): skip getting link status if the last command is running ixl(4): use cv_broadcast to wakeup ioctl and workqueue context. Commands may be issued at the same time from the both context. ixl(4): update link status in workqueue ixl(4): Print device info on attach Without PCIVERBOSE option: ixl0 at pci1 dev 0 function 0: vendor 8086 product 1572 (rev. 0x01) With PCIVERBOSE option: ixl0 at pci1 dev 0 function 0: Intel XL710 SFP+ Ethernet (SFI) (rev. 0x01) OK yamaguchi@ aarch64/GENERIC64: Add ixl(4) Works just fine for little-endian on LX2K with UEFI firmware. Not tested for big-endian yet; attach fails on ROCKPro64 due to host controller problem. OK yamaguchi@
1.99.2.1	02-Aug-2025	perseant	Sync with HEAD
1.2	17-Jun-2022	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.1	10-Dec-2019	yamaguchi	branches: 1.1.6; Ported driver for Intel Ethernet 700 series reviewed by msaitoh and knakahara
1.1.6.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.1.6.1	10-Dec-2019	martin	file if_ixlreg.h was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.8	16-Mar-2022	yamaguchi	ixl: Added support for Intel ethernet X710-T*L
1.7	08-Sep-2020	yamaguchi	Added iavf(4) that is based on OpenBSD's iavf(4) implementation reviewed by msaitoh@n.o and knakahara@n.o
1.6	12-Feb-2020	yamaguchi	branches: 1.6.4; Use Set RSS key command and Set LUT command to enable RSS on X722
1.5	31-Jan-2020	yamaguchi	Add media and flow changes support for ixl(4)
1.4	17-Jan-2020	yamaguchi	Print nvm version of ixl(4) when attaching reviewed by msaitoh
1.3	26-Dec-2019	yamaguchi	branches: 1.3.2; ixl(4) supports IFCAP_CSUM_*_Rx OKed by msaitoh@n.o.
1.2	20-Dec-2019	yamaguchi	ixl(4) supports ETHERCAP_VLAN_HWTAGGING It is enabled by defualt The features is realized by the following operations: - internal switch - use "0b00: Show VLAN, DEI and UP in descriptor" mode - TX - set VLAN tag and IL2TAG1 flag to each descriptor - RX - use 32byte descriptor to use SHOWIV and L2SEL flag included in RX queue configuration - VLAN tags are not stored in descriptor without these config. - get VLAN tags from L2TAG1 field included in RX descriptor and set them to mbuf reviewed by msaitoh and knakahara
1.1	10-Dec-2019	yamaguchi	Ported driver for Intel Ethernet 700 series reviewed by msaitoh and knakahara
1.3.2.2	29-Feb-2020	ad	Sync with head.
1.3.2.1	17-Jan-2020	ad	Sync with head.
1.6.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.6.4.1	12-Feb-2020	martin	file if_ixlvar.h was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.57	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.56	20-Dec-2023	thorpej	branches: 1.56.2; Remove unnecessary <sys/malloc.h>.
1.55	02-Sep-2022	thorpej	Remove unnecessary inclusion of <net/netisr.h>.
1.54	21-Aug-2022	thorpej	jme_ifstart(): Replace "IFQ_DEQUEUE() -> IF_PREPEND() on failure" with "IFQ_POLL() -> IFQ_DEQUEUE() on success (and fatal-to-packet errors)".
1.53	21-Aug-2022	thorpej	Don't allow jme_encap() to modify the head-of-mbuf-chain pointer. Instead, act on fatal packet errors in jme_ifstart().
1.52	21-Aug-2022	thorpej	jme_encap(): Fix up some botched formatting, and don't use (m_head) all over the place (just assign m = m_head earlier). NFCI.
1.51	16-Mar-2022	andvar	s/watchog/watchdog in comment and log message, one wording fix in comment.
1.50	08-May-2021	thorpej	Use pci_compatible_match().
1.49	08-Feb-2020	maxv	branches: 1.49.10; localify
1.48	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.47	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.46	23-Sep-2019	maxv	branches: 1.46.2; A * is missing here. This could cause a use-after-free. Found by the lgtm bot.
1.45	07-Aug-2019	msaitoh	Avoid undefined behavior in jme_mii_write(). Found by kUBSan.
1.44	09-Jul-2019	msaitoh	branches: 1.44.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.43	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.42	28-May-2019	msaitoh	Fix typos. From OpenBSD r1.52.
1.41	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.40	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.39	22-Apr-2019	msaitoh	Remove extra SIOCSIFMEDIA check.
1.38	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.37	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.36	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.35	26-Jun-2018	msaitoh	branches: 1.35.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.34	22-Jun-2018	msaitoh	It's not required to include net/bpfdesc.h. Remove it.
1.33	01-Jun-2018	maxv	Rename M_CSUM_DATA_IPv6_HL -> M_CSUM_DATA_IPv6_IPHL M_CSUM_DATA_IPv6_HL_SET -> M_CSUM_DATA_IPv6_SET Reduces the diff against IPv4. Also, clarify the definitions.
1.32	26-Sep-2017	knakahara	branches: 1.32.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.31	15-Dec-2016	ozaki-r	branches: 1.31.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.30	10-Jun-2016	ozaki-r	branches: 1.30.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.29	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.28	12-Sep-2015	christos	missing splx() found by brainy
1.27	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.26	10-Aug-2014	tls	branches: 1.26.4; Merge tls-earlyentropy branch into HEAD.
1.25	29-Mar-2014	christos	branches: 1.25.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.24	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.23	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.22	30-Mar-2013	christos	branches: 1.22.4; remove trailing whitespace
1.21	22-Jul-2012	matt	branches: 1.21.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.20	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.19	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.18	19-Nov-2011	tls	branches: 1.18.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.17	30-Mar-2011	bouyer	branches: 1.17.4; Some bus_dma(9)-related and memory allocation fixes: - always bus_dmamap_unload() before recycling a receive descriptor - make sure to not sleep from interrupt context: call bus_dmamap_load with M_NOWAIT, and create dma maps BUS_DMA_ALLOCNOW. - if a receive descriptor has a NULL mbuf try to allocate a new one, don't try to receive it.
1.16	09-Jan-2011	kochi	branches: 1.16.2; Try reading MAC addr from register if it fails to read from EEPROM. Copied from FreeBSD driver. Without this my JMC261 doesn't get MAC address properly. OK'ed by bouyer@
1.15	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.14	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.13	19-Jan-2010	pooka	branches: 1.13.2; 1.13.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.12	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.11	27-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Compile test only.
1.10	26-Jun-2009	dyoung	A lot of good it does, but let this build with 'no options INET'.
1.9	18-Apr-2009	tsutsui	branches: 1.9.2; Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.8	18-Mar-2009	cegger	bcopy -> memcpy
1.7	18-Mar-2009	cegger	bzero -> memset
1.6	31-Jan-2009	bouyer	branches: 1.6.2; Under some circunstance, the JMC250A2 (and maybe others) can enter a state where INTA is asserted while no enabled interrupts are pending, which cause a continous strean if interrupts with no way to clear it. Clearing/resetting the enabled interrupt mask in the interrupt handler seems to workaround this issue (this is what the FreeBSD driver does). While there properly return 0 if the interrupt was not for us, and explicitely disable JME_TMCSR in addition to TIMER1 and TIMER2.
1.5	16-Jan-2009	bouyer	branches: 1.5.2; According to the latest datasheet I have (1.09), the transmitter and ckecksum offload clocks needs to be programmed according to the negociated speed; make it so. This makes the sample JME260 I have work.
1.4	16-Oct-2008	abs	branches: 1.4.2; 1.4.4; 1.4.6; Fix compilation if !INET6
1.3	13-Oct-2008	bouyer	Add support for jumbo frames. Some performances tweaks (including better defaults for interrupt mitigation)
1.2	12-Oct-2008	bouyer	Fix reading MAC address from eeprom: don't forget to read the last descriptor. While there, white space fixes.
1.1	11-Oct-2008	bouyer	jme(4), a driver for JMicron Technologies JME250 Gigabit Ethernet and JME260 Fast Ethernet PCI Express controllers. Written with a lot of cut-n-paste from the FreeBSD jme(4) driver. No support for jumbo ethernet frames yet (but should come soon). Thanks to JMicron Technologies for providing me sample boards and documentation for this work.
1.4.6.4	05-Apr-2011	riz	Pull up following revision(s) (requested by bouyer in ticket #1588): sys/dev/pci/if_jme.c: revision 1.17 Some bus_dma(9)-related and memory allocation fixes: - always bus_dmamap_unload() before recycling a receive descriptor - make sure to not sleep from interrupt context: call bus_dmamap_load with M_NOWAIT, and create dma maps BUS_DMA_ALLOCNOW. - if a receive descriptor has a NULL mbuf try to allocate a new one, don't try to receive it.
1.4.6.3	16-Jan-2011	bouyer	Pull up following revision(s) (requested by kochi in ticket #1525): sys/dev/pci/if_jme.c: revision 1.16 Try reading MAC addr from register if it fails to read from EEPROM. Copied from FreeBSD driver. Without this my JMC261 doesn't get MAC address properly. OK'ed by bouyer@
1.4.6.2	02-Feb-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #388): sys/dev/pci/if_jme.c: revision 1.6 Under some circunstance, the JMC250A2 (and maybe others) can enter a state where INTA is asserted while no enabled interrupts are pending, which cause a continous strean if interrupts with no way to clear it. Clearing/resetting the enabled interrupt mask in the interrupt handler seems to workaround this issue (this is what the FreeBSD driver does). While there properly return 0 if the interrupt was not for us, and explicitely disable JME_TMCSR in addition to TIMER1 and TIMER2.
1.4.6.1	17-Jan-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #278): sys/dev/pci/if_jme.c: revision 1.5 sys/dev/pci/if_jmereg.h: revision 1.3 According to the latest datasheet I have (1.09), the transmitter and ckecksum offload clocks needs to be programmed according to the negociated speed; make it so. This makes the sample JME260 I have work.
1.4.4.3	28-Apr-2009	skrll	Sync with HEAD.
1.4.4.2	03-Mar-2009	skrll	Sync with HEAD.
1.4.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.4.2.2	19-Oct-2008	haad	Sync with HEAD.
1.4.2.1	16-Oct-2008	haad	file if_jme.c was added on branch haad-dm on 2008-10-19 22:16:39 +0000
1.5.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.5.2.1	16-Jan-2009	mjf	file if_jme.c was added on branch mjf-devfs2 on 2009-01-17 13:29:00 +0000
1.6.2.2	23-Jul-2009	jym	Sync with HEAD.
1.6.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.9.2.5	11-Aug-2010	yamt	sync with head.
1.9.2.4	11-Mar-2010	yamt	sync with head
1.9.2.3	18-Jul-2009	yamt	sync with head.
1.9.2.2	04-May-2009	yamt	sync with head.
1.9.2.1	18-Apr-2009	yamt	file if_jme.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:57 +0000
1.13.4.3	21-Apr-2011	rmind	sync with head
1.13.4.2	05-Mar-2011	rmind	sync with head
1.13.4.1	30-May-2010	rmind	sync with head
1.13.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.16.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.17.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.17.4.2	30-Oct-2012	yamt	sync with head
1.17.4.1	17-Apr-2012	yamt	sync with head
1.18.2.1	18-Feb-2012	mrg	merge to -current.
1.21.2.3	03-Dec-2017	jdolecek	update from HEAD
1.21.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.21.2.1	23-Jun-2013	tls	resync from head
1.22.4.1	18-May-2014	rmind	sync with head
1.25.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.26.4.5	05-Feb-2017	skrll	Sync with HEAD
1.26.4.4	09-Jul-2016	skrll	Sync with HEAD
1.26.4.3	19-Mar-2016	skrll	Sync with HEAD
1.26.4.2	22-Sep-2015	skrll	Sync with HEAD
1.26.4.1	06-Jun-2015	skrll	Sync with HEAD
1.30.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.31.8.3	10-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1430): sys/dev/pci/if_jme.c: revision 1.46 A * is missing here. This could cause a use-after-free. Found by the lgtm bot.
1.31.8.2	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.31.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.32.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.32.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.32.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.32.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.35.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.35.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.35.2.1	10-Jun-2019	christos	Sync with HEAD
1.44.2.2	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #672): sys/dev/ic/tulip.c: revision 1.198 sys/dev/pci/if_jme.c: revision 1.45 sys/dev/pci/agp.c: revision 1.86 sys/dev/pci/if_lii.c: revision 1.27 sys/dev/acpi/thinkpad_acpi.c: revision 1.47 sys/dev/scsipi/scsipi_base.c: revision 1.183 sys/dev/ic/aic6915reg.h: revision 1.6 Fix undefined behavior in thinkpad_mask_init(). Found by kUBSan. Use unsigned when initialize the transmit completion ring to avoid undefined behavior. Found by kUBSan. Avoid undefined behavior when setting multicast address. found by kUBSan. Use unsigned to avoid undefined behavior in agpattach(). Found by kUBSan. Avoid undefined behavior in jme_mii_write(). Found by kUBSan. Use unsigned to avoid undefined behavior in lii_setmulti(). Use unsigned to avoid undefined behavior in scsipi_{get,put}_tag(). Found by kUBSan.
1.44.2.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #376): sys/dev/pci/if_jme.c: revision 1.46 A * is missing here. This could cause a use-after-free. Found by the lgtm bot.
1.46.2.1	29-Feb-2020	ad	Sync with head.
1.49.10.1	13-May-2021	thorpej	Sync with HEAD.
1.56.2.1	02-Aug-2025	perseant	Sync with HEAD
1.8	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.7	05-Dec-2021	msaitoh	as/aggresive/aggressive/ in comment.
1.6	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.5	14-Nov-2019	msaitoh	Revert previous. Sorry.
1.4	14-Nov-2019	msaitoh	Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.3	16-Jan-2009	bouyer	branches: 1.3.2; 1.3.6; According to the latest datasheet I have (1.09), the transmitter and ckecksum offload clocks needs to be programmed according to the negociated speed; make it so. This makes the sample JME260 I have work.
1.2	13-Oct-2008	bouyer	branches: 1.2.2; 1.2.4; 1.2.6; Add support for jumbo frames. Some performances tweaks (including better defaults for interrupt mitigation)
1.1	11-Oct-2008	bouyer	jme(4), a driver for JMicron Technologies JME250 Gigabit Ethernet and JME260 Fast Ethernet PCI Express controllers. Written with a lot of cut-n-paste from the FreeBSD jme(4) driver. No support for jumbo ethernet frames yet (but should come soon). Thanks to JMicron Technologies for providing me sample boards and documentation for this work.
1.2.6.1	17-Jan-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #278): sys/dev/pci/if_jme.c: revision 1.5 sys/dev/pci/if_jmereg.h: revision 1.3 According to the latest datasheet I have (1.09), the transmitter and ckecksum offload clocks needs to be programmed according to the negociated speed; make it so. This makes the sample JME260 I have work.
1.2.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.2.2.2	19-Oct-2008	haad	Sync with HEAD.
1.2.2.1	13-Oct-2008	haad	file if_jmereg.h was added on branch haad-dm on 2008-10-19 22:16:39 +0000
1.3.6.2	04-May-2009	yamt	sync with head.
1.3.6.1	16-Jan-2009	yamt	file if_jmereg.h was added on branch yamt-nfs-mp on 2009-05-04 08:12:57 +0000
1.3.2.2	17-Jan-2009	mjf	Sync with HEAD.
1.3.2.1	16-Jan-2009	mjf	file if_jmereg.h was added on branch mjf-devfs2 on 2009-01-17 13:29:00 +0000
1.59	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.58	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.57	08-May-2021	thorpej	Use pci_compatible_match().
1.56	20-Sep-2020	nisimura	branches: 1.56.6; - rectify nonsense and bugs around statistics counter managment. - scoop up counter values in one second interval comsolidating with phy_tick. - reformat register definitions for ease to see.
1.55	20-Sep-2020	nisimura	guard statistics gathering callout with #ifdef
1.54	20-Sep-2020	nisimura	be more cautious about error exit path. nuke unused variables
1.53	01-Apr-2020	nisimura	insert missing ETHER_UNLOCK()
1.52	01-Apr-2020	nisimura	switch to use ETHER_F_ALLMULTI flag
1.51	24-Mar-2020	nisimura	nuke <sys/queue.h> and reorder #include
1.50	02-Mar-2020	nisimura	minor knits in mcast handling.
1.49	17-Feb-2020	nisimura	comment mods and code tiddy ups
1.48	04-Feb-2020	skrll	Adopt <net/if_stats.h>
1.47	06-Jan-2020	nisimura	branches: 1.47.2; correct bogus comparison. pointed by msaitoh@
1.46	14-Dec-2019	nisimura	use FC enable bit for 8842 host port
1.45	12-Dec-2019	nisimura	use additional MAC addresses to accept mcast
1.44	03-Dec-2019	nisimura	streamline receive filter logic, work-in-progress.
1.43	29-Nov-2019	nisimura	switch to deferred if_start framework.
1.42	26-Nov-2019	nisimura	- use mii(4) layer to control KSZ8841 builtin PHY. - handle PAUSE flow control properly according to ifconfig(8) mediaopt selection. - some style knits; use aprint(9) and modernise callout(9).
1.41	07-Nov-2019	nisimura	comment touchup
1.40	07-Nov-2019	nisimura	clarify 8842 MAC behaves 100FDX only has no alternative media selection possible.
1.39	06-Nov-2019	nisimura	major rework to fix link control breakage
1.38	29-May-2019	msaitoh	branches: 1.38.2; Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.37	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.36	23-May-2019	msaitoh	-No functional change: - Simplify struct ethercom's pointer near ETHER_FIRST_MULTI(). - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.35	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.34	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.33	18-Jul-2018	sevan	Be consistent among ethernet drivers on the convention for printing ethernet addresses. NFC
1.32	26-Jun-2018	msaitoh	branches: 1.32.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.31	15-Dec-2016	ozaki-r	branches: 1.31.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.30	10-Jun-2016	ozaki-r	branches: 1.30.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.29	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.28	16-Jun-2014	msaitoh	branches: 1.28.4; IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.27	29-Mar-2014	christos	branches: 1.27.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.26	25-Mar-2014	christos	kill sprintf
1.25	08-Nov-2013	nisimura	move EVCNT only variable inside #ifdef
1.24	30-Mar-2013	christos	branches: 1.24.4; remove trailing whitespace
1.23	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.22	05-Apr-2010	joerg	branches: 1.22.8; 1.22.14; 1.22.18; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.21	19-Jan-2010	pooka	branches: 1.21.2; 1.21.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.20	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.19	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.18	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.17	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.16	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.15	26-May-2008	nisimura	branches: 1.15.6; 1.15.12; - assign 2 clause TNF license to the files cited as my ownership. - abandon and remove my copyright notice from the three files were copied in order to adapt HW I don't have; pvr.c, pm.c and stic.c
1.14	10-Apr-2008	cegger	branches: 1.14.2; 1.14.4; 1.14.6; use aprint_*_dev and device_xname
1.13	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.12	07-Feb-2008	dyoung	branches: 1.12.2; 1.12.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.11	16-Nov-2007	nisimura	reduce Tx/Rx maximum number of words (as PDF says) in one DMA transaction to 8. value 32 triggers occational watchdog() Tx timeout when higher system load. This symptom is observed in ipforwarding across two PCI devices case so far, and it remains unidentified what really happens for Tx DMA activity. 16 seems ok, 8 is conservative and heuristic value. may need more adjustment work in other parts.
1.10	19-Oct-2007	ad	branches: 1.10.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.9	19-Oct-2007	nisimura	- fix an error when retriving two 16bit halves of 29bit stat value. - port3 counters are found exist and in fact working as 0x102/0x105.
1.8	19-Oct-2007	nisimura	- add multicast Rx hash table feature. - distinguish 8842 from 8841. 8842 now keeps media selection "auto" and indicates "up 100baseTX-FDX flow" when either of two ports has a valid link. There is no provision to see and control the two this moment and their media selections remain in "auto" all the time. This arrangement is considered acceptable since 8842's external ports are connected with the internal EMAC via managed 3 port Ethernet switch. - 8841 behaves a plain stanadrd 10/100 EMAC with standard media selection feature. - gather MIB statistics counter values with evcnt(8) framework. - increase Tx/Rx DMA DMA burst transfer size from 16 to 32.
1.7	14-Oct-2007	nisimura	branches: 1.7.2; fix typos about a pair of multicast hash table registers.
1.6	14-Oct-2007	nisimura	add multicast filter support; either to use HW hash table lookup filter or to accept any mcast frames.
1.5	07-Oct-2007	nisimura	A typo fix in one of comments and some adjustments of register field names. No functionality change intented. IHAE "IP Header Alignment Enable" feature of RXC register; - careful cross referencing at KSZ8692P/8841P/8842P PDFs indicates IHEA bit works as follows; When the feature is turned on, Rx DMA engine will populate Rx frame data in the Rx memory with +2 or +3 byte swifted to make its IP head field 32bit aligned. The shift amount is recorded inside RDES0 to tell 0, 2 or 3. The automatic alignment is done only when IHAE is enabled _and_ the Rx frame was IP frame. In other cases, RDES0 swift amount field keeps 0. - KSZ8841P document mentions the IHAE bit but its reference link is broken to tell the new RDES0 field. KSZ8842P lacks both. The bit usage of RDES0 23:20 seems different from KSZ8692P, which brings me a vague suspiction of documentation error.
1.4	09-Jul-2007	ad	branches: 1.4.6; 1.4.8; 1.4.10; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.3	04-Mar-2007	christos	branches: 1.3.2; 1.3.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.2	29-Sep-2006	tsutsui	branches: 1.2.4; 1.2.6; - trim CRC off rather than setting M_HASFCS - use appropriate types instead of `unsigned'
1.1	30-Jun-2006	nisimura	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.10; 1.1.12; Add kse(4) driver for Micrel KSZ8842/8841 Ethernet controller.
1.1.12.1	22-Oct-2006	yamt	sync with head
1.1.10.2	09-Sep-2006	rpaulo	sync with head
1.1.10.1	30-Jun-2006	rpaulo	file if_kse.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:17 +0000
1.1.8.1	18-Nov-2006	ad	Sync with head.
1.1.6.2	11-Aug-2006	yamt	sync with head
1.1.6.1	30-Jun-2006	yamt	file if_kse.c was added on branch yamt-pdpolicy on 2006-08-11 15:44:25 +0000
1.1.2.2	13-Jul-2006	gdamore	Merge from HEAD.
1.1.2.1	30-Jun-2006	gdamore	file if_kse.c was added on branch gdamore-uart on 2006-07-13 17:49:27 +0000
1.2.6.1	12-Mar-2007	rmind	Sync with HEAD.
1.2.4.7	17-Mar-2008	yamt	sync with head.
1.2.4.6	11-Feb-2008	yamt	sync with head.
1.2.4.5	07-Dec-2007	yamt	sync with head
1.2.4.4	27-Oct-2007	yamt	sync with head.
1.2.4.3	03-Sep-2007	yamt	sync with head.
1.2.4.2	30-Dec-2006	yamt	sync with head.
1.2.4.1	29-Sep-2006	yamt	file if_kse.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:44 +0000
1.3.4.1	11-Jul-2007	mjf	Sync with head.
1.3.2.3	23-Oct-2007	ad	Sync with head.
1.3.2.2	09-Oct-2007	ad	Sync with head.
1.3.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.4.10.2	18-Oct-2007	yamt	sync with head.
1.4.10.1	14-Oct-2007	yamt	sync with head.
1.4.8.3	23-Mar-2008	matt	sync with HEAD
1.4.8.2	09-Jan-2008	matt	sync with HEAD
1.4.8.1	06-Nov-2007	matt	sync with HEAD
1.4.6.3	21-Nov-2007	joerg	Sync with HEAD.
1.4.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.4.6.1	07-Oct-2007	joerg	Sync with HEAD.
1.7.2.2	18-Nov-2007	bouyer	Sync with HEAD
1.7.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.10.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.10.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.12.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.2.1	24-Mar-2008	keiichi	sync with head.
1.14.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.14.4.4	11-Aug-2010	yamt	sync with head.
1.14.4.3	11-Mar-2010	yamt	sync with head
1.14.4.2	16-May-2009	yamt	sync with head
1.14.4.1	04-May-2009	yamt	sync with head.
1.14.2.1	04-Jun-2008	yamt	sync with head
1.15.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.15.6.1	28-Apr-2009	skrll	Sync with HEAD.
1.21.4.1	30-May-2010	rmind	sync with head
1.21.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.22.18.4	03-Dec-2017	jdolecek	update from HEAD
1.22.18.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.22.18.2	23-Jun-2013	tls	resync from head
1.22.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.22.14.1	09-Nov-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #1189): sys/dev/pci/if_wm.c: revision 1.270 sys/dev/mii/inphy.c: revision 1.53 sys/dev/mii/glxtphy.c: revision 1.24 sys/dev/mii/tlphy.c: revision 1.62 sys/dev/mii/iophy.c: revision 1.37 sys/dev/mii/brgphy.c: revision 1.70 sys/dev/mii/ihphy.c: revision 1.8 sys/dev/mii/bmtphy.c: revision 1.31 sys/dev/mii/urlphy.c: revision 1.30 sys/dev/mii/makphy.c: revision 1.40 sys/dev/mii/qsphy.c: revision 1.48 sys/dev/mii/igphy.c: revision 1.23 sys/dev/mii/nsphy.c: revision 1.58 sys/dev/mii/mvphy.c: revision 1.10 sys/dev/pci/if_txp.c: revision 1.41 sys/dev/mii/nsphy.c: revision 1.59 sys/dev/mii/rlphy.c: revision 1.28 sys/dev/mii/icsphy.c: revision 1.49 sys/dev/mii/rlphy.c: revision 1.29 sys/dev/mii/lxtphy.c: revision 1.49 sys/dev/mii/ciphyreg.h: revision 1.5 sys/dev/mii/nsphyter.c: revision 1.38 sys/dev/mii/sqphy.c: revision 1.50 sys/dev/mii/gentbi.c: revision 1.26 sys/dev/mii/gentbi.c: revision 1.27 sys/dev/mii/tqphy.c: revision 1.39 sys/dev/mii/ikphy.c: revision 1.10 sys/dev/mii/dmphy.c: revision 1.35 sys/dev/mii/amhphy.c: revision 1.20 sys/dev/mii/acphy.c: revision 1.24 sys/dev/mii/ciphy.c: revision 1.25 sys/dev/mii/brgphyreg.h: revision 1.8 sys/dev/mii/ukphy_subr.c: revision 1.12 sys/dev/ic/rtl80x9.c: revision 1.16 sys/arch/mips/adm5120/dev/if_admsw.c: revision 1.12 sys/dev/pci/if_kse.c: revision 1.28 sys/dev/mii/ukphy_subr.c: revision 1.13 sys/dev/mii/mii.h: revision 1.18 sys/dev/mii/gphyter.c: revision 1.29 No functional change: - Fix typo. - Remove trailing white spaces. - Capitalize comments. - Tabify. - KNF. IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.22.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.22.8.1	30-Oct-2012	yamt	sync with head
1.24.4.1	18-May-2014	rmind	sync with head
1.27.2.1	10-Aug-2014	tls	Rebase.
1.28.4.3	05-Feb-2017	skrll	Sync with HEAD
1.28.4.2	09-Jul-2016	skrll	Sync with HEAD
1.28.4.1	19-Mar-2016	skrll	Sync with HEAD
1.30.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.31.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.31.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.32.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.32.2.1	10-Jun-2019	christos	Sync with HEAD
1.38.2.1	10-Nov-2019	martin	Pull up following revision(s) (requested by nisimura in ticket #406): sys/dev/pci/if_kse.c: revision 1.40 sys/dev/pci/if_kse.c: revision 1.41 sys/dev/pci/if_kse.c: revision 1.39 comment touchup - clarify 8842 MAC behaves 100FDX only has no alternative media selection possible. - major rework to fix link control breakage
1.47.2.1	29-Feb-2020	ad	Sync with head.
1.56.6.1	13-May-2021	thorpej	Sync with HEAD.
1.55	25-Jan-2020	thorpej	Retire the le@pci attachment. It has been superseded (and matched at a higher priority) by the pcn(4) driver since NetBSD 1.6.
1.54	25-Apr-2019	msaitoh	branches: 1.54.4; No functional change: - Use __arraycount(). - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.53	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.52	29-Mar-2014	christos	branches: 1.52.28; 1.52.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.51	26-Jul-2011	dyoung	branches: 1.51.2; 1.51.12; 1.51.16; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.50	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.49	28-Apr-2008	martin	branches: 1.49.22; Remove clause 3 and 4 from TNF licenses
1.48	04-Apr-2008	tsutsui	branches: 1.48.2; 1.48.4; Split device_t/softc for le(4) and variants and misc cosmetic changes.
1.47	19-Oct-2007	ad	branches: 1.47.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.46	04-Mar-2007	christos	branches: 1.46.2; 1.46.14; 1.46.16; 1.46.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.45	16-Nov-2006	christos	branches: 1.45.4; __unused removal on arguments; approved by core.
1.44	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.43	11-Dec-2005	christos	branches: 1.43.20; 1.43.22; merge ktrace-lwp.
1.42	27-Feb-2005	perry	branches: 1.42.4; nuke trailing whitespace
1.41	21-Aug-2004	thorpej	branches: 1.41.4; 1.41.6; Use ANSI function decls and make use of static.
1.40	07-Aug-2003	agc	Move UCB-licensed code from 4-clause to 3-clause licence. Patches provided by Joel Baker in PR 22364, verified by myself.
1.39	02-Oct-2002	thorpej	branches: 1.39.6; Add trailing ; to CFATTACH_DECL.
1.38	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.37	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.36	21-Nov-2001	wiz	Explicitly write explicitly without a second e.
1.35	13-Nov-2001	lukem	add RCSID
1.34	08-Jul-2001	thorpej	branches: 1.34.2; Remove unnecessary include files. From Onno van der Linden.
1.33	18-Jun-2001	jdolecek	Convert to use ether_ioctl() for SIOCSIFADDR and SIOCSIFFLAGS.
1.32	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.31	30-May-2001	mrg	use _KERNEL_OPT
1.30	28-Dec-2000	sommerfeld	branches: 1.30.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.29	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.28	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.27	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.26	02-Oct-1998	fvdl	branches: 1.26.12; Add missing field to le_softc structure, missed in previous commit. Name it consistently with the other fields.
1.25	01-Oct-1998	drochner	support media selection / full duplex
1.24	15-Aug-1998	mycroft	Assign my copyrights to TNF.
1.23	21-Jul-1998	drochner	adapt to LANCE driver split; let PCNET/PCI use the 32-bit mode
1.22	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.21	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.20	04-Feb-1998	thorpej	Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC to BUS_DMA_COHERENT.
1.19	06-Jun-1997	thorpej	Pull thorpej-bus-dma branch into mainline.
1.18	13-Apr-1997	cgd	branches: 1.18.2; rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.17	13-Apr-1997	cgd	use pci_map_register().
1.16	17-Mar-1997	thorpej	#include <net/if_media.h>
1.15	15-Mar-1997	is	New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.14	05-Dec-1996	cgd	branches: 1.14.4; update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.13	25-Oct-1996	cgd	update alpha_XXX_dmamap() invocations to match prototype.
1.12	23-Oct-1996	cgd	update drivers to use the new NetBSD/alpha DMA mapping hack, since the old one is not workable with the new bus.h.
1.11	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.10	13-Oct-1996	christos	backout kprintf changes
1.9	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.8	15-Jul-1996	cgd	before #defining the vtophys() hack for the alpha, #undef vtophys since it may be a macro.
1.7	03-Jun-1996	cgd	convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make the code more correct.
1.6	14-May-1996	thorpej	branches: 1.6.4; Don't include <machine/intr.h> twice on the Alpha.
1.5	13-May-1996	mycroft	Use intr.h.
1.4	12-May-1996	thorpej	Squish a straggling inb(), and tweak so that this compiles on the Alpha.
1.3	12-May-1996	thorpej	Fix a think-o (my fault) that caused LANCE register offsets to be wrong. While I'm here, convert driver to use <machine/bus.h> and the new PCI I/O space interface (the new PCI interrupt interface was already in-use). Also, correct a "corrected" printf.
1.2	07-May-1996	christos	- Add an XXX declaration to kvtop - Add a missing argument to printf
1.1	07-May-1996	thorpej	Make the MI LANCE driver standalone, and use cfattach to resolve naming conflicts between bus attachments on ports that can have multiple instances of the LANCE. Add a real PCI front-end for PCnet-PCI Ethernet cards. Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.6.4.1	03-Jun-1996	cgd	pull up from trunk: >convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to >get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make >the code more correct.
1.14.4.2	09-Mar-1997	is	netinet/if_ether.h -> netinet/if_inarp.h
1.14.4.1	21-Feb-1997	is	Transform to the new world order.
1.18.2.2	17-May-1997	thorpej	Update for bus_dmamem_alloc() changes.
1.18.2.1	13-May-1997	thorpej	Convert to use bus dma.
1.26.12.3	05-Jan-2001	bouyer	Sync with HEAD
1.26.12.2	22-Nov-2000	bouyer	Sync with HEAD.
1.26.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.30.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.30.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.30.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.30.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.30.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.34.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.34.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.39.6.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.39.6.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.39.6.3	18-Sep-2004	skrll	Sync with HEAD.
1.39.6.2	25-Aug-2004	skrll	Sync with HEAD.
1.39.6.1	03-Aug-2004	skrll	Sync with HEAD
1.41.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.41.4.1	29-Apr-2005	kent	sync with -current
1.42.4.3	27-Oct-2007	yamt	sync with head.
1.42.4.2	03-Sep-2007	yamt	sync with head.
1.42.4.1	30-Dec-2006	yamt	sync with head.
1.43.22.2	10-Dec-2006	yamt	sync with head.
1.43.22.1	22-Oct-2006	yamt	sync with head
1.43.20.1	18-Nov-2006	ad	Sync with head.
1.45.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.46.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.46.16.1	06-Nov-2007	matt	sync with HEAD
1.46.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.46.2.1	23-Oct-2007	ad	Sync with head.
1.47.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.48.4.1	16-May-2008	yamt	sync with head.
1.48.2.1	18-May-2008	yamt	sync with head.
1.49.22.1	05-Mar-2011	rmind	sync with head
1.51.16.1	18-May-2014	rmind	sync with head
1.51.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.51.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.52.30.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.52.30.1	10-Jun-2019	christos	Sync with HEAD
1.52.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.54.4.1	25-Jan-2020	ad	Sync with head.
1.10	25-Jan-2020	thorpej	Retire the le@pci attachment. It has been superseded (and matched at a higher priority) by the pcn(4) driver since NetBSD 1.6.
1.9	28-Apr-2008	martin	branches: 1.9.88; 1.9.96; Remove clause 3 and 4 from TNF licenses
1.8	02-Oct-1998	fvdl	branches: 1.8.144; 1.8.146; 1.8.148; Add missing field to le_softc structure, missed in previous commit. Name it consistently with the other fields.
1.7	15-Aug-1998	mycroft	Assign my copyrights to TNF.
1.6	21-Jul-1998	drochner	adapt to LANCE driver split; let PCNET/PCI use the 32-bit mode
1.5	06-Jun-1997	thorpej	Pull thorpej-bus-dma branch into mainline.
1.4	15-Mar-1997	is	branches: 1.4.2; New ARP system, supports IPv4 over any hardware link. Some of the stuff (e.g., rarpd, bootpd, dhcpd etc., libsa) still will only support Ethernet. Tcpdump itself should be ok, but libpcap needs lot of work. For the detailed change history, look at the commit log entries for the is-newarp branch.
1.3	21-Oct-1996	thorpej	branches: 1.3.4; New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.2	12-May-1996	thorpej	Fix a think-o (my fault) that caused LANCE register offsets to be wrong. While I'm here, convert driver to use <machine/bus.h> and the new PCI I/O space interface (the new PCI interrupt interface was already in-use). Also, correct a "corrected" printf.
1.1	07-May-1996	thorpej	Make the MI LANCE driver standalone, and use cfattach to resolve naming conflicts between bus attachments on ports that can have multiple instances of the LANCE. Add a real PCI front-end for PCnet-PCI Ethernet cards. Changed struct ifnet to have a pointer to the softc of the underlying device and a printable "external name" (name + unit number), thus eliminating if_name and if_unit. Updated interface to (if_watchdog)() and (if_reset)() to take a struct ifnet *, rather than a unit number.
1.3.4.1	27-Feb-1997	is	Removed #include <route.h> and #include <netisr.h>. They're not needed, and the latter will go away soon.
1.4.2.1	13-May-1997	thorpej	Convert to use bus dma.
1.8.148.1	16-May-2008	yamt	sync with head.
1.8.146.1	18-May-2008	yamt	sync with head.
1.8.144.1	02-Jun-2008	mjf	Sync with HEAD.
1.9.96.1	25-Jan-2020	ad	Sync with head.
1.9.88.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.30	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.29	16-Sep-2022	andvar	s/Paket/Packet/
1.28	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.27	18-Sep-2019	msaitoh	branches: 1.27.2; Use unsigned to avoid undefined behavior in lii_setmulti().
1.26	30-May-2019	msaitoh	branches: 1.26.2; Simplify MII structure initialization and reference. No functional change.
1.25	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.24	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.23	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.22	22-Apr-2019	msaitoh	On drivers which use MII(4) and have hook SIOC[GS]IFMEDIA which just pass to ifmedia_ioctl(), the hook is not required because ether_ioctl has it (if_ethersubr.c rev. 1.160). These drivers don't return ENETRESET in ifmedia_ioctl(), so no functional change.
1.21	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.20	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.19	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.18	26-Jun-2018	msaitoh	branches: 1.18.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.17	15-Dec-2016	ozaki-r	branches: 1.17.8; 1.17.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.16	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.15	10-Jun-2016	ozaki-r	branches: 1.15.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.14	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.13	29-Mar-2014	christos	branches: 1.13.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.12	30-Mar-2013	christos	branches: 1.12.4; remove trailing whitespace
1.11	22-Jul-2012	matt	branches: 1.11.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.10	20-Apr-2011	christos	branches: 1.10.4; PR/41258: Elias Benali: Try to get the ethernet address from the BIOS before failing.
1.9	05-Apr-2010	joerg	branches: 1.9.2; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.8	19-Jan-2010	pooka	branches: 1.8.2; 1.8.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.7	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.6	26-Apr-2009	cegger	fix error handling in lii_attach() ok jmcneill@
1.5	08-Jul-2008	sborrill	branches: 1.5.2; 1.5.4; 1.5.8; Fix multicast (multicast hash table were calculated incorrectly).
1.4	29-Apr-2008	martin	branches: 1.4.2; 1.4.4; 1.4.6; 1.4.8; Convert to new 2 clause license
1.3	06-Apr-2008	mjf	branches: 1.3.2; 1.3.4; 1.3.6; Prevent packet header from being placed outside the TXD buffer. This should stop the "mismatched status and packet" error messages that have been plaguing the EeePC. Fix provided by Joel Sing, thanks!
1.2	31-Mar-2008	mjf	branches: 1.2.2; Register pmf handler.
1.1	29-Mar-2008	cube	Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC. The linux driver was used as documentation. Many thanks to all the people who helped or at least supported me while I was fighting that bug that made the EeePC freeze. At the top of that list, Matt J. Fleming (mjf@) and David Gwynne (from the OpenBSD fame).
1.2.2.4	28-Sep-2008	mjf	Sync with HEAD.
1.2.2.3	02-Jun-2008	mjf	Sync with HEAD.
1.2.2.2	03-Apr-2008	mjf	Sync with HEAD.
1.2.2.1	31-Mar-2008	mjf	file if_lii.c was added on branch mjf-devfs2 on 2008-04-03 12:42:50 +0000
1.3.6.5	11-Aug-2010	yamt	sync with head.
1.3.6.4	11-Mar-2010	yamt	sync with head
1.3.6.3	16-Sep-2009	yamt	sync with head
1.3.6.2	04-May-2009	yamt	sync with head.
1.3.6.1	16-May-2008	yamt	sync with head.
1.3.4.3	20-Aug-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1164): sys/dev/pci/if_lii.c: revision 1.5 Fix multicast (multicast hash table were calculated incorrectly).
1.3.4.2	19-Apr-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1128): sys/dev/pci/if_lii.c: revisions 1.1, 1.3 via patch sys/arch/i386/conf/ALL: revision 1.161 via patch share/man/man4/lii.4: revision 1.1 via patch sys/dev/pci/files.pci: revision 1.303 via patch sys/dev/DEVNAMES: revision 1.238 via patch sys/arch/i386/conf/GENERIC: revision 1.885 via patch distrib/sets/lists/man/mi: revisions 1.1062, 1.1063 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.43 via patch share/man/man4/Makefile: revision 1.460 via patch sys/arch/i386/conf/INSTALL: revision 1.323 via patch sys/dev/pci/if_liireg.h: revision 1.1 via patch Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC.
1.3.4.1	06-Apr-2008	bouyer	file if_lii.c was added on branch netbsd-4 on 2008-04-19 15:57:56 +0000
1.3.2.1	18-May-2008	yamt	sync with head.
1.4.8.1	19-Oct-2008	haad	Sync with HEAD.
1.4.6.1	18-Jul-2008	simonb	Sync with head.
1.4.4.3	04-Sep-2008	skrll	Sync with netbsd-4.
1.4.4.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.4.4.1	29-Apr-2008	skrll	file if_lii.c was added on branch wrstuden-fixsa on 2008-06-03 20:47:24 +0000
1.4.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.5.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.5.4.1	20-May-2011	bouyer	Pull up following revision(s) (requested by dholland in ticket #1612): sys/dev/pci/if_lii.c: revision 1.10 PR/41258: Elias Benali: Try to get the ethernet address from the BIOS before failing.
1.5.2.1	28-Apr-2009	skrll	Sync with HEAD.
1.8.4.2	21-Apr-2011	rmind	sync with head
1.8.4.1	30-May-2010	rmind	sync with head
1.8.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.9.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.10.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.10.4.1	30-Oct-2012	yamt	sync with head
1.11.2.3	03-Dec-2017	jdolecek	update from HEAD
1.11.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.11.2.1	23-Jun-2013	tls	resync from head
1.12.4.1	18-May-2014	rmind	sync with head
1.13.6.3	05-Feb-2017	skrll	Sync with HEAD
1.13.6.2	09-Jul-2016	skrll	Sync with HEAD
1.13.6.1	19-Mar-2016	skrll	Sync with HEAD
1.15.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.17.14.3	26-Jan-2019	pgoyette	Sync with HEAD
1.17.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.17.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.17.8.1	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.18.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.18.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.18.2.1	10-Jun-2019	christos	Sync with HEAD
1.26.2.1	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #672): sys/dev/ic/tulip.c: revision 1.198 sys/dev/pci/if_jme.c: revision 1.45 sys/dev/pci/agp.c: revision 1.86 sys/dev/pci/if_lii.c: revision 1.27 sys/dev/acpi/thinkpad_acpi.c: revision 1.47 sys/dev/scsipi/scsipi_base.c: revision 1.183 sys/dev/ic/aic6915reg.h: revision 1.6 Fix undefined behavior in thinkpad_mask_init(). Found by kUBSan. Use unsigned when initialize the transmit completion ring to avoid undefined behavior. Found by kUBSan. Avoid undefined behavior when setting multicast address. found by kUBSan. Use unsigned to avoid undefined behavior in agpattach(). Found by kUBSan. Avoid undefined behavior in jme_mii_write(). Found by kUBSan. Use unsigned to avoid undefined behavior in lii_setmulti(). Use unsigned to avoid undefined behavior in scsipi_{get,put}_tag(). Found by kUBSan.
1.27.2.1	29-Feb-2020	ad	Sync with head.
1.2	29-Apr-2008	martin	branches: 1.2.4; Convert to new 2 clause license
1.1	29-Mar-2008	cube	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC. The linux driver was used as documentation. Many thanks to all the people who helped or at least supported me while I was fighting that bug that made the EeePC freeze. At the top of that list, Matt J. Fleming (mjf@) and David Gwynne (from the OpenBSD fame).
1.1.8.1	16-May-2008	yamt	sync with head.
1.1.6.2	19-Apr-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1128): sys/dev/pci/if_lii.c: revisions 1.1, 1.3 via patch sys/arch/i386/conf/ALL: revision 1.161 via patch share/man/man4/lii.4: revision 1.1 via patch sys/dev/pci/files.pci: revision 1.303 via patch sys/dev/DEVNAMES: revision 1.238 via patch sys/arch/i386/conf/GENERIC: revision 1.885 via patch distrib/sets/lists/man/mi: revisions 1.1062, 1.1063 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.43 via patch share/man/man4/Makefile: revision 1.460 via patch sys/arch/i386/conf/INSTALL: revision 1.323 via patch sys/dev/pci/if_liireg.h: revision 1.1 via patch Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC.
1.1.6.1	29-Mar-2008	bouyer	file if_liireg.h was added on branch netbsd-4 on 2008-04-19 15:57:56 +0000
1.1.4.1	18-May-2008	yamt	sync with head.
1.1.2.3	02-Jun-2008	mjf	Sync with HEAD.
1.1.2.2	03-Apr-2008	mjf	Sync with HEAD.
1.1.2.1	29-Mar-2008	mjf	file if_liireg.h was added on branch mjf-devfs2 on 2008-04-03 12:42:50 +0000
1.2.4.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.2.4.1	29-Apr-2008	skrll	file if_liireg.h was added on branch wrstuden-fixsa on 2008-06-03 20:47:24 +0000
1.69	12-Dec-2018	maxv	Retire the LMC driver, and its associated lmcconfig tool. LMC has been mentioned repeatedly as a non-MP-safe driver that is hard to maintain, and no one is taking care of it. LMC was removed from OpenBSD three years ago, and from FreeBSD a few months ago.
1.68	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.67	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.66	26-Jun-2018	msaitoh	branches: 1.66.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.65	25-Jun-2018	msaitoh	Move txintr_setup() stuff from lmc_interrupt() and do it in ifnet_start(). Now we can use bpf_mtap() in the TX path. Not tested.
1.64	07-Feb-2018	mrg	branches: 1.64.2; avoid an indentation issue by adding "if (1)".
1.63	24-Jan-2017	ozaki-r	branches: 1.63.6; Defer bpf_mtap in Rx interrupt context to softint bpf_mtap of some drivers is still called in hardware interrupt context. We want to run them in softint as well as bpf_mtap of most drivers (see if_percpuq_softint and if_input). To this end, bpf_mtap_softint mechanism is implemented; it defers bpf_mtap processing to a dedicated softint for a target driver. By using the machanism, we can move bpf_mtap processing to softint without changing target drivers much while it adds some overhead on CPU and memory. Once target drivers are changed to softint-based, we should return to normal bpf_mtap. Proposed on tech-kern and tech-net
1.62	15-Dec-2016	ozaki-r	branches: 1.62.2; Annotate bpf_mtap still running in Rx hardware interrupt with "XXX not in softint"
1.61	10-Jun-2016	ozaki-r	branches: 1.61.2; Avoid storing a pointer of an interface in a mbuf Having a pointer of an interface in a mbuf isn't safe if we remove big kernel locks; an interface object (ifnet) can be destroyed anytime in any packet processing and accessing such object via a pointer is racy. Instead we have to get an object from the interface collection (ifindex2ifnet) via an interface index (if_index) that is stored to a mbuf instead of an pointer. The change provides two APIs: m_{get,put}_rcvif_psref that use psref(9) for sleep-able critical sections and m_{get,put}_rcvif that use pserialize(9) for other critical sections. The change also adds another API called m_get_rcvif_NOMPSAFE, that is NOT MP-safe and for transition moratorium, i.e., it is intended to be used for places where are not planned to be MP-ified soon. The change adds some overhead due to psref to performance sensitive paths, however the overhead is not serious, 2% down at worst. Proposed on tech-kern and tech-net.
1.60	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.59	28-Apr-2016	ozaki-r	Constify rtentry of if_output We no longer need to change rtentry below if_output. The change makes it clear where rtentries are changed (or not) and helps forthcoming locking (os psrefing) rtentries.
1.58	20-Apr-2016	knakahara	IFQ_ENQUEUE refactor (3/3) : eliminate pktattr argument from IFQ_ENQUEUE caller
1.57	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.56	28-Nov-2014	ozaki-r	branches: 1.56.2; Don't call if_free_sadl in ifnet_detach It'll be called in if_detach below.
1.55	05-Jun-2014	rmind	- Implement pktqueue interface for lockless IP input queue. - Replace ipintrq and ip6intrq with the pktqueue mechanism. - Eliminate kernel-lock from ipintr() and ip6intr(). - Some preparation work to push softnet_lock out of ipintr(). Discussed on tech-net.
1.54	29-Mar-2014	christos	branches: 1.54.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.53	30-Mar-2013	christos	branches: 1.53.4; remove trailing whitespace
1.52	29-Oct-2012	chs	fix device/softc split errors.
1.51	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.50	13-Nov-2010	uebayasi	branches: 1.50.8; 1.50.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.49	19-Jan-2010	pooka	branches: 1.49.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.48	16-Sep-2009	dyoung	Pass lmc(4) through 'unifdef -U__FreeBSD__ -U__bsdi__ -U__OpenBSD__ -U__linux__ -D__NetBSD__ -UNETGRAPH' to shorten this driver by >2500 lines. objdump -D does not reveal any differences in the compiled code.
1.47	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.46	02-Feb-2009	tsutsui	branches: 1.46.2; Remove obsolete #if defined(LKM) block. PR kern/40513
1.45	12-Nov-2008	ad	Remove LKMs and switch to the module framework, pass 1. Proposed on tech-kern@.
1.44	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.43	27-Jun-2008	gmcgarry	branches: 1.43.2; 1.43.4; Convert lmc(4) to mutexes, removing ({ }) gcc extension along the way.
1.42	12-Jun-2008	cegger	use device_lookup_private to get softc
1.41	10-Apr-2008	cegger	branches: 1.41.2; 1.41.4; 1.41.6; 1.41.8; use aprint_*_dev and device_xname
1.40	07-Feb-2008	dyoung	branches: 1.40.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.39	10-Jan-2008	dyoung	Straggler: change generic ifmedia_X names to lmc_ifmedia_X.
1.38	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.37	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.36	19-Oct-2007	ad	branches: 1.36.2; 1.36.4; 1.36.6; 1.36.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.35	09-Jul-2007	ad	branches: 1.35.6; 1.35.8; 1.35.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.34	04-Mar-2007	christos	branches: 1.34.2; 1.34.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.33	17-Feb-2007	dyoung	KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous parentheses in return statements. Cosmetic: don't open-code TAILQ_FOREACH(). Cosmetic: change types of variables to avoid oodles of casts: in in6_src.c, avoid casts by changing several route_in6 pointers to struct route pointers. Remove unnecessary casts to caddr_t elsewhere. Pave the way for eliminating address family-specific route caches: soon, struct route will not embed a sockaddr, but it will hold a reference to an external sockaddr, instead. We will set the destination sockaddr using rtcache_setdst(). (I created a stub for it, but it isn't used anywhere, yet.) rtcache_free() will free the sockaddr. I have extracted from rtcache_free() a helper subroutine, rtcache_clear(). rtcache_clear() will "forget" a cached route, but it will not forget the destination by releasing the sockaddr. I use rtcache_clear() instead of rtcache_free() in rtcache_update(), because rtcache_update() is not supposed to forget the destination. Constify: 1 Introduce const accessor for route->ro_dst, rtcache_getdst(). 2 Constify the 'dst' argument to ifnet->if_output(). This led me to constify a lot of code called by output routines. 3 Constify the sockaddr argument to protosw->pr_ctlinput. This led me to constify a lot of code called by ctlinput routines. 4 Introduce const macros for converting from a generic sockaddr to family-specific sockaddrs, e.g., sockaddr_in: satocsin6, satocsin, et cetera.
1.32	16-Nov-2006	christos	branches: 1.32.4; __unused removal on arguments; approved by core.
1.31	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.30	30-Aug-2006	christos	branches: 1.30.2; 1.30.4; complete initializers.
1.29	14-May-2006	elad	integrate kauth.
1.28	20-Apr-2006	christos	New version from David Boggs.
1.27	14-Apr-2006	christos	Coverity CID 1122: Add KASSERT to clarify this NULL deref can't happen.
1.26	10-Mar-2006	christos	branches: 1.26.2; PR/33049: OKANO Takayoshi: Fix the spelling of NetBSD RCS tags.
1.25	13-Feb-2006	martin	branches: 1.25.2; 1.25.4; Protect parts inside #ifdef _LKM that are only supposed to be there when building an LKM.
1.24	11-Feb-2006	christos	new version from David Boggs.
1.23	06-Dec-2005	christos	branches: 1.23.2; 1.23.4; 1.23.6; New lmc driver from David Boggs
1.22	30-May-2005	christos	branches: 1.22.2; - const poisoning - avoid variable shadowing.
1.21	27-Feb-2005	perry	nuke trailing whitespace
1.20	03-May-2003	wiz	branches: 1.20.2; 1.20.10; 1.20.12; DMA, not dma nor Dma.
1.19	20-Jan-2003	simonb	Remove variable that is only assigned too but not referenced.
1.18	04-Jan-2002	martin	Move net/if_sppp.h to net/if_spppvar.h, create a new net/if_sppp.h containing the userland visible thinks (i.e. ioctl definitions). Remove all (both) old ioctls, as they had a brain dead API and made keeping binary compatibility more or less impossible. Replace by several new ioctls. While there, remove any arbitrary limits (resulting from the old, broken ioctls) and allow any length of names and passwords.
1.17	13-Nov-2001	lukem	add RCSID
1.16	27-Aug-2001	eeh	Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.15	19-Jul-2001	itojun	it seems that pp_{up,down} code does not really help, since (1) sppp_ioctl calls lcp.{Up,Down} by itself, and (2) as lmc driver uses interrupt to get the card initialization callback lcp calls from sppp_ioctl does not really work. for chdlc pp_{up,down} are not really necessary so don't bother (#if 0'ed).
1.14	19-Jul-2001	itojun	do not call pp_{up,down} on cisco HDLC mode, it seems that sys/net/if_spppsubr.c does not like it. XXX api clarification needed...
1.13	19-Jul-2001	itojun	- on IFF_UP change, inform sys/net/if_spppsubr.c of the change so that we can run LCP and subsequent PPP negotiation. - suppress too noisy printf() for about AIS event (if you remove the cable, it will bark forever). - sppp_dequeue() can return NULL even when sppp_isempty() is false, so check it and do not deref NULL pointer. - sppp_flush() on LMC interface shutdown. - prepare to change BPF type on the fly (bpf_change_type).
1.12	07-Jul-2001	thorpej	branches: 1.12.2; bzero -> memset
1.11	07-Jul-2001	thorpej	bcopy -> memcpy
1.10	13-Jun-2001	wiz	withough -> without
1.9	12-Jun-2001	wiz	receive, not recieve
1.8	12-Apr-2001	itojun	support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.7	09-Apr-2001	martin	Specify the size of the framing added to packets, so if_sppsubr.c can calculate correct statistics.
1.6	14-Nov-2000	thorpej	branches: 1.6.2; NBPG -> PAGE_SIZE
1.5	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.4	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.3	03-May-2000	thorpej	`typedef void ifnet_ret_t;' isn't valid C -- use #define, instead. Also, don't abuse the C preprocessor.
1.2	04-Dec-1999	ragge	CL* discarding.
1.1	25-Mar-1999	explorer	branches: 1.1.8; 1.1.14; Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.1.14.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.8.3	21-Apr-2001	bouyer	Sync with HEAD
1.1.8.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.6.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.6.2.5	11-Jan-2002	nathanw	More catchup.
1.6.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.6.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.6.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.6.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.12.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.12.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.12.2.1	03-Aug-2001	lukem	update to -current
1.20.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.20.10.1	29-Apr-2005	kent	sync with -current
1.20.2.3	11-Dec-2005	christos	Sync with head.
1.20.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.20.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.22.2.7	11-Feb-2008	yamt	sync with head.
1.22.2.6	21-Jan-2008	yamt	sync with head
1.22.2.5	27-Oct-2007	yamt	sync with head.
1.22.2.4	03-Sep-2007	yamt	sync with head.
1.22.2.3	26-Feb-2007	yamt	sync with head.
1.22.2.2	30-Dec-2006	yamt	sync with head.
1.22.2.1	21-Jun-2006	yamt	sync with head.
1.23.6.2	01-Jun-2006	kardel	Sync with head.
1.23.6.1	22-Apr-2006	simonb	Sync with head.
1.23.4.1	09-Sep-2006	rpaulo	sync with head
1.23.2.1	18-Feb-2006	yamt	sync with head.
1.25.4.3	11-May-2006	elad	sync with head
1.25.4.2	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.25.4.1	19-Apr-2006	elad	sync with head.
1.25.2.3	03-Sep-2006	yamt	sync with head.
1.25.2.2	24-May-2006	yamt	sync with head.
1.25.2.1	13-Mar-2006	yamt	sync with head.
1.26.2.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.30.4.2	10-Dec-2006	yamt	sync with head.
1.30.4.1	22-Oct-2006	yamt	sync with head
1.30.2.1	18-Nov-2006	ad	Sync with head.
1.32.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.32.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.34.4.1	11-Jul-2007	mjf	Sync with head.
1.34.2.2	23-Oct-2007	ad	Sync with head.
1.34.2.1	05-Apr-2007	ad	Compile fixes.
1.35.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.35.8.3	23-Mar-2008	matt	sync with HEAD
1.35.8.2	09-Jan-2008	matt	sync with HEAD
1.35.8.1	06-Nov-2007	matt	sync with HEAD
1.35.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.36.8.3	10-Jan-2008	bouyer	Sync with HEAD
1.36.8.2	08-Jan-2008	bouyer	Sync with HEAD
1.36.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.36.6.1	11-Dec-2007	yamt	sync with head.
1.36.4.1	26-Dec-2007	ad	Sync with head.
1.36.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.40.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.40.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.40.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.41.8.2	27-Jun-2008	simonb	Sync with head.
1.41.8.1	18-Jun-2008	simonb	Sync with head.
1.41.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.41.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.41.4.3	11-Mar-2010	yamt	sync with head
1.41.4.2	16-May-2009	yamt	sync with head
1.41.4.1	04-May-2009	yamt	sync with head.
1.41.2.1	17-Jun-2008	yamt	sync with head.
1.43.4.2	03-Mar-2009	skrll	Sync with HEAD.
1.43.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.43.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.46.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.49.4.1	05-Mar-2011	rmind	sync with head
1.50.18.4	03-Dec-2017	jdolecek	update from HEAD
1.50.18.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.50.18.2	23-Jun-2013	tls	resync from head
1.50.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.50.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.50.8.1	30-Oct-2012	yamt	sync with head
1.53.4.1	18-May-2014	rmind	sync with head
1.54.2.1	10-Aug-2014	tls	Rebase.
1.56.2.4	09-Jul-2016	skrll	Sync with HEAD
1.56.2.3	29-May-2016	skrll	Sync with HEAD
1.56.2.2	22-Apr-2016	skrll	Sync with HEAD
1.56.2.1	19-Mar-2016	skrll	Sync with HEAD
1.61.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.61.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.62.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.63.6.1	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #950): sys/dev/pci/if_lmc.h: revision 1.25 sys/dev/pci/if_lmc.c: revision 1.65 doc/TODO.smpnet: revision 1.20 Move txintr_setup() stuff from lmc_interrupt() and do it in ifnet_start(). Now we can use bpf_mtap() in the TX path. Not tested.
1.64.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.64.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.64.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.66.2.1	10-Jun-2019	christos	Sync with HEAD
1.27	12-Dec-2018	maxv	Retire the LMC driver, and its associated lmcconfig tool. LMC has been mentioned repeatedly as a non-MP-safe driver that is hard to maintain, and no one is taking care of it. LMC was removed from OpenBSD three years ago, and from FreeBSD a few months ago.
1.26	16-Sep-2018	skrll	interrupt has two 'r's fix another typo while I'm here (flsah)
1.25	25-Jun-2018	msaitoh	branches: 1.25.2; Move txintr_setup() stuff from lmc_interrupt() and do it in ifnet_start(). Now we can use bpf_mtap() in the TX path. Not tested.
1.24	24-Jan-2017	ozaki-r	branches: 1.24.6; 1.24.12; Defer bpf_mtap in Rx interrupt context to softint bpf_mtap of some drivers is still called in hardware interrupt context. We want to run them in softint as well as bpf_mtap of most drivers (see if_percpuq_softint and if_input). To this end, bpf_mtap_softint mechanism is implemented; it defers bpf_mtap processing to a dedicated softint for a target driver. By using the machanism, we can move bpf_mtap processing to softint without changing target drivers much while it adds some overhead on CPU and memory. Once target drivers are changed to softint-based, we should return to normal bpf_mtap. Proposed on tech-kern and tech-net
1.23	28-Apr-2016	ozaki-r	branches: 1.23.2; 1.23.4; Constify rtentry of if_output We no longer need to change rtentry below if_output. The change makes it clear where rtentries are changed (or not) and helps forthcoming locking (os psrefing) rtentries.
1.22	06-Sep-2015	dholland	More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.21	27-Oct-2012	chs	branches: 1.21.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.20	13-Mar-2012	elad	branches: 1.20.2; Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.19	05-Apr-2010	joerg	branches: 1.19.8; 1.19.12; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.18	19-Jan-2010	pooka	branches: 1.18.2; 1.18.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.17	16-Sep-2009	dyoung	Pass lmc(4) through 'unifdef -U__FreeBSD__ -U__bsdi__ -U__OpenBSD__ -U__linux__ -D__NetBSD__ -UNETGRAPH' to shorten this driver by >2500 lines. objdump -D does not reveal any differences in the compiled code.
1.16	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.15	12-Nov-2008	ad	branches: 1.15.4; Remove LKMs and switch to the module framework, pass 1. Proposed on tech-kern@.
1.14	08-Sep-2008	gmcgarry	branches: 1.14.2; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.13	27-Jun-2008	gmcgarry	branches: 1.13.2; Convert lmc(4) to mutexes, removing ({ }) gcc extension along the way.
1.12	12-Jun-2008	cegger	use device_lookup_private to get softc
1.11	10-Apr-2008	cegger	branches: 1.11.2; 1.11.4; 1.11.6; 1.11.8; use aprint_*_dev and device_xname
1.10	10-Jan-2008	dyoung	branches: 1.10.6; Straggler: change generic ifmedia_X names to lmc_ifmedia_X.
1.9	04-Mar-2007	christos	branches: 1.9.16; 1.9.22; 1.9.28; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.8	17-Feb-2007	dyoung	KNF: de-__P, bzero -> memset, bcmp -> memcmp. Remove extraneous parentheses in return statements. Cosmetic: don't open-code TAILQ_FOREACH(). Cosmetic: change types of variables to avoid oodles of casts: in in6_src.c, avoid casts by changing several route_in6 pointers to struct route pointers. Remove unnecessary casts to caddr_t elsewhere. Pave the way for eliminating address family-specific route caches: soon, struct route will not embed a sockaddr, but it will hold a reference to an external sockaddr, instead. We will set the destination sockaddr using rtcache_setdst(). (I created a stub for it, but it isn't used anywhere, yet.) rtcache_free() will free the sockaddr. I have extracted from rtcache_free() a helper subroutine, rtcache_clear(). rtcache_clear() will "forget" a cached route, but it will not forget the destination by releasing the sockaddr. I use rtcache_clear() instead of rtcache_free() in rtcache_update(), because rtcache_update() is not supposed to forget the destination. Constify: 1 Introduce const accessor for route->ro_dst, rtcache_getdst(). 2 Constify the 'dst' argument to ifnet->if_output(). This led me to constify a lot of code called by output routines. 3 Constify the sockaddr argument to protosw->pr_ctlinput. This led me to constify a lot of code called by ctlinput routines. 4 Introduce const macros for converting from a generic sockaddr to family-specific sockaddrs, e.g., sockaddr_in: satocsin6, satocsin, et cetera.
1.7	04-Jan-2007	elad	branches: 1.7.2; Consistent usage of KAUTH_GENERIC_ISSUSER.
1.6	27-Aug-2006	christos	branches: 1.6.2; Grr. Don't define _KERNEL! Most of the kernel files do #ifdef _KERNEL
1.5	21-Jul-2006	ad	- Use the LWP cached credentials where sane. - Minor cosmetic changes.
1.4	14-May-2006	elad	branches: 1.4.6; integrate kauth.
1.3	20-Apr-2006	christos	New version from David Boggs.
1.2	11-Feb-2006	christos	branches: 1.2.2; 1.2.4; 1.2.6; new version from David Boggs.
1.1	06-Dec-2005	christos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; New lmc driver from David Boggs
1.1.8.2	01-Jun-2006	kardel	Sync with head.
1.1.8.1	22-Apr-2006	simonb	Sync with head.
1.1.6.1	09-Sep-2006	rpaulo	sync with head
1.1.4.1	18-Feb-2006	yamt	sync with head.
1.1.2.2	11-Dec-2005	christos	Sync with head.
1.1.2.1	06-Dec-2005	christos	file if_lmc.h was added on branch ktrace-lwp on 2005-12-11 10:28:58 +0000
1.2.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.2.4.3	11-May-2006	elad	sync with head
1.2.4.2	10-Mar-2006	elad	generic_authorize() -> kauth_authorize_generic().
1.2.4.1	08-Mar-2006	elad	Adapt to kernel authorization KPI.
1.2.2.3	03-Sep-2006	yamt	sync with head.
1.2.2.2	11-Aug-2006	yamt	sync with head
1.2.2.1	24-May-2006	yamt	sync with head.
1.4.6.6	21-Jan-2008	yamt	sync with head
1.4.6.5	03-Sep-2007	yamt	sync with head.
1.4.6.4	26-Feb-2007	yamt	sync with head.
1.4.6.3	30-Dec-2006	yamt	sync with head.
1.4.6.2	21-Jun-2006	yamt	sync with head.
1.4.6.1	14-May-2006	yamt	file if_lmc.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:04 +0000
1.6.2.1	12-Jan-2007	ad	Sync with head.
1.7.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.7.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.9.28.1	10-Jan-2008	bouyer	Sync with HEAD
1.9.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.16.1	23-Mar-2008	matt	sync with HEAD
1.10.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.10.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.10.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.10.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.11.8.2	27-Jun-2008	simonb	Sync with head.
1.11.8.1	18-Jun-2008	simonb	Sync with head.
1.11.6.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.11.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.11.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.11.4.4	11-Aug-2010	yamt	sync with head.
1.11.4.3	11-Mar-2010	yamt	sync with head
1.11.4.2	16-May-2009	yamt	sync with head
1.11.4.1	04-May-2009	yamt	sync with head.
1.11.2.1	17-Jun-2008	yamt	sync with head.
1.13.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.13.2.1	19-Oct-2008	haad	Sync with HEAD.
1.14.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.15.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.4.1	30-May-2010	rmind	sync with head
1.18.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.19.12.1	05-Apr-2012	mrg	sync to latest -current.
1.19.8.2	30-Oct-2012	yamt	sync with head
1.19.8.1	17-Apr-2012	yamt	sync with head
1.20.2.2	03-Dec-2017	jdolecek	update from HEAD
1.20.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.21.14.3	05-Feb-2017	skrll	Sync with HEAD
1.21.14.2	29-May-2016	skrll	Sync with HEAD
1.21.14.1	22-Sep-2015	skrll	Sync with HEAD
1.23.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.23.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.24.12.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.24.12.2	30-Sep-2018	pgoyette	Ssync with HEAD
1.24.12.1	28-Jul-2018	pgoyette	Sync with HEAD
1.24.6.1	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #950): sys/dev/pci/if_lmc.h: revision 1.25 sys/dev/pci/if_lmc.c: revision 1.65 doc/TODO.smpnet: revision 1.20 Move txintr_setup() stuff from lmc_interrupt() and do it in ifnet_start(). Now we can use bpf_mtap() in the TX path. Not tested.
1.25.2.1	10-Jun-2019	christos	Sync with HEAD
1.11	06-Dec-2005	christos	New lmc driver from David Boggs
1.10	27-Feb-2005	perry	branches: 1.10.4; nuke trailing whitespace
1.9	04-Jan-2002	martin	branches: 1.9.16; 1.9.24; 1.9.26; Move net/if_sppp.h to net/if_spppvar.h, create a new net/if_sppp.h containing the userland visible thinks (i.e. ioctl definitions). Remove all (both) old ioctls, as they had a brain dead API and made keeping binary compatibility more or less impossible. Replace by several new ioctls. While there, remove any arbitrary limits (resulting from the old, broken ioctls) and allow any length of names and passwords.
1.8	13-Nov-2001	lukem	add RCSID
1.7	27-Aug-2001	eeh	Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.6	13-Jun-2001	wiz	branches: 1.6.2; withough -> without
1.5	12-Apr-2001	itojun	support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.4	28-Jun-2000	mrg	branches: 1.4.2; remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.3	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.2	03-May-2000	thorpej	`typedef void ifnet_ret_t;' isn't valid C -- use #define, instead. Also, don't abuse the C preprocessor.
1.1	25-Mar-1999	explorer	branches: 1.1.8; Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.1.8.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.4.2.4	11-Jan-2002	nathanw	More catchup.
1.4.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.4.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.6.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.9.26.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.24.1	29-Apr-2005	kent	sync with -current
1.9.16.2	11-Dec-2005	christos	Sync with head.
1.9.16.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.4.1	21-Jun-2006	yamt	sync with head.
1.18	06-Dec-2005	christos	New lmc driver from David Boggs
1.17	30-May-2005	christos	branches: 1.17.2; - const poisoning - avoid variable shadowing.
1.16	27-Feb-2005	perry	nuke trailing whitespace
1.15	22-Feb-2003	tsutsui	branches: 1.15.2; 1.15.10; 1.15.12; hz -> Hz
1.14	25-Nov-2002	thorpej	Fix a strict-alias problem uncovered by GCC 3.3.
1.13	04-Jan-2002	martin	Move net/if_sppp.h to net/if_spppvar.h, create a new net/if_sppp.h containing the userland visible thinks (i.e. ioctl definitions). Remove all (both) old ioctls, as they had a brain dead API and made keeping binary compatibility more or less impossible. Replace by several new ioctls. While there, remove any arbitrary limits (resulting from the old, broken ioctls) and allow any length of names and passwords.
1.12	13-Nov-2001	lukem	add RCSID
1.11	01-Nov-2001	uch	missing NBPFILTER > 0
1.10	27-Aug-2001	eeh	branches: 1.10.4; Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.9	19-Jul-2001	itojun	sync with latest Linux driver, avoid framing error on LMC1200 (T1/E1). from hal@vailsys.com
1.8	19-Jul-2001	itojun	readability changes (s/0/LMC_LINK_DOWN/, whitespace)
1.7	19-Jul-2001	itojun	- on IFF_UP change, inform sys/net/if_spppsubr.c of the change so that we can run LCP and subsequent PPP negotiation. - suppress too noisy printf() for about AIS event (if you remove the cable, it will bark forever). - sppp_dequeue() can return NULL even when sppp_isempty() is false, so check it and do not deref NULL pointer. - sppp_flush() on LMC interface shutdown. - prepare to change BPF type on the fly (bpf_change_type).
1.6	12-Jun-2001	wiz	branches: 1.6.2; receive, not recieve
1.5	12-Apr-2001	itojun	support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.4	28-Jun-2000	mrg	branches: 1.4.2; remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.3	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.2	03-May-2000	thorpej	`typedef void ifnet_ret_t;' isn't valid C -- use #define, instead. Also, don't abuse the C preprocessor.
1.1	25-Mar-1999	explorer	branches: 1.1.8; Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.1.8.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.7	11-Dec-2002	thorpej	Sync with HEAD.
1.4.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.4.2.5	11-Jan-2002	nathanw	More catchup.
1.4.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.4.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.4.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.6.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.6.2.1	03-Aug-2001	lukem	update to -current
1.10.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.15.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.15.10.1	29-Apr-2005	kent	sync with -current
1.15.2.3	11-Dec-2005	christos	Sync with head.
1.15.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.15.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.17.2.1	21-Jun-2006	yamt	sync with head.
1.23	06-Dec-2005	christos	New lmc driver from David Boggs
1.22	27-Feb-2005	perry	branches: 1.22.4; nuke trailing whitespace
1.21	01-Apr-2003	thorpej	branches: 1.21.2; 1.21.10; 1.21.12; Use PAGE_SIZE rather than NBPG.
1.20	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.19	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.18	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.17	09-May-2002	matt	Print out interrupt we attached to.
1.16	04-Jan-2002	martin	Move net/if_sppp.h to net/if_spppvar.h, create a new net/if_sppp.h containing the userland visible thinks (i.e. ioctl definitions). Remove all (both) old ioctls, as they had a brain dead API and made keeping binary compatibility more or less impossible. Replace by several new ioctls. While there, remove any arbitrary limits (resulting from the old, broken ioctls) and allow any length of names and passwords.
1.15	13-Nov-2001	lukem	add RCSID
1.14	27-Aug-2001	eeh	Get rid of debug printf.
1.13	27-Aug-2001	augustss	Make it compile.
1.12	27-Aug-2001	eeh	Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.11	19-Jul-2001	itojun	- on IFF_UP change, inform sys/net/if_spppsubr.c of the change so that we can run LCP and subsequent PPP negotiation. - suppress too noisy printf() for about AIS event (if you remove the cable, it will bark forever). - sppp_dequeue() can return NULL even when sppp_isempty() is false, so check it and do not deref NULL pointer. - sppp_flush() on LMC interface shutdown. - prepare to change BPF type on the fly (bpf_change_type).
1.10	07-Jul-2001	thorpej	branches: 1.10.2; bcopy -> strcpy
1.9	13-Jun-2001	wiz	withough -> without
1.8	12-Apr-2001	itojun	support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.7	12-Jan-2001	itojun	branches: 1.7.2; match T1 card. from freebsd driver, it should just work. i'll test it as soon as i get the card handy.
1.6	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.5	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.4	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.3	03-May-2000	thorpej	`typedef void ifnet_ret_t;' isn't valid C -- use #define, instead. Also, don't abuse the C preprocessor.
1.2	12-May-1999	explorer	branches: 1.2.2; just say no to __BROKEN_INDIRECT_CONFIG
1.1	25-Mar-1999	explorer	branches: 1.1.4; Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.1.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.2.2.4	21-Apr-2001	bouyer	Sync with HEAD
1.2.2.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.2.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.7	20-Jun-2002	nathanw	Catch up to -current.
1.7.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.7.2.5	11-Jan-2002	nathanw	More catchup.
1.7.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.7.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.7.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.10.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.10.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.10.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.10.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.10.2.1	03-Aug-2001	lukem	update to -current
1.21.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.10.1	29-Apr-2005	kent	sync with -current
1.21.2.2	11-Dec-2005	christos	Sync with head.
1.21.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.22.4.1	21-Jun-2006	yamt	sync with head.
1.6	06-Dec-2005	christos	New lmc driver from David Boggs
1.5	09-May-2002	matt	branches: 1.5.10; 1.5.26; LMC_CTL_LEN1 and LMC_CTL_LEN2 extracted each other's fields. Stop doing that.
1.4	27-Aug-2001	eeh	Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.3	12-Apr-2001	itojun	branches: 1.3.2; support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.2	25-Mar-1999	explorer	branches: 1.2.8; 1.2.20; wrap kernel bits in shared kernel/userland files
1.1	25-Mar-1999	explorer	Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.2.20.3	20-Jun-2002	nathanw	Catch up to -current.
1.2.20.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.20.1	21-Jun-2001	nathanw	Catch up to -current.
1.2.8.1	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.5.26.1	21-Jun-2006	yamt	sync with head.
1.5.10.1	11-Dec-2005	christos	Sync with head.
1.6	06-Dec-2005	christos	New lmc driver from David Boggs
1.5	08-Jul-2003	itojun	branches: 1.5.16; function prototype must not have variable name
1.4	27-Aug-2001	eeh	branches: 1.4.20; Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.3	12-Apr-2001	itojun	branches: 1.3.2; support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.2	25-Mar-1999	explorer	branches: 1.2.8; 1.2.20; wrap kernel bits in shared kernel/userland files
1.1	25-Mar-1999	explorer	Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.2.20.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.20.1	21-Jun-2001	nathanw	Catch up to -current.
1.2.8.1	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.4.20.4	11-Dec-2005	christos	Sync with head.
1.4.20.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.20.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.20.1	03-Aug-2004	skrll	Sync with HEAD
1.5.16.1	21-Jun-2006	yamt	sync with head.
1.13	06-Dec-2005	christos	New lmc driver from David Boggs
1.12	27-Feb-2005	perry	branches: 1.12.4; nuke trailing whitespace
1.11	01-Oct-2002	thorpej	branches: 1.11.6; 1.11.14; 1.11.16; Remove uneecessary extern decl of de_ca.
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	31-May-2002	thorpej	* Add __unused___ to lmc_system_errors, since it is not used by everything which includes if_lmcvar.h. * #if 0 lmc_status_bits, since it's not used by anything.
1.8	27-Aug-2001	eeh	branches: 1.8.12; Make the LMC driver bus_dma and endian clean (big-endian mode does not appear to work on LMC cards).
1.7	13-Jun-2001	wiz	branches: 1.7.2; withough -> without
1.6	12-Apr-2001	itojun	support T1 card (LMC1200) properly. PR 12331 (patch from openbsd). change DLT type to cisco HDLC, as lmc driver is hardwired to cisco HDLC on driver attach. XXX we may need to revisit this, as if_spppsubr may want to alter this later.
1.5	12-Dec-2000	thorpej	branches: 1.5.2; Adapt to bpfattach() changes, and further centralize the bpfattach() and bpfdetach() calls into link-type subroutines where possible.
1.4	06-Sep-2000	thorpej	Cast the argument to vtophys() to vaddr_t.
1.3	03-May-2000	thorpej	`typedef void ifnet_ret_t;' isn't valid C -- use #define, instead. Also, don't abuse the C preprocessor.
1.2	11-May-1999	thorpej	branches: 1.2.2; Set the data link type to DLT_PPP_SERIAL so that tcpdump can do something useful with our captured frames.
1.1	25-Mar-1999	explorer	branches: 1.1.4; Commit files for the LAN Media Corporation high speed serial/HSSI/DS3 driver
1.1.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.2.2.3	21-Apr-2001	bouyer	Sync with HEAD
1.2.2.2	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.5.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.5.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.5.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.7.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.7.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.8.12.1	20-Jun-2002	gehenna	catch up with -current.
1.11.16.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.14.1	29-Apr-2005	kent	sync with -current
1.11.6.2	11-Dec-2005	christos	Sync with head.
1.11.6.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.4.1	21-Jun-2006	yamt	sync with head.
1.9	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.8	08-May-2021	thorpej	Use pci_compatible_match().
1.7	09-Dec-2018	jdolecek	branches: 1.7.16; use pci_intr_establish_xname() everywhere
1.6	02-Feb-2017	nonaka	branches: 1.6.12; 1.6.14; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.5	29-Mar-2014	christos	branches: 1.5.6; 1.5.10; 1.5.14; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.4	30-Mar-2013	christos	branches: 1.4.4; remove trailing whitespace
1.3	05-Aug-2012	degroote	branches: 1.3.2; 1.3.4; Add pmf(9) handler to malo(4) wireless driver
1.2	30-Jul-2012	degroote	Add missing $NetBSD$ tags and __KERNEL_RCSID().
1.1	30-Jul-2012	degroote	Add malo(4)@pci driver for Marvell Libertas wireless adaptor Ported from OpenBSD Known issues : - contrary to OpenBSD one, only support pci at the moment, because I don't have the necessary hardware to test PCMCIA / CARDUS Marvell Card - not connected to pmf(9) (unable to test it)
1.3.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.4.2	30-Oct-2012	yamt	sync with head
1.3.4.1	05-Aug-2012	yamt	file if_malo_pci.c was added on branch yamt-pagecache on 2012-10-30 17:21:30 +0000
1.3.2.3	03-Dec-2017	jdolecek	update from HEAD
1.3.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.3.2.1	23-Jun-2013	tls	resync from head
1.4.4.1	18-May-2014	rmind	sync with head
1.5.14.1	21-Apr-2017	bouyer	Sync with HEAD
1.5.10.1	20-Mar-2017	pgoyette	Sync with HEAD
1.5.6.1	05-Feb-2017	skrll	Sync with HEAD
1.6.14.1	10-Jun-2019	christos	Sync with HEAD
1.6.12.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7.16.1	13-May-2021	thorpej	Sync with HEAD.
1.29	07-Jul-2024	msaitoh	mcx(4): Match on ConnectX-6 virtual functions, too. From OpenBSD.
1.28	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.27	11-Apr-2024	andvar	branches: 1.27.2; mcx(4): enforce full-duplex mark in mcx_media_status(), when link is up. LACP protocol requires full-duplex to be enabled for lagg(4) to work, however mcx(4) was not setting this capability making it to fail. Fixes PR kern/58124. OK'd by msaitoh@
1.26	26-Oct-2023	msaitoh	mcx(4): Add Mellanox ConnectX-6 Lx from OpenBSD.
1.25	28-Aug-2022	skrll	branches: 1.25.4; Empty if_watchdog functions are useless and now waste resources. Let's remove them.
1.24	07-Jul-2022	skrll	Trailing whitespace
1.23	05-Nov-2021	andvar	mcx(4): ensure that RQT size is always a power of two. Fixes PR kern/56484. Thanks riastradh for the power of two patch :). OK: mrg, jmcneill, riastradh.
1.22	26-Sep-2021	jmcneill	Enable checksum offload features by default.
1.21	26-Sep-2021	jmcneill	We are not mapping registers prefetchable so no need for explicit bs barriers.
1.20	25-Sep-2021	jmcneill	Sync with OpenBSD r1.101: When processing a received packet, only sync the amount of bytes mcx(4) told us has arrived. The DMA map's mapsize on RX packets is the length of the allocated buffer. For mcx(4), this can be more than around 9000 bytes, as each buffer will be at least as big as the maximum supported MTU. There's no need to sync the whole buffer, if it's only a small packet.
1.19	20-Sep-2021	jmcneill	Use coherent DMA mappings when possible.
1.18	24-Jul-2021	andvar	s/structres/structures + s/whenver/whenever
1.17	20-Feb-2021	jmcneill	branches: 1.17.4; Sync with OpenBSD r1.99: move the rearming of the cq after the refill of the rq. this is the only real diff we have left outstanding on a box that experienced rx lockups. since adding this change it's been happy for the last 4 weeks and counting so far.
1.16	05-Feb-2021	jmcneill	mcx: Add if_transmit callback to enable more than one tx queue
1.15	30-Jan-2021	jmcneill	Sync with OpenBSD r1.98. Lots of improvements including checksum offload, hardware vlan tagging, and support for multiple receive queues.
1.14	25-May-2020	jmcneill	branches: 1.14.2; Add __KERNEL_RCSID
1.13	24-Apr-2020	jmcneill	mcx: sync with OpenBSD sys/dev/pci/if_mcx.c r1.44 1.44: Fix typo which could lead into a double free 1.43: Commands that create objects return a 24 bit object ID, so mask off the high 8 bits of the value we extract, in case the firmware leaves junk there. Hrvoje Popovski has seen this with newer firmware on a ConnectX 5 card, which now works properly. 1.42: Increase the completion queue size to prevent overflow. Under reasonably unlikely circumstances - lots of single-fragment packets being sent, a significant number of packets being received, while the interrupt handler was unable to process the completion queue - the completion queue could overflow, which would result in the interface locking up. 1.41: Check if we've reached the end of the current mailbox before writing past the end of it, rather than after. Now we can actually allocate queues big enough to need multiple mailboxes. 1.40: Don't call mcx_intr() from mcx_cmdq_poll(); this was a leftover from early development that I forgot about, but turns out to be a potential race with the actual interrupt handler. 1.39: fix previous: use the correct offset for sq/rq creation, and don't reset the mbox counter to 0 after calculating it. 1.38: Add a helper function for writing physical addresses for queues into command queue mailboxes, and use this for all queue setup commands. Previously we just assumed the addresses would fit in the first mailbox, which is currently true but may not be for much longer. 1.37: (skipped) 1.36: The event queue consumer counter also needs to be unsigned like the others. 1.35: try to make if_baudrate look plausible. this updates the eth proto capability map so it records the baudrate against the different link types and their media, and then reads it when the link state changes. 1.34: (skipped)
1.12	15-Mar-2020	thorpej	branches: 1.12.2; 1.12.4; Define and implement a locking protocol for the ifmedia / mii layers: - MP-safe drivers provide a mutex to ifmedia that is used to serialize access to media-related structures / hardware regsiters. Converted drivers use the new ifmedia_init_with_lock() function for this. The new name is provided to ease the transition. - Un-converted drivers continue to call ifmedia_init(), which will supply a compatibility lock to be used instead. Several media-related entry points must be aware of this compatibility lock, and are able to acquire it recursively a limited number of times, if needed. This is a SPIN mutex with priority IPL_NET. - This same lock is used to serialize access to PHY registers and other MII-related data structures. The PHY drivers are modified to acquire and release the lock, as needed, and assert the lock is held as a diagnostic aid. The "usbnet" framework has had an overhaul of its internal locking protocols to fit in with the media / mii changes, and the drivers adapted. USB wifi drivers have been changed to provide their own adaptive mutex to the ifmedia later via a new ieee80211_media_init_with_lock() function. This is required because the USB drivers need an adaptive mutex. Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv. mcx also now calls ifmedia_init_with_lock() because it needs to also use an adaptive mutex. The mcx driver still needs to be fully converted to NET_MPSAFE.
1.11	29-Feb-2020	thorpej	Use the 64-bit PCI DMA tag if available. Otherwise, we are needlessly allocating (and using) bounce buffers on sytems with >4GB of RAM. XXX pullup-9
1.10	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.9	29-Nov-2019	msaitoh	branches: 1.9.2; Set if_baudrate.
1.8	28-Nov-2019	msaitoh	- Change IFM_10G_T(10GBase-T) entry to IFM_10G_CR1(10GBASE-CR1) - Add the following entries: 20GBASE-KR2 56GBASE-R4 100GBASE-LR4 100BaseTX 1000baseT 10GBase-T
1.7	26-Nov-2019	jmcneill	Fix IFF_ALLMULTI handling.
1.6	18-Nov-2019	nonaka	Fixed incorrect m_defrag return value check.
1.5	17-Oct-2019	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.4	28-Sep-2019	tnn	len -> m_len
1.3	28-Sep-2019	tnn	mcx(4): fix rx mbuf DMA overrun pullup-9
1.2	24-Sep-2019	msaitoh	Use IFM_50G_SR2.
1.1	21-Sep-2019	jmcneill	branches: 1.1.2; Add support for Mellanox 5th generation ethernet devices. This driver supports the ConnectX-4 Lx EN, ConnectX-4 EN, ConnectX-5 EN, and ConnectX-6 EN adapters. Driver ported from OpenBSD.
1.1.2.10	28-Apr-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #858): sys/dev/pci/if_mcx.c: revision 1.13 mcx: sync with OpenBSD sys/dev/pci/if_mcx.c r1.44 1.44: Fix typo which could lead into a double free 1.43: Commands that create objects return a 24 bit object ID, so mask off the high 8 bits of the value we extract, in case the firmware leaves junk there. Hrvoje Popovski has seen this with newer firmware on a ConnectX 5 card, which now works properly. 1.42: Increase the completion queue size to prevent overflow. Under reasonably unlikely circumstances - lots of single-fragment packets being sent, a significant number of packets being received, while the interrupt handler was unable to process the completion queue - the completion queue could overflow, which would result in the interface locking up. 1.41: Check if we've reached the end of the current mailbox before writing past the end of it, rather than after. Now we can actually allocate queues big enough to need multiple mailboxes. 1.40: Don't call mcx_intr() from mcx_cmdq_poll(); this was a leftover from early development that I forgot about, but turns out to be a potential race with the actual interrupt handler. 1.39: fix previous: use the correct offset for sq/rq creation, and don't reset the mbox counter to 0 after calculating it. 1.38: Add a helper function for writing physical addresses for queues into command queue mailboxes, and use this for all queue setup commands. Previously we just assumed the addresses would fit in the first mailbox, which is currently true but may not be for much longer. 1.37: (skipped) 1.36: The event queue consumer counter also needs to be unsigned like the others. 1.35: try to make if_baudrate look plausible. this updates the eth proto capability map so it records the baudrate against the different link types and their media, and then reads it when the link state changes. 1.34: (skipped)
1.1.2.9	01-Mar-2020	martin	Pull up following revision(s) (requested by thorpej in ticket #747): sys/dev/pci/if_mcx.c: revision 1.11 Use the 64-bit PCI DMA tag if available. Otherwise, we are needlessly allocating (and using) bounce buffers on sytems with >4GB of RAM. XXX pullup-9
1.1.2.8	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.1.2.7	26-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #651): sys/dev/pci/if_mcx.c: revision 1.8 sys/dev/pci/if_mcx.c: revision 1.9 - Change IFM_10G_T(10GBase-T) entry to IFM_10G_CR1(10GBASE-CR1) - Add the following entries: 20GBASE-KR2 56GBASE-R4 100GBASE-LR4 100BaseTX 1000baseT 10GBase-T Set if_baudrate.
1.1.2.6	26-Nov-2019	martin	Pull up following revision(s) (requested by jmcneill in ticket #486): sys/dev/pci/if_mcx.c: revision 1.7 Fix IFF_ALLMULTI handling.
1.1.2.5	18-Nov-2019	martin	Pull up following revision(s) (requested by nonaka in ticket #441): sys/dev/hyperv/if_hvn.c: revision 1.7 sys/dev/pci/if_mcx.c: revision 1.6 Fixed incorrect m_defrag return value check.
1.1.2.4	29-Sep-2019	martin	Pull up following revision(s) (requested by tnn in ticket #266): sys/dev/pci/if_mcx.c: revision 1.3 sys/dev/pci/if_mcx.c: revision 1.4 mcx(4): fix rx mbuf DMA overrun pullup-9 - len -> m_len
1.1.2.3	26-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #247): sys/dev/pci/if_mcx.c: revision 1.2 Use IFM_50G_SR2.
1.1.2.2	23-Sep-2019	martin	Pull up following revision(s) (requested by jmcneill in ticket #230): sys/dev/pci/files.pci: revision 1.415 distrib/sets/lists/man/mi: revision 1.1655 sys/dev/pci/if_mcx.c: revision 1.1 share/man/man4/Makefile: revision 1.683 share/man/man4/mcx.4: revision 1.1 sys/arch/evbarm/conf/GENERIC64: revision 1.107 sys/dev/pci/files.pci: revision 1.414 Add support for Mellanox 5th generation ethernet devices. This driver supports the ConnectX-4 Lx EN, ConnectX-4 EN, ConnectX-5 EN, and ConnectX-6 EN adapters. Driver ported from OpenBSD. add mcx attachment. Correct mcx comment, drop mii dependency
1.1.2.1	21-Sep-2019	martin	file if_mcx.c was added on branch netbsd-9 on 2019-09-23 07:04:40 +0000
1.9.2.1	29-Feb-2020	ad	Sync with head.
1.12.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.12.4.1	15-Mar-2020	martin	file if_mcx.c was added on branch phil-wifi on 2020-04-13 08:04:26 +0000
1.12.2.1	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.14.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.17.4.1	01-Aug-2021	thorpej	Sync with HEAD.
1.25.4.2	18-Apr-2024	martin	Pull up following revision(s) (requested by andvar in ticket #661): sys/dev/pci/if_mcx.c: revision 1.27 mcx(4): enforce full-duplex mark in mcx_media_status(), when link is up. LACP protocol requires full-duplex to be enabled for lagg(4) to work, however mcx(4) was not setting this capability making it to fail. Fixes PR kern/58124. OK'd by msaitoh@
1.25.4.1	03-Nov-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #447): sys/dev/pci/pcidevs: revision 1.1498 sys/dev/pci/if_mcx.c: revision 1.26 share/man/man4/mcx.4: revision 1.3 pcidevs: Add Mellanox ConnectX-6 Lx. mcx(4): Add Mellanox ConnectX-6 Lx from OpenBSD.
1.27.2.1	02-Aug-2025	perseant	Sync with HEAD
1.119	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.118	20-Sep-2021	jmcneill	Use coherent DMA mappings when possible.
1.117	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.116	08-May-2021	thorpej	branches: 1.116.6; Use pci_compatible_match().
1.115	24-Apr-2021	thorpej	branches: 1.115.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.114	04-Oct-2020	jakllsch	branches: 1.114.4; device_printf() instead of aprint_error_dev() where appropriate
1.113	11-May-2020	jakllsch	msk(4): ack Status BMU IRQ only if we're (apparently) caught up with it Possibility for acking an unprocessed Status BMU IRQ still exists, and remains handled by the watchdog. Nevertheless this should reduce the likelyhood that path is taken.
1.112	11-May-2020	jakllsch	msk(4): simplify interrupt return value creation This also corrects a problem where we would not claim we handled an interrupt that was ours.
1.111	11-May-2020	jakllsch	msk(4): moderate no interrupts globally The global interrupt moderator on mskc(4) is mostly redundant with the Status Unit's tx and rx interrupt coalescing for our purpouses Revisits interrupt moderation change in 1.108.
1.110	08-May-2020	jakllsch	Use sysctl_teardown() upon detach of mskc(4).
1.109	08-May-2020	jakllsch	msk(4): Simplify setting the first List Element owner-bit last. We don't actually need to do this, as the BMU won't proceed to the new descriptor until we advance the ring-end pointer on the chip, but perhaps it has value as paranoia.
1.108	30-Apr-2020	jakllsch	msk(4): actually moderate interrupts from the Yukon 2 Status BMU Previously the interrupt moderation enable register was being programmed with sk(4)-style enable bits, none of which matched the significant interrupt sources of the Yukon 2 we enable.
1.107	30-Apr-2020	jakllsch	msk(4): during attach print intrstr only once
1.106	30-Apr-2020	jakllsch	msk(4): rework rx descriptor loading to support multiple segments This paves the way to replace the driver-internal jumbo frame rx buffer with other recieve buffers (for example MCLGET/MEXTMALLOC) in the future.
1.105	29-Apr-2020	jakllsch	msk(4): Simply keep a ring of (tx) dmamaps, rather than a linked list
1.104	29-Apr-2020	jakllsch	msk(4): don't keep array of pointers to each ring descriptor With 512 descriptors each in 2 rings this saves 4KiB (LP32) or 8KiB (LP64) per interface.
1.103	28-Apr-2020	jakllsch	Use correct-semantic byteorder(9) functions to provide letoh*() in msk(4)
1.102	28-Apr-2020	jakllsch	Remove unused cargo cult include of <dev/mii/brgphyreg.h> in if_msk.c.
1.101	26-Apr-2020	jakllsch	Remove written-only sk_next member of struct (m)sk_chain.
1.100	22-Apr-2020	jdolecek	show the 'interrupting at ...' info on boot
1.99	18-Apr-2020	jakllsch	msk(4): Avoid bus_dmamap_destroy() in msk_stop() bus_dmamap_destroy() can not be executed in soft interrupt context, and msk_stop() can be called in soft interrupt context. As such, move creation and destruction of tx dmamaps to attach() and detach() functions.
1.98	04-Feb-2020	thorpej	branches: 1.98.4; Use ifmedia_fini().
1.97	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.96	01-Dec-2019	maxv	branches: 1.96.2; minor adjustments, to avoid warnings on debug builds
1.95	28-Nov-2019	maxv	localify
1.94	15-Nov-2019	msaitoh	Make Yukon EX, FE+, SUPR stable. The code is mainly taken from FreeBSD. At least, this change made my own Yukon EX machine (HP ProBook 4501s) much stable than before.
1.93	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.92	17-Oct-2019	msaitoh	Fix order of m_freem(). Found by kASan. OK'd by jdolecek and mrg.
1.91	03-Jun-2019	msaitoh	branches: 1.91.2; I think it's not required to call msk_init_yukon() in msk_attach(). Fixes PR kern/54261 reported by David shao.
1.90	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.89	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.88	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.87	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.86	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.85	27-Jan-2019	pgoyette	Merge the [pgoyette-compat] branch
1.84	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.83	08-Jan-2019	msaitoh	KNF. No functional change.
1.82	27-Dec-2018	mrg	- initialise sk_rx_hiaddr and sk_tx_hiaddr - use MSK_ADDR_LO() and MSK_ADDR_HI() - expand a debug message - remove unused softc members
1.81	08-Dec-2018	bouyer	Don't try to handle interrupts for a detached device. Should fix panic on shutdown reported by Stefan Hertenberger on current-users.
1.80	13-Nov-2018	jdolecek	whitespace fix, NFC
1.79	21-Oct-2018	jmcneill	Enable MSI support where available.
1.78	14-Sep-2018	jakllsch	msk(4): add 64-bit DMA support portions of this change set provided by mrg@
1.77	12-Sep-2018	jakllsch	if_msk: detach/reattach fixes - Fill in sc->sk_bsize upon pci_mapreg_map() so that we'll correctly bus_space_unmap() on detach. - Clear sc->sk_intrhand after interrupt disestablish, to prevent disestablishing unestablished interrupt.
1.76	23-Aug-2018	maxv	Fix buffer overflow, detected by kASan. [ 1.044878] kASan: Unauthorized Access In 0xffffffff804ec7e2: Addr 0xffffffff818a51e4 [2 bytes, read] [ 1.044878] #0 0xffffffff804ec7e2 in mskc_probe <netbsd> [ 1.044878] #1 0xffffffff80e92a77 in mapply <netbsd> [ 1.044878] #2 0xffffffff80e92e5f in config_search_loc <netbsd> [ 1.044878] #3 0xffffffff80e93fb5 in config_found_sm_loc <netbsd> [ 1.044878] #4 0xffffffff802ca9ea in pci_probe_device <netbsd> [ 1.044878] #5 0xffffffff802cad97 in pci_enumerate_bus <netbsd> [ 1.044878] #6 0xffffffff802caf00 in pcirescan <netbsd> [ 1.044878] #7 0xffffffff802cb1ee in pciattach <netbsd> [ 1.044878] #8 0xffffffff80e93e5b in config_attach_loc <netbsd> [ 1.044878] #9 0xffffffff80e93fce in config_found_sm_loc <netbsd> [ 1.044878] #10 0xffffffff80271212 in mp_pci_scan <netbsd> [ 1.044878] #11 0xffffffff8022d9ee in mainbus_attach <netbsd> [ 1.044878] #12 0xffffffff80e93e5b in config_attach_loc <netbsd> [ 1.044878] #13 0xffffffff8021e38b in cpu_configure <netbsd> [ 1.044878] #14 0xffffffff814a7068 in main <netbsd>
1.75	10-Jul-2018	jdolecek	replace code using SK_WA_4109 workaround flag by explicit conditional on affected type ids and revisions, so it's properly applied whenever new types are added; taken with FreeBSD contrary to FreeBSD, keep the IWM workaround for all XL revs - my XL rev. A3 seems to consistently have slightly (~2%) better Rx performance with the 0x04 value
1.74	10-Jul-2018	jdolecek	minor fix for earlier sync with OpenBSD if_msk.c; remove duplicit COMA handling, add also workaround code for Yukon EC Ultra and EX, and remove a write-only assignment
1.73	10-Jul-2018	jdolecek	fix comment typo
1.72	10-Jul-2018	jdolecek	add callout for re-filling RX ring - it's invoked when the RX refill code can't get enough buffers; fixes possible stall when the RXSTAT interrupt arrives while no buffers are available idea from OpenBSD if_msk.c rev. 1.131 as part of this remove the confusingly commented re-fill code from msk_jfree() and instead just invoke the callout if needed; also refill in msk_intr() only when running low instead of every RX intr; this simplifies the code, and gives small RX performance boost on my hardware (6%: ~38 MB/s -> ~40 MB/s) tested on SK-9E22 (Yukon-2 XL rev. A3) by me, and 88E8071 (Yukon-2 Extreme rev. B0) by Jogn Halfpenny; unfortunately still doesn't resolve the PR kern/53301 stall
1.71	04-Jul-2018	jdolecek	use pci_intr_establish_xname()
1.70	03-Jul-2018	jdolecek	fix detach code and do it on shutdown
1.69	03-Jul-2018	jdolecek	attach the rnd source only once even with dual-port adapters
1.68	02-Jul-2018	jdolecek	sync OpenBSD if_msk.c revs 1.66-1.79 skipping framework changes and already present changes: - Make Yukon-2 FE+ (88E8040, 88E8042) work - 88E8057 is a Yukon 2 Ultra 2, and this might make it work on a little shiny green box in Japan - Prevent null pointer dereference that could happen when we process an RX interrupt that was queued while stopping the interface. - Remove holdover XMAC II writes/reads inherited from sk as they aren't required and cause problems like the 88E8072 hard locking a system when enabling macsec bypass. - Bypass macsec on extreme/supreme based chips. Makes my 88E8072 work. - Add detach support to a few more drivers, and in others do the neccessary operations in the detach function in the right order. towards resolution of PR kern/53301 but will need more work tested by John Halfpenny on another 8040/Yukon-2 FE+ and reported improving behaviour (system hang to stall) so committing as stopgap; also tested on my SK-9E22 (Yukon-2 XL), there works without any issues both before and after
1.67	26-Jun-2018	msaitoh	branches: 1.67.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.66	21-Jun-2018	msaitoh	Simplify & white space fix. No functional change.
1.65	14-Jun-2018	msaitoh	Remove extra tabs. No functional change.
1.64	13-Jun-2018	jdolecek	fix device name in printf when not matched
1.63	13-Jun-2018	jdolecek	sync and adapt with OpenBSD if_msk.c rev. 1.43-1.65 (modulo some cosmetics, stuff we already had, or stuff which does not apply due to different frameworks): - Add support for fiber on 88E8021/22 - Avoid unnecessary resets. This should make fiber 88E8021/22 work completely - Only probe phy's at address 0; gets rid of bogus phy's on the Marvell Yukon 88E8036 - Remove an unused function msk_setfilt() which was copied from sk(4) - Make msk(4) detachable - Simplify the combination use of pci_mapreg_type()/pci_mapreg_map()
1.62	13-Jun-2018	jdolecek	revert rev. 1.29 'Only tick mii if the phy had a status change', as neither OpenBSD nor FreeBSD drivers have this conditional
1.61	12-Jun-2018	jdolecek	move SK_HASH_BITS to if_skreg.h and reuse in if_msk.c (from OpenBSD)
1.60	12-Jun-2018	jdolecek	match three more variants - DGE550T_B1, 8040T, 8042 sync more sk_type conditionals from OpenBSD
1.59	12-Jun-2018	jdolecek	fix whitespace
1.58	11-Jun-2018	jdolecek	Factual correction: actually old code also used number of octents, but set sc_if->sk_tx_ramend to invalid value, and assumed 128k buffer when SK_EPROM0 read returned zero (no code change)
1.57	11-Jun-2018	jdolecek	adopt tx/rx buffer size computation from OpenBSD rev. 1.51 to fix "random" msk failures the crucial fix is that the start/end adresses are computed and written to hw registers in number of octets (i.e. size / sizeof(u_int64_t)), as that is actually what the hardware expects fixes PR kern/36454 (which had fix which highlighted the trouble), and seems likely to also fix kern/35711 and port-i386/42514
1.56	09-Jun-2018	jdolecek	port basic match support for several more Yukon variants from OpenBSD, together with some basic conditionals I've found in that driver; only compile-tested, so very likely needs further work to actually work adresses also PR kern/43507
1.55	20-Oct-2017	christos	branches: 1.55.2; update entry for Marvel Yukon 8058, and fix URL for pci ids (Rocky Hotas)
1.54	15-Dec-2016	ozaki-r	branches: 1.54.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.53	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.52	06-Nov-2016	christos	msk_init is too heavyweight and loses link. Provide a specialized way to enter and exit promiscuous mode so that we don't lose seconds of traffic each time we run tcpdump.
1.51	10-Jun-2016	ozaki-r	branches: 1.51.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.50	01-Jun-2016	pgoyette	Commit patch from t-hash in PR kern/48309
1.49	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.48	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.47	10-Oct-2014	christos	branches: 1.47.2; PR/49270: Hauke Fath: Samsung NC10 Marvell Yukon ethernet not supported
1.46	10-Aug-2014	tls	Merge tls-earlyentropy branch into HEAD.
1.45	29-Mar-2014	christos	branches: 1.45.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.44	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.43	30-Mar-2013	christos	branches: 1.43.4; remove trailing whitespace
1.42	04-Aug-2012	riastradh	branches: 1.42.2; Fix null pointer dereference in msk_rxeof. This triggers when processing an RX interrupt that was queued while stopping the interface, which caused my machine to panic last night. In this case, just drop the packet. From OpenBSD's if_msk.c rev. 1.71.
1.41	22-Jul-2012	matt	Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.40	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.39	02-Feb-2012	tls	branches: 1.39.2; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.38	24-Apr-2011	plunky	branches: 1.38.4; 1.38.8; use static inline for msk_rxvalid(), it is not referenced elsewhere (avoids C99 vs GNU semantics for external inline functions)
1.37	16-Apr-2011	jakllsch	Make previous compile.
1.36	16-Apr-2011	jakllsch	Ensure PCI_COMMAND_MASTER_ENABLE.
1.35	05-Apr-2010	joerg	branches: 1.35.2; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.34	30-Mar-2010	tnn	yinterrupt -> interrupt
1.33	24-Feb-2010	dyoung	branches: 1.33.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.32	19-Jan-2010	pooka	branches: 1.32.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.31	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.30	24-Dec-2009	christos	- split the sk_dev - don't program the interrupt mitigation to 0, only to set it to 1000 later. - don't be verbose when setting the interrupt mitigation during reset, it messes up the attach line and does not convey useful information.
1.29	14-Dec-2009	matt	Only tick mii if the phy had a status change.
1.28	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.27	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.26	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.25	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.24	18-Mar-2009	cegger	bzero -> memset
1.23	18-Mar-2009	cegger	bcmp -> memcmp
1.22	18-Nov-2008	chris	branches: 1.22.4; Fix Yukon EC Ultra cold power up issue. For the EC Ultra it is necessary to update some extra registers during reset. Without doing so causes the system to hang at boot. The only workaround I found was to PXE boot before booting into NetBSD. This change is based on the code from FreeBSD's if_msk.c. Specifically the msk_phy_power function. Also add an splnet/splx across mii_tick. This matches most other network drivers. Change posted for review on 3rd Oct 2008 to tech-net. No feedback received.
1.21	20-Jun-2008	cube	branches: 1.21.2; 1.21.4; 1.21.6; Apply a similar patch as what was just applied to sk(4): > Use a mutex instead of splvm() to protect the list of jubo-ready mbufs, as > done with nfe(4) a while ago. > > Issue reported by Gary Duzan, who kindly fixed the patch I had sent him. Lars Nordlund noted that such a change made things a lot better with his msk(4).
1.20	28-May-2008	joerg	branches: 1.20.2; Add PMF hooks. Fixes PR 38675.
1.19	27-May-2008	dyoung	For all controllers but type SK_YUKON_FE, set the jumbo frames capability, ETHERCAP_JUMBO_MTU. Rely on ether_ioctl() to enforce the MTU range implied by the ethernet capabilities. Should fix kern/38748. While I'm here, use some standard ethernet constants instead of msk-specific constants.
1.18	10-Apr-2008	cegger	branches: 1.18.2; 1.18.4; 1.18.6; use aprint_*_dev and device_xname
1.17	28-Mar-2008	kiyohara	The status buffer must synchronize. If your CPU is cache writeback, synchronous processing is necessary. However, the size of a status buffer will overwrite the memory while synchronously processing it because it is very small. We clear the flag SK_Y2_STOPC_OWN at more late.
1.16	07-Feb-2008	dyoung	branches: 1.16.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.15	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.14	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.13	15-Nov-2007	manu	branches: 1.13.2; 1.13.4; 1.13.6; It's 88C055, not 88CO55.
1.12	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.11	22-Oct-2007	manu	branches: 1.11.2; Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.10	19-Jul-2007	dsl	branches: 1.10.4; 1.10.6; 1.10.10; 1.10.12; include sys/cdefs.h
1.9	09-Jul-2007	ad	branches: 1.9.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.8	04-Mar-2007	christos	branches: 1.8.2; 1.8.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.7	31-Jan-2007	msaitoh	branches: 1.7.2; Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.6	30-Jan-2007	msaitoh	apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs
1.5	30-Jan-2007	msaitoh	sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41
1.4	05-Jan-2007	msaitoh	Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set.
1.3	16-Nov-2006	christos	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; __unused removal on arguments; approved by core.
1.2	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.1	09-Sep-2006	riz	branches: 1.1.2; 1.1.4; 1.1.6; Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.1.6.2	10-Dec-2006	yamt	sync with head.
1.1.6.1	22-Oct-2006	yamt	sync with head
1.1.4.1	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/if_skreg.h: revision 1.8 sys/dev/pci/if_msk.c: revision 1.1 sys/dev/pci/files.pci: revision 1.268 sys/dev/pci/if_mskvar.h: revision 1.1 Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.1.2.2	14-Sep-2006	yamt	sync with head.
1.1.2.1	09-Sep-2006	yamt	file if_msk.c was added on branch yamt-pdpolicy on 2006-09-14 12:31:33 +0000
1.3.8.2	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.3.8.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.3.6.8	11-Feb-2008	yamt	sync with head.
1.3.6.7	21-Jan-2008	yamt	sync with head
1.3.6.6	15-Nov-2007	yamt	sync with head.
1.3.6.5	27-Oct-2007	yamt	sync with head.
1.3.6.4	03-Sep-2007	yamt	sync with head.
1.3.6.3	26-Feb-2007	yamt	sync with head.
1.3.6.2	30-Dec-2006	yamt	sync with head.
1.3.6.1	16-Nov-2006	yamt	file if_msk.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:45 +0000
1.3.4.4	15-Nov-2007	pavel	Backout accidental commit to the release branch.
1.3.4.3	15-Nov-2007	manu	It's not 88CO55, it's 88C055
1.3.4.2	04-Nov-2007	xtraeme	Pull up following revision(s) (requested by manu in ticket #954): sys/dev/pci/pcidevs_data.h: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/if_msk.c: revision 1.11 sys/dev/pci/pcidevs: revision 1.907 Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.3.4.1	26-Aug-2007	liamjfoy	Pull up following revision(s) (requested by riz in ticket #816): sys/dev/pci/if_skreg.h: revision 1.10 sys/dev/pci/if_skreg.h: revision 1.11 sys/dev/pci/pcidevs: revision 1.866 sys/dev/pci/if_skreg.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.36 sys/dev/pci/if_msk.c: revision 1.4 sys/dev/pci/if_msk.c: revision 1.5 sys/dev/pci/if_msk.c: revision 1.6 sys/dev/pci/if_msk.c: revision 1.7 sys/dev/pci/if_mskvar.h: revision 1.2 sys/dev/pci/if_mskvar.h: revision 1.3 Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set. add some msk devices sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41 sync with if_skreg.h rev. 1.41 need more work? apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs fix bit definitions for the RX FIFO Flush mode Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.3.2.4	01-Feb-2007	ad	Sync with head.
1.3.2.3	12-Jan-2007	ad	Sync with head.
1.3.2.2	18-Nov-2006	ad	Sync with head.
1.3.2.1	16-Nov-2006	ad	file if_msk.c was added on branch newlock2 on 2006-11-18 21:34:30 +0000
1.7.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.8.4.1	11-Jul-2007	mjf	Sync with head.
1.8.2.3	01-Sep-2007	ad	Update for pool_cache API changes.
1.8.2.2	20-Aug-2007	ad	Sync with HEAD.
1.8.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.9.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.10.12.2	19-Jul-2007	dsl	include sys/cdefs.h
1.10.12.1	19-Jul-2007	dsl	file if_msk.c was added on branch matt-mips64 on 2007-07-19 22:04:23 +0000
1.10.10.3	18-Nov-2007	bouyer	Sync with HEAD
1.10.10.2	13-Nov-2007	bouyer	Sync with HEAD
1.10.10.1	25-Oct-2007	bouyer	Sync with HEAD.
1.10.6.4	23-Mar-2008	matt	sync with HEAD
1.10.6.3	09-Jan-2008	matt	sync with HEAD
1.10.6.2	08-Nov-2007	matt	sync with -HEAD
1.10.6.1	06-Nov-2007	matt	sync with HEAD
1.10.4.3	21-Nov-2007	joerg	Sync with HEAD.
1.10.4.2	11-Nov-2007	joerg	Sync with HEAD.
1.10.4.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.11.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.11.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.13.6.2	20-Jan-2008	bouyer	Sync with HEAD
1.13.6.1	13-Dec-2007	bouyer	Sync with HEAD
1.13.4.1	11-Dec-2007	yamt	sync with head.
1.13.2.1	26-Dec-2007	ad	Sync with head.
1.16.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.16.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.16.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.16.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.18.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.18.4.5	11-Aug-2010	yamt	sync with head.
1.18.4.4	11-Mar-2010	yamt	sync with head
1.18.4.3	16-Sep-2009	yamt	sync with head
1.18.4.2	16-May-2009	yamt	sync with head
1.18.4.1	04-May-2009	yamt	sync with head.
1.18.2.1	04-Jun-2008	yamt	sync with head
1.20.2.1	27-Jun-2008	simonb	Sync with head.
1.21.6.1	20-Nov-2008	snj	branches: 1.21.6.1.4; Pull up following revision(s) (requested by chris in ticket #84): sys/dev/pci/if_skreg.h: revision 1.13 sys/dev/pci/if_msk.c: revision 1.22 Fix Yukon EC Ultra cold power up issue. For the EC Ultra it is necessary to update some extra registers during reset. Without doing so causes the system to hang at boot. The only workaround I found was to PXE boot before booting into NetBSD. This change is based on the code from FreeBSD's if_msk.c. Specifically the msk_phy_power function. Also add an splnet/splx across mii_tick. This matches most other network drivers. Change posted for review on 3rd Oct 2008 to tech-net. No feedback received.
1.21.6.1.4.1	09-Nov-2009	cliff	- in msk_tick, only call mii_tick() if status has changed
1.21.4.2	28-Apr-2009	skrll	Sync with HEAD.
1.21.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.21.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.22.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.32.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.33.2.3	31-May-2011	rmind	sync with head
1.33.2.2	21-Apr-2011	rmind	sync with head
1.33.2.1	30-May-2010	rmind	sync with head
1.35.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.38.8.1	18-Feb-2012	mrg	merge to -current.
1.38.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.38.4.2	30-Oct-2012	yamt	sync with head
1.38.4.1	17-Apr-2012	yamt	sync with head
1.39.2.1	12-Aug-2012	martin	Pull up following revision(s) (requested by riastradh in ticket #480): sys/dev/pci/if_msk.c: revision 1.42 Fix null pointer dereference in msk_rxeof. This triggers when processing an RX interrupt that was queued while stopping the interface, which caused my machine to panic last night. In this case, just drop the packet. From OpenBSD's if_msk.c rev. 1.71.
1.42.2.3	03-Dec-2017	jdolecek	update from HEAD
1.42.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.42.2.1	23-Jun-2013	tls	resync from head
1.43.4.1	18-May-2014	rmind	sync with head
1.45.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.47.2.5	05-Feb-2017	skrll	Sync with HEAD
1.47.2.4	05-Dec-2016	skrll	Sync with HEAD
1.47.2.3	09-Jul-2016	skrll	Sync with HEAD
1.47.2.2	19-Mar-2016	skrll	Sync with HEAD
1.47.2.1	06-Jun-2015	skrll	Sync with HEAD
1.51.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.54.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.54.8.1	08-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #616): sys/dev/pci/if_msk.c: revision 1.55 sys/dev/pci/pcidevs: revision 1.1299 update entry for Marvel Yukon 8058, and fix URL for pci ids (Rocky Hotas)
1.55.2.8	26-Jan-2019	pgoyette	Sync with HEAD
1.55.2.7	18-Jan-2019	pgoyette	Synch with HEAD
1.55.2.6	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.55.2.5	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.55.2.4	30-Sep-2018	pgoyette	Ssync with HEAD
1.55.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.55.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.55.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.67.2.4	21-Apr-2020	martin	Sync with HEAD
1.67.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.67.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.67.2.1	10-Jun-2019	christos	Sync with HEAD
1.91.2.2	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #450): sys/dev/pci/if_msk.c: revision 1.94 sys/dev/pci/if_skreg.h: revision 1.27 Make Yukon EX, FE+, SUPR stable. The code is mainly taken from FreeBSD. At least, this change made my own Yukon EX machine (HP ProBook 4501s) much stable than before.
1.91.2.1	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #374): sys/dev/pci/if_msk.c: revision 1.92 Fix order of m_freem(). Found by kASan. OK'd by jdolecek and mrg.
1.96.2.1	29-Feb-2020	ad	Sync with head.
1.98.4.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.98.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.114.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.115.2.1	13-May-2021	thorpej	Sync with HEAD.
1.116.6.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.25	29-Apr-2020	jakllsch	msk(4): Simply keep a ring of (tx) dmamaps, rather than a linked list
1.24	29-Apr-2020	jakllsch	msk(4): don't keep array of pointers to each ring descriptor With 512 descriptors each in 2 rings this saves 4KiB (LP32) or 8KiB (LP64) per interface.
1.23	26-Apr-2020	jakllsch	Remove written-only sk_next member of struct (m)sk_chain.
1.22	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.21	27-Dec-2018	mrg	- initialise sk_rx_hiaddr and sk_tx_hiaddr - use MSK_ADDR_LO() and MSK_ADDR_HI() - expand a debug message - remove unused softc members
1.20	21-Oct-2018	jmcneill	Enable MSI support where available.
1.19	14-Sep-2018	jakllsch	msk(4): add 64-bit DMA support portions of this change set provided by mrg@
1.18	10-Jul-2018	jdolecek	add the RX refill callout (forgot to commit with if_msk.c change)
1.17	03-Jul-2018	jdolecek	fix detach code and do it on shutdown
1.16	03-Jul-2018	jdolecek	attach the rnd source only once even with dual-port adapters
1.15	02-Jul-2018	jdolecek	sync OpenBSD if_msk.c revs 1.66-1.79 skipping framework changes and already present changes: - Make Yukon-2 FE+ (88E8040, 88E8042) work - 88E8057 is a Yukon 2 Ultra 2, and this might make it work on a little shiny green box in Japan - Prevent null pointer dereference that could happen when we process an RX interrupt that was queued while stopping the interface. - Remove holdover XMAC II writes/reads inherited from sk as they aren't required and cause problems like the 88E8072 hard locking a system when enabling macsec bypass. - Bypass macsec on extreme/supreme based chips. Makes my 88E8072 work. - Add detach support to a few more drivers, and in others do the neccessary operations in the detach function in the right order. towards resolution of PR kern/53301 but will need more work tested by John Halfpenny on another 8040/Yukon-2 FE+ and reported improving behaviour (system hang to stall) so committing as stopgap; also tested on my SK-9E22 (Yukon-2 XL), there works without any issues both before and after
1.14	13-Jun-2018	jdolecek	branches: 1.14.2; sync and adapt with OpenBSD if_msk.c rev. 1.43-1.65 (modulo some cosmetics, stuff we already had, or stuff which does not apply due to different frameworks): - Add support for fiber on 88E8021/22 - Avoid unnecessary resets. This should make fiber 88E8021/22 work completely - Only probe phy's at address 0; gets rid of bogus phy's on the Marvell Yukon 88E8036 - Remove an unused function msk_setfilt() which was copied from sk(4) - Make msk(4) detachable - Simplify the combination use of pci_mapreg_type()/pci_mapreg_map()
1.13	11-Jun-2018	jdolecek	adopt tx/rx buffer size computation from OpenBSD rev. 1.51 to fix "random" msk failures the crucial fix is that the start/end adresses are computed and written to hw registers in number of octets (i.e. size / sizeof(u_int64_t)), as that is actually what the hardware expects fixes PR kern/36454 (which had fix which highlighted the trouble), and seems likely to also fix kern/35711 and port-i386/42514
1.12	13-Apr-2015	riastradh	branches: 1.12.16; Convert sys/dev to use <sys/rndsource.h>.
1.11	14-Oct-2012	msaitoh	branches: 1.11.14; struct device * -> device_t change.
1.10	02-Feb-2012	tls	branches: 1.10.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.9	19-Nov-2011	tls	branches: 1.9.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.8	24-Dec-2009	christos	branches: 1.8.12; - split the sk_dev - don't program the interrupt mitigation to 0, only to set it to 1000 later. - don't be verbose when setting the interrupt mitigation during reset, it messes up the attach line and does not convey useful information.
1.7	20-Jun-2008	cube	Apply a similar patch as what was just applied to sk(4): > Use a mutex instead of splvm() to protect the list of jubo-ready mbufs, as > done with nfe(4) a while ago. > > Issue reported by Gary Duzan, who kindly fixed the patch I had sent him. Lars Nordlund noted that such a change made things a lot better with his msk(4).
1.6	28-Apr-2008	martin	branches: 1.6.2; 1.6.4; Remove clause 3 and 4 from TNF licenses
1.5	28-Mar-2008	kiyohara	branches: 1.5.2; 1.5.4; The status buffer must synchronize. If your CPU is cache writeback, synchronous processing is necessary. However, the size of a status buffer will overwrite the memory while synchronously processing it because it is very small. We clear the flag SK_Y2_STOPC_OWN at more late.
1.4	04-Mar-2007	christos	branches: 1.4.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	30-Jan-2007	msaitoh	branches: 1.3.2; apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs
1.2	30-Jan-2007	msaitoh	sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41
1.1	09-Sep-2006	riz	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; 1.1.12; 1.1.14; Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.1.14.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.1.12.4	03-Sep-2007	yamt	sync with head.
1.1.12.3	26-Feb-2007	yamt	sync with head.
1.1.12.2	30-Dec-2006	yamt	sync with head.
1.1.12.1	09-Sep-2006	yamt	file if_mskvar.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:45 +0000
1.1.10.1	26-Aug-2007	liamjfoy	Pull up following revision(s) (requested by riz in ticket #816): sys/dev/pci/if_skreg.h: revision 1.10 sys/dev/pci/if_skreg.h: revision 1.11 sys/dev/pci/pcidevs: revision 1.866 sys/dev/pci/if_skreg.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.36 sys/dev/pci/if_msk.c: revision 1.4 sys/dev/pci/if_msk.c: revision 1.5 sys/dev/pci/if_msk.c: revision 1.6 sys/dev/pci/if_msk.c: revision 1.7 sys/dev/pci/if_mskvar.h: revision 1.2 sys/dev/pci/if_mskvar.h: revision 1.3 Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set. add some msk devices sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41 sync with if_skreg.h rev. 1.41 need more work? apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs fix bit definitions for the RX FIFO Flush mode Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.1.8.2	01-Feb-2007	ad	Sync with head.
1.1.8.1	18-Nov-2006	ad	Sync with head.
1.1.4.1	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/if_skreg.h: revision 1.8 sys/dev/pci/if_msk.c: revision 1.1 sys/dev/pci/files.pci: revision 1.268 sys/dev/pci/if_mskvar.h: revision 1.1 Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.1.2.2	14-Sep-2006	yamt	sync with head.
1.1.2.1	09-Sep-2006	yamt	file if_mskvar.h was added on branch yamt-pdpolicy on 2006-09-14 12:31:33 +0000
1.3.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.4.36.3	29-Jun-2008	mjf	Sync with HEAD.
1.4.36.2	02-Jun-2008	mjf	Sync with HEAD.
1.4.36.1	03-Apr-2008	mjf	Sync with HEAD.
1.5.4.3	11-Mar-2010	yamt	sync with head
1.5.4.2	04-May-2009	yamt	sync with head.
1.5.4.1	16-May-2008	yamt	sync with head.
1.5.2.1	18-May-2008	yamt	sync with head.
1.6.4.1	27-Jun-2008	simonb	Sync with head.
1.6.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.8.12.2	30-Oct-2012	yamt	sync with head
1.8.12.1	17-Apr-2012	yamt	sync with head
1.9.2.1	18-Feb-2012	mrg	merge to -current.
1.10.6.2	03-Dec-2017	jdolecek	update from HEAD
1.10.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.11.14.1	06-Jun-2015	skrll	Sync with HEAD
1.12.16.5	18-Jan-2019	pgoyette	Synch with HEAD
1.12.16.4	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.12.16.3	30-Sep-2018	pgoyette	Ssync with HEAD
1.12.16.2	28-Jul-2018	pgoyette	Sync with HEAD
1.12.16.1	25-Jun-2018	pgoyette	Sync with HEAD
1.14.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.14.2.1	10-Jun-2019	christos	Sync with HEAD
1.22	08-May-2021	thorpej	Use pci_compatible_match().
1.21	09-Dec-2018	jdolecek	branches: 1.21.16; use pci_intr_establish_xname() everywhere
1.20	07-Jul-2016	msaitoh	branches: 1.20.16; 1.20.18; KNF. Remove extra spaces. No functional change.
1.19	29-Mar-2014	christos	branches: 1.19.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	27-Oct-2012	chs	branches: 1.18.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.17	30-Jan-2012	drochner	branches: 1.17.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.16	26-Jul-2011	dyoung	branches: 1.16.2; 1.16.6; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.15	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.14	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.13	28-Apr-2008	martin	branches: 1.13.14; Remove clause 3 and 4 from TNF licenses
1.12	10-Apr-2008	cegger	branches: 1.12.2; 1.12.4; use aprint_*_dev and device_xname
1.11	12-Mar-2008	dyoung	Use device_t and accessors.
1.10	19-Oct-2007	ad	branches: 1.10.12; 1.10.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.9	16-Nov-2006	christos	branches: 1.9.8; 1.9.22; 1.9.24; 1.9.28; __unused removal on arguments; approved by core.
1.8	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.7	11-Dec-2005	christos	branches: 1.7.20; 1.7.22; merge ktrace-lwp.
1.6	27-Feb-2005	perry	branches: 1.6.4; nuke trailing whitespace
1.5	21-Aug-2004	thorpej	branches: 1.5.4; 1.5.6; Use ANSI function decls and make use of static.
1.4	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.3	15-Oct-2003	simonb	Fix obvious typos in a call to pci_mapreg_map().
1.2	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.1	07-Nov-2002	martin	branches: 1.1.2; 1.1.8; Add a driver for the Myson Technology MTD803 3-in-1 Fast Ethernet Controller, provided by Peter Bex in PR 18675.
1.1.8.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.8.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.8.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.8.2	25-Aug-2004	skrll	Sync with HEAD.
1.1.8.1	03-Aug-2004	skrll	Sync with HEAD
1.1.2.2	11-Nov-2002	nathanw	Catch up to -current
1.1.2.1	07-Nov-2002	nathanw	file if_mtd_pci.c was added on branch nathanw_sa on 2002-11-11 22:11:06 +0000
1.5.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.5.4.1	29-Apr-2005	kent	sync with -current
1.6.4.3	17-Mar-2008	yamt	sync with head.
1.6.4.2	27-Oct-2007	yamt	sync with head.
1.6.4.1	30-Dec-2006	yamt	sync with head.
1.7.22.2	10-Dec-2006	yamt	sync with head.
1.7.22.1	22-Oct-2006	yamt	sync with head
1.7.20.1	18-Nov-2006	ad	Sync with head.
1.9.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.9.24.2	23-Mar-2008	matt	sync with HEAD
1.9.24.1	06-Nov-2007	matt	sync with HEAD
1.9.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.9.8.1	23-Oct-2007	ad	Sync with head.
1.10.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.10.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.12.1	24-Mar-2008	keiichi	sync with head.
1.12.4.3	11-Mar-2010	yamt	sync with head
1.12.4.2	16-May-2009	yamt	sync with head
1.12.4.1	16-May-2008	yamt	sync with head.
1.12.2.1	18-May-2008	yamt	sync with head.
1.13.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.6.1	18-Feb-2012	mrg	merge to -current.
1.16.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.16.2.2	30-Oct-2012	yamt	sync with head
1.16.2.1	17-Apr-2012	yamt	sync with head
1.17.6.3	03-Dec-2017	jdolecek	update from HEAD
1.17.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.17.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.18.2.1	18-May-2014	rmind	sync with head
1.19.6.1	09-Jul-2016	skrll	Sync with HEAD
1.20.18.1	10-Jun-2019	christos	Sync with HEAD
1.20.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.21.16.1	13-May-2021	thorpej	Sync with HEAD.
1.40	27-Jan-2019	pgoyette	Merge the [pgoyette-compat] branch
1.39	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.38	15-Jul-2018	maxv	Retire ipkdb entirely. The option was removed from the config files yesterday. ok kamil christos
1.37	29-Mar-2014	christos	branches: 1.37.28; 1.37.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.36	26-Jul-2011	dyoung	branches: 1.36.2; 1.36.12; 1.36.16; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.35	17-May-2011	dyoung	PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.34	06-May-2009	cegger	branches: 1.34.4; 1.34.6; struct device * -> device_t, no functional changes intended.
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	28-Apr-2008	martin	branches: 1.32.14; Remove clause 3 and 4 from TNF licenses
1.31	12-Mar-2008	cube	branches: 1.31.2; 1.31.4; Split device_t and softc for the NE2000 Ethernet chip and all its variants and attachments. Use device_t accessors, correct types, and ANSIfy when appropriate.
1.30	19-Oct-2007	ad	branches: 1.30.12; 1.30.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.29	16-Nov-2006	christos	branches: 1.29.8; 1.29.22; 1.29.24; 1.29.28; __unused removal on arguments; approved by core.
1.28	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.27	11-Dec-2005	christos	branches: 1.27.20; 1.27.22; merge ktrace-lwp.
1.26	21-Aug-2004	thorpej	branches: 1.26.12; Use ANSI function decls and make use of static.
1.25	13-Feb-2004	wiz	RealTek -> Realtek.
1.24	02-Oct-2002	thorpej	branches: 1.24.6; Add trailing ; to CFATTACH_DECL.
1.23	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.22	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.21	13-Nov-2001	lukem	add RCSID
1.20	08-Jul-2001	thorpej	branches: 1.20.2; Remove unnecessary include files. From Onno van der Linden.
1.19	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.18	12-Feb-2001	thorpej	branches: 1.18.2; Adjust the way that media is initialized on DP8390-compatible chips. The dp8390_softc now has media_init and media_fini function pointers that do the work.
1.17	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.16	22-Mar-2000	ws	Make IPKDB working again. Add support for i386 debugging and pci-based ne2000 boards.
1.15	06-Mar-2000	mark	Support the W89C940F alternative ID.
1.14	09-Nov-1998	thorpej	branches: 1.14.10; Const poisoning.
1.13	31-Oct-1998	thorpej	Use the rtl80x9 common code for media selection.
1.12	31-Oct-1998	thorpej	Move the RTL8029 register definitions to <dev/ic/rtl80x9reg.h>, and add the registers/bits present in the RTL8019 (ISA version).
1.11	28-Oct-1998	thorpej	Add support for media selection to the PCI front-end of the NE2000 driver. Currently supports: RealTek 8029: 10base2, 10baseT, 10baseT-FDX, auto detect (but not FDX).
1.10	27-Oct-1998	thorpej	Add register definitions for PCI NE2000-compatible extenstions, allowing control over e.g. full-duplex mode, media, etc. Currently, only extensions for the RealTek 8029 are defined.
1.9	27-Oct-1998	thorpej	Cosmetic changes in product lookup, to match other drivers.
1.8	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.7	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.6	05-May-1998	thorpej	Add 5 more NE-2000 compatible PCI Ethernet interfaces. XXX Note that some of these may need revision in pcidevs.
1.5	10-Apr-1998	thorpej	Add support for the VIA Technologies VT86C926, from Kazuki Sakamoto, PR #5256.
1.4	27-Oct-1997	thorpej	Oops, missed one in previous.
1.3	27-Oct-1997	thorpej	Update for new pcidevs entry for Winbond NE2000-compatible Ethernet ASIC.
1.2	14-Oct-1997	thorpej	branches: 1.2.2; Pulldown from marc-pcmcia branch: PCI front-end for the "ne" driver.
1.1	14-Oct-1997	thorpej	branches: 1.1.2; file if_ne_pci.c was initially added on branch marc-pcmcia.
1.1.2.2	14-Oct-1997	thorpej	Mark this interface as enabled right from the get-go.
1.1.2.1	14-Oct-1997	thorpej	Split off NE2000 support from the ISA "ed" driver, and create a new "ne" driver that deals with NE2000s and compatibles on ISA, PCI, and PCMCIA busses. This driver uses the generic DP8390 core chipset back-end.
1.2.2.1	27-Oct-1997	thorpej	Update from trunk: Update for new Winbond NE2000-compatible Ethernet ASIC description.
1.14.10.3	12-Mar-2001	bouyer	Sync with HEAD.
1.14.10.2	05-Jan-2001	bouyer	Sync with HEAD
1.14.10.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.18.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.18.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.18.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.18.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.20.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.20.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.24.6.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.24.6.3	18-Sep-2004	skrll	Sync with HEAD.
1.24.6.2	25-Aug-2004	skrll	Sync with HEAD.
1.24.6.1	03-Aug-2004	skrll	Sync with HEAD
1.26.12.3	17-Mar-2008	yamt	sync with head.
1.26.12.2	27-Oct-2007	yamt	sync with head.
1.26.12.1	30-Dec-2006	yamt	sync with head.
1.27.22.2	10-Dec-2006	yamt	sync with head.
1.27.22.1	22-Oct-2006	yamt	sync with head
1.27.20.1	18-Nov-2006	ad	Sync with head.
1.29.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.29.24.2	23-Mar-2008	matt	sync with HEAD
1.29.24.1	06-Nov-2007	matt	sync with HEAD
1.29.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.29.8.1	23-Oct-2007	ad	Sync with head.
1.30.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.30.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.30.12.1	24-Mar-2008	keiichi	sync with head.
1.31.4.2	16-May-2009	yamt	sync with head
1.31.4.1	16-May-2008	yamt	sync with head.
1.31.2.1	18-May-2008	yamt	sync with head.
1.32.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.34.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.34.4.1	31-May-2011	rmind	sync with head
1.36.16.1	18-May-2014	rmind	sync with head
1.36.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.36.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.30.1	10-Jun-2019	christos	Sync with HEAD
1.37.28.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.37.28.1	28-Jul-2018	pgoyette	Sync with HEAD
1.2	31-Oct-1998	thorpej	Move the RTL8029 register definitions to <dev/ic/rtl80x9reg.h>, and add the registers/bits present in the RTL8019 (ISA version).
1.1	27-Oct-1998	thorpej	Add register definitions for PCI NE2000-compatible extenstions, allowing control over e.g. full-duplex mode, media, etc. Currently, only extensions for the RealTek 8029 are defined.
1.81	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.80	22-Jan-2022	martin	branches: 1.80.10; Explicitly cast a __MASK(40) to bus_addr_t (for 32bit bus_addr_t archs)
1.79	22-Jan-2022	skrll	Ensure bus_dmatag_subregion is called with an inclusive max_addr everywhere.
1.78	13-Mar-2020	msaitoh	Improve error check: - We check PHY register read error correctly (timeout and NFE_PHY_ERROR), so don't check NFE_PHY_DATA register's value with 0xffffffff or 0. At least, some registers may have 0. - Check NFE_PHY_ERROR bit in nfe_miibus_writereg(). - Improve debug printf
1.77	08-Mar-2020	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.76	01-Mar-2020	thorpej	Add explicit casts in bus_dmatag_subregion() calls to avoid overflow / truncation warnings on 32-bit platforms.
1.75	01-Mar-2020	thorpej	- Range-limit our DMA tag to 40-bit. - Eliminate __LP64__ ifdefs.
1.74	04-Feb-2020	thorpej	Use ifmedia_fini().
1.73	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.72	13-Sep-2019	msaitoh	branches: 1.72.2; if_flags is neither int nor short. It's unsigned short.
1.71	09-Jul-2019	msaitoh	branches: 1.71.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.70	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.69	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.68	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.67	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.66	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.65	26-Jun-2018	msaitoh	branches: 1.65.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.64	26-Sep-2017	knakahara	branches: 1.64.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.63	15-Dec-2016	ozaki-r	branches: 1.63.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.62	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.61	10-Jun-2016	ozaki-r	branches: 1.61.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.60	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.59	29-Mar-2014	christos	branches: 1.59.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.58	30-Mar-2013	christos	branches: 1.58.4; remove trailing whitespace
1.57	23-Sep-2012	chs	use 64-bit DMA where possible.
1.56	22-Jul-2012	matt	branches: 1.56.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.55	30-Jan-2012	drochner	branches: 1.55.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.54	23-Jan-2011	tsutsui	branches: 1.54.4; 1.54.8; Pull the following fix from OpenBSD: http://www.openbsd.org/cgi-bin/cvsweb/src/sys/dev/pci/if_nfe.c#rev1.97 >> Some nfe(4)/rlphy(4) combos don't work, because the PHY responds to all >> addresses on the mii bus. As a countereasure, only attach the first PHY we >> encounter. It is very unlikely we're going to ever see nfe(4) with multiple >> PHYs. The same is probably true for any modern NIC. Should suppress bogus ukphy30 noted in PR kern/42950. No bad side effect on my nfe(4) which didn't get ghosts: --- nfe0 at pci0 dev 5 function 0: NVIDIA nForce3 Ethernet #4 (rev. 0xa2) : rlphy0 at nfe0 phy 1: RTL8201L 10/100 media interface, rev. 1 rlphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto
1.53	03-Nov-2010	jakllsch	branches: 1.53.2; 1.53.4; Make nfe(4) detachable.
1.52	02-Nov-2010	jakllsch	Explicitly set PCI_COMMAND_MASTER_ENABLE.
1.51	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.50	24-Feb-2010	dyoung	branches: 1.50.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.49	19-Jan-2010	pooka	branches: 1.49.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.48	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.47	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.46	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.45	14-Jun-2009	cegger	use __arraycount
1.44	18-Mar-2009	cegger	bcmp -> memcmp
1.43	01-Mar-2009	cegger	bzero -> memset per request from martin@ bcopy -> memcpy XXX Would changing bcmp to memcpy be a functional change regarding its return value?
1.42	01-Mar-2009	cegger	when nfe_attach() fails, free all resources including interrupt map, interrupt handler and dma maps. Also prevent call to bus_dmamap_destroy() when bus_dmamap_create() failed. The nfe_free_* routines assume, the map pointers are NULL but they are actually undefined. Make the assumption true by making them NULL in the error path of bus_dmamap_create(). All together, this fixes crashes when nfe_attach() fails. ok martin@
1.41	01-Mar-2009	cegger	MCP79 supports JUMBO frames. Information from Linux driver. Patch presented on current-users@ http://mail-index.netbsd.org/current-users/2009/02/11/msg008037.html
1.40	12-Feb-2009	cegger	flow-control support for MCP65 chip. XXX other chips may also have flow-control, but only on MCP65 tested. ok cube@
1.39	18-Jan-2009	cegger	branches: 1.39.2; Bracket IRQ masking around nfe_intr() in nfe_init() rather inside nfe_intr(). Proposed by dyoung@ in PR 40345. Should fix PR 40345.
1.38	16-Dec-2008	christos	replace bitmask_snprintf(9) with snprintb(3)
1.37	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.36	17-Jun-2008	cube	branches: 1.36.2; 1.36.4; 1.36.6; Use MUTEX_DEFAULT to initialise the mutex, per martin@'s input.
1.35	25-May-2008	jmcneill	branches: 1.35.2; Based on changes to nfe_attach in r1.27, wake newer chips from powerdown mode on resume as well as at attach time.
1.34	20-Apr-2008	cube	branches: 1.34.2; 1.34.4; Restore jumbo frame support and protect access to the DMA space for received frames with a mutex. Tested by Paul Goyette, Martin Husemann and myself.
1.33	17-Apr-2008	christos	branches: 1.33.2; join line for symmetry.
1.32	17-Apr-2008	xtraeme	Make this build again... HI XTOS.
1.31	17-Apr-2008	christos	sync with openbsd; fixes observed corruption issues.
1.30	26-Mar-2008	cube	Split device_t and softc, and related cosmetic changes.
1.29	24-Feb-2008	isaki	Minor style fix.
1.28	07-Feb-2008	dyoung	branches: 1.28.2; 1.28.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.27	26-Jan-2008	tsutsui	Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.26	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.25	17-Dec-2007	tsutsui	Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576.
1.24	09-Dec-2007	jmcneill	branches: 1.24.2; Merge jmcneill-pm branch.
1.23	14-Nov-2007	xtraeme	branches: 1.23.2; 1.23.4; Add the MCP{67,73} products on nfe_attach().
1.22	14-Nov-2007	xtraeme	Attach to NVIDIA MCP67/73 Ethernet controllers.
1.21	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.20	19-Oct-2007	ad	branches: 1.20.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.19	24-Sep-2007	cube	branches: 1.19.2; - If the chip doesn't support jumbo frames, don't use bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but bus_dmamap_load for the mbuf storage space instead. - If the chip supports jumbo frames + keep track of which RX descriptor uses which jumbo mbuf buffer, so that we can rewrite the physaddr field of the descriptor later, as it might be partially overwritten by the hw + when we're out of jumbo mbufs, and if the packet is small enough, copy it into a cluster mbuf Those changes make my nfe(4) stable in both cases (defining NFE_NO_JUMBO for the first one).
1.18	24-Sep-2007	cube	From OpenBSD, rev 1.72: In nfe_start() do a fast return if IFF_OACTIVE is set, in this case we need a Tx interrupt to clean up the DMA ring before if_start can be properly called.
1.17	01-Sep-2007	dyoung	branches: 1.17.2; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.16	09-Jul-2007	ad	branches: 1.16.2; 1.16.6; 1.16.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.15	04-Mar-2007	christos	branches: 1.15.2; 1.15.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.14	28-Feb-2007	tsutsui	Apply patches which may fix RX stall problem on nfe(4): - make DMA descriptors volatile to avoid possible unintended reordering which might cause some race conditions - process interrupts until all NFE_IRQ_WANTED bits are handled and also put misc fixes: - return 1 and call nfe_start() in nfe_intr() only if any own interrupts are actually handled - use bus_dmamap_load_mbuf(9) for RX mbufs rather than bus_dmamap_load(9) with mtod(9) and MCLBYTES - check sc->txq.queued to see if TX descriptors are queued or handled in nfe_start() and nfe_txeof() - use proper BUS_DMASYNC_{PRE,POST} ops - prepare and use NFE_[RT]X_NEXTDESC() macro - rename NFE_TX_TCP_CSUM to NFE_TX_TCP_UDP_CSUM since it also enables hardware udp4csum-tx for UDP4 packets - some minor optimization - misc KNF Tested and confirmed by matthew green by "to send >25MB/sec to nfe0 for over one hour," and also tested by me (with light TRX load on 100baseTX though) for a month.
1.13	09-Jan-2007	tsutsui	branches: 1.13.2; Enable hardware checksum support by hardware info taken from FreeBSD. Tested on nForce3 250 and nForce4 by Chuck Silvers, Cesar Catrian Carreno, and me.
1.12	05-Jan-2007	jmcneill	A few changes to make this work on the Microsoft Xbox. Should also help with other users who have been experiencing watchdog timeouts: * Mask all interrupts while servicing a tx or rx interrupt. * On init, clear IRQ status registers (workaround for buggy netbooters).
1.11	01-Jan-2007	tsutsui	Pull a TX timeout fix from OpenBSD: > Defer setting of the valid bit in the first TX descriptor after > all descriptors have been setup. Otherwise, hardware may start > processing descriptors faster than us and crap out. > Fixes "watchdog timeout" errors. > > Original idea from Matthew Dillon @DragonFly.
1.10	01-Jan-2007	tsutsui	Minor style fixes in nfe_attach: - print PCI device name and revision - print interrupt and Ethernet address like other devices Before: --- nfe0 at pci0 dev 5 function 0LKLN: Picked IRQ 20 with weight 1 : ioapic0 pin 20 (irq 9), address xx:xx:xx:xx:xx:xx After: --- nfe0 at pci0 dev 5 function 0: NVIDIA nForce3 ethernet #4 (rev. 0xa2) LKLN: Picked IRQ 20 with weight 1 nfe0: interrupting at ioapic0 pin 20 (irq 9) nfe0: Ethernet address xx:xx:xx:xx:xx:xx (note "Picked IRQ" message is logged by aprint_verbose(9) in acpi(4))
1.9	27-Dec-2006	alc	remove the KASSERT() introduced in last commit and use VLAN_OUTPUT_TAG() macro. tested by martin@ pointed out and ok by pavel@, also ok from christos@
1.8	25-Dec-2006	alc	CID-3667: check if 'mtag != NULL' before dereferencing it
1.7	16-Nov-2006	christos	branches: 1.7.2; __unused removal on arguments; approved by core.
1.6	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.5	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.4	03-Sep-2006	xtraeme	branches: 1.4.2; 1.4.4; 1.4.6; Add support for MCP61/65 controllers... from obsd.
1.3	26-Mar-2006	chs	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.12; 1.3.14; work around another mbuf leak, the hardware doesn't seem to work quite the way that the driver expects.
1.2	16-Mar-2006	chs	branches: 1.2.2; handle another difference between the BSDs: free the original mbuf in the jumbo-free callback too.
1.1	12-Mar-2006	chs	branches: 1.1.2; add nfe driver and manpage from OpenBSD.
1.1.2.4	03-Sep-2006	yamt	sync with head.
1.1.2.3	01-Apr-2006	yamt	sync with head.
1.1.2.2	13-Mar-2006	yamt	sync with head.
1.1.2.1	12-Mar-2006	yamt	file if_nfe.c was added on branch yamt-pdpolicy on 2006-03-13 09:07:26 +0000
1.2.2.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.3.14.1	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #112): sys/dev/pci/if_nfe.c: revision 1.4 sys/dev/pci/pcidevs: revision 1.824 - Add nForce MCP61/65 LAN controllers... from obsd. - Improve description in some nForce controllers as well. Add support for MCP61/65 controllers... from obsd.
1.3.12.11	27-Feb-2008	yamt	sync with head.
1.3.12.10	11-Feb-2008	yamt	sync with head.
1.3.12.9	04-Feb-2008	yamt	sync with head.
1.3.12.8	21-Jan-2008	yamt	sync with head
1.3.12.7	15-Nov-2007	yamt	sync with head.
1.3.12.6	27-Oct-2007	yamt	sync with head.
1.3.12.5	03-Sep-2007	yamt	sync with head.
1.3.12.4	26-Feb-2007	yamt	sync with head.
1.3.12.3	30-Dec-2006	yamt	sync with head.
1.3.12.2	21-Jun-2006	yamt	sync with head.
1.3.12.1	26-Mar-2006	yamt	file if_nfe.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:04 +0000
1.3.6.2	22-Apr-2006	simonb	Sync with head.
1.3.6.1	26-Mar-2006	simonb	file if_nfe.c was added on branch simonb-timecounters on 2006-04-22 11:39:14 +0000
1.3.4.4	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): sys/dev/pci/if_nfe.c: revision 1.3 work around another mbuf leak, the hardware doesn't seem to work quite the way that the driver expects.
1.3.4.3	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): sys/dev/pci/if_nfe.c: revision 1.2 handle another difference between the BSDs: free the original mbuf in the jumbo-free callback too.
1.3.4.2	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): share/man/man4/Makefile: revision 1.377 sys/dev/pci/if_nfe.c: revision 1.1 sys/dev/pci/if_nfereg.h: revision 1.1 share/man/man4/nfe.4: revision 1.1 sys/dev/pci/if_nfevar.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.863 add nfe driver and manpage from OpenBSD.
1.3.4.1	26-Mar-2006	snj	file if_nfe.c was added on branch netbsd-3 on 2006-04-20 03:11:14 +0000
1.3.2.2	19-Apr-2006	elad	sync with head.
1.3.2.1	26-Mar-2006	elad	file if_nfe.c was added on branch elad-kernelauth on 2006-04-19 03:25:35 +0000
1.4.6.2	10-Dec-2006	yamt	sync with head.
1.4.6.1	22-Oct-2006	yamt	sync with head
1.4.4.2	09-Sep-2006	rpaulo	sync with head
1.4.4.1	03-Sep-2006	rpaulo	file if_nfe.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:17 +0000
1.4.2.2	12-Jan-2007	ad	Sync with head.
1.4.2.1	18-Nov-2006	ad	Sync with head.
1.7.2.3	23-Feb-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1061): sys/dev/pci/if_nfe.c: revision 1.22, 1.23, 1.25, 1.27 sys/dev/pci/pcidevs: revision 1.914 sys/dev/pci/if_nfereg.h: revision 1.5, 1.6 share/man/man4/nfe.4: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.5, 1.6 Add NVIDIA MCP7[37] devices, from OpenBSD. Attach to NVIDIA MCP67/73 Ethernet controllers. Add the MCP{67,73} products on nfe_attach(). Sync with reality. Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576. Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.7.2.2	07-Oct-2007	pavel	Pull up following revisions (requested by cube in ticket #920): sys/dev/pci/if_nfe.c: revision 1.14, 1.18-1.19 sys/dev/pci/if_nfereg.h: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.3 Apply patches which may fix RX stall problem on nfe(4): - make DMA descriptors volatile to avoid possible unintended reordering which might cause some race conditions - process interrupts until all NFE_IRQ_WANTED bits are handled and also put misc fixes: - return 1 and call nfe_start() in nfe_intr() only if any own interrupts are actually handled - use bus_dmamap_load_mbuf(9) for RX mbufs rather than bus_dmamap_load(9) with mtod(9) and MCLBYTES - check sc->txq.queued to see if TX descriptors are queued or handled in nfe_start() and nfe_txeof() - use proper BUS_DMASYNC_{PRE,POST} ops - prepare and use NFE_[RT]X_NEXTDESC() macro - rename NFE_TX_TCP_CSUM to NFE_TX_TCP_UDP_CSUM since it also enables hardware udp4csum-tx for UDP4 packets - some minor optimization - misc KNF Tested and confirmed by matthew green by "to send >25MB/sec to nfe0 for over one hour," and also tested by me (with light TRX load on 100baseTX though) for a month. From OpenBSD, rev 1.72: In nfe_start() do a fast return if IFF_OACTIVE is set, in this case we need a Tx interrupt to clean up the DMA ring before if_start can be properly called. - If the chip doesn't support jumbo frames, don't use bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but bus_dmamap_load for the mbuf storage space instead. - If the chip supports jumbo frames + keep track of which RX descriptor uses which jumbo mbuf buffer, so that we can rewrite the physaddr field of the descriptor later, as it might be partially overwritten by the hw + when we're out of jumbo mbufs, and if the packet is small enough, copy it into a cluster mbuf Those changes make my nfe(4) stable in both cases (defining NFE_NO_JUMBO for the first one).
1.7.2.1	19-Jan-2007	bouyer	branches: 1.7.2.1.2; Pull up following revision(s) (requested by tsutsui in ticket #370): sys/dev/pci/if_nfe.c: revision 1.8 - 1.13 sys/dev/pci/if_nfereg.h: revision 1.2 - 1.3 several bug fixes and hardware checksum support for nfs(4).
1.7.2.1.2.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.7.2.1.2.1	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.13.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.15.4.1	11-Jul-2007	mjf	Sync with head.
1.15.2.4	23-Oct-2007	ad	Sync with head.
1.15.2.3	09-Oct-2007	ad	Sync with head.
1.15.2.2	01-Sep-2007	ad	Update for pool_cache API changes.
1.15.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.16.8.4	23-Mar-2008	matt	sync with HEAD
1.16.8.3	09-Jan-2008	matt	sync with HEAD
1.16.8.2	08-Nov-2007	matt	sync with -HEAD
1.16.8.1	06-Nov-2007	matt	sync with HEAD
1.16.6.8	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.16.6.7	14-Nov-2007	joerg	Sync with HEAD.
1.16.6.6	11-Nov-2007	joerg	Sync with HEAD.
1.16.6.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.16.6.4	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.6.3	05-Oct-2007	joerg	Convert nfe(4) to PNP power management.
1.16.6.2	02-Oct-2007	joerg	Sync with HEAD.
1.16.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.16.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.17.2.1	06-Oct-2007	yamt	sync with head.
1.19.2.3	18-Nov-2007	bouyer	Sync with HEAD
1.19.2.2	13-Nov-2007	bouyer	Sync with HEAD
1.19.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.20.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.20.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.20.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.23.4.1	11-Dec-2007	yamt	sync with head.
1.23.2.1	26-Dec-2007	ad	Sync with head.
1.24.2.2	20-Jan-2008	bouyer	Sync with HEAD
1.24.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.28.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.28.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.28.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.28.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.28.2.1	24-Mar-2008	keiichi	sync with head.
1.33.2.2	04-Jun-2008	yamt	sync with head
1.33.2.1	18-May-2008	yamt	sync with head.
1.34.4.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.34.2.5	11-Aug-2010	yamt	sync with head.
1.34.2.4	11-Mar-2010	yamt	sync with head
1.34.2.3	16-Sep-2009	yamt	sync with head
1.34.2.2	20-Jun-2009	yamt	sync with head
1.34.2.1	04-May-2009	yamt	sync with head.
1.35.2.1	18-Jun-2008	simonb	Sync with head.
1.36.6.1	02-Mar-2009	snj	Pull up following revision(s) (requested by cegger in ticket #540): sys/dev/pci/if_nfe.c: revision 1.42 when nfe_attach() fails, free all resources including interrupt map, interrupt handler and dma maps. Also prevent call to bus_dmamap_destroy() when bus_dmamap_create() failed. The nfe_free_* routines assume, the map pointers are NULL but they are actually undefined. Make the assumption true by making them NULL in the error path of bus_dmamap_create(). All together, this fixes crashes when nfe_attach() fails. ok martin@
1.36.4.3	28-Apr-2009	skrll	Sync with HEAD.
1.36.4.2	03-Mar-2009	skrll	Sync with HEAD.
1.36.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.36.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.39.2.2	23-Jul-2009	jym	Sync with HEAD.
1.39.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.49.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.49.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.50.2.2	05-Mar-2011	rmind	sync with head
1.50.2.1	30-May-2010	rmind	sync with head
1.53.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.53.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.54.8.1	18-Feb-2012	mrg	merge to -current.
1.54.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.54.4.2	30-Oct-2012	yamt	sync with head
1.54.4.1	17-Apr-2012	yamt	sync with head
1.55.2.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #685): sys/dev/pci/if_nfe.c: revision 1.57 use 64-bit DMA where possible.
1.56.2.4	03-Dec-2017	jdolecek	update from HEAD
1.56.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.56.2.2	23-Jun-2013	tls	resync from head
1.56.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.58.4.1	18-May-2014	rmind	sync with head
1.59.6.3	05-Feb-2017	skrll	Sync with HEAD
1.59.6.2	09-Jul-2016	skrll	Sync with HEAD
1.59.6.1	19-Mar-2016	skrll	Sync with HEAD
1.61.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.63.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.63.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.64.2.3	26-Jan-2019	pgoyette	Sync with HEAD
1.64.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.64.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.65.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.65.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.65.2.1	10-Jun-2019	christos	Sync with HEAD
1.71.2.1	19-Mar-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #786): sys/dev/pci/if_nfe.c: revision 1.77 sys/dev/pci/if_nfe.c: revision 1.78 sys/dev/mii/atphy.c: revision 1.28 sys/dev/mii/atphy.c: revision 1.29 sys/dev/mii/miidevs: revision 1.166 Use unsigned to avoid undefined behavior. Found by kUBSan. 0x001374 is non-bitreversed value of Attansic OUI(0x00c82e). Attansic/Atheros correctly uses ID1 and ID2 register, so delete all 0x001374 related entries. Improve error check: - We check PHY register read error correctly (timeout and NFE_PHY_ERROR), so don't check NFE_PHY_DATA register's value with 0xffffffff or 0. At least, some registers may have 0. - Check NFE_PHY_ERROR bit in nfe_miibus_writereg(). - Improve debug printf Fix a bug that atphy(4) doesn't work with Attansic L2 rev. 1. Reported by Rocky Hotas. - On ASUS M2N-MX SE Plus (NVIDIA MCP61 with Attansic L2 rev. 1), changing debug port 0x29's value makes the next PHY read fail with error. Read any register to ignore this problem if the PHY is Attansic L2 revision 1. I don't know if this problem is from L2 rev. 1 itself or from the combination because I have only one machine which has L2 rev. "1". At least, ASUS eee pc 900 (Attansic L2 rev. "2") has no this problem. - Add comment. AR8021 document has no description about the power saving control register(debug port 0x29). - Add comment. AR8031 document says the lower 14 bits are reserved and the default value is 0x36d0. Shouldn't we clear those bits? - I have no document neither L1(F1) nor L2(F2), so I don't know whether the debug port access is correct or not. Tested with the following machines: - ASUS P5B SE, L1 rev. 5, age(4) - ASUS K50IJ, L1 rev. 9, ale(4) - ASUS eee pc 900, L2 rev. 2, lii(4) - ASUS M2N-MX SE Plus, L2 rev. 1, nfe(4) - Intel DP55WB, 82578(AR8021 rev. 2), wm(4) - Dell inspiron 14z, AR0835 rev. 9, alc(4)
1.72.2.1	29-Feb-2020	ad	Sync with head.
1.80.10.1	02-Aug-2025	perseant	Sync with HEAD
1.8	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.7	17-Apr-2008	christos	branches: 1.7.92; sync with openbsd; fixes observed corruption issues.
1.6	26-Jan-2008	tsutsui	branches: 1.6.6; Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.5	17-Dec-2007	tsutsui	Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576.
1.4	28-Feb-2007	tsutsui	branches: 1.4.16; 1.4.22; 1.4.24; 1.4.28; Apply patches which may fix RX stall problem on nfe(4): - make DMA descriptors volatile to avoid possible unintended reordering which might cause some race conditions - process interrupts until all NFE_IRQ_WANTED bits are handled and also put misc fixes: - return 1 and call nfe_start() in nfe_intr() only if any own interrupts are actually handled - use bus_dmamap_load_mbuf(9) for RX mbufs rather than bus_dmamap_load(9) with mtod(9) and MCLBYTES - check sc->txq.queued to see if TX descriptors are queued or handled in nfe_start() and nfe_txeof() - use proper BUS_DMASYNC_{PRE,POST} ops - prepare and use NFE_[RT]X_NEXTDESC() macro - rename NFE_TX_TCP_CSUM to NFE_TX_TCP_UDP_CSUM since it also enables hardware udp4csum-tx for UDP4 packets - some minor optimization - misc KNF Tested and confirmed by matthew green by "to send >25MB/sec to nfe0 for over one hour," and also tested by me (with light TRX load on 100baseTX though) for a month.
1.3	09-Jan-2007	tsutsui	branches: 1.3.2; Enable hardware checksum support by hardware info taken from FreeBSD. Tested on nForce3 250 and nForce4 by Chuck Silvers, Cesar Catrian Carreno, and me.
1.2	05-Jan-2007	jmcneill	Fix incorrect definition of wake-on-LAN magic number, from FreeBSD.
1.1	12-Mar-2006	chs	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.10; 1.1.16; 1.1.20; 1.1.22; 1.1.26; add nfe driver and manpage from OpenBSD.
1.1.26.3	23-Feb-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1061): sys/dev/pci/if_nfe.c: revision 1.22, 1.23, 1.25, 1.27 sys/dev/pci/pcidevs: revision 1.914 sys/dev/pci/if_nfereg.h: revision 1.5, 1.6 share/man/man4/nfe.4: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.5, 1.6 Add NVIDIA MCP7[37] devices, from OpenBSD. Attach to NVIDIA MCP67/73 Ethernet controllers. Add the MCP{67,73} products on nfe_attach(). Sync with reality. Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576. Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.1.26.2	07-Oct-2007	pavel	Pull up following revisions (requested by cube in ticket #920): sys/dev/pci/if_nfe.c: revision 1.14, 1.18-1.19 sys/dev/pci/if_nfereg.h: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.3 Apply patches which may fix RX stall problem on nfe(4): - make DMA descriptors volatile to avoid possible unintended reordering which might cause some race conditions - process interrupts until all NFE_IRQ_WANTED bits are handled and also put misc fixes: - return 1 and call nfe_start() in nfe_intr() only if any own interrupts are actually handled - use bus_dmamap_load_mbuf(9) for RX mbufs rather than bus_dmamap_load(9) with mtod(9) and MCLBYTES - check sc->txq.queued to see if TX descriptors are queued or handled in nfe_start() and nfe_txeof() - use proper BUS_DMASYNC_{PRE,POST} ops - prepare and use NFE_[RT]X_NEXTDESC() macro - rename NFE_TX_TCP_CSUM to NFE_TX_TCP_UDP_CSUM since it also enables hardware udp4csum-tx for UDP4 packets - some minor optimization - misc KNF Tested and confirmed by matthew green by "to send >25MB/sec to nfe0 for over one hour," and also tested by me (with light TRX load on 100baseTX though) for a month. From OpenBSD, rev 1.72: In nfe_start() do a fast return if IFF_OACTIVE is set, in this case we need a Tx interrupt to clean up the DMA ring before if_start can be properly called. - If the chip doesn't support jumbo frames, don't use bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but bus_dmamap_load for the mbuf storage space instead. - If the chip supports jumbo frames + keep track of which RX descriptor uses which jumbo mbuf buffer, so that we can rewrite the physaddr field of the descriptor later, as it might be partially overwritten by the hw + when we're out of jumbo mbufs, and if the packet is small enough, copy it into a cluster mbuf Those changes make my nfe(4) stable in both cases (defining NFE_NO_JUMBO for the first one).
1.1.26.1	19-Jan-2007	bouyer	branches: 1.1.26.1.2; Pull up following revision(s) (requested by tsutsui in ticket #370): sys/dev/pci/if_nfe.c: revision 1.8 - 1.13 sys/dev/pci/if_nfereg.h: revision 1.2 - 1.3 several bug fixes and hardware checksum support for nfs(4).
1.1.26.1.2.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.1.26.1.2.1	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.1.22.2	09-Sep-2006	rpaulo	sync with head
1.1.22.1	12-Mar-2006	rpaulo	file if_nfereg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:17 +0000
1.1.20.1	12-Jan-2007	ad	Sync with head.
1.1.16.6	04-Feb-2008	yamt	sync with head.
1.1.16.5	21-Jan-2008	yamt	sync with head
1.1.16.4	03-Sep-2007	yamt	sync with head.
1.1.16.3	26-Feb-2007	yamt	sync with head.
1.1.16.2	21-Jun-2006	yamt	sync with head.
1.1.16.1	12-Mar-2006	yamt	file if_nfereg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:04 +0000
1.1.10.2	22-Apr-2006	simonb	Sync with head.
1.1.10.1	12-Mar-2006	simonb	file if_nfereg.h was added on branch simonb-timecounters on 2006-04-22 11:39:14 +0000
1.1.8.2	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): share/man/man4/Makefile: revision 1.377 sys/dev/pci/if_nfe.c: revision 1.1 sys/dev/pci/if_nfereg.h: revision 1.1 share/man/man4/nfe.4: revision 1.1 sys/dev/pci/if_nfevar.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.863 add nfe driver and manpage from OpenBSD.
1.1.8.1	12-Mar-2006	snj	file if_nfereg.h was added on branch netbsd-3 on 2006-04-20 03:11:14 +0000
1.1.6.2	19-Apr-2006	elad	sync with head.
1.1.6.1	12-Mar-2006	elad	file if_nfereg.h was added on branch elad-kernelauth on 2006-04-19 03:25:35 +0000
1.1.2.2	13-Mar-2006	yamt	sync with head.
1.1.2.1	12-Mar-2006	yamt	file if_nfereg.h was added on branch yamt-pdpolicy on 2006-03-13 09:07:26 +0000
1.3.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.4.28.1	02-Jan-2008	bouyer	Sync with HEAD
1.4.24.1	26-Dec-2007	ad	Sync with head.
1.4.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.16.2	23-Mar-2008	matt	sync with HEAD
1.4.16.1	09-Jan-2008	matt	sync with HEAD
1.6.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.7.92.1	10-Jun-2019	christos	Sync with HEAD
1.12	01-Mar-2020	thorpej	- Range-limit our DMA tag to 40-bit. - Eliminate __LP64__ ifdefs.
1.11	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.10	03-Nov-2010	jakllsch	branches: 1.10.60; Make nfe(4) detachable.
1.9	20-Apr-2008	cube	branches: 1.9.22; 1.9.24; Restore jumbo frame support and protect access to the DMA space for received frames with a mutex. Tested by Paul Goyette, Martin Husemann and myself.
1.8	17-Apr-2008	christos	branches: 1.8.2; sync with openbsd; fixes observed corruption issues.
1.7	26-Mar-2008	cube	Split device_t and softc, and related cosmetic changes.
1.6	26-Jan-2008	tsutsui	branches: 1.6.6; Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.5	17-Dec-2007	tsutsui	Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576.
1.4	09-Dec-2007	jmcneill	branches: 1.4.2; Merge jmcneill-pm branch.
1.3	24-Sep-2007	cube	branches: 1.3.4; 1.3.6; 1.3.8; - If the chip doesn't support jumbo frames, don't use bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but bus_dmamap_load for the mbuf storage space instead. - If the chip supports jumbo frames + keep track of which RX descriptor uses which jumbo mbuf buffer, so that we can rewrite the physaddr field of the descriptor later, as it might be partially overwritten by the hw + when we're out of jumbo mbufs, and if the packet is small enough, copy it into a cluster mbuf Those changes make my nfe(4) stable in both cases (defining NFE_NO_JUMBO for the first one).
1.2	04-Mar-2007	christos	branches: 1.2.2; 1.2.14; 1.2.16; 1.2.18; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.1	12-Mar-2006	chs	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.10; 1.1.16; 1.1.22; 1.1.26; 1.1.28; 1.1.32; add nfe driver and manpage from OpenBSD.
1.1.32.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.1.32.1	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.1.28.1	12-Mar-2007	rmind	Sync with HEAD.
1.1.26.2	23-Feb-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1061): sys/dev/pci/if_nfe.c: revision 1.22, 1.23, 1.25, 1.27 sys/dev/pci/pcidevs: revision 1.914 sys/dev/pci/if_nfereg.h: revision 1.5, 1.6 share/man/man4/nfe.4: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.5, 1.6 Add NVIDIA MCP7[37] devices, from OpenBSD. Attach to NVIDIA MCP67/73 Ethernet controllers. Add the MCP{67,73} products on nfe_attach(). Sync with reality. Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576. Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.1.26.1	07-Oct-2007	pavel	Pull up following revisions (requested by cube in ticket #920): sys/dev/pci/if_nfe.c: revision 1.14, 1.18-1.19 sys/dev/pci/if_nfereg.h: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.3 Apply patches which may fix RX stall problem on nfe(4): - make DMA descriptors volatile to avoid possible unintended reordering which might cause some race conditions - process interrupts until all NFE_IRQ_WANTED bits are handled and also put misc fixes: - return 1 and call nfe_start() in nfe_intr() only if any own interrupts are actually handled - use bus_dmamap_load_mbuf(9) for RX mbufs rather than bus_dmamap_load(9) with mtod(9) and MCLBYTES - check sc->txq.queued to see if TX descriptors are queued or handled in nfe_start() and nfe_txeof() - use proper BUS_DMASYNC_{PRE,POST} ops - prepare and use NFE_[RT]X_NEXTDESC() macro - rename NFE_TX_TCP_CSUM to NFE_TX_TCP_UDP_CSUM since it also enables hardware udp4csum-tx for UDP4 packets - some minor optimization - misc KNF Tested and confirmed by matthew green by "to send >25MB/sec to nfe0 for over one hour," and also tested by me (with light TRX load on 100baseTX though) for a month. From OpenBSD, rev 1.72: In nfe_start() do a fast return if IFF_OACTIVE is set, in this case we need a Tx interrupt to clean up the DMA ring before if_start can be properly called. - If the chip doesn't support jumbo frames, don't use bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but bus_dmamap_load for the mbuf storage space instead. - If the chip supports jumbo frames + keep track of which RX descriptor uses which jumbo mbuf buffer, so that we can rewrite the physaddr field of the descriptor later, as it might be partially overwritten by the hw + when we're out of jumbo mbufs, and if the packet is small enough, copy it into a cluster mbuf Those changes make my nfe(4) stable in both cases (defining NFE_NO_JUMBO for the first one).
1.1.22.2	09-Sep-2006	rpaulo	sync with head
1.1.22.1	12-Mar-2006	rpaulo	file if_nfevar.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:17 +0000
1.1.16.6	04-Feb-2008	yamt	sync with head.
1.1.16.5	21-Jan-2008	yamt	sync with head
1.1.16.4	27-Oct-2007	yamt	sync with head.
1.1.16.3	03-Sep-2007	yamt	sync with head.
1.1.16.2	21-Jun-2006	yamt	sync with head.
1.1.16.1	12-Mar-2006	yamt	file if_nfevar.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:04 +0000
1.1.10.2	22-Apr-2006	simonb	Sync with head.
1.1.10.1	12-Mar-2006	simonb	file if_nfevar.h was added on branch simonb-timecounters on 2006-04-22 11:39:14 +0000
1.1.8.2	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): share/man/man4/Makefile: revision 1.377 sys/dev/pci/if_nfe.c: revision 1.1 sys/dev/pci/if_nfereg.h: revision 1.1 share/man/man4/nfe.4: revision 1.1 sys/dev/pci/if_nfevar.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.863 add nfe driver and manpage from OpenBSD.
1.1.8.1	12-Mar-2006	snj	file if_nfevar.h was added on branch netbsd-3 on 2006-04-20 03:11:14 +0000
1.1.6.2	19-Apr-2006	elad	sync with head.
1.1.6.1	12-Mar-2006	elad	file if_nfevar.h was added on branch elad-kernelauth on 2006-04-19 03:25:35 +0000
1.1.2.2	13-Mar-2006	yamt	sync with head.
1.1.2.1	12-Mar-2006	yamt	file if_nfevar.h was added on branch yamt-pdpolicy on 2006-03-13 09:07:26 +0000
1.2.18.1	06-Oct-2007	yamt	sync with head.
1.2.16.3	23-Mar-2008	matt	sync with HEAD
1.2.16.2	09-Jan-2008	matt	sync with HEAD
1.2.16.1	06-Nov-2007	matt	sync with HEAD
1.2.14.2	05-Oct-2007	joerg	Convert nfe(4) to PNP power management.
1.2.14.1	02-Oct-2007	joerg	Sync with HEAD.
1.2.2.1	09-Oct-2007	ad	Sync with head.
1.3.8.1	11-Dec-2007	yamt	sync with head.
1.3.6.1	26-Dec-2007	ad	Sync with head.
1.3.4.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.6.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.6.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.8.2.1	18-May-2008	yamt	sync with head.
1.9.24.1	05-Mar-2011	rmind	sync with head
1.9.22.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.10.60.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.10.60.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.33	01-Aug-2023	andvar	s/diable/disable/ in comments.
1.32	24-May-2022	andvar	fix various typos in comments, docs and log messages.
1.31	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.30	14-Jul-2016	msaitoh	branches: 1.30.16; 1.30.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.29	29-Mar-2014	christos	branches: 1.29.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.28	27-Oct-2012	chs	branches: 1.28.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.27	26-Jul-2011	dyoung	branches: 1.27.2; 1.27.12; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.26	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.25	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.24	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.23	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.22	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.21	10-Apr-2008	cegger	branches: 1.21.4; 1.21.12; 1.21.18; use aprint_*_dev and device_xname
1.20	19-Oct-2007	ad	branches: 1.20.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.19	16-Nov-2006	christos	branches: 1.19.8; 1.19.22; 1.19.24; 1.19.28; __unused removal on arguments; approved by core.
1.18	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.17	29-Mar-2006	thorpej	branches: 1.17.8; 1.17.10; Use device_cfdata().
1.16	11-Dec-2005	christos	branches: 1.16.4; 1.16.6; 1.16.8; 1.16.10; 1.16.12; merge ktrace-lwp.
1.15	30-May-2005	christos	branches: 1.15.2; - const poisoning - avoid variable shadowing.
1.14	04-Feb-2005	perry	de-__P
1.13	02-Nov-2003	wiz	branches: 1.13.8; 1.13.10; boundary, not boundry. Inspired by Tom Cosgrove.
1.12	03-May-2003	wiz	branches: 1.12.2; DMA, not dma nor Dma.
1.11	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.10	06-Jan-2003	wiz	interrupt with two rs.
1.9	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.8	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.7	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.6	13-Nov-2001	lukem	add RCSID
1.5	14-Aug-2001	mrg	add (u_long) casts for sparc64.
1.4	28-Dec-2000	sommerfeld	branches: 1.4.2; 1.4.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.3	04-Jan-2000	chopps	Add support for setting the clock and adapt the code to the new more generic ic/hd56470.c core which now supports the isa version of the same card.
1.2	15-Feb-1999	hubertf	branches: 1.2.8; RCS ID police
1.1	26-Jul-1998	explorer	add ntwo* at pci? function ?, a T1 network driver.
1.2.8.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.4.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.4.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.4.2.4	07-Jan-2003	thorpej	Sync with HEAD.
1.4.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.4.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.12.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.12.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.12.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.12.2.1	03-Aug-2004	skrll	Sync with HEAD
1.13.10.1	12-Feb-2005	yamt	sync with head.
1.13.8.1	29-Apr-2005	kent	sync with -current
1.15.2.3	27-Oct-2007	yamt	sync with head.
1.15.2.2	30-Dec-2006	yamt	sync with head.
1.15.2.1	21-Jun-2006	yamt	sync with head.
1.16.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.16.10.1	19-Apr-2006	elad	sync with head.
1.16.8.1	01-Apr-2006	yamt	sync with head.
1.16.6.1	22-Apr-2006	simonb	Sync with head.
1.16.4.1	09-Sep-2006	rpaulo	sync with head
1.17.10.2	10-Dec-2006	yamt	sync with head.
1.17.10.1	22-Oct-2006	yamt	sync with head
1.17.8.1	18-Nov-2006	ad	Sync with head.
1.19.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.19.24.1	06-Nov-2007	matt	sync with HEAD
1.19.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.19.8.1	23-Oct-2007	ad	Sync with head.
1.20.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.21.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.21.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.21.4.3	11-Mar-2010	yamt	sync with head
1.21.4.2	16-May-2009	yamt	sync with head
1.21.4.1	04-May-2009	yamt	sync with head.
1.27.12.3	03-Dec-2017	jdolecek	update from HEAD
1.27.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.27.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.27.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.27.2.1	30-Oct-2012	yamt	sync with head
1.28.2.1	18-May-2014	rmind	sync with head
1.29.6.1	05-Oct-2016	skrll	Sync with HEAD
1.30.18.1	10-Jun-2019	christos	Sync with HEAD
1.30.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.6	06-Jul-2024	andvar	Fix various typos in comments: s/defininitions/definitions/ s/ininitialise/initialise/ s/collasped/collapsed/ s/optionaly/optionally/
1.5	11-Dec-2005	christos	branches: 1.5.202; merge ktrace-lwp.
1.4	27-Feb-2005	perry	nuke trailing whitespace
1.3	04-Jan-2000	chopps	branches: 1.3.28; 1.3.36; 1.3.38; Add support for setting the clock and adapt the code to the new more generic ic/hd56470.c core which now supports the isa version of the same card.
1.2	15-Feb-1999	hubertf	branches: 1.2.8; RCS ID police
1.1	26-Jul-1998	explorer	add ntwo* at pci? function ?, a T1 network driver.
1.2.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.38.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.3.36.1	29-Apr-2005	kent	sync with -current
1.3.28.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.202.1	02-Aug-2025	perseant	Sync with HEAD
1.81	04-Oct-2025	thorpej	Use device_getprop_bool().
1.80	10-Nov-2024	mlelstv	Add MBUFTRACE
1.79	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.78	10-Feb-2024	andvar	branches: 1.78.2; s/alloted/allotted/ in comments.
1.77	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.76	16-Mar-2020	thorpej	Do the "have transmit resources available" check before checking that there is a packet in the send queue.
1.75	15-Mar-2020	thorpej	Don't bother with IFF_OACTIVE.
1.74	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.73	29-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.72	11-Oct-2019	msaitoh	branches: 1.72.2; Use unsigned to avoid undefined behavior. Found by kUBSan.
1.71	28-May-2019	msaitoh	branches: 1.71.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.70	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.69	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.68	22-Apr-2019	msaitoh	In drivers which use MII(4) and have hook SIOC[GS]IFMEDIA which just pass to ifmedia_ioctl(), the hook is not required because ether_ioctl has it (if_ethersubr.c rev. 1.160). These drivers don't return ENETRESET in ifmedia_ioctl(), so no functional change.
1.67	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.66	02-Dec-2018	jdolecek	use pci_intr_establish_xname()
1.65	26-Jun-2018	msaitoh	branches: 1.65.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.64	15-Dec-2016	ozaki-r	branches: 1.64.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.63	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.62	10-Jun-2016	ozaki-r	branches: 1.62.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.61	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.60	27-Apr-2015	christos	fix mbuf leak on failure (Brainy)
1.59	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.58	10-Aug-2014	tls	branches: 1.58.2; 1.58.4; Merge tls-earlyentropy branch into HEAD.
1.57	29-Mar-2014	christos	branches: 1.57.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	30-Mar-2013	christos	branches: 1.56.4; remove trailing whitespace
1.55	22-Jul-2012	matt	branches: 1.55.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.54	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.53	19-Nov-2011	tls	branches: 1.53.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.52	13-Nov-2010	uebayasi	branches: 1.52.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.51	30-Apr-2010	hubertf	silence "boot -z"
1.50	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.49	19-Jan-2010	pooka	branches: 1.49.2; 1.49.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.48	06-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested Am79c973 PCnet-FAST III.
1.47	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.46	04-Apr-2008	tsutsui	branches: 1.46.4; 1.46.14; 1.46.18; Split device_t/softc with misc related changes.
1.45	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.44	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.43	20-Jan-2008	jmmv	branches: 1.43.2; 1.43.6; Now that the driver goes through the mii_ifmedia_change indirection instead of directly calling the pcn_79c970_mediachange function, we must initialize sc->sc_mii.mii_ifp so that this last function is passed a valid parameter. Fixes a panic in pcn_79c970_mediachange that arose immediately when trying to use this interface due to a NULL pointer dereference. Hi dyoung@!
1.42	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.41	19-Oct-2007	ad	branches: 1.41.2; 1.41.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.40	29-Aug-2007	dyoung	branches: 1.40.4; Constify: LLADDR() -> CLLADDR().
1.39	09-Jul-2007	ad	branches: 1.39.2; 1.39.6; 1.39.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.38	04-Mar-2007	christos	branches: 1.38.2; 1.38.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.37	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.36	16-Nov-2006	christos	branches: 1.36.4; __unused removal on arguments; approved by core.
1.35	15-Nov-2006	jdarrow	Fix typo (enaddr[3] should be enaddr[2]), and add note to comment, both in VMware bug-workaround section.
1.34	01-Nov-2006	thorpej	Detect if we're talking to a VMware virtual interface, and, if so, limit the number of Tx segments to 8 to work around a VMware bug.
1.33	31-Oct-2006	thorpej	In the pcn driver: - Remove the PCN_NO_PROM option. Instead, query the am79c970-no-eeprom property, and read the MAC address from the CSRs if that property is TRUE. In the ibmnws port: - Implement device_register(). - In device_register(), set the am79c970-no-eeprom property for the built-in Ethernet.
1.32	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.31	17-Jun-2006	christos	branches: 1.31.4; 1.31.6; re-factor the pci powestate api. reviewed by gimpy
1.30	14-Jun-2006	tsutsui	branches: 1.30.2; Call pcn_reset() in pcn_shutdown() so that my 7248 (prep) firmware can netboot even after reboot without powercycle.
1.29	22-Feb-2006	garbled	branches: 1.29.2; 1.29.8; IBM makes a pcn card that shows up as: Trident Microsystems 4DWAVE DX (ethernet network, revision 0x25) Unfortunately, this is an autri sound card. Special case the match routines for both drivers so if_pcn properly picks it up, and autri doesn't.
1.28	24-Dec-2005	perry	branches: 1.28.2; 1.28.4; 1.28.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.27	11-Dec-2005	christos	merge ktrace-lwp.
1.26	07-May-2005	is	branches: 1.26.2; As discussed on tech-net: don't suppress any PHYs to enable multiple external PHYs. The internal PHY at MII (nr 31) will be the last instance, so it won't be the default instance when external PHYs are present.
1.25	27-Feb-2005	perry	branches: 1.25.2; nuke trailing whitespace
1.24	30-Oct-2004	thorpej	branches: 1.24.4; 1.24.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.23	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.22	25-Oct-2003	christos	branches: 1.22.4; Fix uninitialized variable warnings
1.21	19-Oct-2003	matt	Add PCN_NO_PROM flag. (read the macaddr from the chip assuming the "BIOS" has properly written it). From John Gordon.
1.20	21-Jul-2003	jdolecek	add rnd hooks
1.19	30-Mar-2003	jdolecek	branches: 1.19.2; sprinkle some const qualifiers
1.18	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.17	23-Oct-2002	perry	fix lint warnings about things being stringified under cpp -traditional
1.16	22-Oct-2002	thorpej	Fix typos in media configuration register init. From patch posted to tech-kern by Erik Anggard <erik.anggard@packetfront.com>. May address PR 16346.
1.15	21-Oct-2002	fair	Change the "dontcare bits" argument to ifmedia_init() to IFM_IMASK so that it is possible select PHY instances other than the first one (instance zero), if there is more than one PHY attached.
1.14	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.13	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.12	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.11	04-Sep-2002	thorpej	Add support for memory-mapped PCnet-PCI chips.
1.10	02-Aug-2002	thorpej	Fix a typo which caused the wrong register to be updated when the media is not full-duplex.
1.9	03-May-2002	thorpej	branches: 1.9.2; 1.9.4; Bump the number of Tx DMA segments from 8 to 16 (the zero-copy socket code sometimes sees more than 8).
1.8	29-Jan-2002	nisimura	DANAS (disable autonegotiation) bit of BCR32 takes its effect when set, not cleared.
1.7	29-Nov-2001	thorpej	Fix typo in copyright notice.
1.6	27-Nov-2001	onoe	Set ONES (must be ones) bits in tmd1 of tx descriptors. This fix avoids VMware 3.0/WinXP to crash.
1.5	13-Nov-2001	lukem	add RCSID
1.4	02-Sep-2001	enami	branches: 1.4.4; 1.4.6; Fix an obvious typo found while reading the source.
1.3	28-Aug-2001	thorpej	- Swap the bytes OR'd into init_ladrf[], for big-endian machines. - Make sure to sync the init block. Pointed out by Izumi Tsutsui.
1.2	28-Aug-2001	thorpej	Print "auto-FDX" for the second "autoselect" instance. Pointed out by Izumi Tsutsui.
1.1	27-Aug-2001	thorpej	New driver for the AMD PCnet-PCI family of Ethernet chips. This driver uses direct DMA to mbufs (like other PCI network drivers, and unlike the old "le at pci" driver), and also supports communication with the MII-connected PHYs on the 10/100 boards.
1.4.6.12	29-Dec-2002	thorpej	Sync with HEAD.
1.4.6.11	11-Nov-2002	nathanw	Catch up to -current
1.4.6.10	22-Oct-2002	thorpej	Sync with HEAD.
1.4.6.9	18-Oct-2002	nathanw	Catch up to -current.
1.4.6.8	17-Sep-2002	nathanw	Catch up to -current.
1.4.6.7	13-Aug-2002	nathanw	Catch up to -current.
1.4.6.6	20-Jun-2002	nathanw	Catch up to -current.
1.4.6.5	28-Feb-2002	nathanw	Catch up to -current.
1.4.6.4	08-Jan-2002	nathanw	Catch up to -current.
1.4.6.3	14-Nov-2001	nathanw	Catch up to -current.
1.4.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.4.6.1	02-Sep-2001	nathanw	file if_pcn.c was added on branch nathanw_sa on 2001-09-21 22:35:56 +0000
1.4.4.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.4.4.6	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.4.4.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.4.4.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.4.4.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.4.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.4.4.1	02-Sep-2001	thorpej	file if_pcn.c was added on branch kqueue on 2001-09-13 01:15:55 +0000
1.9.4.3	30-Nov-2002	he	Pull up revision 1.11 (requested by thorpej in ticket #767): Add support for memory-mapped PCnet-PCI chips.
1.9.4.2	23-Oct-2002	lukem	Pull up revision 1.16 (requested by thorpej in ticket #940): Fix typos in media configuration register init. From patch posted to tech-kern by Erik Anggard <erik.anggard@packetfront.com>. May address PR 16346.
1.9.4.1	02-Aug-2002	lukem	Pull up revision 1.10 (requested by thorpej in ticket #605): Fix a typo which caused the wrong register to be updated when the media is not full-duplex.
1.9.2.1	29-Aug-2002	gehenna	catch up with -current.
1.19.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.19.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.19.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.19.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.19.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.19.2.1	03-Aug-2004	skrll	Sync with HEAD
1.22.4.1	24-Jan-2005	he	Pull up revision 1.24 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.24.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.24.4.1	29-Apr-2005	kent	sync with -current
1.25.2.2	15-Nov-2006	bouyer	Pull up following revision(s) (requested by thorpej in ticket #1586): sys/dev/pci/if_pcn.c: revision 1.35 Fix typo (enaddr[3] should be enaddr[2]), and add note to comment, both in VMware bug-workaround section.
1.25.2.1	11-Nov-2006	bouyer	Pull up following revision(s) (requested by thorpej in ticket #1577): sys/dev/pci/if_pcn.c: revision 1.34 Detect if we're talking to a VMware virtual interface, and, if so, limit the number of Tx segments to 8 to work around a VMware bug.
1.26.2.8	24-Mar-2008	yamt	sync with head.
1.26.2.7	17-Mar-2008	yamt	sync with head.
1.26.2.6	21-Jan-2008	yamt	sync with head
1.26.2.5	27-Oct-2007	yamt	sync with head.
1.26.2.4	03-Sep-2007	yamt	sync with head.
1.26.2.3	26-Feb-2007	yamt	sync with head.
1.26.2.2	30-Dec-2006	yamt	sync with head.
1.26.2.1	21-Jun-2006	yamt	sync with head.
1.28.6.1	22-Apr-2006	simonb	Sync with head.
1.28.4.1	09-Sep-2006	rpaulo	sync with head
1.28.2.1	01-Mar-2006	yamt	sync with head.
1.29.8.1	19-Jun-2006	chap	Sync with head.
1.29.2.1	26-Jun-2006	yamt	sync with head.
1.30.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.31.6.2	10-Dec-2006	yamt	sync with head.
1.31.6.1	22-Oct-2006	yamt	sync with head
1.31.4.1	18-Nov-2006	ad	Sync with head.
1.36.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.36.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.38.4.1	11-Jul-2007	mjf	Sync with head.
1.38.2.3	23-Oct-2007	ad	Sync with head.
1.38.2.2	09-Oct-2007	ad	Sync with head.
1.38.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.39.8.2	23-Mar-2008	matt	sync with HEAD
1.39.8.1	06-Nov-2007	matt	sync with HEAD
1.39.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.39.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.39.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.40.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.41.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.41.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.43.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.43.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.43.2.1	24-Mar-2008	keiichi	sync with head.
1.46.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.46.14.1	20-Nov-2010	riz	Pull up following revision(s) (requested by hubertf in ticket #1385): sys/dev/pci/if_pcn.c: revision 1.51 silence "boot -z"
1.46.4.4	11-Aug-2010	yamt	sync with head.
1.46.4.3	11-Mar-2010	yamt	sync with head
1.46.4.2	16-Sep-2009	yamt	sync with head
1.46.4.1	16-May-2009	yamt	sync with head
1.49.4.2	05-Mar-2011	rmind	sync with head
1.49.4.1	30-May-2010	rmind	sync with head
1.49.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.49.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.52.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.52.8.2	30-Oct-2012	yamt	sync with head
1.52.8.1	17-Apr-2012	yamt	sync with head
1.53.2.1	18-Feb-2012	mrg	merge to -current.
1.55.2.3	03-Dec-2017	jdolecek	update from HEAD
1.55.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.2.1	23-Jun-2013	tls	resync from head
1.56.4.1	18-May-2014	rmind	sync with head
1.57.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.58.4.4	05-Feb-2017	skrll	Sync with HEAD
1.58.4.3	09-Jul-2016	skrll	Sync with HEAD
1.58.4.2	19-Mar-2016	skrll	Sync with HEAD
1.58.4.1	06-Jun-2015	skrll	Sync with HEAD
1.58.2.1	30-Jul-2015	martin	Pull up following revision(s) (requested by maxv in ticket #889): sys/dev/pci/if_bge.c: revision 1.283 sys/dev/pci/if_pcn.c: revision 1.60 Fix mbuf leaks in error branches (found by the Brainy code scanner)
1.62.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.64.14.3	26-Jan-2019	pgoyette	Sync with HEAD
1.64.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.64.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.65.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.65.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.65.2.1	10-Jun-2019	christos	Sync with HEAD
1.71.2.1	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.72.2.1	29-Feb-2020	ad	Sync with head.
1.78.2.1	02-Aug-2025	perseant	Sync with HEAD
1.3	04-Sep-2002	thorpej	Add support for memory-mapped PCnet-PCI chips.
1.2	29-Nov-2001	thorpej	branches: 1.2.10; Fix typo in copyright notice.
1.1	27-Aug-2001	thorpej	branches: 1.1.4; 1.1.6; New driver for the AMD PCnet-PCI family of Ethernet chips. This driver uses direct DMA to mbufs (like other PCI network drivers, and unlike the old "le at pci" driver), and also supports communication with the MII-connected PHYs on the 10/100 boards.
1.1.6.4	17-Sep-2002	nathanw	Catch up to -current.
1.1.6.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.6.1	27-Aug-2001	nathanw	file if_pcnreg.h was added on branch nathanw_sa on 2001-09-21 22:35:57 +0000
1.1.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.4.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.4.1	27-Aug-2001	thorpej	file if_pcnreg.h was added on branch kqueue on 2001-09-13 01:15:55 +0000
1.2.10.1	30-Nov-2002	he	Pull up revision 1.3 (requested by thorpej in ticket #767): Add support for memory-mapped PCnet-PCI chips.
1.26	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.25	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.24	25-Jul-2017	maya	branches: 1.24.2; 1.24.4; enable rt2860 power management code adjust to fit netbsd: make suspend,resume functions match desired pmf* prototypes remove wakeup/activate wrapper functions avoid jumping to NULL on resume by initializing if_stop the problem machine has other issues on resume, so there might be further issues, but it's an improvement over a jump to NULL. tested by Riccardo Mottola
1.23	06-Jul-2016	christos	branches: 1.23.10; more cards, sort.
1.22	27-Apr-2016	christos	Sync with OpenBSD (add rt2860). Handle 32/64 memory bars, some power management, more device matches.
1.21	29-Mar-2014	christos	branches: 1.21.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.20	18-Feb-2012	drochner	branches: 1.20.2; 1.20.4; split device_t/softc
1.19	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.18	26-Jul-2011	dyoung	branches: 1.18.2; 1.18.6; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.17	10-May-2011	dyoung	Constify pci_attach_args.
1.16	10-May-2011	dyoung	Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.15	19-Jan-2010	pooka	branches: 1.15.4; 1.15.6; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.14	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.13	08-Aug-2009	matt	If we've had to enable MEM access, make sure our flags reflect that so that pci_mapreg_map won't fail because it didn't know we did it.
1.12	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.11	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.10	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.9	29-Apr-2008	scw	branches: 1.9.14; Pull in some more changes from OpenBSD: - replace rssadapt(9) with amrr for automatic rate control. - don't blindly IFQ_DEQUEUE() then drop a Tx packet if there are no available Tx resources. - move default MAC/BBP/RF settings from rt2661.c to rt2661reg.h. - enable packet bursting when operating as a STA. - implement new ic_updateslot() callback. - in hostap mode, we defer update of the slot time until all associated STAs are notified with updated beacons. - 802.11a uses a 16 microseconds short interframe space. - Fix rt2661_set_macaddr() so that we don't override the "unicast to me" flag in RT2661_MAC_CSR3 when setting the MAC address. - fix index of ERP information element in beacons. Add a couple of tweaks of my own: - The RX/TX BUSY flag should be the last thing written to a descriptor. - Check and service any additional h/w interrupts before returning from the isr. Tested in STA, AP, and Monitor modes. Tested with WEP, WPA, and WPA2 crypto. Additional testing by xtraeme@
1.8	10-Apr-2008	cegger	branches: 1.8.2; 1.8.4; use aprint_*_dev and device_xname
1.7	19-Oct-2007	ad	branches: 1.7.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.6	16-Nov-2006	christos	branches: 1.6.8; 1.6.22; 1.6.24; 1.6.28; __unused removal on arguments; approved by core.
1.5	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.4	06-Jun-2006	rpaulo	branches: 1.4.6; 1.4.8; Attachment framework for the rt2561.c and rt2661.c drivers. From OpenBSD.
1.3	11-Dec-2005	christos	branches: 1.3.4; 1.3.6; 1.3.8; 1.3.14; merge ktrace-lwp.
1.2	04-Jul-2005	drochner	branches: 1.2.2; 1.2.8; update PCI/Cardbus ral wlan driver, and adapt to new FreeBSD/NetBSD 80211 framework, from FUKAUMI Naoki per mail to tech-net
1.1	01-Jul-2005	drochner	add drivers for Ralink RT2500-based wireless adapters, written by Damien Bergamini, ported and submitted by FUKAUMI Naoki per PR kern/30449 I've modified the USB "ural" driver for recent changes to the NetBSD ieee80211 framework, possibly not completely, but with an ASUS wireless adapter I'm getting some signs of life. Didn't care about pci/cardbus for now, hopefully someone with hardware will do it.
1.2.8.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.8.1	04-Jul-2005	skrll	file if_ral_pci.c was added on branch ktrace-lwp on 2005-11-10 14:06:01 +0000
1.2.2.3	27-Oct-2007	yamt	sync with head.
1.2.2.2	30-Dec-2006	yamt	sync with head.
1.2.2.1	21-Jun-2006	yamt	sync with head.
1.3.14.1	19-Jun-2006	chap	Sync with head.
1.3.8.1	26-Jun-2006	yamt	sync with head.
1.3.6.1	07-Jun-2006	kardel	Sync with head.
1.3.4.1	09-Sep-2006	rpaulo	sync with head
1.4.8.2	10-Dec-2006	yamt	sync with head.
1.4.8.1	22-Oct-2006	yamt	sync with head
1.4.6.1	18-Nov-2006	ad	Sync with head.
1.6.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.6.24.1	06-Nov-2007	matt	sync with HEAD
1.6.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.6.8.1	23-Oct-2007	ad	Sync with head.
1.7.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.8.4.4	11-Mar-2010	yamt	sync with head
1.8.4.3	19-Aug-2009	yamt	sync with head.
1.8.4.2	16-May-2009	yamt	sync with head
1.8.4.1	16-May-2008	yamt	sync with head.
1.8.2.1	18-May-2008	yamt	sync with head.
1.9.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.15.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.15.4.1	31-May-2011	rmind	sync with head
1.18.6.2	24-Feb-2012	mrg	sync to -current.
1.18.6.1	18-Feb-2012	mrg	merge to -current.
1.18.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.18.2.1	17-Apr-2012	yamt	sync with head
1.20.4.1	18-May-2014	rmind	sync with head
1.20.2.2	03-Dec-2017	jdolecek	update from HEAD
1.20.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.21.6.3	28-Aug-2017	skrll	Sync with HEAD
1.21.6.2	09-Jul-2016	skrll	Sync with HEAD
1.21.6.1	29-May-2016	skrll	Sync with HEAD
1.23.10.1	01-Aug-2017	snj	Pull up following revision(s) (requested by maya in ticket #162): sys/dev/pci/if_ral_pci.c: revision 1.24 sys/dev/ic/rt2860.c: revision 1.28 sys/dev/ic/rt2860var.h: revision 1.5 enable rt2860 power management code adjust to fit netbsd: make suspend,resume functions match desired pmf* prototypes remove wakeup/activate wrapper functions avoid jumping to NULL on resume by initializing if_stop the problem machine has other issues on resume, so there might be further issues, but it's an improvement over a jump to NULL. tested by Riccardo Mottola
1.24.4.1	10-Jun-2019	christos	Sync with HEAD
1.24.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7	21-Jan-2005	yamt	- switch to dev/rtl8169.c from pci/if_re.c. - enable re@cardbus. both of re@pci and re@cardbus seem to work for me.
1.6	30-Oct-2004	thorpej	branches: 1.6.4; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.5	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.4	01-Jun-2004	mrg	branches: 1.4.2; 1.4.4; #include <sys/device.h> call ether_ifdetach() and if_detach() if attaching fails...
1.3	30-May-2004	toshii	rcsid
1.2	30-May-2004	toshii	Fix brokenness I introduced while debugging. - initialize rxlenmask when 8139C+ - remove meaningless memsets.
1.1	28-May-2004	toshii	Add re(4) RealTek 8139C+/8169/8169S/8110S PCI ethernet adapter driver, from FreeBSD. I haven't tested this code very well, but it seems to work fairly well for me.
1.4.4.7	24-Jan-2005	skrll	Sync with HEAD.
1.4.4.6	02-Nov-2004	skrll	Sync with HEAD.
1.4.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.4.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.4.4.2	03-Aug-2004	skrll	Sync with HEAD
1.4.4.1	01-Jun-2004	skrll	file if_re.c was added on branch ktrace-lwp on 2004-08-03 10:49:08 +0000
1.4.2.5	21-Jun-2004	tron	branches: 1.4.2.5.2; Pull up revision 1.4 (requested by toshii in ticket #517): call ether_ifdetach() and if_detach() if attaching fails...
1.4.2.4	21-Jun-2004	tron	Pull up revision 1.2 (requested by toshii in ticket #517): Fix brokenness I introduced while debugging. - initialize rxlenmask when 8139C+ - remove meaningless memsets.
1.4.2.3	21-Jun-2004	tron	Pull up revision 1.3 (requested by toshii in ticket #517): rcsid
1.4.2.2	21-Jun-2004	tron	Pull up revision 1.1 (requested by toshii in ticket #517): Add re(4) RealTek 8139C+/8169/8169S/8110S PCI ethernet adapter driver, from FreeBSD. I haven't tested this code very well, but it seems to work fairly well for me.
1.4.2.1	01-Jun-2004	tron	file if_re.c was added on branch netbsd-2-0 on 2004-06-21 16:41:45 +0000
1.4.2.5.2.2	01-Dec-2005	tron	Apply patch (requested by riz in ticket #10160): Match two candidates. From Francois Romieu in private mail.
1.4.2.5.2.1	01-Dec-2005	tron	Pull up following revision(s) (requested by riz in ticket #10156): sys/dev/pci/if_sk.c: revision 1.18 via patch sys/dev/pci/if_re.c: patch Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.6.4.1	29-Apr-2005	kent	sync with -current
1.53	28-Feb-2024	jakllsch	Add Killer E2600, adapted from PR 57969
1.52	18-Feb-2020	msaitoh	branches: 1.52.24; Whitespace fix. No functional change.
1.51	30-Jan-2020	jmcneill	Prefer memory space register mapping over IO space. Make an exception for RTL8169SC, as the FreeBSD driver says memory mapped IO doesn't work there.
1.50	14-Nov-2019	msaitoh	branches: 1.50.2; Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.49	13-Nov-2018	jdolecek	branches: 1.49.4; use pci_intr_establish_xname()
1.48	18-May-2017	jakllsch	branches: 1.48.2; 1.48.8; 1.48.10; Re-enable RTKQ_IM_HW. Timeout issues appear to be fixed since src/sys/dev/ic/rtl8169.c r1.151
1.47	01-May-2017	jakllsch	Disable interrupt moderation, as it is it causes watchdog timeouts on some chips/systems.
1.46	19-Apr-2017	jmcneill	branches: 1.46.2; Performance improvements for PCIe and 8168 based devices: - When using the countdown timer for interrupt moderation on PCIe devices, use a timer rate value based on a 125MHz PCIe reference clock instead of 33 MHz. - For 8168 based devices, ditch the countdown timer and instead use the (undocumented) hardware interrupt moderation feature. - Support TSOv4 on 8168D and later devices.
1.45	14-Dec-2015	jakllsch	branches: 1.45.2; 1.45.4; Switch PCI re(4) attachment from pci_intr_map() to pci_intr_alloc()/pci_intr_release(). This enables MSI where available.
1.44	03-May-2015	matt	Deal with 64-bit BARs
1.43	29-Mar-2014	christos	branches: 1.43.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.42	30-Mar-2013	christos	branches: 1.42.4; remove trailing whitespace
1.41	30-Jan-2012	drochner	branches: 1.41.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.40	27-Jul-2010	jakllsch	branches: 1.40.8; 1.40.12; Use 64-bit DMA tag (if available) on PCI-E re(4) NICs.
1.39	27-Jul-2010	jakllsch	Make the Realtek PCI network interfaces detachable. Use __arraycount to avoid null table entries. Miscellaneous other touch-ups in bus front-ends.
1.38	02-Sep-2009	tsutsui	branches: 1.38.2; 1.38.4; Move pmf(9) calls to MI re_attach() and re_detach() from each backend.
1.37	29-Aug-2009	tsutsui	Remove unnecessary #includes.
1.36	25-Apr-2009	tsutsui	Remove suffix "B" from rtk_name of PCI_PRODUCT_REALTEK_RT8168 devices. All 8168/8111 variants (8168/8168B/8168C/8168CP/8168D/8111B/8111C/8111CP) have the same PCI device ID.
1.35	23-Aug-2008	tnn	branches: 1.35.2; 1.35.4; 1.35.8; 1.35.12; Add support for the Realtek 8102E/8102EL PCIe 10/100 Ethernet devices. From FreeBSD.
1.34	25-Apr-2008	tsutsui	branches: 1.34.2; 1.34.4; 1.34.8; Split device_t/softc for re(4) and rtk(4).
1.33	10-Apr-2008	cegger	branches: 1.33.2; use aprint_*_dev and device_xname
1.32	11-Dec-2007	lukem	branches: 1.32.8; use __KERNEL_RCSID()
1.31	09-Dec-2007	jmcneill	branches: 1.31.2; Merge jmcneill-pm branch.
1.30	19-Oct-2007	ad	branches: 1.30.4; 1.30.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.29	07-Aug-2007	simonb	branches: 1.29.2; 1.29.6; Fix a tyop.
1.28	19-Jul-2007	dsl	branches: 1.28.4; 1.28.6; include sys/cdefs.h
1.27	21-Mar-2007	tsutsui	branches: 1.27.4; Cleanup handling of quikrs on each RealTek chip: - replace rtk_type member in rtk_softc which has chip types with new rtk_quirk that represents quirks on each chip: - RTKQ_8129 doesn't have internal MII (used in rtk(4)) - RTKQ_8139CPLUS has different register layout (for re(4)) - RTKQ_8169NONS (original 8169) requires some settings on init - RTKQ_PCIE requires different settings in setmulti so that we don't have to check each hwrev values or types everywhere and newer variants will also work without changes if they don't have other quirks (sc_rev is unchenged for now for reference to the Realtek's driver) - don't check hwrev register in re_pci_match() but check only PCI_VENDER(), PCI_PRODUCT() and PCI_REVISION() so that we no longer have to map pci space there - add a new HWREV value for another 8168 variant - try to map PCI mem space more properly - remove (probably unneeded) ifp->if_baudrate initialization Tested on a newer 8168 variant by Dennis den Brok on tech-kern, and also tested on 8139 and 8169C on macppc, and 8139C+ on landisk by me.
1.26	05-Mar-2007	tsutsui	branches: 1.26.2; 1.26.4; 1.26.6; Also check both I/O and memory space on unmap. Pointer out by yamt.
1.25	03-Mar-2007	tsutsui	Add an entry to match hwrev RTK_HWREV_8100E_SPIN2 too. XXX: should we use PCI_REVISION() rather than hwrev for 8139C+ as OpenBSD does?
1.24	21-Feb-2007	thorpej	Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.23	04-Feb-2007	tsutsui	branches: 1.23.2; Add an 8169SC variant to re_devs. From Darrin B.Jewell.
1.22	04-Feb-2007	tsutsui	Try to map both PCI I/O and memory space.
1.21	24-Nov-2006	christos	branches: 1.21.2; fix capitalization of NetBSD; from Zapher
1.20	18-Nov-2006	tsutsui	Rename re_pci_probe() -> re_pci_match().
1.19	18-Nov-2006	tsutsui	- sync HWREV definitions with FreeBSD - recognize 8169SC/8100E/8101E (untested, but won't hurt working devices) XXX: 8100E/8101E don't support 1000M; need some flag to set if_baudrate
1.18	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.17	03-Nov-2006	tsutsui	Pull another 8139C+ fix from FreeBSD if_re.c rev 1.68: Perform hardware diagnostic only on the original RTL8169, which was the only device that really needed it. (i.e. a possible hardware bug when the NIC was put on a 64bit PCI slot) Tested with on-board 8139C+ by Brian A. Seklecki. Note this change might also fix PR kern/34952 (because re_diag() is no longer called on 8169S/8110S), but I'm not sure if the re(4) chip was properly initialized in such case.
1.16	28-Oct-2006	christos	PR/34933: Sami Kantoluoto: RealTek 8168B found from Asus P5B motherboard is not recognized
1.15	27-Oct-2006	tsutsui	Use common macro rather than local one.
1.14	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.13	18-Jun-2006	christos	branches: 1.13.4; 1.13.6; PR/33761: dieter roelants: NetBSD doesn't recognize a RealTek 8168B nic
1.12	25-Dec-2005	rpaulo	branches: 1.12.4; 1.12.8; 1.12.14; 1.12.16; PR/32386: Dawid Szymanski (arhea). Add support for the 8169SB chipset.
1.11	23-Nov-2005	riz	Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.10	15-Jun-2005	cube	branches: 1.10.2; 1.10.8; Match two candidates. From Francois Romieu in private mail.
1.9	09-May-2005	martin	Add support for Corega CG-LAPCIGT Gigabit Ethernet. From PR kern/30173.
1.8	27-Feb-2005	perry	branches: 1.8.2; nuke trailing whitespace
1.7	23-Feb-2005	yamt	re_pci_probe: don't use i/o space unless RE_USEIOSPACE.
1.6	13-Feb-2005	jdolecek	constify
1.5	22-Jan-2005	briggs	branches: 1.5.2; Put the headers in a more standard order with machine-specific and dev/bus-specific headers after the sys and net headers. This allows it to actually compile on macppc, among other things.
1.4	18-Jan-2005	yamt	- add rcsid. - restore Bill Paul's copyright notice.
1.3	15-Jan-2005	kanaoka	branches: 1.3.2; 1.3.4; - Make sure D0 state if device has Power Managmet capability. - Use aprint_*. - Change some cosmetic. XXX not tested.
1.2	13-Jan-2005	kanaoka	- Move re_diag() from bus independend code to PCI attachment code. XXX: I tested cardbus device,but not PCI device.
1.1	23-Dec-2004	jonathan	Snapshot of incomplete-but-working split of re(4) driver into a bus-independent backend, with PCI and CardBus attachment code. The committed code has two serious bugs: 1. The driver makes no attempt to recover resources when a (Cardbus) instance is removed; bus resources are leaked. 2. In testing with a NetGear GA-511, the Cardbus card never responded to a reset/wakeup if the card is powered down after attachment. So for now, leave cardbus instances powered up at attachment (insertion, or at boot if a card is already present). That aside, it acutally works on my GA-511. Committed as-is despite the bugs, after repeated requests to make the code available for further testing. Also requires sys/dev/mii/miidevs rev 1.54 -> 1.55, and consequent regen of miidevs{,_data}.h.
1.3.4.7	11-Dec-2005	christos	Sync with head.
1.3.4.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.3.4.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.3.4.4	15-Feb-2005	skrll	Sync with HEAD.
1.3.4.3	24-Jan-2005	skrll	Sync with HEAD.
1.3.4.2	17-Jan-2005	skrll	Sync with HEAD.
1.3.4.1	15-Jan-2005	skrll	file if_re_pci.c was added on branch ktrace-lwp on 2005-01-17 19:31:24 +0000
1.3.2.1	29-Apr-2005	kent	sync with -current
1.5.2.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.2.7	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1852): sys/dev/ic/rtl81x9var.h: revision 1.39 sys/dev/cardbus/if_re_cardbus.c: revision 1.13 sys/dev/pci/if_re_pci.c: revision 1.27 via patch sys/dev/pci/if_rtk_pci.c: revision 1.32 sys/dev/cardbus/if_rtk_cardbus.c: revision 1.32 sys/dev/ic/rtl8169.c: revision 1.84, 1.85 sys/dev/ic/rtl81x9reg.h: revision 1.27 sys/dev/ic/rtl81x9.c: revision 1.72 Cleanup handling of quikrs on each RealTek chip: Pull a fix for PCIe variants from FreeBSD
1.8.2.6	10-Mar-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1680): sys/dev/pci/if_re_pci.c: revision 1.25-1.26 Add an entry to match hwrev RTK_HWREV_8100E_SPIN2 too. Also check both I/O and memory space on unmap. Pointer out by yamt.
1.8.2.5	03-Mar-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1673): distrib/sets/lists/man/mi 1.844, 1.871 share/man/man4/Makefile 1.372, 1.379 via patch share/man/man4/re.4 1.8, 1.9, 1.11, 1.12 share/man/man4/rgephy.4 1.1 share/man/man4/rlphy.4 1.1, 1.2 sys/arch/amd64/conf/GENERIC 1.79, 1.89 sys/arch/amd64/conf/INSTALL 1.45, 1.49 sys/arch/i386/conf/GENERIC 1.717, 1.747 sys/arch/i386/conf/GENERIC_LAPTOP 1.164 sys/arch/i386/conf/INSTALL 1.277, 1.283 sys/arch/i386/conf/INSTALL_LAPTOP 1.96, 1.98 sys/arch/i386/conf/XEN2_DOM0 1.22 via patch sys/arch/macppc/conf/GENERIC 1.220, 1.246 sys/arch/macppc/conf/INSTALL 1.96, 1.100 sys/dev/cardbus/if_re_cardbus.c 1.10 sys/dev/cardbus/if_rtk_cardbus.c 1.29-1.31 sys/dev/ic/rtl8169.c 1.14, 1.20, 1.24, 1.25, 1.28-1.61, 1.63, 1.64-1.81 via patch sys/dev/ic/rtl81x9.c 1.52, 1.54-1.63, 1.65, 1.67-1.70 sys/dev/ic/rtl81x9reg.h 1.15-1.26 sys/dev/ic/rtl81x9var.h 1.19, 1.21-1.37 sys/dev/mii/files.mii 1.38 sys/dev/mii/miidevs 1.62, 1.64 sys/dev/mii/rgephy.c 1.10, 1.13-1.16 sys/dev/mii/rlphy.c 1.1, 1.6, 1.7, 1.11 via patch sys/dev/pci/if_re_pci.c 1.13, 1.15-1.17, 1.19-1.23 sys/dev/pci/if_rtk_pci.c 1.25, 1.28, 1.29, 1.31 sys/dev/pci/pcidevs 1.851, 1.852 on re(4): - improve stability (I believe ;-) - add a workaround for hardware ip4csum-tx bug - support newer chips (8169SB/SC, PCIe based 8168 etc.) - fix 8139C+ support - enable hardware VLAN - misc bus_dma(9) fix (which makes re(4) work on mips ports) on rtk(4): - fix kern/31348 - fix possible panic on dreamcast
1.8.2.4	29-Dec-2005	riz	Pull up following revision(s) (requested by rpaulo in ticket #1070): sys/dev/pci/if_re_pci.c: revision 1.12 sys/dev/ic/rtl81x9reg.h: revision 1.14 PR/32386: Dawid Szymanski (arhea). Add support for the 8169SB chipset. From FreeBSD.
1.8.2.3	24-Nov-2005	tron	Pull up following revision(s) (requested by riz in ticket #993): sys/dev/pci/if_sk.c: revision 1.18 sys/dev/pci/if_re_pci.c: revision 1.11 Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.8.2.2	24-Nov-2005	tron	Pull up following revision(s) (requested by riz in ticket #995): sys/dev/pci/if_re_pci.c: revision 1.10 Match two candidates. From Francois Romieu in private mail.
1.8.2.1	10-May-2005	tron	Pull up revision 1.9 (requested by martin in ticket #273): Add support for Corega CG-LAPCIGT Gigabit Ethernet. From PR kern/30173.
1.10.8.1	29-Nov-2005	yamt	sync with head.
1.10.2.6	21-Jan-2008	yamt	sync with head
1.10.2.5	27-Oct-2007	yamt	sync with head.
1.10.2.4	03-Sep-2007	yamt	sync with head.
1.10.2.3	26-Feb-2007	yamt	sync with head.
1.10.2.2	30-Dec-2006	yamt	sync with head.
1.10.2.1	21-Jun-2006	yamt	sync with head.
1.12.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.12.14.1	22-Jun-2006	chap	Complete a sync sys/ with head.
1.12.8.3	15-Sep-2006	yamt	revert unrelated changes erroneously slipped in.
1.12.8.2	26-Jun-2006	yamt	sync with head.
1.12.8.1	24-May-2006	yamt	sync with head.
1.12.4.1	09-Sep-2006	rpaulo	sync with head
1.13.6.2	10-Dec-2006	yamt	sync with head.
1.13.6.1	22-Oct-2006	yamt	sync with head
1.13.4.3	09-Feb-2007	ad	Sync with HEAD.
1.13.4.2	12-Jan-2007	ad	Sync with head.
1.13.4.1	18-Nov-2006	ad	Sync with head.
1.21.2.6	18-Aug-2009	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1339): sys/dev/ic/rtl8169.c: revisions 1.107, 1.114, 1.115, 1.116, 1.117 (via patch), 1.118 (via patch), 1.119, 1.121 sys/dev/ic/rtl81x9reg.h: revisions 1.36, 1.37, 1.38, 1.39 sys/dev/ic/rtl81x9var.h: revision 1.47 sys/dev/mii/rgephy.c: revisions 1.16, 1.18, 1.19, 1.27 (via patch) sys/dev/mii/rgephyreg.h: revision 1.3 sys/dev/pci/if_re_pci.c: revision 1.36 - add support for RTL8211B(L) to rgephy(4) - add a wakeup instruction for rgephy(4) on newer re(4) chips - detect RTL8169CP, RTL8168D/8111D, and RTL8103E variants - fix rgephy(4) problem on RTL8111D reported on current-users: http://mail-index.NetBSD.org/current-users/2009/04/12/msg008977.html http://mail-index.NetBSD.org/current-users/2009/04/19/msg009096.html - pull more quirk handling from FreeBSD - fix RX hwcksum for DESCV2 chips for PR kern/40605 - remove "B" suffix from RTL8168 device names in attach message
1.21.2.5	31-Mar-2009	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1297): src/sys/dev/ic/rtl8169.c 1.104 (via patch), 1.105 (via patch), 1.106 (via patch), 1.108, 1.109 (via patch), 1.110, 1.111, 1.112, 1.113 sys/dev/ic/rtl81x9reg.h: revision 1.30 - 1.35 sys/dev/ic/rtl81x9var.h: revision 1.42 - 1.45 sys/dev/pci/if_re_pci.c: revision 1.35 Add support for 8168C/8111C/8102E chips. Fix MAC reset issue (PR kern/41009) Add/fix HW checksum support (PR kern/40955)
1.21.2.4	16-May-2007	jdc	Pull up revision 1.27 via patch (requested by tsutsui in ticket #640). Cleanup handling of quikrs on each RealTek chip: - replace rtk_type member in rtk_softc which has chip types with new rtk_quirk that represents quirks on each chip: - RTKQ_8129 doesn't have internal MII (used in rtk(4)) - RTKQ_8139CPLUS has different register layout (for re(4)) - RTKQ_8169NONS (original 8169) requires some settings on init - RTKQ_PCIE requires different settings in setmulti so that we don't have to check each hwrev values or types everywhere and newer variants will also work without changes if they don't have other quirks (sc_rev is unchenged for now for reference to the Realtek's driver) - don't check hwrev register in re_pci_match() but check only PCI_VENDER(), PCI_PRODUCT() and PCI_REVISION() so that we no longer have to map pci space there - add a new HWREV value for another 8168 variant - try to map PCI mem space more properly - remove (probably unneeded) ifp->if_baudrate initialization Tested on a newer 8168 variant by Dennis den Brok on tech-kern, and also tested on 8139 and 8169C on macppc, and 8139C+ on landisk by me.
1.21.2.3	10-Mar-2007	bouyer	branches: 1.21.2.3.2; Pull up following revision(s) (requested by tsutsui in ticket #502): sys/dev/pci/if_re_pci.c: revision 1.25-1.26 Add an entry to match hwrev RTK_HWREV_8100E_SPIN2 too. Also check both I/O and memory space on unmap. Pointer out by yamt.
1.21.2.2	10-Feb-2007	tron	Pull up following revision(s) (requested by tsutsui in ticket #437): sys/dev/pci/if_re_pci.c: revision 1.23 Add an 8169SC variant to re_devs. From Darrin B.Jewell.
1.21.2.1	10-Feb-2007	tron	Pull up following revision(s) (requested by tsutsui in ticket #437): sys/dev/pci/if_re_pci.c: revision 1.22 Try to map both PCI I/O and memory space.
1.21.2.3.2.1	03-Jun-2007	wrstuden	Catch up with now-somewhat-dated netbsd-4. These changes took longer than expected for me to actually get around to merging.
1.23.2.3	24-Mar-2007	yamt	sync with head.
1.23.2.2	12-Mar-2007	rmind	Sync with HEAD.
1.23.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.26.6.1	29-Mar-2007	reinoud	Pullup to -current
1.26.4.1	11-Jul-2007	mjf	Sync with head.
1.26.2.3	23-Oct-2007	ad	Sync with head.
1.26.2.2	20-Aug-2007	ad	Sync with HEAD.
1.26.2.1	10-Apr-2007	ad	Sync with head.
1.27.4.1	15-Aug-2007	skrll	Sync with HEAD.
1.28.6.2	19-Jul-2007	dsl	include sys/cdefs.h
1.28.6.1	19-Jul-2007	dsl	file if_re_pci.c was added on branch matt-mips64 on 2007-07-19 22:04:23 +0000
1.28.4.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.28.4.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.28.4.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.28.4.2	05-Oct-2007	joerg	Remove the power management nonsense in the PCI versions of rtk(4) and re(4). It is not really likely to work anyway. Use PNP power management for the cardbus attachment as well and remove the power management handlers from the chipset code.
1.28.4.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.29.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.29.2.2	09-Jan-2008	matt	sync with HEAD
1.29.2.1	06-Nov-2007	matt	sync with HEAD
1.30.6.1	11-Dec-2007	yamt	sync with head.
1.30.4.1	26-Dec-2007	ad	Sync with head.
1.31.2.1	13-Dec-2007	bouyer	Sync with HEAD
1.32.8.2	28-Sep-2008	mjf	Sync with HEAD.
1.32.8.1	02-Jun-2008	mjf	Sync with HEAD.
1.33.2.1	18-May-2008	yamt	sync with head.
1.34.8.1	19-Oct-2008	haad	Sync with HEAD.
1.34.4.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.34.2.3	11-Aug-2010	yamt	sync with head.
1.34.2.2	16-Sep-2009	yamt	sync with head
1.34.2.1	04-May-2009	yamt	sync with head.
1.35.12.1	21-Apr-2010	matt	sync to netbsd-5
1.35.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.35.4.1	19-Jun-2009	snj	Pull up following revision(s) (requested by tsutsui in ticket #821): sys/dev/ic/rtl8169.c: revisions 1.107, 1.114-1.119, 1.121 sys/dev/ic/rtl81x9reg.h: revisions 1.36-1.39 sys/dev/ic/rtl81x9var.h: revision 1.47 sys/dev/mii/rgephy.c: revision 1.27 via patch sys/dev/pci/if_re_pci.c: revision 1.36 remove extra semicolons. -- Add HWREV values of RTL8168CP and RTL8168D. From FreeBSD. XXX: needs more quirk handling after wakeup for newer chips. -- Add HWREV of RTL8102EL variant. From FUKAUMI Naoki. -- Assume an unknown HWREV chip has the same features with the latest one. -- Remove suffix "B" from rtk_name of PCI_PRODUCT_REALTEK_RT8168 devices. All 8168/8111 variants (8168/8168B/8168C/8168CP/8168D/8111B/8111C/8111CP) have the same PCI device ID. -- Remove magic reset sequence except wakeup for rev 2 chips which breaks 8111D. Problem reported and fix confirmed by Thomas Bieg on current-users. Also tested on 8111C (no bad side effect) by several users privately. -- Pull some changes for newer chips from FreeBSD: - pull MACSTAT and CMDSTOP quirks for 8168/8111 chips - always set CPLUSCMD_PCI_MRW on reset - set VLANSTRIP and RXCSUM_ENB bits on CPLUS register per if_capenable Tested on 8111C and 8111D by several users, and no bad side effect on my old 8169S. -- Remove unused sc_rev settings (all quirks are handled by sc_quirk) and merge HWREV cases which have the same quirks. -- - rename RTK_HWREV_8102EL_SPIN2 -> RTK_HWREV_8103E - add a HWREV value for 8168DP Per Realtek's Linux drivers. -- Two fixes for RX hwcsum on DESCV2 chips: * On checking TCPv4/UDPv4 RX checksum on DESCV2 chips, also check RE_RDESC_VLANCTL_IPV4 bit because those DESCV2 chips may also recognize IPv6 packets and set RE_PROTOID_TCPIP or RE_PROTOID_UDPIP bits for TCPv6/UDPv6 packets. This may fix PR kern/40605. * According to Realtek's Linux driver, DESCV2 chips don't set RE_PROTOID_IP for non-TCP/UDP IP packets (set only RE_RDESC_VLANCTL_IPV[46]) so remove PROTOID check for IPv4 RX cheksum on DESCV2 chips.
1.35.2.1	28-Apr-2009	skrll	Sync with HEAD.
1.38.4.1	05-Mar-2011	rmind	sync with head
1.38.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.40.12.1	18-Feb-2012	mrg	merge to -current.
1.40.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.40.8.1	17-Apr-2012	yamt	sync with head
1.41.6.3	03-Dec-2017	jdolecek	update from HEAD
1.41.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.41.6.1	23-Jun-2013	tls	resync from head
1.42.4.1	18-May-2014	rmind	sync with head
1.43.6.3	28-Aug-2017	skrll	Sync with HEAD
1.43.6.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.43.6.1	06-Jun-2015	skrll	Sync with HEAD
1.45.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.45.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.46.2.2	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.46.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.48.10.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.48.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.48.10.1	10-Jun-2019	christos	Sync with HEAD
1.48.8.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.48.2.1	19-Nov-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1449: sys/dev/pci/if_re_pci.c 1.50 share/man/man4/re.4 1.17-1.18 Add D-Link DGE-530T Rev. C,D and TP-Link TG-3468 v2,v3. Add note about UDP checksum offload feature to the document.
1.49.4.1	19-Nov-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #451: sys/dev/pci/if_re_pci.c 1.50 (patch) share/man/man4/re.4 1.17-1.18 (patch) Add D-Link DGE-530T Rev. C,D and TP-Link TG-3468 v2,v3. Add note about UDP checksum offload feature to the document.
1.50.2.1	29-Feb-2020	ad	Sync with head.
1.52.24.1	24-Aug-2024	martin	Pull up following revision(s) (requested by jakllsch in ticket #805): sys/dev/pci/pcidevs: revision 1.1503 sys/dev/pci/if_re_pci.c: revision 1.53 Add (Realtek/)Killer E2600 Ethernet Add Killer E2600, adapted from PR 57969
1.42	29-Oct-2025	pgoyette	One more attempt to revert
1.41	29-Oct-2025	pgoyette	Undo recent changes since they seem to have broken something. Will reexamine moving forward from last time we synched with OpenBSD - January 20, 2021
1.40	28-Oct-2025	pgoyette	Update microcode, initialization and reset behavior. From OpenBSD, if_rge.c rev 1.23 and if_rgereg.h rev 1.10. XXX In the same commit, two chip variants were de-supported, with an expectation that those two variants were never shipped in any product. It's not clear that this is true, so I did NOT remove support for them.
1.39	27-Oct-2025	pgoyette	Update RTL8125 microcode. From openbsd, if_rge.c rev 1.13, if_rgereg.h rev 1.7
1.38	26-Oct-2025	pgoyette	We only suppoprt the MAC_CFG2_8126 (hwrev 0x649000000). Disable the others. Based on driver code from Realtek, all three of these are "8126A" models ...
1.37	25-Oct-2025	pgoyette	Clean up interrupt handling and add rge_detach(). Disable recognition of 8126-B until we know more about how to drive it.
1.36	23-Oct-2025	pgoyette	Recognize the RT8126 Rev B. XXX Assume for now that it works the same as Rev A. Neither XXX Free- nor OpenBSD give any clues on Rev B details.
1.35	21-Oct-2025	pgoyette	Add MODULE glue
1.34	04-Feb-2025	jmcneill	rge(4): Add support for RTL8126 Patch from FUKAUMI Naoki in PR# 59044
1.33	10-Nov-2024	mlelstv	Add MBUFTRACE
1.32	29-Jun-2024	riastradh	branches: 1.32.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.31	18-Jan-2024	msaitoh	rge(4): Print HW revision.
1.30	21-Dec-2023	skrll	G/C sc_task and #include "opt_net_mpsafe.h"
1.29	16-Dec-2023	mlelstv	- handle stuck transmitter (descriptor still owned) - restart send queue after transmit - count output packets - use deferred start Should fix PR 57694
1.28	19-Oct-2023	mrg	rge: properly handle mbuf allocation failures in rx interrupts several changes that should fix crashes seen after an mbuf allocation failure: - create RX ring dma maps with BUS_DMA_ALLOCNOW, so that any future bus_dmamap_load*() call will succeed. this avoids one error case in rge_newbuf(), that similar cases in eg wm(4) actually call panic for now (i think this idea can be copied into wm(4) as well.) - extract the RX descriptor set into a common function that both rge_newbuf() and rge_rxeof() can both use. it's almost identical to the old rge_discard_rxbuf() except it also sets the rge_addr (this is needed for the newbuf case.) - move the bus_dmamap_unload() into rge_newbuf(), so that the existing mbuf will remain mapped until a new mbuf is allocated. (this part is what should fix crashes seen by wiz and Chavdar, as the unload follow by sync is what triggers the assert in x86 bus_dma. without the assert, it will would have shortly triggered a page fault.) remove the assignment to NULL for the rxq mbuf pointer, it is required for reload. - add a couple of missing if_statinc() calls. tested on amd64 and arm64.
1.27	09-Oct-2023	riastradh	rge(4): callout_stop can't be right here; just do callout_halt. XXX pullup-10
1.26	05-Oct-2023	mrg	rge(4): check for all errors in rx buffer allocation should fix a crash seen by by Chavdar Ivanov reported on current-users. move the rx and tx list clean up into their own functions, and call the rx clean up function from the init function if something fails. this should fix a potential leak in this case, and generally frees up memory that won't be used without a successful init phase again. slight application of 'static', much more could be done.
1.25	21-Dec-2022	nonaka	Update the Rx descriptor based on the vendor driver for Linux. This fixes a panic on RTL8125. Patch from OpenBSD if_rge.c r1.20, if_rgereg.h r1.8. Tested by msaitoh@n.o.
1.24	24-Sep-2022	thorpej	branches: 1.24.4; Remove unnecessary include of <sys/malloc.h>.
1.23	07-Jul-2022	skrll	Trailing whitespace
1.22	11-Oct-2021	msaitoh	Avoid unaligned access in rge_get_macaddr(). Found by kUBSan.
1.21	11-Oct-2021	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.20	11-Oct-2021	msaitoh	Add pmf_device_register().
1.19	08-May-2021	thorpej	Use pci_compatible_match().
1.18	02-Mar-2021	knakahara	branches: 1.18.4; Fix build failure for options NET_MPSAFE.
1.17	01-Mar-2021	jakllsch	Update rge(4) from older OpenBSD, finish porting. Should consider merging this all into re(4) and rgephy(4) someday. Some cleanup tasks remain here.
1.16	27-Jan-2021	jakllsch	if_rge.c: avoid aprint_error*() abuse
1.15	20-Jan-2021	jakllsch	Sync with OpenBSD if_rge.c r1.4 """ Change users of IFQ_SET_MAXLEN() and IFQ_IS_EMPTY() to use the "new" API. """
1.14	30-May-2020	sevan	branches: 1.14.2; Use device_xname() to access dv_xname
1.13	30-May-2020	sevan	Switch from printf to aprintf_error/normal
1.12	30-May-2020	sevan	Initialise sc->sc_dev, otherwise unnamed interfaces show up and things go downhill from there.
1.11	20-May-2020	sevan	use device_private() instead of casting self as it doesn't work here. Bump rcs tag which was missed in r1.9
1.10	30-Apr-2020	sevan	Sync with r1.3 from OpenBSD "Use correct bit mask to promptly exit rge_set_phy_power() and rge_exit_oob() timeout loops. Spotted by CID 1491296 and 1491309."
1.9	29-Feb-2020	thorpej	branches: 1.9.4; Use the 64-bit PCI tag, if available.
1.8	27-Feb-2020	sevan	Cast to make up for the difference in type between our bus_dma_map(9) and theirs. Where we expect kvap to be of void *, they expect caddr_t (typedef char ).
1.7	27-Feb-2020	sevan	Add an implementation of MCLGETI. Borrowed from sys/dev/pci/if_bwfm_pci.c
1.6	13-Feb-2020	sevan	redundant include
1.5	04-Feb-2020	skrll	Trailing whitespace
1.4	04-Feb-2020	skrll	Adopt <net/if_stats.h>
1.3	12-Jan-2020	sevan	branches: 1.3.2; Start to switch from OpenBSD timeout(9) API to callout(9) and workqueue(9) Replace functions with NetBSD counterparts.
1.2	11-Jan-2020	sevan	Start porting to NetBSD. (WIP, incomplete)
1.1	11-Jan-2020	sevan	Import rge(4) from OpenBSD for Realtek's RTL8125 based 2.5GbE network cards.
1.3.2.3	29-Feb-2020	ad	Sync with head.
1.3.2.2	17-Jan-2020	ad	Sync with head.
1.3.2.1	12-Jan-2020	ad	file if_rge.c was added on branch ad-namecache on 2020-01-17 21:47:31 +0000
1.9.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.9.4.1	29-Feb-2020	martin	file if_rge.c was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.14.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.18.4.1	13-May-2021	thorpej	Sync with HEAD.
1.24.4.6	02-Jun-2025	martin	ticket #1124: minor fix to make it compile on this branch
1.24.4.5	01-Jun-2025	martin	Pull up following revision(s) (requested by jmcneill in ticket #1124): sys/dev/pci/if_rge.c: revision 1.34 sys/dev/pci/if_rgereg.h: revision 1.9 share/man/man4/rge.4: revision 1.5 rge(4): Add support for RTL8126 Patch from FUKAUMI Naoki in PR# 59044
1.24.4.4	22-Aug-2024	martin	Pull up following revision(s) (requested by mrg in ticket #782): sys/dev/pci/if_rge.c: revision 1.29 - handle stuck transmitter (descriptor still owned) - restart send queue after transmit - count output packets - use deferred start Should fix PR 57694
1.24.4.3	22-Oct-2023	martin	Pull up following revision(s) (requested by mrg in ticket #434): sys/dev/pci/if_rge.c: revision 1.26 sys/dev/pci/if_rge.c: revision 1.28 rge(4): check for all errors in rx buffer allocation should fix a crash seen by by Chavdar Ivanov reported on current-users. move the rx and tx list clean up into their own functions, and call the rx clean up function from the init function if something fails. this should fix a potential leak in this case, and generally frees up memory that won't be used without a successful init phase again. slight application of 'static', much more could be done. rge: properly handle mbuf allocation failures in rx interrupts several changes that should fix crashes seen after an mbuf allocation failure: - create RX ring dma maps with BUS_DMA_ALLOCNOW, so that any future bus_dmamap_load*() call will succeed. this avoids one error case in rge_newbuf(), that similar cases in eg wm(4) actually call panic for now (i think this idea can be copied into wm(4) as well.) - extract the RX descriptor set into a common function that both rge_newbuf() and rge_rxeof() can both use. it's almost identical to the old rge_discard_rxbuf() except it also sets the rge_addr (this is needed for the newbuf case.) - move the bus_dmamap_unload() into rge_newbuf(), so that the existing mbuf will remain mapped until a new mbuf is allocated. (this part is what should fix crashes seen by wiz and Chavdar, as the unload follow by sync is what triggers the assert in x86 bus_dma. without the assert, it will would have shortly triggered a page fault.) remove the assignment to NULL for the rxq mbuf pointer, it is required for reload. - add a couple of missing if_statinc() calls. tested on amd64 and arm64.
1.24.4.2	14-Oct-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #414): sys/dev/pci/if_rge.c: revision 1.27 rge(4): callout_stop can't be right here; just do callout_halt.
1.24.4.1	21-Dec-2022	martin	Pull up following revision(s) (requested by nonaka in ticket #14): sys/dev/pci/if_rge.c: revision 1.25 sys/dev/pci/if_rgereg.h: revision 1.7 Update the Rx descriptor based on the vendor driver for Linux. This fixes a panic on RTL8125. Patch from OpenBSD if_rge.c r1.20, if_rgereg.h r1.8. Tested by msaitoh@n.o.
1.32.2.1	02-Aug-2025	perseant	Sync with HEAD
1.14	29-Oct-2025	pgoyette	One more attempt to revert
1.13	29-Oct-2025	pgoyette	Undo recent changes since they seem to have broken something. Will reexamine moving forward from last time we synched with OpenBSD - January 20, 2021
1.12	28-Oct-2025	pgoyette	Update microcode, initialization and reset behavior. From OpenBSD, if_rge.c rev 1.23 and if_rgereg.h rev 1.10. XXX In the same commit, two chip variants were de-supported, with an expectation that those two variants were never shipped in any product. It's not clear that this is true, so I did NOT remove support for them.
1.11	27-Oct-2025	pgoyette	Update RTL8125 microcode. From openbsd, if_rge.c rev 1.13, if_rgereg.h rev 1.7
1.10	25-Oct-2025	pgoyette	Clean up interrupt handling and add rge_detach(). Disable recognition of 8126-B until we know more about how to drive it.
1.9	04-Feb-2025	jmcneill	rge(4): Add support for RTL8126 Patch from FUKAUMI Naoki in PR# 59044
1.8	21-Dec-2023	skrll	branches: 1.8.2; G/C sc_task and #include "opt_net_mpsafe.h"
1.7	21-Dec-2022	nonaka	Update the Rx descriptor based on the vendor driver for Linux. This fixes a panic on RTL8125. Patch from OpenBSD if_rge.c r1.20, if_rgereg.h r1.8. Tested by msaitoh@n.o.
1.6	28-Aug-2022	skrll	branches: 1.6.4; Trailing whitespace
1.5	01-Mar-2021	jakllsch	Update rge(4) from older OpenBSD, finish porting. Should consider merging this all into re(4) and rgephy(4) someday. Some cleanup tasks remain here.
1.4	30-May-2020	sevan	branches: 1.4.2; Use device_t for sc_dev
1.3	12-Jan-2020	sevan	branches: 1.3.2; 1.3.6; Start to switch from OpenBSD timeout(9) API to callout(9) and workqueue(9) Replace functions with NetBSD counterparts.
1.2	11-Jan-2020	sevan	Start porting to NetBSD. (WIP, incomplete)
1.1	11-Jan-2020	sevan	Import rge(4) from OpenBSD for Realtek's RTL8125 based 2.5GbE network cards.
1.3.6.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.3.6.1	12-Jan-2020	martin	file if_rgereg.h was added on branch phil-wifi on 2020-04-08 14:08:09 +0000
1.3.2.2	17-Jan-2020	ad	Sync with head.
1.3.2.1	12-Jan-2020	ad	file if_rgereg.h was added on branch ad-namecache on 2020-01-17 21:47:31 +0000
1.4.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.6.4.2	01-Jun-2025	martin	Pull up following revision(s) (requested by jmcneill in ticket #1124): sys/dev/pci/if_rge.c: revision 1.34 sys/dev/pci/if_rgereg.h: revision 1.9 share/man/man4/rge.4: revision 1.5 rge(4): Add support for RTL8126 Patch from FUKAUMI Naoki in PR# 59044
1.6.4.1	21-Dec-2022	martin	Pull up following revision(s) (requested by nonaka in ticket #14): sys/dev/pci/if_rge.c: revision 1.25 sys/dev/pci/if_rgereg.h: revision 1.7 Update the Rx descriptor based on the vendor driver for Linux. This fixes a panic on RTL8125. Patch from OpenBSD if_rge.c r1.20, if_rgereg.h r1.8. Tested by msaitoh@n.o.
1.8.2.1	02-Aug-2025	perseant	Sync with HEAD
1.11	10-May-2000	thorpej	These are no longer used.
1.10	30-Mar-2000	augustss	Remove register declarations.
1.9	23-Mar-2000	mycroft	Rather than guessing at the location of the PMCSR, use the pointer returned by pci_get_capability(). (This is, after all, a standardized interface...)
1.8	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.7	06-Mar-2000	thorpej	No longer necessary to futz with ifp->if_baudrate here.
1.6	02-Feb-2000	thorpej	Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass flags down from the parent to child vi mii_attach().
1.5	02-Feb-2000	thorpej	Bring some order to the chaos which was the MII code function naming "conventions".
1.4	12-Nov-1999	thorpej	Call mii_down() as appropriate.
1.3	04-Nov-1999	thorpej	Adapt to mii_phy_probe() change.
1.2	20-Aug-1999	sommerfeld	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.10; clean up attach messages slightly.
1.1	27-Jun-1999	drochner	branches: 1.1.2; add driver for realtek 8129/8139, from freebsd, with mii stuff kicked out
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	27-Jun-1999	thorpej	file if_rl.c was added on branch chs-ubc2 on 1999-07-01 23:34:07 +0000
1.2.10.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.2.6.2	20-Oct-1999	he	Pull up revisions 1.1-1.2 + patch (requested by bouyer): Back-port of the Realtek ethernet driver.
1.2.6.1	20-Aug-1999	he	file if_rl.c was added on branch netbsd-1-4 on 1999-10-20 23:39:22 +0000
1.2.4.1	15-Nov-1999	fvdl	Sync with -current
1.2.2.1	20-Nov-2000	bouyer	Remove files that are no longer on the trunck, and commit Makefile which I forgot in the batch of commits.
1.8	10-May-2000	haya	Move if_rl_pci.c:1.7 to if_rtk_pci.c.
1.7	01-May-2000	tsutsui	- Move some initialization into MI rl_attach() from each bus attachment. - Use proper CARDBUS_* macro in if_rl_cardbus.c - Use product ID read from PCI configuration space rather than from EEPROM in if_rl_pci.c. Patch sent from Masanori Kanaoka <kanaoka@ann.hi-ho.ne.jp> and some cosmetic changes by me.
1.6	30-Apr-2000	tsutsui	Modify EEPROM read functions and handle also 9356 EEPROM. Fixes kern/9861.
1.5	26-Apr-2000	tsutsui	- Allocate mbuf dynamically on Tx and use bus_dmamap_load_mbuf(9). - Call bus_dmamap_sync(9) as appropriate. - Leave only register declarations in rtl81x9reg.h and split other stuff into rtl81x9var.h. bus_dma(9) code mostly taken from thorpej's if_rtp.c.
1.4	25-Apr-2000	tsutsui	Fix two endian problems. rtk now works on macppc. - Use le32toh() on reading rxstat from the Rx DMA buffer. - rl_read_eeprom() should return EEPROM values in little endian when caller requires byte-stream data. bus_dma(9) fixes are still on-going. XXX rtk on macppc seems to make tons of CRC errors... XXX Does it work fine on i386 or others?
1.3	24-Apr-2000	tsutsui	Merge changes of sys/pci/if_rl.c rev. 1.8->1.10. (from kern/9870) Also remove unused pbase, vbase and PCI_CAP_PWRMGMT declarations.
1.2	19-Apr-2000	haya	Changes the name of RealTek driver. The new name is `rtk'. This used to be called `rl' and it conflict with RL vax disks, canonical and historical unix driver name. This changes are minimal: it only changes the name of RealTek driver. The filename of source code and a lot of the letter `rl' in source files should be changed shortly.
1.1	10-Apr-2000	haya	Incorporate the changes of RL81x9 driver provided by M. Kanaoka <kanaoka@ann.hi-ho.ne.jp>. He separated the driver into IC specific portion and bus attachment portion and added cardbus attachent.
1.3	10-May-2000	thorpej	These are no longer used.
1.2	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.1	27-Jun-1999	drochner	branches: 1.1.2; 1.1.4; 1.1.8; add driver for realtek 8129/8139, from freebsd, with mii stuff kicked out
1.1.8.2	20-Oct-1999	he	Pull up revision 1.1 (requested by boyuer): Back-port of the Realtek ethernet driver.
1.1.8.1	27-Jun-1999	he	file if_rlreg.h was added on branch netbsd-1-4 on 1999-10-20 23:44:52 +0000
1.1.4.1	20-Nov-2000	bouyer	Remove files that are no longer on the trunck, and commit Makefile which I forgot in the batch of commits.
1.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.1.2.1	27-Jun-1999	thorpej	file if_rlreg.h was added on branch chs-ubc2 on 1999-07-01 23:34:08 +0000
1.48	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.47	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.46	09-May-2015	christos	branches: 1.46.16; 1.46.18; PR/49889: Maxim Tsyplakov: support for D-Link DFE520TX
1.45	29-Mar-2014	christos	branches: 1.45.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.44	30-Mar-2013	christos	branches: 1.44.4; remove trailing whitespace
1.43	30-Jan-2012	drochner	branches: 1.43.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.42	02-Nov-2010	jakllsch	branches: 1.42.8; 1.42.12; Explicitly set PCI_COMMAND_MASTER_ENABLE, in case the platform firmware didn't.
1.41	27-Jul-2010	jakllsch	Make the Realtek PCI network interfaces detachable. Use __arraycount to avoid null table entries. Miscellaneous other touch-ups in bus front-ends.
1.40	26-Nov-2009	njoly	branches: 1.40.2; 1.40.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.39	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.38	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.37	23-Aug-2008	tsutsui	branches: 1.37.8; Also match Realtek 8138, 8139D, and 8100.
1.36	25-Apr-2008	tsutsui	branches: 1.36.2; 1.36.4; 1.36.8; Split device_t/softc for re(4) and rtk(4).
1.35	09-Dec-2007	jmcneill	branches: 1.35.10; 1.35.12; Merge jmcneill-pm branch.
1.34	06-Nov-2007	uwe	branches: 1.34.2; 1.34.4; Use device_t and device_private(). Convert attach to use aprint_*.
1.33	19-Oct-2007	ad	branches: 1.33.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.32	21-Mar-2007	tsutsui	branches: 1.32.8; 1.32.10; 1.32.14; Cleanup handling of quikrs on each RealTek chip: - replace rtk_type member in rtk_softc which has chip types with new rtk_quirk that represents quirks on each chip: - RTKQ_8129 doesn't have internal MII (used in rtk(4)) - RTKQ_8139CPLUS has different register layout (for re(4)) - RTKQ_8169NONS (original 8169) requires some settings on init - RTKQ_PCIE requires different settings in setmulti so that we don't have to check each hwrev values or types everywhere and newer variants will also work without changes if they don't have other quirks (sc_rev is unchenged for now for reference to the Realtek's driver) - don't check hwrev register in re_pci_match() but check only PCI_VENDER(), PCI_PRODUCT() and PCI_REVISION() so that we no longer have to map pci space there - add a new HWREV value for another 8168 variant - try to map PCI mem space more properly - remove (probably unneeded) ifp->if_baudrate initialization Tested on a newer 8168 variant by Dennis den Brok on tech-kern, and also tested on 8139 and 8169C on macppc, and 8139C+ on landisk by me.
1.31	25-Feb-2007	tsutsui	branches: 1.31.4; 1.31.6; 1.31.8; Try to map both PCI I/O and memory space and remove #ifdef RTK_USEIOSPACE. Tested on dreamcast, landisk, and macppc with 8139C on PCI slot.
1.30	16-Nov-2006	christos	branches: 1.30.2; 1.30.4; 1.30.6; __unused removal on arguments; approved by core.
1.29	09-Nov-2006	tsutsui	Print device revision on attach.
1.28	27-Oct-2006	tsutsui	Use common macro rather than local one.
1.27	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.26	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.25	01-Sep-2006	uwe	branches: 1.25.2; 1.25.4; s/dreamcast/sh3/ to cover landisk.
1.24	09-Mar-2006	he	Add recognition of Nortel Baystack 21 as an rtk8139 variant. Rest of information submitted in PR#33044.
1.23	11-Dec-2005	christos	branches: 1.23.4; 1.23.6; 1.23.8; 1.23.10; merge ktrace-lwp.
1.22	27-Jan-2005	jmcneill	branches: 1.22.4; 1.22.6; Add rtk_pci powerhook, from Kentaro A. Kurahone
1.21	21-Aug-2004	thorpej	branches: 1.21.4; 1.21.6; Use ANSI function decls and make use of static.
1.20	13-Feb-2004	wiz	branches: 1.20.4; RealTek -> Realtek.
1.19	25-Oct-2003	fvdl	Add definitions for 8139C+ and 8169 chips. Not used yet. From FreeBSD.
1.18	29-Jan-2003	kanaoka	branches: 1.18.2; Set RTK_ENABLE flag when it attached. This fixed multicast problem. Found by FUKAUMI Naoki <naoki@fukaumi.org>.
1.17	23-Dec-2002	tsutsui	Don't use PCI_PRODUCT_DELTA_8139 (0x1360) for args of pci_conf_read() and pci_conf_write(); use PCI_INTERRUPT_REG (0x3c) instead. (How does it work?)
1.16	23-Dec-2002	tsutsui	Fix typo in printf message.
1.15	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.14	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.13	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.12	06-Feb-2002	tsutsui	branches: 1.12.10; TAB/space cosmetics
1.11	04-Feb-2002	tsutsui	Remove chip description which is also mentioned in dev/ic/rtl81x9.c.
1.10	13-Nov-2001	lukem	add RCSID
1.9	18-Jul-2001	thorpej	Costmetic change.
1.8	08-Jul-2001	thorpej	branches: 1.8.2; Remove unnecessary include files. From Onno van der Linden.
1.7	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.6	11-Jun-2001	tron	Add support for the D-Link DFE-530TX+. Patch supplied by Gabriel Rosenkoetter in PR kern/13161.
1.5	31-Jan-2001	thorpej	branches: 1.5.2; Match the SEGA Dreamcast Broadband Adapter, and use PCI memory space on that device (XXX should be done differently than it is).
1.4	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.3	19-May-2000	tsutsui	branches: 1.3.4; 1.3.6; - Change RL_* -> RTK_* - Free bus_dma resources if attach fails - Add detach and power management code Patch from Masanori Kanaoka <kanaoka@ann.hi-ho.ne.jp> in kern/10156.
1.2	15-May-2000	thorpej	Symbol namespace cleanup: rl_ -> rtk_
1.1	10-May-2000	haya	Move if_rl_pci.c:1.7 to if_rtk_pci.c.
1.3.6.4	11-Feb-2001	bouyer	Sync with HEAD.
1.3.6.3	05-Jan-2001	bouyer	Sync with HEAD
1.3.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.6.1	19-May-2000	bouyer	file if_rtk_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:23 +0000
1.3.4.1	11-Jun-2001	he	Pull up revision 1.6 (requested by tron): Support the D-Link DFE-530TX+. Fixes PR#13161.
1.5.2.6	29-Dec-2002	thorpej	Sync with HEAD.
1.5.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.5.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.5.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.5.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.5.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.8.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.8.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.1	03-Aug-2001	lukem	update to -current
1.12.10.1	28-Apr-2003	tron	Pull up revision 1.18 (requested by kanaoka in ticket #1134): Set RTK_ENABLE flag when it attached. This fixed multicast problem. Found by FUKAUMI Naoki <naoki@fukaumi.org>.
1.18.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.18.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.18.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.18.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.18.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.4.1	17-Mar-2006	riz	Pull up following revision(s) (requested by hw in ticket #10364): sys/dev/pci/if_rtk_pci.c: revision 1.24 Add recognition of Nortel Baystack 21 as an rtk8139 variant. Rest of information submitted in PR#33044.
1.21.6.1	12-Feb-2005	yamt	sync with head.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.22.6.7	21-Jan-2008	yamt	sync with head
1.22.6.6	15-Nov-2007	yamt	sync with head.
1.22.6.5	27-Oct-2007	yamt	sync with head.
1.22.6.4	03-Sep-2007	yamt	sync with head.
1.22.6.3	26-Feb-2007	yamt	sync with head.
1.22.6.2	30-Dec-2006	yamt	sync with head.
1.22.6.1	21-Jun-2006	yamt	sync with head.
1.22.4.3	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1852): sys/dev/ic/rtl81x9var.h: revision 1.39 sys/dev/cardbus/if_re_cardbus.c: revision 1.13 sys/dev/pci/if_re_pci.c: revision 1.27 via patch sys/dev/pci/if_rtk_pci.c: revision 1.32 sys/dev/cardbus/if_rtk_cardbus.c: revision 1.32 sys/dev/ic/rtl8169.c: revision 1.84, 1.85 sys/dev/ic/rtl81x9reg.h: revision 1.27 sys/dev/ic/rtl81x9.c: revision 1.72 Cleanup handling of quikrs on each RealTek chip: Pull a fix for PCIe variants from FreeBSD
1.22.4.2	03-Mar-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1673): distrib/sets/lists/man/mi 1.844, 1.871 share/man/man4/Makefile 1.372, 1.379 via patch share/man/man4/re.4 1.8, 1.9, 1.11, 1.12 share/man/man4/rgephy.4 1.1 share/man/man4/rlphy.4 1.1, 1.2 sys/arch/amd64/conf/GENERIC 1.79, 1.89 sys/arch/amd64/conf/INSTALL 1.45, 1.49 sys/arch/i386/conf/GENERIC 1.717, 1.747 sys/arch/i386/conf/GENERIC_LAPTOP 1.164 sys/arch/i386/conf/INSTALL 1.277, 1.283 sys/arch/i386/conf/INSTALL_LAPTOP 1.96, 1.98 sys/arch/i386/conf/XEN2_DOM0 1.22 via patch sys/arch/macppc/conf/GENERIC 1.220, 1.246 sys/arch/macppc/conf/INSTALL 1.96, 1.100 sys/dev/cardbus/if_re_cardbus.c 1.10 sys/dev/cardbus/if_rtk_cardbus.c 1.29-1.31 sys/dev/ic/rtl8169.c 1.14, 1.20, 1.24, 1.25, 1.28-1.61, 1.63, 1.64-1.81 via patch sys/dev/ic/rtl81x9.c 1.52, 1.54-1.63, 1.65, 1.67-1.70 sys/dev/ic/rtl81x9reg.h 1.15-1.26 sys/dev/ic/rtl81x9var.h 1.19, 1.21-1.37 sys/dev/mii/files.mii 1.38 sys/dev/mii/miidevs 1.62, 1.64 sys/dev/mii/rgephy.c 1.10, 1.13-1.16 sys/dev/mii/rlphy.c 1.1, 1.6, 1.7, 1.11 via patch sys/dev/pci/if_re_pci.c 1.13, 1.15-1.17, 1.19-1.23 sys/dev/pci/if_rtk_pci.c 1.25, 1.28, 1.29, 1.31 sys/dev/pci/pcidevs 1.851, 1.852 on re(4): - improve stability (I believe ;-) - add a workaround for hardware ip4csum-tx bug - support newer chips (8169SB/SC, PCIe based 8168 etc.) - fix 8139C+ support - enable hardware VLAN - misc bus_dma(9) fix (which makes re(4) work on mips ports) on rtk(4): - fix kern/31348 - fix possible panic on dreamcast
1.22.4.1	17-Mar-2006	tron	Pull up following revision(s) (requested by he in ticket #1203): sys/dev/pci/if_rtk_pci.c: revision 1.24 Add recognition of Nortel Baystack 21 as an rtk8139 variant. Rest of information submitted in PR#33044.
1.23.10.1	19-Apr-2006	elad	sync with head.
1.23.8.2	03-Sep-2006	yamt	sync with head.
1.23.8.1	13-Mar-2006	yamt	sync with head.
1.23.6.1	22-Apr-2006	simonb	Sync with head.
1.23.4.1	09-Sep-2006	rpaulo	sync with head
1.25.4.2	10-Dec-2006	yamt	sync with head.
1.25.4.1	22-Oct-2006	yamt	sync with head
1.25.2.1	18-Nov-2006	ad	Sync with head.
1.30.6.1	03-Jun-2007	wrstuden	Catch up with now-somewhat-dated netbsd-4. These changes took longer than expected for me to actually get around to merging.
1.30.4.2	24-Mar-2007	yamt	sync with head.
1.30.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.30.2.1	16-May-2007	jdc	Pull up revision 1.32 (requested by tsutsui in ticket #640). Cleanup handling of quikrs on each RealTek chip: - replace rtk_type member in rtk_softc which has chip types with new rtk_quirk that represents quirks on each chip: - RTKQ_8129 doesn't have internal MII (used in rtk(4)) - RTKQ_8139CPLUS has different register layout (for re(4)) - RTKQ_8169NONS (original 8169) requires some settings on init - RTKQ_PCIE requires different settings in setmulti so that we don't have to check each hwrev values or types everywhere and newer variants will also work without changes if they don't have other quirks (sc_rev is unchenged for now for reference to the Realtek's driver) - don't check hwrev register in re_pci_match() but check only PCI_VENDER(), PCI_PRODUCT() and PCI_REVISION() so that we no longer have to map pci space there - add a new HWREV value for another 8168 variant - try to map PCI mem space more properly - remove (probably unneeded) ifp->if_baudrate initialization Tested on a newer 8168 variant by Dennis den Brok on tech-kern, and also tested on 8139 and 8169C on macppc, and 8139C+ on landisk by me.
1.31.8.1	29-Mar-2007	reinoud	Pullup to -current
1.31.6.1	11-Jul-2007	mjf	Sync with head.
1.31.4.2	23-Oct-2007	ad	Sync with head.
1.31.4.1	10-Apr-2007	ad	Sync with head.
1.32.14.2	13-Nov-2007	bouyer	Sync with HEAD
1.32.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.32.10.2	09-Jan-2008	matt	sync with HEAD
1.32.10.1	06-Nov-2007	matt	sync with HEAD
1.32.8.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.32.8.5	06-Nov-2007	joerg	Sync with HEAD.
1.32.8.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.32.8.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.32.8.2	05-Oct-2007	joerg	Remove the power management nonsense in the PCI versions of rtk(4) and re(4). It is not really likely to work anyway. Use PNP power management for the cardbus attachment as well and remove the power management handlers from the chipset code.
1.32.8.1	05-Oct-2007	joerg	Convert rtk/pci to PNP power management.
1.33.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.33.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.34.4.1	11-Dec-2007	yamt	sync with head.
1.34.2.1	26-Dec-2007	ad	Sync with head.
1.35.12.1	18-May-2008	yamt	sync with head.
1.35.10.2	28-Sep-2008	mjf	Sync with HEAD.
1.35.10.1	02-Jun-2008	mjf	Sync with HEAD.
1.36.8.1	19-Oct-2008	haad	Sync with HEAD.
1.36.4.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.36.2.5	11-Aug-2010	yamt	sync with head.
1.36.2.4	11-Mar-2010	yamt	sync with head
1.36.2.3	16-Sep-2009	yamt	sync with head
1.36.2.2	16-May-2009	yamt	sync with head
1.36.2.1	04-May-2009	yamt	sync with head.
1.37.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.40.4.1	05-Mar-2011	rmind	sync with head
1.40.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.40.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.42.12.1	18-Feb-2012	mrg	merge to -current.
1.42.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.42.8.1	17-Apr-2012	yamt	sync with head
1.43.6.3	03-Dec-2017	jdolecek	update from HEAD
1.43.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.43.6.1	23-Jun-2013	tls	resync from head
1.44.4.1	18-May-2014	rmind	sync with head
1.45.6.1	06-Jun-2015	skrll	Sync with HEAD
1.46.18.1	10-Jun-2019	christos	Sync with HEAD
1.46.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.25	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.24	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.23	29-Mar-2014	christos	branches: 1.23.28; 1.23.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.22	30-Sep-2012	dsl	branches: 1.22.2; Fix debug build
1.21	23-Sep-2012	chs	reformat product array to match openbsd.
1.20	26-Jul-2011	dyoung	branches: 1.20.2; 1.20.12; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.19	26-Jan-2011	dyoung	Make oodles of mainly cosmetic changes that make rtw(4)'s PCI attachment resemble its CardBus attachment very, very closely: slightly more than 24 lines are different. Alas, I cannot commit the CardBus part of this change, yet, because I have to finish my overhaul of CardBus resource handling, first.
1.18	04-Mar-2010	dyoung	branches: 1.18.2; 1.18.4; 1.18.6; Make cosmetic changes in order to reduce differences with sys/dev/cardbus/if_rtw_cardbus.c: remove an unnecessary #include. Change a few cut & paste instances of ADM8211 to RTL8180. Make the suspend & resume functions static, add declarations for them at the top of the file, and move the functions themselves to the bottom.
1.17	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.16	08-Jan-2010	dyoung	branches: 1.16.2; Expand PMF_FN_* macros.
1.15	16-Sep-2009	dyoung	In pmf(9), improve the implementation of device self-suspension and make suspension by self, by drvctl(8), and by ACPI system sleep play nice together. Start solidifying some temporary API changes. 1. Extract a new header file, <sys/device_if.h>, from <sys/device.h> and #include it from <sys/pmf.h> instead of <sys/device.h> to break the circular dependency between <sys/device.h> and <sys/pmf.h>. 2. Introduce pmf_qual_t, an aggregate of qualifications on a PMF suspend/resume call. Start to replace instances of PMF_FN_PROTO, PMF_FN_ARGS, et cetera, with a pmf_qual_t. 3. Introduce the notion of a "suspensor," an entity that holds a device in suspension. More than one suspensor may hold a device at once. A device stays suspended as long as at least one suspensor holds it. A device resumes when the last suspensor releases it. Currently, the kernel defines three suspensors, 3a the system-suspensor: for system suspension, initiated by 'sysctl -w machdep.sleep_state=3', by lid closure, by power-button press, et cetera, 3b the drvctl-suspensor: for device suspension by /dev/drvctl ioctl, e.g., drvctl -S sip0. 3c the system self-suspensor: for device drivers that suspend themselves and their children. Several drivers for network interfaces put the network device to sleep while it is not administratively up, that is, after the kernel calls if_stop(, 1). The self-suspensor should not be used directly. See the description of suspensor delegates, below. A suspensor can have one or more "delegates". A suspensor can release devices that its delegates hold suspended. Right now, only the system self-suspensor has delegates. For each device that a self-suspending driver attaches, it creates the device's self-suspensor, a delegate of the system self-suspensor. Suspensors stop a system-wide suspend/resume cycle from waking devices that the operator put to sleep with drvctl before the cycle. They also help self-suspension to work more simply, safely, and in accord with expectations. 4. Add the notion of device activation level, devact_level_t, and a routine for checking the current activation level, device_activation(). Current activation levels are DEVACT_LEVEL_BUS, DEVACT_LEVEL_DRIVER, and DEVACT_LEVEL_CLASS, which respectively indicate that the device's bus is active, that the bus and device are active, and that the bus, device, and the functions of the device's class (network, audio) are active. Suspend/resume calls can be qualified with a devact_level_t. The power-management framework treats a devact_level_t that qualifies a device suspension as the device's current activation level; it only runs hooks to reduce the activation level from the presumed current level to the fully suspended state. The framework treats a devact_level_t qualifying device resumption as the target activation level; it only runs hooks to raise the activation level to the target. 5. Use pmf_qual_t, devact_level_t, and self-suspensors in several drivers. 6. Temporarily add an unused power-management workqueue that I will remove or replace, soon.
1.14	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.13	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.12	28-Apr-2008	martin	branches: 1.12.14; Remove clause 3 and 4 from TNF licenses
1.11	21-Mar-2008	dyoung	branches: 1.11.2; 1.11.4; pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.10	12-Mar-2008	dyoung	Use device_t and its accessors throughout. Use aprint_*_dev(). Improve PMF-ability. Add a 'flags' argument to suspend/resume handlers and callers such as pmf_system_suspend(). Define a flag, PMF_F_SELF, which indicates to PMF that a device is suspending/resuming itself. Add helper routines, pmf_device_suspend_self(dev) and pmf_device_resume_self(dev), that call pmf_device_suspend(dev, PMF_F_SELF) and pmf_device_resume(dev, PMF_F_SELF), respectively. Use PMF_F_SELF to suspend/resume self in ath(4), audio(4), rtw(4), and sip(4). In ath(4) and in rtw(4), replace the icky sc_enable/sc_disable callbacks, provided by the bus front-end, with self-suspension/resumption. Also, clean up the bus front-ends. Make sure that the interrupt handler is disestablished during suspension. Get rid of driver-private flags (e.g., RTW_F_ENABLED, ath_softc->sc_invalid); use device_is_active()/device_has_power() calls, instead. In the network-class suspend handler, call if_stop(, 0) instead of if_stop(, 1), because the latter is superfluous (bus- and driver-suspension hooks will 'disable' the NIC), and it may cause recursion. In the network-class resume handler, prevent infinite recursion through if_init() by getting out early if we are self-suspending (PMF_F_SELF). rtw(4) improvements: Destroy rtw(4) callouts when we detach it. Make rtw at pci detachable. Print some more information with the "rx frame too long" warning. Remove activate() methods: Get rid of rtw_activate() and ath_activate(). The device activate() methods are not good for much these days. Make ath at cardbus resume with crypto functions intact: Introduce a boolean device property, "pmf-powerdown". If pmf-powerdown is present and false, it indicates that a bus back-end should not remove power from a device. Honor this property in cardbus_child_suspend(). Set this property to 'false' in ath_attach(), since removing power from an ath at cardbus seems to lobotomize the WPA crypto engine. XXX Should the pmf-powerdown property propagate toward the root of the device tree? Miscellaneous ath(4) changes: Warn if ath(4) tries to write crypto keys to suspended hardware. Reduce differences between FreeBSD and NetBSD in ath(4) multicast filter setup. Make ath_printrxbuf() print an rx descriptor's status & key index, to help debug crypto errors. Shorten a staircase in ath_ioctl(). Don't check for ieee80211_ioctl() return code ERESTART, it never happens.
1.9	21-Dec-2007	dyoung	branches: 1.9.2; 1.9.6; Do not embed a struct device in rtw_softc any longer. Register both pci and cardbus attachments with CFATTACH_DECL_NEW(). Access the softc through the device_t using device_private(). While I'm here, change a couple of KASSERT()s about the Rx buffer length to a warning.
1.8	19-Oct-2007	ad	branches: 1.8.4; 1.8.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.7	01-Sep-2006	dyoung	branches: 1.7.12; 1.7.26; 1.7.28; 1.7.32; Remove the declaration of an unused local variable 'reg'.
1.6	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.5	28-Apr-2006	rpaulo	branches: 1.5.2; 1.5.4; Use aprint family of functions.
1.4	04-Dec-2005	christos	branches: 1.4.4; 1.4.6; 1.4.8; 1.4.10; 1.4.12; PR/32181: Matthias-Christian Ott: Add support for the Belkin F5D6001 pci device
1.3	22-Jun-2005	dyoung	branches: 1.3.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.2	27-Feb-2005	perry	nuke trailing whitespace
1.1	26-Sep-2004	dyoung	branches: 1.1.2; 1.1.6; 1.1.8; Add Cardbus, PCI bus front-ends for RTL8180 802.11b MAC/baseband.
1.1.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	29-Apr-2005	kent	sync with -current
1.1.2.5	11-Dec-2005	christos	Sync with head.
1.1.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.2	19-Oct-2004	skrll	Sync with HEAD
1.1.2.1	26-Sep-2004	skrll	file if_rtw_pci.c was added on branch ktrace-lwp on 2004-10-19 15:56:59 +0000
1.3.2.6	24-Mar-2008	yamt	sync with head.
1.3.2.5	17-Mar-2008	yamt	sync with head.
1.3.2.4	21-Jan-2008	yamt	sync with head
1.3.2.3	27-Oct-2007	yamt	sync with head.
1.3.2.2	30-Dec-2006	yamt	sync with head.
1.3.2.1	21-Jun-2006	yamt	sync with head.
1.4.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.4.10.1	11-May-2006	elad	sync with head
1.4.8.3	03-Sep-2006	yamt	sync with head.
1.4.8.2	26-Jun-2006	yamt	sync with head.
1.4.8.1	24-May-2006	yamt	sync with head.
1.4.6.1	01-Jun-2006	kardel	Sync with head.
1.4.4.1	09-Sep-2006	rpaulo	sync with head
1.5.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.5.2.1	19-Jun-2006	chap	Sync with head.
1.7.32.1	25-Oct-2007	bouyer	Sync with HEAD.
1.7.28.3	23-Mar-2008	matt	sync with HEAD
1.7.28.2	09-Jan-2008	matt	sync with HEAD
1.7.28.1	06-Nov-2007	matt	sync with HEAD
1.7.26.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.7.12.1	23-Oct-2007	ad	Sync with head.
1.8.8.1	02-Jan-2008	bouyer	Sync with HEAD
1.8.4.1	26-Dec-2007	ad	Sync with head.
1.9.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.9.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.9.2.1	24-Mar-2008	keiichi	sync with head.
1.11.4.4	11-Mar-2010	yamt	sync with head
1.11.4.3	16-Sep-2009	yamt	sync with head
1.11.4.2	16-May-2009	yamt	sync with head
1.11.4.1	16-May-2008	yamt	sync with head.
1.11.2.1	18-May-2008	yamt	sync with head.
1.12.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.18.6.1	08-Feb-2011	bouyer	Sync with HEAD
1.18.4.1	06-Jun-2011	jruoho	Sync with HEAD.
1.18.2.1	05-Mar-2011	rmind	sync with head
1.20.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.20.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.20.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.20.2.1	30-Oct-2012	yamt	sync with head
1.22.2.1	18-May-2014	rmind	sync with head
1.23.30.1	10-Jun-2019	christos	Sync with HEAD
1.23.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.21	01-Aug-2023	mrg	fix simple mis-matched function prototype and definitions. most of these are like, eg void foo(int[2]); with either of these void foo(int) { ... } void foo(int[]) { ... } in some cases (such as stat or utimes calls found in our header files), we now match standard definition from opengroup. found by GCC 12.
1.20	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.19	30-Jan-2020	thorpej	branches: 1.19.10; Adopt <net/if_stats.h>.
1.18	09-Dec-2018	jdolecek	branches: 1.18.6; use pci_intr_establish_xname() everywhere
1.17	29-Jun-2018	thorpej	- Unify the "rtwn" and "urtwn" register definitions and initialization tables into common header files shared by both drivers. - Fix some register and bit definitions, from OpenBSD and verified against the vendor driver. - Add some additional register / descriptor defnitions, from OpenBSD.
1.16	26-Jun-2018	msaitoh	branches: 1.16.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.15	06-Apr-2018	macallan	in rtwn_attach(): return; before unconditionally running into fail:
1.14	08-Feb-2018	dholland	branches: 1.14.2; Typos.
1.13	23-Oct-2017	msaitoh	branches: 1.13.2; If if_initialize() failed in the attach function, free resources and return.
1.12	18-May-2017	nonaka	branches: 1.12.2; Fixed m is not freed if m_defrag() fails.
1.11	02-Feb-2017	nonaka	branches: 1.11.4; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.10	24-Jan-2017	nonaka	rtwn(4): increase the number of tx queue required to clear the full mask.
1.9	08-Dec-2016	ozaki-r	branches: 1.9.2; Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.8	10-Jun-2016	ozaki-r	branches: 1.8.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.7	26-May-2016	ozaki-r	Introduce M_CLEARCTX and use it instead of open-coding rcvif No functional change.
1.6	26-May-2016	ozaki-r	Use M_GETCTX No functional change.
1.5	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.4	06-Feb-2016	riastradh	Missed a spot.
1.3	06-Feb-2016	riastradh	Avoid shadowing global `max' from libkern.
1.2	06-Nov-2015	nonaka	Always use pci_intr_alloc(9)/pci_intr_release(9).
1.1	27-Aug-2015	nonaka	branches: 1.1.2; Added rtwn(4) for Realtek RTL8188CE/RTL8192CE PCIe 802.11b/g/n wireless network devices. Ported from OpenBSD.
1.1.2.8	28-Aug-2017	skrll	Sync with HEAD
1.1.2.7	05-Feb-2017	skrll	Sync with HEAD
1.1.2.6	09-Jul-2016	skrll	Sync with HEAD
1.1.2.5	29-May-2016	skrll	Sync with HEAD
1.1.2.4	19-Mar-2016	skrll	Sync with HEAD
1.1.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	27-Aug-2015	skrll	file if_rtwn.c was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.8.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.8.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.9.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.11.4.1	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.12.2.2	09-Apr-2018	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #719): sys/dev/pci/if_rtwn.c: revision 1.15 in rtwn_attach(): return; before unconditionally running into fail:
1.12.2.1	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.13.2.2	03-Dec-2017	jdolecek	update from HEAD
1.13.2.1	23-Oct-2017	jdolecek	file if_rtwn.c was added on branch tls-maxphys on 2017-12-03 11:37:08 +0000
1.14.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.14.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.14.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.16.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.16.2.1	10-Jun-2019	christos	Sync with HEAD
1.18.6.1	29-Feb-2020	ad	Sync with head.
1.19.10.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.6	05-Oct-2019	mrg	remove __packed attribute from struct ieee80211_radiotap_header and all the structures that include it. this should not change anything while avoiding packed vs alignment warnings from GCC 8, and potentially pessimised code generation due to the packed marker (there are no misaligned members, just that the per-device parts may end unaligned.) all consumers of these members are done from the properly aligned packet members directly, or, as a union with a 64 byte member, also properly aligned. codegen didn't appear to change, except for the definition of sizeof(struct driver_[rt]x_radiotap_header) in debug info, which is not directly used anywhere.
1.5	29-Jun-2018	thorpej	- Unify the "rtwn" and "urtwn" register definitions and initialization tables into common header files shared by both drivers. - Fix some register and bit definitions, from OpenBSD and verified against the vendor driver. - Add some additional register / descriptor defnitions, from OpenBSD.
1.4	02-Feb-2017	nonaka	branches: 1.4.12; 1.4.14; 1.4.16; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.3	24-Jan-2017	nonaka	rtwn(4): increase the number of tx queue required to clear the full mask.
1.2	06-Nov-2015	nonaka	branches: 1.2.2; 1.2.4; Always use pci_intr_alloc(9)/pci_intr_release(9).
1.1	27-Aug-2015	nonaka	branches: 1.1.2; Added rtwn(4) for Realtek RTL8188CE/RTL8192CE PCIe 802.11b/g/n wireless network devices. Ported from OpenBSD.
1.1.2.4	05-Feb-2017	skrll	Sync with HEAD
1.1.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	27-Aug-2015	skrll	file if_rtwnreg.h was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.2.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.2.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.4.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.4.16.1	10-Jun-2019	christos	Sync with HEAD
1.4.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.4.12.2	03-Dec-2017	jdolecek	update from HEAD
1.4.12.1	02-Feb-2017	jdolecek	file if_rtwnreg.h was added on branch tls-maxphys on 2017-12-03 11:37:08 +0000
1.23	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.22	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.21	07-Jul-2016	msaitoh	branches: 1.21.16; 1.21.18; KNF. Remove extra spaces. No functional change.
1.20	29-Mar-2014	christos	branches: 1.20.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.19	27-Oct-2012	chs	branches: 1.19.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.18	26-Nov-2009	njoly	branches: 1.18.12; 1.18.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.17	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.16	28-Apr-2008	martin	branches: 1.16.14; Remove clause 3 and 4 from TNF licenses
1.15	10-Apr-2008	cegger	branches: 1.15.2; 1.15.4; use aprint_*_dev and device_xname
1.14	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.13	19-Oct-2007	ad	branches: 1.13.12; 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	17-Jun-2006	christos	branches: 1.10.4; 1.10.6; re-factor the pci powestate api. reviewed by gimpy
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.8; 1.9.14; 1.9.16; merge ktrace-lwp.
1.8	21-Aug-2004	thorpej	branches: 1.8.12; Use ANSI function decls and make use of static.
1.7	23-Dec-2002	tsutsui	branches: 1.7.2; Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	13-Nov-2001	lukem	add RCSID
1.2	08-Jul-2001	thorpej	branches: 1.2.2; Remove unnecessary include files. From Onno van der Linden.
1.1	18-Jun-2001	thorpej	branches: 1.1.2; Add support for the Adaptec AIC-6915 10/100 Ethernet. This is a 64-bit PCI chip, available in 1, 2, and 4 port models.
1.1.2.6	29-Dec-2002	thorpej	Sync with HEAD.
1.1.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.1.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.1	18-Jun-2001	nathanw	file if_sf_pci.c was added on branch nathanw_sa on 2001-06-21 20:04:49 +0000
1.2.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.1	25-Aug-2004	skrll	Sync with HEAD.
1.8.12.4	24-Mar-2008	yamt	sync with head.
1.8.12.3	27-Oct-2007	yamt	sync with head.
1.8.12.2	30-Dec-2006	yamt	sync with head.
1.8.12.1	21-Jun-2006	yamt	sync with head.
1.9.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.9.14.1	19-Jun-2006	chap	Sync with head.
1.9.8.1	26-Jun-2006	yamt	sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.6.2	10-Dec-2006	yamt	sync with head.
1.10.6.1	22-Oct-2006	yamt	sync with head
1.10.4.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.2	23-Mar-2008	matt	sync with HEAD
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.12.1	24-Mar-2008	keiichi	sync with head.
1.15.4.3	11-Mar-2010	yamt	sync with head
1.15.4.2	16-May-2009	yamt	sync with head
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.22.3	03-Dec-2017	jdolecek	update from HEAD
1.18.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.18.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.18.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.18.12.1	30-Oct-2012	yamt	sync with head
1.19.2.1	18-May-2014	rmind	sync with head
1.20.6.1	09-Jul-2016	skrll	Sync with HEAD
1.21.18.1	10-Jun-2019	christos	Sync with HEAD
1.21.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.193	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.192	29-Jun-2024	riastradh	branches: 1.192.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.191	10-Feb-2024	andvar	s/alloted/allotted/ in comments.
1.190	02-Jun-2023	andvar	follow the steps of Andrew Doran (ad) commit and fix more s/loose/lose/ typos. also s/beyound/beyond/ and few others along the way, mainly in comments.
1.189	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.188	11-Sep-2022	ryo	Packets larger than 1518 bytes (1522 bytes if VLAN tag is included) are no longer received. Previously, those packets were received as truncated (incomplete) packets.
1.187	06-Aug-2022	riastradh	sip(4): Tidy up DMA syncs. - No membar_producer in sip_init_rxdesc -- use bus_dmamap_sync with BUS_DMASYNC_PREWRITE to order updates to the DMA descriptors. - Omit needless membar_producer in sip_init_txdesc -- the hardware will not look at any of these descriptors until we set CMDSTS_OWN on the first one in the sequence, which is done later in the caller, sipcom_start. - In gsip_rxintr, make sure to read cmdsts _before_ extsts, by separating them with BUS_DMASYNC_PREREAD. Otherwise, the CPU might reorder the loads and read a stale extsts first before witnessing an updated cmdsts with the CMDSTS_OWN bit that transfers ownership of the rx packet to us.
1.186	24-May-2022	andvar	fix various typos in comment, documentation and log messages.
1.185	16-Feb-2022	andvar	fix various typos, mainly in comments.
1.184	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.183	21-Oct-2021	andvar	fix various typos, mainly in comments, but also in man pages and log messages.
1.182	16-Mar-2020	thorpej	Do the "have transmit resources available" check before checking that there is a packet in the send queue.
1.181	15-Mar-2020	thorpej	Don't bother with IFF_OACTIVE.
1.180	13-Mar-2020	thorpej	Adjust the logic for enabling the 64-bit data path when a 64-bit slot is detected: - If DATA64_EN isn't set in CFG after a reset, don't use 64-bit data path at all (it's been disabled by an EEPROM setting). - Provide a hook for force-disabling the 64-bit data path. - Otherwise, perform the "known 64-bit cards" check as done previously (because dodgy-vendor-EEPROM-settings still applies).
1.179	08-Mar-2020	thorpej	Support 64-bit DMA addressing on the DP83820, used only when a 64-bit DMA tag is available.
1.178	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.177	04-Feb-2020	thorpej	Use ifmedia_fini().
1.176	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.175	13-Sep-2019	msaitoh	branches: 1.175.2; if_flags is neither int nor short. It's unsigned short.
1.174	09-Jul-2019	msaitoh	Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.173	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.172	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.171	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.170	26-Apr-2019	msaitoh	Fix a bug that read value of MII_EXTSR register on TBI mode always fails. This bug was added in rev. 1.169.
1.169	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.168	26-Jun-2018	msaitoh	branches: 1.168.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.167	26-Sep-2017	knakahara	branches: 1.167.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.166	10-May-2017	msaitoh	branches: 1.166.2; Use pci_intr_establish_xname().
1.165	15-Dec-2016	ozaki-r	branches: 1.165.6; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.164	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.163	14-Jul-2016	msaitoh	branches: 1.163.2; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.162	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.161	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.160	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.159	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.158	10-Aug-2014	tls	branches: 1.158.4; Merge tls-earlyentropy branch into HEAD.
1.157	29-Mar-2014	christos	branches: 1.157.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.156	30-Mar-2013	christos	branches: 1.156.4; remove trailing whitespace
1.155	23-Sep-2012	chs	match some more devices.
1.154	22-Jul-2012	matt	branches: 1.154.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.153	02-Feb-2012	tls	branches: 1.153.2; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.152	12-Dec-2011	jakllsch	Using BUS_DMA_NOCACHE for bus_dmamem_map() causes issues on (at least) sparc64.
1.151	19-Nov-2011	tls	branches: 1.151.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.150	15-Oct-2011	dholland	branches: 1.150.2; "return f();" where both f and the enclosing function return void is sometimes convenient. But, it's not standard, so don't do it.
1.149	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.148	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.147	24-Feb-2010	dyoung	branches: 1.147.2; Use device_suspensor_t.
1.146	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.145	19-Jan-2010	pooka	branches: 1.145.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.144	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.143	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.142	16-Sep-2009	dyoung	In pmf(9), improve the implementation of device self-suspension and make suspension by self, by drvctl(8), and by ACPI system sleep play nice together. Start solidifying some temporary API changes. 1. Extract a new header file, <sys/device_if.h>, from <sys/device.h> and #include it from <sys/pmf.h> instead of <sys/device.h> to break the circular dependency between <sys/device.h> and <sys/pmf.h>. 2. Introduce pmf_qual_t, an aggregate of qualifications on a PMF suspend/resume call. Start to replace instances of PMF_FN_PROTO, PMF_FN_ARGS, et cetera, with a pmf_qual_t. 3. Introduce the notion of a "suspensor," an entity that holds a device in suspension. More than one suspensor may hold a device at once. A device stays suspended as long as at least one suspensor holds it. A device resumes when the last suspensor releases it. Currently, the kernel defines three suspensors, 3a the system-suspensor: for system suspension, initiated by 'sysctl -w machdep.sleep_state=3', by lid closure, by power-button press, et cetera, 3b the drvctl-suspensor: for device suspension by /dev/drvctl ioctl, e.g., drvctl -S sip0. 3c the system self-suspensor: for device drivers that suspend themselves and their children. Several drivers for network interfaces put the network device to sleep while it is not administratively up, that is, after the kernel calls if_stop(, 1). The self-suspensor should not be used directly. See the description of suspensor delegates, below. A suspensor can have one or more "delegates". A suspensor can release devices that its delegates hold suspended. Right now, only the system self-suspensor has delegates. For each device that a self-suspending driver attaches, it creates the device's self-suspensor, a delegate of the system self-suspensor. Suspensors stop a system-wide suspend/resume cycle from waking devices that the operator put to sleep with drvctl before the cycle. They also help self-suspension to work more simply, safely, and in accord with expectations. 4. Add the notion of device activation level, devact_level_t, and a routine for checking the current activation level, device_activation(). Current activation levels are DEVACT_LEVEL_BUS, DEVACT_LEVEL_DRIVER, and DEVACT_LEVEL_CLASS, which respectively indicate that the device's bus is active, that the bus and device are active, and that the bus, device, and the functions of the device's class (network, audio) are active. Suspend/resume calls can be qualified with a devact_level_t. The power-management framework treats a devact_level_t that qualifies a device suspension as the device's current activation level; it only runs hooks to reduce the activation level from the presumed current level to the fully suspended state. The framework treats a devact_level_t qualifying device resumption as the target activation level; it only runs hooks to raise the activation level to the target. 5. Use pmf_qual_t, devact_level_t, and self-suspensors in several drivers. 6. Temporarily add an unused power-management workqueue that I will remove or replace, soon.
1.141	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.140	05-May-2009	dyoung	Change struct cfdata * to cfdata_t. No functional change intended.
1.139	05-May-2009	cegger	device_t/softc split tested by and ok dyoung@
1.138	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.137	27-Mar-2009	dyoung	If a sip(4) is powered down, then so is its PHY, so don't try to read media status from the PHY.
1.136	07-Mar-2009	dyoung	Don't try to write the registers of the NIC to stop it if it is already powered down. This change prevents PCI bus exceptions (indicated by NMI) when, for example, I detach the whole device tree on a Soekris net4521: com1: detached NMI ... going to debugger Stopped in pid 0.1 (system) at netbsd:sipcom_stop+0x4d: pushl $0 db{0}> continue NMI ... going to debugger Stopped in pid 0.1 (system) at netbsd:sipcom_stop+0x149: addl $0x10,%e sp db{0}> continue sip0: detached (On the AMD Elan SC520, the NMI occurs at the instruction after the write or, if the write was posted, at either that instruction or some later one.)
1.135	07-Nov-2008	dyoung	branches: 1.135.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.134	05-May-2008	dyoung	branches: 1.134.6; 1.134.8; Make PCI power re-activation DTRT: pass pci_activate_null instead of NULL to pci_activate() Call the variable where we store the PCI Command & Status Register 'csr' instead of 'pmreg'. In sipcom_intr(), get out if the device is inactive. Otherwise sip(4) may inadvertently try to handle a shared interrupt before the driver state is completely set up.
1.133	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.132	10-Apr-2008	cegger	branches: 1.132.2; 1.132.4; use aprint_*_dev and device_xname
1.131	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.130	12-Mar-2008	dyoung	Use device_t and its accessors throughout. Use aprint_*_dev(). Improve PMF-ability. Add a 'flags' argument to suspend/resume handlers and callers such as pmf_system_suspend(). Define a flag, PMF_F_SELF, which indicates to PMF that a device is suspending/resuming itself. Add helper routines, pmf_device_suspend_self(dev) and pmf_device_resume_self(dev), that call pmf_device_suspend(dev, PMF_F_SELF) and pmf_device_resume(dev, PMF_F_SELF), respectively. Use PMF_F_SELF to suspend/resume self in ath(4), audio(4), rtw(4), and sip(4). In ath(4) and in rtw(4), replace the icky sc_enable/sc_disable callbacks, provided by the bus front-end, with self-suspension/resumption. Also, clean up the bus front-ends. Make sure that the interrupt handler is disestablished during suspension. Get rid of driver-private flags (e.g., RTW_F_ENABLED, ath_softc->sc_invalid); use device_is_active()/device_has_power() calls, instead. In the network-class suspend handler, call if_stop(, 0) instead of if_stop(, 1), because the latter is superfluous (bus- and driver-suspension hooks will 'disable' the NIC), and it may cause recursion. In the network-class resume handler, prevent infinite recursion through if_init() by getting out early if we are self-suspending (PMF_F_SELF). rtw(4) improvements: Destroy rtw(4) callouts when we detach it. Make rtw at pci detachable. Print some more information with the "rx frame too long" warning. Remove activate() methods: Get rid of rtw_activate() and ath_activate(). The device activate() methods are not good for much these days. Make ath at cardbus resume with crypto functions intact: Introduce a boolean device property, "pmf-powerdown". If pmf-powerdown is present and false, it indicates that a bus back-end should not remove power from a device. Honor this property in cardbus_child_suspend(). Set this property to 'false' in ath_attach(), since removing power from an ath at cardbus seems to lobotomize the WPA crypto engine. XXX Should the pmf-powerdown property propagate toward the root of the device tree? Miscellaneous ath(4) changes: Warn if ath(4) tries to write crypto keys to suspended hardware. Reduce differences between FreeBSD and NetBSD in ath(4) multicast filter setup. Make ath_printrxbuf() print an rx descriptor's status & key index, to help debug crypto errors. Shorten a staircase in ath_ioctl(). Don't check for ieee80211_ioctl() return code ERESTART, it never happens.
1.129	11-Mar-2008	dyoung	Use device_t and accessors.
1.128	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.127	07-Feb-2008	dyoung	branches: 1.127.2; 1.127.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.126	19-Jan-2008	dyoung	Let ether_ioctl() handle default SIOCSIFMEDIA/SIOCGIFMEDIA actions.
1.125	10-Jan-2008	dyoung	Re-use ether_mediachange(), ether_mediastatus(). Stop using shutdownhook_establish(9).
1.124	16-Dec-2007	dyoung	Convert many macros to inline subroutines. Some of the subroutines probably should not be inline.
1.123	15-Dec-2007	dyoung	sip_copy_small and gsip_copy_small should not be static any longer. Make it so.
1.122	15-Dec-2007	he	Only declare sip_copy_small under __NO_STRICT_ALIGNMENT, to afoid a "defined but not used" warning turned into error by compiler options.
1.121	15-Dec-2007	dyoung	Take care to release all resources. Now I can 'drvctl -d sip0; drvctl -r pci0' to to detach / re-attach a sip0.
1.120	15-Dec-2007	dyoung	Finish making the sip(4) and gsip(4) drivers share compiled code.
1.119	15-Dec-2007	dyoung	Attach sip to 10/100 variants, gsip to gigabit variants. Thanks Izumi Tsutsui for the bug report.
1.118	14-Dec-2007	dogcow	fix KDASSERTs by defining ifp
1.117	14-Dec-2007	dyoung	Simplify suspend/resume using device_has_power().
1.116	14-Dec-2007	dyoung	Take a step toward sharing compiled code between sip(4) and gsip(4): get rid of SIP_DECL() and reduce #ifdef DP83820 code. Next step is to move a bunch of shared code to a new file (if_sipcom.c, say) and compile it once. While I am here, add suspend/resume handling to sip(4) and to gsip(4). Tested with the NatSemi sip(4) on the Soekris net4521. I don't have any gsip(4) to test with, and it seems that the few holders of gsip(4) in the world keep them in their attic, anyway.
1.115	19-Oct-2007	ad	branches: 1.115.2; 1.115.4; 1.115.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.114	29-Aug-2007	dyoung	branches: 1.114.4; Constify: LLADDR() -> CLLADDR().
1.113	09-Jul-2007	ad	branches: 1.113.2; 1.113.6; 1.113.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.112	04-Mar-2007	yamt	branches: 1.112.2; 1.112.4; fix fallout from caddr_t changes.
1.111	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.110	16-Nov-2006	christos	branches: 1.110.4; __unused removal on arguments; approved by core.
1.109	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.108	17-Jun-2006	christos	branches: 1.108.4; 1.108.6; re-factor the pci powestate api. reviewed by gimpy
1.107	18-Apr-2006	pavel	branches: 1.107.2; 1.107.4; VLAN tags were sent byte-swapped by the gsip driver, because the driver forgets to convert them from the host to the network byte order. Use bswap16 to convert them. (Not htons, because on a big-endian machine, they are in the correct byte order initially, but then, they are byte-swapped by a htole32 call when written to the transmit descriptor. So byte-swapping is needed in this case too, to compensate for this htole32 call.) For a similar reason, tags were seen byte-swapped when received on a big-endian machine. Replace ntohs by bswap16 in the input path too. (Again, it is needed to compensate for a le32toh call when the receive descriptor is read.) Fixes PR 32644. Tested on 3.0/i386, 3.0/sgimips and current/alpha. OK by martin@.
1.106	10-Apr-2006	pavel	In rev. 1.98, the ioctl method of the (g)sip drivers was optimized for SIOCSIFFLAGS: it compares the new flags with the old flags and avoids reset if there are only certain changes. This was done to fix PR 29126. It does not take into account, though, that there is other state which can change and SIOCSIFFLAGS is called to inform about it. Namely, if_capenable, ec_capenable and ec_nvlans. For all three, the _init method must program the hardware specially. Not doing it resulted in: - VLAN frames getting truncated - hw checksumming not working - outgoing VLAN frames not being tagged when they should - incoming VLAN frames being treated as untagged. Fix by keeping all the old state in the softc and initializing the hardware if any of it changes. Tested on gsip. Also tested by Nino Dehne and Martin J. Laubach on sip, thanks. Fixes PRs 32900 and 33216. Approved by martin@ .
1.105	27-Mar-2006	bouyer	MCLAIM() mbufs allocated here.
1.104	07-Feb-2006	thorpej	branches: 1.104.2; 1.104.4; 1.104.6; In the receive interrupt handler, compute the packet length before resetting m to point to the first buffer in the packet. The bug would cause the length of jumbo frames to be incorrect.
1.103	11-Dec-2005	christos	branches: 1.103.2; 1.103.4; 1.103.6; merge ktrace-lwp.
1.102	02-May-2005	yamt	branches: 1.102.2; split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.101	27-Feb-2005	perry	branches: 1.101.2; nuke trailing whitespace
1.100	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition
1.99	06-Feb-2005	cube	Initialize 'error' in all cases in ioctl handler, otherwise it doesn't compile (and of course might return garbage). [hi kim!]
1.98	06-Feb-2005	kim	If the interface is up and running, only modify the receive filter when setting promiscuous or debug mode. This avoids resetting the chip unnecessarily. Fixes PR kern/29126.
1.97	30-Jan-2005	thorpej	- Fix some logic errors in multi-descriptor packet reception case for DP83820. - Eliminate use of M_HASFCS.
1.96	30-Oct-2004	thorpej	branches: 1.96.4; 1.96.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.95	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.94	15-May-2004	thorpej	Add PAUSE-related event counters for sip(4) and gsip(4), slightly modified from a patch supplied by HITOSHI Osada.
1.93	15-May-2004	thorpej	Patch from HITOSHI Osada: * On the DP83820, don't set PCR_PS_DA if the PAUSE packet address is not registered in the multicast filter.
1.92	15-May-2004	thorpej	Patch from HITOSHI Osada: * SiS900 chips require the address of the PAUSE packet to be registered in the multicast filter.
1.91	09-May-2004	fair	Two changes: 1. ifdef out the restriction that the SiS 900 has only one PHY This is demonstrably false; the SiS 960 super south bridge in PR 18590 has a SiS 900 rev 1 core in it. 2. bitbang the MII for all versions of the SiS 900; this is the only way that the PHYs on this system answer. Also, I suspect that SIS900_REV_960 constant in if_sipreg.h is incorrectly labelled - there were later revisions of the super south bridge (e.g. the 961, 962, and 963), and I suspect the SiS 900 revision code there refers to one of those.
1.90	22-Apr-2004	enami	If defined(DP83820), (always) drop IFF_OACTIVE if we got txintr. Otherwise, the driver simply stuck once we face tx resource shortage.
1.89	11-Apr-2004	thorpej	Flow control support for DP83820 and SiS900. From HITOSHI Osada.
1.88	11-Apr-2004	thorpej	Improvements to DP83820 support, from HITOSHI Osada: - Fix jumbo frame support. - Disable interrupts at the chip in sip_intr(). - Disable checksum offloading if MTU > 8109 - the hardware is broken in this case. - Use the interrupt hold-off timer. - Correct the Tx FIFO size. - Add Pause Control/Status register definitions.
1.87	11-Jan-2004	cube	branches: 1.87.2; The limitation for TX/RX DMA of rev. 900B and 635 os the SiS chips also applies to rev. 0x91 for 96x chipsets. Needed to fix PR 24043, but more stressing testing has to be performed before closing it.
1.86	05-Dec-2003	cube	Remove bitbang code that was taken from FreeBSD to support recent SiS chipsets and use mii_bitbang interface instead. Reflect sip dependency in the config file. Support for SiS96x needs broader testing.
1.85	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.84	03-Dec-2003	cube	o Add support for accessing the PHY through MDIO for recent SiS chips o Add support for the recent SiS96x chipsets that have a new revision. That includes a new bit of code to access the EEPROM, since it is shared with the ieee1394 controller on those chipsets. Mostly taken from FreeBSD (rev. 1.62 and 1.64 of sys/pci/if_sip.c). I tried to make the code look less ugly, but couldn't invent documentation. Fix PR #23481. Thanks to Stephane ENGEL <sengel AT melshake DOT com> for the report and the cheerful testing.
1.83	29-Oct-2003	mycroft	Fix the real cause of the uninitialized warning -- we were looking for the VLAN tag in the wrong place!
1.82	25-Oct-2003	christos	Fix uninitialized variable warnings
1.81	30-Sep-2003	martin	Fix typo (DP83020 -> DP83820) from HITOSHI Osada in PR kern/23023.
1.80	25-Aug-2003	itojun	KNF
1.79	15-Aug-2003	itojun	accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.78	23-Mar-2003	thorpej	branches: 1.78.2; Add a work-around for the "short cable problem" that some DP83815 revisions have, as discussed on the soekris-tech mailing list a while ago, whereby one can experience excessive recieve erros when using < 30m cables. The patch detects overflow in a DSP filter parameter, and corrects it by writing a known good value.
1.77	13-Mar-2003	briggs	Detect SMC EZ Card as 64-bit. Patch from Pavel Cahyna in kern/20680.
1.76	17-Jan-2003	itojun	switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag implementation. it will simplify porting across *bsd (such as kame/altq), and make us more synchronized. from Joel Wilsson
1.75	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.74	24-Nov-2002	scw	Fix uninitialised variable warnings.
1.73	17-Oct-2002	fair	Change the "dontcare" bits argument of ifmedia_init() to IFM_IMASK, so that PHY instance is not siginificant in ifmedia_match(). This is done to support multiple PHYs on the MII. Without this change, ifmedia_set() would panic the system when no PHYs were matched. I ran into this on an AMD EasyNow PC, which is built around SiS system chips with an embedded SiS 900 core, and an external AMD Am79c901 PHY, which presents two PHYs on the MII: one for HomePNA, and one for standard 10base-T. The 10base-T PHY ends up with instance number 1...
1.72	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.71	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.70	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.69	26-Aug-2002	thorpej	* The Netgear GA-621 is a 64-bit card, so add it to the 64-bit quirk table. * We want to hardwire BMSR_EXTSTAT, not BMSR_EXTCAP, when reading the TBI BMSR. * Fetch the GPIO bits from the GPIOR register after an auto-load, rather than reading from the EEPROM directly.
1.68	26-Aug-2002	itojun	need to set wantinit variable
1.67	26-Aug-2002	itojun	suppress some of debugging output (IFF_DEBUG will re-enable it). PR 18069
1.66	21-Aug-2002	itojun	use RND_ENABLED() to avoid unneeded function call. should help if_gsip case.
1.65	20-Aug-2002	itojun	attach random number source.
1.64	16-Aug-2002	thorpej	* Hard-code EXTSR_1000XFDX\|EXTSR_1000XHDX for reads of the MII_EXTSR in the TBI case. * Force BMSR_ANEG \| BMSR_EXTCAP to be returned for reads of the MII_BMSR in the TBI case.
1.63	10-Aug-2002	thorpej	Preliminary support for the ten-bit interface on the DP83820. This code needs more testing, and more bug fixing.
1.62	10-Aug-2002	thorpej	Don't treat the "reset complete" interrupts as errors. Doing so causes us to go into a reset/interrupt/reset/... loop.
1.61	11-Jul-2002	thorpej	Add 64-bit quirk for the Accton EN1407-T/Planex GN-1000TE. IDs provided by SAITOH Masanobu <msaitoh@netbsd.org>.
1.60	30-Jun-2002	thorpej	Add a table of known-64-bit DP83820-based cards. Use this table to enable 64-bit data transfers on 64-bit cards when plugged into a 64-bit slot. Right know the Asante GigaNIX is listed in that table. Sigh, there is an EEPROM bit that can be used to detect 64-bit vs 32-bit cards. Unfortunately, at least 2 vendors of 32-bit cards fail to clear the "DATA64_EN" bit in the EEPROM, which causes the card to lose badly, because it still manages to detect that it's plugged into a 64-bit PCI slot. Yay, stupid hardware vendors.
1.59	30-Jun-2002	thorpej	Load configuration data from the EEPROM on the DP83820 differently: rather than grovel the EEPROM directly, initiate an "EEPROM load" in the PCI test register, and fetch the values from the CFG register.
1.58	30-Jun-2002	thorpej	Update the TODO list: We have some Tx interrupt mitigation now, so we need to do Rx interrupt mitigation next.
1.57	30-Jun-2002	thorpej	Be more aggressive in giving descriptors to the chip in the transmit path: Instead of waiting for the if_snd queue to be drained before giving ownership of the frist descriptor to the chip, do it after sync'ing all the descriptors for a single packet.
1.56	30-Jun-2002	thorpej	Implement a sliding interrupt delay window for Tx interrupts.
1.55	30-Jun-2002	thorpej	* Give symbolic names to the CFG bits in the EEPROM. * Get CFG_M64ADDR, CFG_T64ADDR, and CFG_DATA64_EN from the EEPROM. Note, we still disable CFG_M64ADDR and CFG_T64ADDR later (XXX need PCI bus capability flags for these). * Print a message if we're in a 64-bit slot and 64-bit data is disabled in the EEPROM. Make sure CFG_DATA64_EN is disabled if we're not in a 64-bit slot.
1.54	01-Jun-2002	lukem	SIMPLEQ rototill: - implement SIMPLEQ_REMOVE(head, elm, type, field). whilst it's O(n), this mirrors the functionality of SLIST_REMOVE() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE() - remove the unnecessary elm arg from SIMPLEQ_REMOVE_HEAD(). this mirrors the functionality of SLIST_REMOVE_HEAD() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE_HEAD() - remove notes about SIMPLEQ not supporting arbitrary element removal - use SIMPLEQ_FOREACH() instead of home-grown for loops - use SIMPLEQ_EMPTY() appropriately - use SIMPLEQ_*() instead of accessing sqh_first,sqh_last,sqe_next directly - reorder manual page; be consistent about how the types are listed - other minor cleanups
1.53	28-May-2002	tron	Set initial transmit drain threshold to 1504 to avoid the problem described in PR kern/16070. Change approved by Jason Thorpe. XXX We'll should try to find a better adaptive scheme for the next NetBSD release.
1.52	03-May-2002	thorpej	branches: 1.52.2; 1.52.4; Bump the number of Tx DMA segments from 8 to 16 (the zero-copy socket code sometimes sees more than 8).
1.51	27-Mar-2002	briggs	Erf. Actually use the symbolic name for the 630ET instead of a constant.
1.50	27-Mar-2002	briggs	Print the chip revision in the attach function. Treat revision 0x84 of the SiS like other SIS630 variants. Call revision 0x84 the 630ET (name from OpenBSD).
1.49	24-Mar-2002	is	use the right endian ether_crc32 function in the multicast hash table computation
1.48	28-Feb-2002	thorpej	If the upper layer says Write/Invalidate is okay, then make sure it's enabled in the PCI CSR.
1.47	28-Feb-2002	thorpej	Change the default Rx FIFO drain threshold to 128 bytes; this reduces the likelihood of Rx FIFO overruns on machines with wimpy PCI busses, while still allowing the interface to burst with a reasonable size.
1.46	28-Feb-2002	thorpej	* On the DP83820, make sure to make the Tx DMA map large enough for jumbo frames. * Work around broken PXE firmware on some boards, which leave the ROM BAR enabled even after the PXE stack has been unloaded. * Set up the initial values for sc_tx_fill_thresh, sc_tx_drain_thresh, and sc_rx_drain_thresh in sip_attach(), rather than in sip_init().
1.45	09-Feb-2002	thorpej	Fix problems associated with the SiS 635/735 on-board Ethernet, from Stephen Degler <sdegler@degler.net>, port-i386/15261.
1.44	20-Dec-2001	thorpej	Fetch the MAC address from the NVRAM on the SiS 630 built-in Ethernet. Based on PR 14665, Stephen Borrill <netbsd@precedence.co.uk>.
1.43	13-Nov-2001	lukem	add RCSID
1.42	23-Jul-2001	thorpej	Make sure to initialize the all Rx descriptors properly in *_init() even if mbufs for them are already allocated.
1.41	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.40	08-Jul-2001	thorpej	branches: 1.40.2; Some cosmetic shuffling.
1.39	08-Jul-2001	thorpej	Fix a very, very dumb mistake that broke multicast on the dp83815 and dp83820.
1.38	07-Jul-2001	thorpej	Change the structure of dp83815_set_filter ever so slightly.
1.37	07-Jul-2001	thorpej	bcmp -> memcmp
1.36	07-Jul-2001	thorpej	Add support for jumbo Ethernet frames on the DP83820.
1.35	30-Jun-2001	thorpej	Split the receive interrupt handler into two versions, one for the SiS900/DP83815, one for the DP83820. In preparation for changes to the DP83820 version for handling jumbo Ethernet frames.
1.34	18-Jun-2001	simonb	Make error checking/reporting a little more correct.
1.33	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.32	03-Jun-2001	thorpej	Clarify a historical comment.
1.31	02-Jun-2001	thorpej	Implement support for IP/TCP/UDP checksum offloading provided by network interfaces. This works by pre-computing the pseudo-header checksum and caching it, delaying the actual checksum to ip_output() if the hardware cannot perform the sum for us. In-bound checksums can either be fully-checked by hardware, or summed up for final verification by software. This method was modeled after how this is done in FreeBSD, although the code is significantly different in most places. We don't delay checksums for IPv6/TCP, but we do take advantage of the cached pseudo-header checksum. Note: hardware-assisted checksumming defaults to "off". It is enabled with ifconfig(8). See the manual page for details. Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet, 3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
1.30	18-May-2001	thorpej	- Add some instrumentation which can be enabled by defining SIP_EVENT_COUNTERS. - Decrease SIP_NTXSEGS to 8 and increase SIP_TXQUEUELEN to 256. This gives us 256 transmit jobs with only twice the descriptor memory footprint. This prevents stalling transmissions due to lack of transmit jobs under heavy load. - Increase SIP_NRXDESC to 128 so we don't run out of receive descriptors under extreme load. - As recommended by the manual, leave one descriptor in the transmit ring with OWN clear to prevent wrap-around. - Even though the manual claims that driver software must initialize TXDP if there are no pending transmissions when a new transmission is to be initialized, doing so causes serious performance degredation on the DP83820 under heavy load. Not initializing TXDP also increases performance on the DP83815. All of this gets us up to ~90Mb/s on both the DP83820 and DP83815 when connected to a 100base network.
1.29	18-May-2001	thorpej	Add a driver for the National Semiconductor DP83820 Gigabit Ethernet chip. This is found on the NetGear GA-622 and Asante FriendlyNet GigaNIX.
1.28	15-May-2001	thorpej	Encapsulate name declarations in macros in preparation for adding conditionally-compiled code for the DP83820.
1.27	24-Mar-2001	briggs	There is no need to accept all ARP packets (regardless of destination address) on the DP83815. Noted by Shiva Shenoy <shiva@riverstonenet.com> in private correspondence.
1.26	09-Mar-2001	briggs	More fixes for the DP83815... - Set the destination address register properly for "perfect match" mode in the receive filter setup. - Do not enable multicast receipt unless we are configured for some multicast. - Use the "recommended settings" (which set undocumented registers and documented-as-reserved fields) for the silicon revision 302h (not 203h, as documented in one of the two places in the manual) because the documentation is unclear and because those settings fix the card's behavior in "perfect match" mode. Without those settings, the card was generating random CRC/invalid symbol errors and generally not working unless it was set to be promiscuous. With these changes, this week's version of the Netgear FA311 works for me.
1.25	09-Mar-2001	briggs	Add a variant hook to read the ethernet MAC address from EEPROM differently for the SiS 900 and the DP83815. Also add some configuration suggested by the DP83815 documentation for one model of the DP83815.
1.24	06-Feb-2001	thorpej	branches: 1.24.2; Gack, fix a stupid bug in the DP83815 multicast setup code.
1.23	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.22	19-Dec-2000	thorpej	Fix a problem with the ALTQ changes that can cause bogus memory refernces. Problem reported by Luke Mewburn.
1.21	14-Dec-2000	thorpej	ALTQ'ify.
1.20	15-Nov-2000	thorpej	Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
1.19	28-Oct-2000	tsutsui	Increase initial Tx drain threshold to 512 bytes. My macppc complains about transmit FIFO underrun every time.
1.18	15-Oct-2000	thorpej	Don't trim off the FCS, pass it up.
1.17	11-Oct-2000	thorpej	Use ether_ioctl().
1.16	01-Oct-2000	thorpej	Move the check for "promisc + unicast + not for us" into ether_input(), and change Ethernet drivers to always pass all received frames to ether_input() (with a few exceptions, which are documented in the code).
1.15	20-Sep-2000	thorpej	Preliminary support for the National Semiconductor DP83815 10/100 Ethernet chip, which is based on the SiS 900 core, but with differences in the PHY and filter programming interface, and different WoL and 802.3x flow control programming. This chip appears on new NetGear FA-312TX cards. Thank goodness they're finally ditching the LiteOn chips.
1.14	12-Aug-2000	tsutsui	- Check CR_RST rather than ISR_TXRCMP and ISR_RXRCMP in sip_reset() - Add more delay after reset - "Big endian mode" on SiS 900 seems broken so don't enable it; use htole32()/le32toh() instead - Don't cast u_int16_t array to u_int8_t pointer - Add some delay() to FILTER_EMIT macro in sip_set_filter() Now SiS 900 works on my macppc.
1.13	04-Aug-2000	tsutsui	Add one more delay() to sip_read_eeprom().
1.12	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.11	12-May-2000	thorpej	branches: 1.11.4; Use ether_crc32_le().
1.10	23-Mar-2000	mycroft	Rather than guessing at the location of the PMCSR, use the pointer returned by pci_get_capability(). (This is, after all, a standardized interface...)
1.9	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.8	06-Mar-2000	thorpej	No longer necessary to futz with ifp->if_baudrate here.
1.7	02-Feb-2000	thorpej	Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass flags down from the parent to child vi mii_attach().
1.6	02-Feb-2000	thorpej	Bring some order to the chaos which was the MII code function naming "conventions".
1.5	31-Jan-2000	thorpej	Add support for the SiS 7016.
1.4	12-Nov-1999	thorpej	Call mii_down() as appropriate.
1.3	04-Nov-1999	thorpej	Adapt to mii_phy_probe() change.
1.2	03-Aug-1999	thorpej	branches: 1.2.2; 1.2.4; 1.2.8; Be a little nicer about memory usage: * Don't allocate receive buffers until the interface is actually brought up, and release all of them if the interface is taken down. * Add a knob (defaults to off) which will copy an incoming packet to a single header mbuf if it is small enough to fit in one, rather than burning an entire cluster on it. Note that this change will be mostly moot if/when sbcompress() it changed to handle compressing clusters.
1.1	01-Jun-1999	thorpej	branches: 1.1.2; Device driver for the Silicon Integrated Systems SiS900 10/100 Ethernet chip, found on the IBM Netstation.
1.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.2.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.2.4.1	15-Nov-1999	fvdl	Sync with -current
1.2.2.6	27-Mar-2001	bouyer	Sync with HEAD.
1.2.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.2.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.11.4.9	08-Sep-2003	msaitoh	Pull up revision 1.79 via patch (requested by itojun in ticket #69): Accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.11.4.8	27-Mar-2002	he	Pull up revision 1.49 (via patch, requested by is): Fix multicast reception on DP83815. Fixes PR#15998.
1.11.4.7	09-Mar-2002	he	Apply patch (requested by bouyer): Fix an uninitialized pointer in sip_start(), which might eventually lead to a crash, especially if bpf is used.
1.11.4.6	27-Oct-2001	he	Apply patch (requested by briggs): Adjust receive and transmit descriptor list structures, sip_start(), and sip_txintr() to be more similar to the current development version. Fixes periodic device timeouts. Tested by briggs with a Netgear FA311TX on i386.
1.11.4.5	13-Mar-2001	he	Pull up revisions 1.25-1.26 (requested by briggs): Fix support for the DP83815 so that it works properly with the current crop of Netgear FA311 10/100 cards.
1.11.4.4	26-Feb-2001	he	Pull up revision 1.24 (requested by thorpej): Fix a bug in the DP83815 multicast setup code.
1.11.4.3	30-Oct-2000	tv	Pullup 1.19 [thorpej/tsutsui]: Increase initial Tx drain threshold to 512 bytes. My macppc complains about transmit FIFO underrun every time.
1.11.4.2	28-Sep-2000	thorpej	Sync with trunk: Add support for the DP83815 10/100 Ethernet, found on new NetGear cards.
1.11.4.1	12-Aug-2000	tsutsui	Pull up rev 1.13 and 1.14 (approved by thorpej): revision 1.13: - Add one more delay() to sip_read_eeprom() revision 1.14: - Check CR_RST rather than ISR_TXRCMP and ISR_RXRCMP in sip_reset() - Add more delay after reset - "Big endian mode" on SiS 900 seems broken so don't enable it; use htole32()/le32toh() instead - Don't cast u_int16_t array to u_int8_t pointer - Add some delay() to FILTER_EMIT macro in sip_set_filter()
1.24.2.16	17-Jan-2003	thorpej	Sync with HEAD.
1.24.2.15	29-Dec-2002	thorpej	Sync with HEAD.
1.24.2.14	11-Dec-2002	thorpej	Sync with HEAD.
1.24.2.13	11-Nov-2002	nathanw	Catch up to -current
1.24.2.12	18-Oct-2002	nathanw	Catch up to -current.
1.24.2.11	27-Aug-2002	nathanw	Catch up to -current.
1.24.2.10	13-Aug-2002	nathanw	Catch up to -current.
1.24.2.9	01-Aug-2002	nathanw	Catch up to -current.
1.24.2.8	20-Jun-2002	nathanw	Catch up to -current.
1.24.2.7	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.24.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.24.2.5	08-Jan-2002	nathanw	Catch up to -current.
1.24.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.24.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.24.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.24.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.40.2.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.40.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.40.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.40.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.40.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.40.2.1	03-Aug-2001	lukem	update to -current
1.52.4.10	02-Oct-2003	tron	Pull up revision 1.81 (requested by martin in ticket #1490): Fix typo (DP83020 -> DP83820) from HITOSHI Osada in PR kern/23023.
1.52.4.9	15-Aug-2003	tron	Pull up revision 1.79 (requested by itojun in ticket #1413): accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.52.4.8	15-Jun-2003	tron	Pull up revision 1.78 (requested by thorpej in ticket #1228): Add a work-around for the "short cable problem" that some DP83815 revisions have, as discussed on the soekris-tech mailing list a while ago, whereby one can experience excessive recieve erros when using < 30m cables. The patch detects overflow in a DSP filter parameter, and corrects it by writing a known good value.
1.52.4.7	21-Nov-2002	he	Pull up revision 1.62 (requested by tls in ticket #1007): Do not treat the ``reset complete'' interrupts as errors. Doing so causes us to go into a reset/interrupt/reset loop.
1.52.4.6	01-Nov-2002	tron	Pull up revision 1.60 (requested by thorpej in ticket #415): Add a table of known-64-bit DP83820-based cards. Use this table to enable 64-bit data transfers on 64-bit cards when plugged into a 64-bit slot. Right know the Asante GigaNIX is listed in that table. Sigh, there is an EEPROM bit that can be used to detect 64-bit vs 32-bit cards. Unfortunately, at least 2 vendors of 32-bit cards fail to clear the "DATA64_EN" bit in the EEPROM, which causes the card to lose badly, because it still manages to detect that it's plugged into a 64-bit PCI slot. Yay, stupid hardware vendors.
1.52.4.5	01-Nov-2002	tron	Pull up revision 1.59 (requested by thorpej in ticket #414): Load configuration data from the EEPROM on the DP83820 differently: rather than grovel the EEPROM directly, initiate an "EEPROM load" in the PCI test register, and fetch the values from the CFG register.
1.52.4.4	01-Nov-2002	tron	Pull up revision 1.58 (requested by thorpej in ticket #413): Update the TODO list: We have some Tx interrupt mitigation now, so we need to do Rx interrupt mitigation next.
1.52.4.3	01-Nov-2002	tron	Pull up revision 1.56 (requested by thorpej in ticket #412): Implement a sliding interrupt delay window for Tx interrupts.
1.52.4.2	01-Nov-2002	tron	Pull up revision 1.55 (requested by thorpej in ticket #411): * Give symbolic names to the CFG bits in the EEPROM. * Get CFG_M64ADDR, CFG_T64ADDR, and CFG_DATA64_EN from the EEPROM. Note, we still disable CFG_M64ADDR and CFG_T64ADDR later (XXX need PCI bus capability flags for these). * Print a message if we're in a 64-bit slot and 64-bit data is disabled in the EEPROM. Make sure CFG_DATA64_EN is disabled if we're not in a 64-bit slot.
1.52.4.1	28-May-2002	tv	Pull up revision 1.53 (requested by tron in ticket #50): Set initial transmit drain threshold to 1504 to avoid the problem described in PR kern/16070. Change approved by Jason Thorpe. XXX We'll should try to find a better adaptive scheme for the next NetBSD release.
1.52.2.4	29-Aug-2002	gehenna	catch up with -current.
1.52.2.3	15-Jul-2002	gehenna	catch up with -current.
1.52.2.2	20-Jun-2002	gehenna	catch up with -current.
1.52.2.1	30-May-2002	gehenna	Catch up with -current.
1.78.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.78.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.78.2.8	06-Feb-2005	skrll	Sync with HEAD.
1.78.2.7	06-Feb-2005	skrll	Sync with HEAD.
1.78.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.78.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.78.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.78.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.78.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.78.2.1	03-Aug-2004	skrll	Sync with HEAD
1.87.2.1	28-May-2004	tron	branches: 1.87.2.1.2; Pull up revision 1.91 (requested by fair in ticket #296): Two changes: 1. ifdef out the restriction that the SiS 900 has only one PHY This is demonstrably false; the SiS 960 super south bridge in PR 18590 has a SiS 900 rev 1 core in it. 2. bitbang the MII for all versions of the SiS 900; this is the only way that the PHYs on this system answer. Also, I suspect that SIS900_REV_960 constant in if_sipreg.h is incorrectly labelled - there were later revisions of the super south bridge (e.g. the 961, 962, and 963), and I suspect the SiS 900 revision code there refers to one of those.
1.87.2.1.2.1	24-Jan-2005	he	Pull up revision 1.96 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.96.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.96.6.1	12-Feb-2005	yamt	sync with head.
1.96.4.1	29-Apr-2005	kent	sync with -current
1.101.2.3	21-Apr-2006	tron	Pull up following revision(s) (requested by pavel in ticket #1280): sys/dev/pci/if_sip.c: revision 1.106 In rev. 1.98, the ioctl method of the (g)sip drivers was optimized for SIOCSIFFLAGS: it compares the new flags with the old flags and avoids reset if there are only certain changes. This was done to fix PR 29126. It does not take into account, though, that there is other state which can change and SIOCSIFFLAGS is called to inform about it. Namely, if_capenable, ec_capenable and ec_nvlans. For all three, the _init method must program the hardware specially. Not doing it resulted in: - VLAN frames getting truncated - hw checksumming not working - outgoing VLAN frames not being tagged when they should - incoming VLAN frames being treated as untagged. Fix by keeping all the old state in the softc and initializing the hardware if any of it changes. Tested on gsip. Also tested by Nino Dehne and Martin J. Laubach on sip, thanks. Fixes PRs 32900 and 33216. Approved by martin@ .
1.101.2.2	21-Apr-2006	tron	Pull up following revision(s) (requested by pavel in ticket #1279): sys/dev/pci/if_sip.c: revision 1.107 VLAN tags were sent byte-swapped by the gsip driver, because the driver forgets to convert them from the host to the network byte order. Use bswap16 to convert them. (Not htons, because on a big-endian machine, they are in the correct byte order initially, but then, they are byte-swapped by a htole32 call when written to the transmit descriptor. So byte-swapping is needed in this case too, to compensate for this htole32 call.) For a similar reason, tags were seen byte-swapped when received on a big-endian machine. Replace ntohs by bswap16 in the input path too. (Again, it is needed to compensate for a le32toh call when the receive descriptor is read.) Fixes PR 32644. Tested on 3.0/i386, 3.0/sgimips and current/alpha. OK by martin@.
1.101.2.1	08-Feb-2006	tron	Pull up following revision(s) (requested by thorpej in ticket #1161): sys/dev/pci/if_wm.c: revision 1.111 sys/dev/pci/if_sip.c: revision 1.104 In the receive interrupt handler, compute the packet length before resetting m to point to the first buffer in the packet. The bug would cause the length of jumbo frames to be incorrect.
1.102.2.8	24-Mar-2008	yamt	sync with head.
1.102.2.7	17-Mar-2008	yamt	sync with head.
1.102.2.6	11-Feb-2008	yamt	sync with head.
1.102.2.5	21-Jan-2008	yamt	sync with head
1.102.2.4	27-Oct-2007	yamt	sync with head.
1.102.2.3	03-Sep-2007	yamt	sync with head.
1.102.2.2	30-Dec-2006	yamt	sync with head.
1.102.2.1	21-Jun-2006	yamt	sync with head.
1.103.6.1	22-Apr-2006	simonb	Sync with head.
1.103.4.1	09-Sep-2006	rpaulo	sync with head
1.103.2.1	18-Feb-2006	yamt	sync with head.
1.104.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.104.6.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.104.4.1	19-Apr-2006	elad	sync with head.
1.104.2.4	26-Jun-2006	yamt	sync with head.
1.104.2.3	24-May-2006	yamt	sync with head.
1.104.2.2	11-Apr-2006	yamt	sync with head
1.104.2.1	01-Apr-2006	yamt	sync with head.
1.107.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.107.2.1	19-Jun-2006	chap	Sync with head.
1.108.6.2	10-Dec-2006	yamt	sync with head.
1.108.6.1	22-Oct-2006	yamt	sync with head
1.108.4.1	18-Nov-2006	ad	Sync with head.
1.110.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.112.4.1	11-Jul-2007	mjf	Sync with head.
1.112.2.3	23-Oct-2007	ad	Sync with head.
1.112.2.2	09-Oct-2007	ad	Sync with head.
1.112.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.113.8.3	23-Mar-2008	matt	sync with HEAD
1.113.8.2	09-Jan-2008	matt	sync with HEAD
1.113.8.1	06-Nov-2007	matt	sync with HEAD
1.113.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.113.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.113.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.114.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.115.8.3	20-Jan-2008	bouyer	Sync with HEAD
1.115.8.2	10-Jan-2008	bouyer	Sync with HEAD
1.115.8.1	02-Jan-2008	bouyer	Sync with HEAD
1.115.4.1	26-Dec-2007	ad	Sync with head.
1.115.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.127.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.127.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.127.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.127.2.1	24-Mar-2008	keiichi	sync with head.
1.132.4.6	11-Aug-2010	yamt	sync with head.
1.132.4.5	11-Mar-2010	yamt	sync with head
1.132.4.4	16-Sep-2009	yamt	sync with head
1.132.4.3	16-May-2009	yamt	sync with head
1.132.4.2	04-May-2009	yamt	sync with head.
1.132.4.1	16-May-2008	yamt	sync with head.
1.132.2.1	18-May-2008	yamt	sync with head.
1.134.8.2	28-Apr-2009	skrll	Sync with HEAD.
1.134.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.134.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.135.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.145.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.147.2.2	05-Mar-2011	rmind	sync with head
1.147.2.1	30-May-2010	rmind	sync with head
1.150.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.150.2.2	30-Oct-2012	yamt	sync with head
1.150.2.1	17-Apr-2012	yamt	sync with head
1.151.2.1	18-Feb-2012	mrg	merge to -current.
1.153.2.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #691): sys/dev/pci/aac_pci.c: revision 1.34 sys/dev/pci/if_sip.c: revision 1.155 sys/dev/pci/if_tlp_pci.c: revision 1.122 sys/dev/pci/if_an_pci.c: revision 1.34 match some more devices.
1.154.2.4	03-Dec-2017	jdolecek	update from HEAD
1.154.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.154.2.2	23-Jun-2013	tls	resync from head
1.154.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.156.4.1	18-May-2014	rmind	sync with head
1.157.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.158.4.6	28-Aug-2017	skrll	Sync with HEAD
1.158.4.5	05-Feb-2017	skrll	Sync with HEAD
1.158.4.4	05-Oct-2016	skrll	Sync with HEAD
1.158.4.3	09-Jul-2016	skrll	Sync with HEAD
1.158.4.2	19-Mar-2016	skrll	Sync with HEAD
1.158.4.1	06-Jun-2015	skrll	Sync with HEAD
1.163.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.165.6.1	11-May-2017	pgoyette	Sync with HEAD
1.166.2.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.167.2.2	26-Jan-2019	pgoyette	Sync with HEAD
1.167.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.168.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.168.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.168.2.1	10-Jun-2019	christos	Sync with HEAD
1.175.2.1	29-Feb-2020	ad	Sync with head.
1.192.2.1	02-Aug-2025	perseant	Sync with HEAD
1.22	02-Feb-2024	andvar	fix various typos in comments.
1.21	08-Mar-2020	thorpej	Support 64-bit DMA addressing on the DP83820, used only when a 64-bit DMA tag is available.
1.20	08-Feb-2018	dholland	branches: 1.20.4; Typos.
1.19	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.18	15-Dec-2007	dyoung	branches: 1.18.6; 1.18.8; 1.18.10; Finish making the sip(4) and gsip(4) drivers share compiled code.
1.17	15-Dec-2007	dyoung	Convert some #ifdef DP83820 to comments /* DP83820 only */ where that will neither make a functional difference nor cause a name collision.
1.16	14-Dec-2007	dyoung	Take a step toward sharing compiled code between sip(4) and gsip(4): get rid of SIP_DECL() and reduce #ifdef DP83820 code. Next step is to move a bunch of shared code to a new file (if_sipcom.c, say) and compile it once. While I am here, add suspend/resume handling to sip(4) and to gsip(4). Tested with the NatSemi sip(4) on the Soekris net4521. I don't have any gsip(4) to test with, and it seems that the few holders of gsip(4) in the world keep them in their attic, anyway.
1.15	11-Dec-2005	christos	branches: 1.15.46; 1.15.56; 1.15.60; merge ktrace-lwp.
1.14	11-Apr-2004	thorpej	branches: 1.14.12; Improvements to DP83820 support, from HITOSHI Osada: - Fix jumbo frame support. - Disable interrupts at the chip in sip_intr(). - Disable checksum offloading if MTU > 8109 - the hardware is broken in this case. - Use the interrupt hold-off timer. - Correct the Tx FIFO size. - Add Pause Control/Status register definitions.
1.13	03-Dec-2003	cube	o Add support for accessing the PHY through MDIO for recent SiS chips o Add support for the recent SiS96x chipsets that have a new revision. That includes a new bit of code to access the EEPROM, since it is shared with the ieee1394 controller on those chipsets. Mostly taken from FreeBSD (rev. 1.62 and 1.64 of sys/pci/if_sip.c). I tried to make the code look less ugly, but couldn't invent documentation. Fix PR #23481. Thanks to Stephane ENGEL <sengel AT melshake DOT com> for the report and the cheerful testing.
1.12	15-Aug-2003	itojun	accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.11	30-Jun-2002	thorpej	branches: 1.11.6; * Give symbolic names to the CFG bits in the EEPROM. * Get CFG_M64ADDR, CFG_T64ADDR, and CFG_DATA64_EN from the EEPROM. Note, we still disable CFG_M64ADDR and CFG_T64ADDR later (XXX need PCI bus capability flags for these). * Print a message if we're in a 64-bit slot and 64-bit data is disabled in the EEPROM. Make sure CFG_DATA64_EN is disabled if we're not in a 64-bit slot.
1.10	27-Mar-2002	briggs	branches: 1.10.2; 1.10.4; Print the chip revision in the attach function. Treat revision 0x84 of the SiS like other SIS630 variants. Call revision 0x84 the 630ET (name from OpenBSD).
1.9	09-Feb-2002	thorpej	Fix problems associated with the SiS 635/735 on-board Ethernet, from Stephen Degler <sdegler@degler.net>, port-i386/15261.
1.8	20-Dec-2001	thorpej	Fetch the MAC address from the NVRAM on the SiS 630 built-in Ethernet. Based on PR 14665, Stephen Borrill <netbsd@precedence.co.uk>.
1.7	18-May-2001	thorpej	branches: 1.7.2; Add a driver for the National Semiconductor DP83820 Gigabit Ethernet chip. This is found on the NetGear GA-622 and Asante FriendlyNet GigaNIX.
1.6	15-May-2001	thorpej	Define the register/descriptor differences between the DP83815 and the DP83820. The DP83820 definitions are wrapped in #ifdef DP83820, since there are enough differences that a single binary object would not be able to efficiently drive both the DP83815 and DP83820.
1.5	09-Mar-2001	briggs	Goes along with rev. 1.25/1.26 of if_sip.c -- forgot to commit.
1.4	21-Sep-2000	thorpej	branches: 1.4.2; Some more register definitions for the DP83815.
1.3	20-Sep-2000	thorpej	Add additional register defintions for the National Semiconductor DP83815, which is based on the SiS900 core.
1.2	31-Jan-2000	thorpej	branches: 1.2.4; Add support for the SiS 7016.
1.1	01-Jun-1999	thorpej	branches: 1.1.2; 1.1.4; Device driver for the Silicon Integrated Systems SiS900 10/100 Ethernet chip, found on the IBM Netstation.
1.1.4.2	12-Mar-2001	bouyer	Sync with HEAD.
1.1.4.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.2.4.3	08-Sep-2003	msaitoh	Pull up revision 1.12 via patch (requested by itojun in ticket #69): Accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.2.4.2	13-Mar-2001	he	Pull up revision 1.5 (requested by briggs): Fix support for the DP83815 so that it works properly with the current crop of Netgear FA311 10/100 cards.
1.2.4.1	28-Sep-2000	thorpej	Sync with trunk: Add support for the DP83815 10/100 Ethernet, found on new NetGear cards.
1.4.2.6	01-Aug-2002	nathanw	Catch up to -current.
1.4.2.5	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.4.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.4.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.4.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.4.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.7.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.7.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.2.2	16-Mar-2002	jdolecek	Catch up with -current.
1.7.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.10.4.2	15-Aug-2003	tron	Pull up revision 1.12 (requested by itojun in ticket #1413): accept 1518-byte frames (needed for vlan). Valtteri Vuorikoski
1.10.4.1	01-Nov-2002	tron	Pull up revision 1.11 (requested by thorpej in ticket #411): * Give symbolic names to the CFG bits in the EEPROM. * Get CFG_M64ADDR, CFG_T64ADDR, and CFG_DATA64_EN from the EEPROM. Note, we still disable CFG_M64ADDR and CFG_T64ADDR later (XXX need PCI bus capability flags for these). * Print a message if we're in a 64-bit slot and 64-bit data is disabled in the EEPROM. Make sure CFG_DATA64_EN is disabled if we're not in a 64-bit slot.
1.10.2.1	15-Jul-2002	gehenna	catch up with -current.
1.11.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.11.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.11.6.1	03-Aug-2004	skrll	Sync with HEAD
1.14.12.1	21-Jan-2008	yamt	sync with head
1.15.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.15.56.1	26-Dec-2007	ad	Sync with head.
1.15.46.1	09-Jan-2008	matt	sync with HEAD
1.18.10.1	16-May-2008	yamt	sync with head.
1.18.8.1	18-May-2008	yamt	sync with head.
1.18.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.20.4.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.114	04-Oct-2025	thorpej	Add a shared function to query the common properties used for configuring an Ethernet address.
1.113	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.112	23-May-2024	andvar	branches: 1.112.2; s/separare/separate/ and s/separete/separate/ in comments.
1.111	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.110	03-May-2022	andvar	fix various typos, mainly s/trasfering/transferring/ and s/theese/these/.
1.109	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.108	08-May-2021	thorpej	branches: 1.108.6; Use pci_compatible_match().
1.107	24-Apr-2021	thorpej	branches: 1.107.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.106	02-Jul-2020	msaitoh	branches: 1.106.4; prop_data_data_nocopy -> prop_data_value
1.105	24-Feb-2020	rin	0x%p --> %p for non-external codes.
1.104	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.103	27-Dec-2019	msaitoh	branches: 1.103.2; s/is is/is/ in comment.
1.102	28-Nov-2019	maxv	localify
1.101	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.100	03-Jun-2019	msaitoh	KNF. No functional change.
1.99	03-Jun-2019	msaitoh	Make new sk_unreset_{xmac,yukon}() and use them in sk_attach(). Fixes PR kern/54267 reported by martin.
1.98	30-May-2019	msaitoh	Simplify MII structure initialization and reference. No functional change.
1.97	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.96	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.95	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.94	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.93	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.92	08-Jan-2019	msaitoh	Whitespace fix. No functional change.
1.91	08-Jan-2019	msaitoh	u_int{8,16,32,64}_t -> uint{8,16,32,64}_t. No functional change.
1.90	15-Nov-2018	maxv	Remove the 'copy' argument from m_devget(), unused. While here rename off0->off.
1.89	04-Jul-2018	jdolecek	use pci_intr_establish_xname()
1.88	03-Jul-2018	jdolecek	attach the rnd source only once even with dual-port adapters
1.87	26-Jun-2018	msaitoh	branches: 1.87.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.86	12-Jun-2018	jdolecek	move SK_HASH_BITS to if_skreg.h and reuse in if_msk.c (from OpenBSD)
1.85	15-Dec-2016	ozaki-r	branches: 1.85.8; 1.85.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.84	14-Dec-2016	christos	Tidy up and make it look like the other drivers.
1.83	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.82	10-Jun-2016	ozaki-r	branches: 1.82.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.81	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.80	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.79	10-Dec-2014	christos	don't give up on mii tick before the interface is marked up.
1.78	10-Aug-2014	tls	branches: 1.78.4; Merge tls-earlyentropy branch into HEAD.
1.77	29-Mar-2014	christos	branches: 1.77.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.76	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.75	13-Sep-2013	martin	#ifdef a variable like their use
1.74	11-Jun-2013	msaitoh	branches: 1.74.2; Remove an extra mii_pollstat() call. The PHY status should be updated by adjacent mii_tick() call. I suspect that this mii_pollstat() call was added to do workaround for broken MII_TICK code. A lot of MII PHY drivers had bugs in MII_TICK and those bugs were fixed.
1.73	30-Mar-2013	christos	remove trailing whitespace
1.72	22-Jul-2012	matt	branches: 1.72.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.71	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.70	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.69	29-May-2011	phx	branches: 1.69.4; 1.69.8; The Genesis interface provides a register in the device specific config space to automatically revert the byte order in all descriptors. Make sure this feature is inactive. Otherwise the driver is unable to receive or transmit any frame.
1.68	26-Jul-2010	jym	branches: 1.68.2; Add PAE to ALL kernel, so that most paddr_t format string errors get caught during compilation. While here, fix the compilation for ALL.
1.67	16-May-2010	phx	Make netboot generate a BTINFO_NET bootinfo node for the Synology sk(4) NIC. It will pass the MAC address, which is read from Flash ROM, into the kernel. The kernel creates a "mac-address" device-property, which is used by sk(4), when given, before reading the MAC from its EEPROM.
1.66	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.65	24-Feb-2010	dyoung	branches: 1.65.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.64	19-Jan-2010	pooka	branches: 1.64.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.63	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.62	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.61	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.60	23-Apr-2009	kefren	Add pmf hooks. Proposed two days ago on tech-kern@, no objections. Tested on amd64 with Asus P5Q-E mainboard with onboard Yukon Lite
1.59	18-Mar-2009	cegger	bzero -> memset
1.58	18-Mar-2009	cegger	bcmp -> memcmp
1.57	13-Feb-2009	bouyer	More printf format fixes.
1.56	12-Feb-2009	cegger	Yukon chips support flow-control. Tested on Yukon Lite chip. ok cube@
1.55	07-Nov-2008	dyoung	branches: 1.55.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.54	09-Sep-2008	cegger	branches: 1.54.2; 1.54.4; 1.54.6; make an i386/ALL kernel build
1.53	08-Sep-2008	christos	more device_private, pointed by cube.
1.52	08-Sep-2008	christos	avoid using casts by referencing the proper struct members.
1.51	08-Sep-2008	christos	cf_attach_decl_new
1.50	20-Jun-2008	cube	branches: 1.50.2; Use a mutex instead of splvm() to protect the list of jubo-ready mbufs, as done with nfe(4) a while ago. Issue reported by Gary Duzan, who kindly fixed the patch I had sent him.
1.49	28-Apr-2008	martin	branches: 1.49.2; 1.49.4; Remove clause 3 and 4 from TNF licenses
1.48	10-Apr-2008	cegger	branches: 1.48.2; 1.48.4; use aprint_*_dev and device_xname
1.47	07-Feb-2008	dyoung	branches: 1.47.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.46	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.45	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.44	01-Dec-2007	jmcneill	branches: 1.44.2; 1.44.4; 1.44.6; aprintify, on behalf of xtraeme
1.43	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.42	19-Jul-2007	dsl	branches: 1.42.4; 1.42.6; 1.42.10; 1.42.12; 1.42.14; include sys/cdefs.h
1.41	09-Jul-2007	ad	branches: 1.41.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.40	06-Jul-2007	briggs	Print an error and fail the sk_attach() if we have an unknown chip type (instead of panicing). Also check to see if we have a Yukon2 type, and if we do, complain about that explicitly earlier (and still fail the attach, but refer to msk(4)). It would be better if we didn't have Yukon2 devices match in the sk(4) device table, but in case people change the hardware without changing device IDs, we might as well catch it.
1.39	06-Jul-2007	briggs	Remove DLink DGE560T as it (in at least some versions) uses a Yukon EC chipset, which is not handled by this driver, but is handled by the msk(4) driver. The DGE560T_2 entry is left alone for now--it might also need to go to msk(4).
1.38	13-Mar-2007	msaitoh	Apply if_msk.c rev. 1.4 to if_sk.c: > Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" > > Original commit message: > > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > > continuing when busy flag set.
1.37	04-Mar-2007	christos	branches: 1.37.2; 1.37.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.36	30-Jan-2007	msaitoh	branches: 1.36.2; sync with if_skreg.h rev. 1.41 need more work?
1.35	16-Nov-2006	christos	branches: 1.35.2; __unused removal on arguments; approved by core.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	10-Sep-2006	riz	branches: 1.33.2; g/c unused SK_{SET,CLR}BIT macros
1.32	25-Aug-2006	riz	branches: 1.32.2; In sk_watchdog(), attempt to reclaim pending TX descriptors before resetting the hardware. Should fix occasional watchdog timeouts seen with this driver. From FreeBSD, via OpenBSD.
1.31	25-Aug-2006	riz	Cleanup, with no functional changes: - remove unused code - KNF - ANSI function declarations - replace printf() with aprint_error() except in debug functions - a few minor indentation/whitespace changes
1.30	22-Aug-2006	riz	Vendor GALILEO is now MARVELL.
1.29	22-Aug-2006	riz	Add byteswapping where appropriate, to allow things to work on (at least some) bigendian systems - I tested on macppc. From brad@openbsd .
1.28	20-Aug-2006	riz	Some cleanup in skc_attach(): - does not need splnet() - disestablish the interrupt and unmap bus space on failures.
1.27	20-Aug-2006	riz	Make this driver compile when SK_USEIOSPACE is defined - I have no evidence that this is actually needed except for the existence of the code itself, but if it's going to be here, it should compile. Tested briefly on my ASUS motherboard with built-in sk interface.
1.26	31-May-2006	riz	branches: 1.26.4; Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.25	17-Apr-2006	christos	branches: 1.25.2; Fix typo. Pointed out by brad.
1.24	14-Apr-2006	christos	Coviery CID 1108: Avoid NULL pointer deref.
1.23	14-Apr-2006	christos	Coverity CID 1105: Avoid NULL deref.
1.22	28-Mar-2006	riz	Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months.
1.21	25-Feb-2006	wiz	branches: 1.21.2; 1.21.4; 1.21.6; Fix some typos.
1.20	14-Feb-2006	riz	Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.19	11-Feb-2006	rpaulo	PR 31304: Contribute to the random pool. From Rhialto.
1.18	23-Nov-2005	riz	branches: 1.18.2; 1.18.4; 1.18.6; Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.17	19-Nov-2005	riz	From FreeBSD (appropriate pieces of revs 1.90 and 1.92): Only clear the IFF_OACTIVE flag when we have a chance of being able to queue a packet to the hardware, instead of when the hardware queue is empty, and fix up handling and prodding of the tx. These fixes clear up an occasional "sk0: watchdog timeout" from the on-board ethernet on my Asus A8V motherboard. OK christos@
1.16	11-Sep-2005	xtraeme	branches: 1.16.6; Fix from OpenBSD: rev 1.56: Don't initialize the card (and start an autonegotiation!) every time the IP address changes. Makes 'dhclient sk0' invocations way faster and more consistant. i.e. one DHCPREQUEST elicts the DHCPACK. Fix from FreeBSD: rev 1.109: Solve "No PHY found" problem for more Yukon Lite variants. These changes fixed the problem on my sk(4) trying to get an IP via dhclient(8).
1.15	30-May-2005	christos	branches: 1.15.2; - const poisoning - avoid variable shadowing.
1.14	27-Feb-2005	perry	branches: 1.14.2; nuke trailing whitespace
1.13	23-Jan-2005	fredb	branches: 1.13.2; Recognize Belkin Gigabit Desktop Network PCI card.
1.12	30-Oct-2004	thorpej	branches: 1.12.4; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.11	26-Sep-2004	skd	Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.10	24-May-2004	kleink	branches: 1.10.2; Fix a glitch in the initial import: The XMAC hash function is actually based on CRC32-_LE_. From Christian Weisgerber in private mail.
1.9	24-May-2004	kleink	In sk_init_yukon(), don't open-code clearing the multicast filter but go via the normal filter setup path instead. The old behaviour only worked ok as long as the init function was exclusively called at attach time, when there was no configuration to be lost. From Christian Weisgerber in private mail.
1.8	07-May-2004	kleink	Add an additional multicast hash function for Yukon.
1.7	28-Jan-2004	chs	branches: 1.7.2; also match the version of this that's built into my Asus A7N8X-E Deluxe motherboard.
1.6	17-Dec-2003	tls	Add support for some more Marvell-based adapters to the sk driver: D-Link DGE-530T Linksys EG1032v2 Linksys EG1064v2 The Linux driver treats all of these the same. Tested with a DGE-530T; it works, though performance on at least one of my systems is atrocious. Fix setting of ethernet capabilities into interface capabilites word in attach routine. Manual page: note that jumbo support doesn't work and checksum support doesn't exist.
1.5	30-Oct-2003	briggs	Make sure we can transmit larger packets correctly. Don't initialize extra fields of the ifp. Use strcpy() instead of bcopy() for copying the device name. Handle if_timer a bit better.
1.4	25-Oct-2003	chs	NULL -> 0
1.3	15-Oct-2003	briggs	Make sure descriptors get bus_dmamap_sync()ed. Restructure a little to make this a little easier. Also try to avoid allocating things at reset time.
1.2	08-Sep-2003	thorpej	Fix a printf format.
1.1	26-Aug-2003	jdolecek	Add driver for SysKonnect SK-9821 and 3COM 3C940 gigabit ethernet boards From FreeBSD via OpenBSD, port to NetBSD done by Stephen Degler Changes relative to submitted version: * yukonreg.h and xmaciireg.h merged into if_skreg.h * bhack[] constified + other small editing changes * use 'Ethernet address' rather than 'address' in attach message XXX completely untested by me, needs further cleanup Driver provided in PR kern/22511 by Stephen Degler
1.7.2.4	07-Jan-2005	jdc	Pull up revision 1.11 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.7.2.3	25-May-2004	jmc	branches: 1.7.2.3.2; Pullup rev 1.10 (requested by kleink in ticket #384) Fix a glitch in the initial import: The XMAC hash function is actually based on CRC32-_LE_.
1.7.2.2	25-May-2004	jmc	Pullup rev 1.9 (requested by kleink in ticket #382) In sk_init_yukon(), don't open-code clearing the multicast filter but go via the normal filter setup path instead. The old behaviour only worked ok as long as the init function was exclusively called at attach time, when there was no configuration to be lost.
1.7.2.1	11-May-2004	tron	Pull up revision 1.8 (requested by kleink in ticket #286): Add an additional multicast hash function for Yukon.
1.7.2.3.2.11	11-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #10693): sys/arch/cobalt/pci/pchb.c: revision 1.9 via patch sys/dev/pci/if_sk.c: revision 1.30 via patch sys/arch/evbmips/malta/pci/pchb.c: revision 1.9 via patch Vendor GALILEO is now MARVELL.
1.7.2.3.2.10	07-Sep-2006	rpaulo	Pull up following revision(s) (requested by riz in ticket #10680): sys/dev/pci/if_skvar.h: revision 1.10 share/man/man4/sk.4: revision 1.9 sys/dev/pci/if_sk.c: revision 1.22 Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months. Jumbo frames now work under NetBSD, so note it.
1.7.2.3.2.9	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #10636): sys/dev/pci/if_skreg.h: revision 1.7 sys/dev/pci/if_sk.c: revision 1.26 Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.7.2.3.2.8	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #10403): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.7.2.3.2.7	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #10402): sys/dev/pci/if_sk.c: revision 1.19 sys/dev/pci/if_skvar.h: revision 1.8 PR 31304: Contribute to the random pool. From Rhialto.
1.7.2.3.2.6	01-Dec-2005	tron	Pull up following revision(s) (requested by riz in ticket #10156): sys/dev/pci/if_sk.c: revision 1.18 via patch sys/dev/pci/if_re.c: patch Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.7.2.3.2.5	01-Dec-2005	tron	Pull up following revision(s) (requested by riz in ticket #10154): sys/dev/pci/if_sk.c: revision 1.17 From FreeBSD (appropriate pieces of revs 1.90 and 1.92): Only clear the IFF_OACTIVE flag when we have a chance of being able to queue a packet to the hardware, instead of when the hardware queue is empty, and fix up handling and prodding of the tx. These fixes clear up an occasional "sk0: watchdog timeout" from the on-board ethernet on my Asus A8V motherboard. OK christos@
1.7.2.3.2.4	13-Sep-2005	riz	Pull up following revision(s) (requested by xtraeme in ticket #5841): sys/dev/pci/if_sk.c: revision 1.16 Fix from OpenBSD: rev 1.56: Don't initialize the card (and start an autonegotiation!) every time the IP address changes. Makes 'dhclient sk0' invocations way faster and more consistant. i.e. one DHCPREQUEST elicts the DHCPACK. Fix from FreeBSD: rev 1.109: Solve "No PHY found" problem for more Yukon Lite variants. These changes fixed the problem on my sk(4) trying to get an IP via dhclient(8).
1.7.2.3.2.3	17-Apr-2005	tron	Pull up revision 1.13 (requested by fredb in ticket #1120): Recognize Belkin Gigabit Desktop Network PCI card.
1.7.2.3.2.2	24-Jan-2005	he	Pull up revision 1.12 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.7.2.3.2.1	07-Jan-2005	jdc	Pull up revision 1.11 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.10.2.10	11-Dec-2005	christos	Sync with head.
1.10.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.7	24-Jan-2005	skrll	Sync with HEAD.
1.10.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.10.2.5	19-Oct-2004	skrll	Sync with HEAD
1.10.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.2	03-Aug-2004	skrll	Sync with HEAD
1.10.2.1	24-May-2004	skrll	file if_sk.c was added on branch ktrace-lwp on 2004-08-03 10:49:09 +0000
1.12.4.1	29-Apr-2005	kent	sync with -current
1.13.2.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.2.7	03-Nov-2006	tron	Pull up following revision(s) (requested by riz in ticket #1480): sys/dev/pci/if_skvar.h: revision 1.10 sys/dev/pci/if_sk.c: revision 1.22 Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months.
1.14.2.6	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #1350): sys/dev/pci/if_skreg.h: revision 1.7 sys/dev/pci/if_sk.c: revision 1.26 Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.14.2.5	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #1230): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.14.2.4	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #1229): sys/dev/pci/if_sk.c: revision 1.19 sys/dev/pci/if_skvar.h: revision 1.8 PR 31304: Contribute to the random pool. From Rhialto.
1.14.2.3	24-Nov-2005	tron	Pull up following revision(s) (requested by riz in ticket #993): sys/dev/pci/if_sk.c: revision 1.18 sys/dev/pci/if_re_pci.c: revision 1.11 Special-case the Linksys EG1032 cards using PCI subsystem ID; rev. 2 uses the sk(4) driver, while rev. 3 needs re(4). Mostly from brad@openbsd.
1.14.2.2	21-Nov-2005	tron	Pull up following revision(s) (requested by riz in ticket #979): sys/dev/pci/if_sk.c: revision 1.17 From FreeBSD (appropriate pieces of revs 1.90 and 1.92): Only clear the IFF_OACTIVE flag when we have a chance of being able to queue a packet to the hardware, instead of when the hardware queue is empty, and fix up handling and prodding of the tx. These fixes clear up an occasional "sk0: watchdog timeout" from the on-board ethernet on my Asus A8V motherboard. OK christos@
1.14.2.1	13-Sep-2005	tron	Pull up following revision(s) (requested by xtraeme in ticket #765): sys/dev/pci/if_sk.c: revision 1.16 Fix from OpenBSD: rev 1.56: Don't initialize the card (and start an autonegotiation!) every time the IP address changes. Makes 'dhclient sk0' invocations way faster and more consistant. i.e. one DHCPREQUEST elicts the DHCPACK. Fix from FreeBSD: rev 1.109: Solve "No PHY found" problem for more Yukon Lite variants. These changes fixed the problem on my sk(4) trying to get an IP via dhclient(8).
1.15.2.8	11-Feb-2008	yamt	sync with head.
1.15.2.7	21-Jan-2008	yamt	sync with head
1.15.2.6	07-Dec-2007	yamt	sync with head
1.15.2.5	15-Nov-2007	yamt	sync with head.
1.15.2.4	03-Sep-2007	yamt	sync with head.
1.15.2.3	26-Feb-2007	yamt	sync with head.
1.15.2.2	30-Dec-2006	yamt	sync with head.
1.15.2.1	21-Jun-2006	yamt	sync with head.
1.16.6.2	29-Nov-2005	yamt	sync with head.
1.16.6.1	22-Nov-2005	yamt	sync with head.
1.18.6.2	01-Jun-2006	kardel	Sync with head.
1.18.6.1	22-Apr-2006	simonb	Sync with head.
1.18.4.1	09-Sep-2006	rpaulo	sync with head
1.18.2.2	01-Mar-2006	yamt	sync with head.
1.18.2.1	18-Feb-2006	yamt	sync with head.
1.21.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.21.6.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.21.4.1	19-Apr-2006	elad	sync with head.
1.21.2.5	14-Sep-2006	yamt	sync with head.
1.21.2.4	03-Sep-2006	yamt	sync with head.
1.21.2.3	26-Jun-2006	yamt	sync with head.
1.21.2.2	24-May-2006	yamt	sync with head.
1.21.2.1	01-Apr-2006	yamt	sync with head.
1.25.2.1	19-Jun-2006	chap	Sync with head.
1.26.4.2	08-Sep-2006	rpaulo	Pull up following revision(s) (requested by riz in ticket #140): sys/dev/pci/if_sk.c: revision 1.31 sys/dev/pci/if_sk.c: revision 1.32 sys/dev/pci/if_sk.c: revision 1.27 sys/dev/pci/if_sk.c: revision 1.28 sys/dev/pci/if_sk.c: revision 1.29 Make this driver compile when SK_USEIOSPACE is defined - I have no evidence that this is actually needed except for the existence of the code itself, but if it's going to be here, it should compile. Tested briefly on my ASUS motherboard with built-in sk interface. Some cleanup in skc_attach(): - does not need splnet() - disestablish the interrupt and unmap bus space on failures. Add byteswapping where appropriate, to allow things to work on (at least some) bigendian systems - I tested on macppc. From brad@openbsd . Cleanup, with no functional changes: - remove unused code - KNF - ANSI function declarations - replace printf() with aprint_error() except in debug functions - a few minor indentation/whitespace changes In sk_watchdog(), attempt to reclaim pending TX descriptors before resetting the hardware. Should fix occasional watchdog timeouts seen with this driver. From FreeBSD, via OpenBSD.
1.26.4.1	08-Sep-2006	rpaulo	Pull up following revision(s) (requested by riz in ticket #139): sys/arch/cobalt/pci/pchb.c: revision 1.9 sys/dev/pci/if_sk.c: revision 1.30 sys/arch/cobalt/pci/pci_machdep.c: revision 1.24 sys/dev/pci/pcidevs: revision 1.820 sys/arch/evbmips/malta/pci/pchb.c: revision 1.9 Rename vendor GALILEO -> vendor MARVELL . Marvell bought Galileo years ago, we have little existing stuff in the tree, and this will make it easier to keep certain network drivers in sync with other OSes. OK matt@, uwe@ Vendor GALILEO is now MARVELL.
1.32.2.2	01-Feb-2007	ad	Sync with head.
1.32.2.1	18-Nov-2006	ad	Sync with head.
1.33.2.2	10-Dec-2006	yamt	sync with head.
1.33.2.1	22-Oct-2006	yamt	sync with head
1.35.2.2	26-Aug-2007	liamjfoy	Pull up following revision(s) (requested by riz in ticket #816): sys/dev/pci/if_skreg.h: revision 1.10 sys/dev/pci/if_skreg.h: revision 1.11 sys/dev/pci/pcidevs: revision 1.866 sys/dev/pci/if_skreg.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.36 sys/dev/pci/if_msk.c: revision 1.4 sys/dev/pci/if_msk.c: revision 1.5 sys/dev/pci/if_msk.c: revision 1.6 sys/dev/pci/if_msk.c: revision 1.7 sys/dev/pci/if_mskvar.h: revision 1.2 sys/dev/pci/if_mskvar.h: revision 1.3 Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set. add some msk devices sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41 sync with if_skreg.h rev. 1.41 need more work? apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs fix bit definitions for the RX FIFO Flush mode Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.35.2.1	26-Mar-2007	jdc	branches: 1.35.2.1.2; Pull up revision 1.38 (requested by msaitoh in ticket #519). Apply if_msk.c rev. 1.4 to if_sk.c: > Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" > > Original commit message: > > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > > continuing when busy flag set.
1.35.2.1.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.36.2.2	24-Mar-2007	yamt	sync with head.
1.36.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.37.4.1	11-Jul-2007	mjf	Sync with head.
1.37.2.5	01-Sep-2007	ad	Update for pool_cache API changes.
1.37.2.4	20-Aug-2007	ad	Sync with HEAD.
1.37.2.3	15-Jul-2007	ad	Sync with head.
1.37.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.37.2.1	13-Mar-2007	ad	Sync with head.
1.41.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.42.14.2	19-Jul-2007	dsl	include sys/cdefs.h
1.42.14.1	19-Jul-2007	dsl	file if_sk.c was added on branch matt-mips64 on 2007-07-19 22:04:24 +0000
1.42.12.4	18-Feb-2008	mjf	Sync with HEAD.
1.42.12.3	27-Dec-2007	mjf	Sync with HEAD.
1.42.12.2	08-Dec-2007	mjf	Sync with HEAD.
1.42.12.1	19-Nov-2007	mjf	Sync with HEAD.
1.42.10.1	13-Nov-2007	bouyer	Sync with HEAD
1.42.6.3	23-Mar-2008	matt	sync with HEAD
1.42.6.2	09-Jan-2008	matt	sync with HEAD
1.42.6.1	08-Nov-2007	matt	sync with -HEAD
1.42.4.2	03-Dec-2007	joerg	Sync with HEAD.
1.42.4.1	11-Nov-2007	joerg	Sync with HEAD.
1.44.6.2	20-Jan-2008	bouyer	Sync with HEAD
1.44.6.1	13-Dec-2007	bouyer	Sync with HEAD
1.44.4.1	11-Dec-2007	yamt	sync with head.
1.44.2.1	26-Dec-2007	ad	Sync with head.
1.47.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.47.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.47.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.47.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.48.4.5	11-Aug-2010	yamt	sync with head.
1.48.4.4	11-Mar-2010	yamt	sync with head
1.48.4.3	16-Sep-2009	yamt	sync with head
1.48.4.2	04-May-2009	yamt	sync with head.
1.48.4.1	16-May-2008	yamt	sync with head.
1.48.2.1	18-May-2008	yamt	sync with head.
1.49.4.1	27-Jun-2008	simonb	Sync with head.
1.49.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.49.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.50.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.50.2.1	19-Oct-2008	haad	Sync with HEAD.
1.54.6.1	26-Jul-2009	snj	branches: 1.54.6.1.2; Pull up following revision(s) (requested by dholland in ticket #874): sys/dev/ic/ciss.c: revision 1.15 sys/dev/isa/isadma.c: revision 1.59 sys/dev/usb/ulpt.c: revision 1.82 sys/dev/ic/tcic2.c: revision 1.31 sys/dev/isa/if_ntwoc_isa.c: revision 1.18 sys/dev/pci/if_sk.c: revision 1.57 sys/dev/pcmcia/if_cnw.c: revision 1.46 More printf format fixes.
1.54.6.1.2.1	21-Apr-2010	matt	sync to netbsd-5
1.54.4.1	26-Jul-2009	snj	Pull up following revision(s) (requested by dholland in ticket #874): sys/dev/ic/ciss.c: revision 1.15 sys/dev/isa/isadma.c: revision 1.59 sys/dev/usb/ulpt.c: revision 1.82 sys/dev/ic/tcic2.c: revision 1.31 sys/dev/isa/if_ntwoc_isa.c: revision 1.18 sys/dev/pci/if_sk.c: revision 1.57 sys/dev/pcmcia/if_cnw.c: revision 1.46 More printf format fixes.
1.54.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.54.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.54.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.55.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.64.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.64.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.65.2.3	31-May-2011	rmind	sync with head
1.65.2.2	05-Mar-2011	rmind	sync with head
1.65.2.1	30-May-2010	rmind	sync with head
1.68.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.69.8.1	18-Feb-2012	mrg	merge to -current.
1.69.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.69.4.2	30-Oct-2012	yamt	sync with head
1.69.4.1	17-Apr-2012	yamt	sync with head
1.72.2.3	03-Dec-2017	jdolecek	update from HEAD
1.72.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.72.2.1	23-Jun-2013	tls	resync from head
1.74.2.1	18-May-2014	rmind	sync with head
1.77.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.78.4.5	05-Feb-2017	skrll	Sync with HEAD
1.78.4.4	09-Jul-2016	skrll	Sync with HEAD
1.78.4.3	19-Mar-2016	skrll	Sync with HEAD
1.78.4.2	06-Jun-2015	skrll	Sync with HEAD
1.78.4.1	06-Apr-2015	skrll	Sync with HEAD
1.82.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.85.14.5	26-Jan-2019	pgoyette	Sync with HEAD
1.85.14.4	18-Jan-2019	pgoyette	Synch with HEAD
1.85.14.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.85.14.2	28-Jul-2018	pgoyette	Sync with HEAD
1.85.14.1	25-Jun-2018	pgoyette	Sync with HEAD
1.85.8.1	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.87.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.87.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.87.2.1	10-Jun-2019	christos	Sync with HEAD
1.103.2.1	29-Feb-2020	ad	Sync with head.
1.106.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.107.2.1	13-May-2021	thorpej	Sync with HEAD.
1.108.6.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.112.2.1	02-Aug-2025	perseant	Sync with HEAD
1.34	16-Feb-2025	jakllsch	Add additional Yukon2 definitions relating to hardware VLAN tagging and queue interrupt information
1.33	09-Feb-2024	andvar	branches: 1.33.2; fix spelling mistakes, mainly in comments and log messages.
1.32	02-Feb-2024	andvar	fix various typos in comments.
1.31	23-Feb-2022	andvar	fix various typos in comments, mainly immediatly/immediately/, as well shared and recently fixed typos in OpenBSD code by Jonathan Grey.
1.30	02-Aug-2021	andvar	fix various typos in comments and log messages.
1.29	08-May-2020	jakllsch	if_skreg.h: remove duplicate or triplicate identical #defines
1.28	26-Apr-2020	jakllsch	Use __aligned(8) with naturally-aligned-yet-__packed msk(4) descriptors. Allows compiler to generate more efficent struct member accesses.
1.27	15-Nov-2019	msaitoh	Make Yukon EX, FE+, SUPR stable. The code is mainly taken from FreeBSD. At least, this change made my own Yukon EX machine (HP ProBook 4501s) much stable than before.
1.26	05-Mar-2019	msaitoh	branches: 1.26.4; Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.25	14-Sep-2018	jakllsch	msk(4): add 64-bit DMA support portions of this change set provided by mrg@
1.24	11-Jul-2018	kre	Build fix: Supply missing symbol definition. Value taken from OpenBSD (and confirmed with FreeBSD) (Feel free to fix/reformat as appropriate...)
1.23	10-Jul-2018	kre	Whitespace fixes. NFC.
1.22	04-Jul-2018	jdolecek	sync SK_IMTIMER_TICKS_YUKON value with OpenBSD (likely pasto)
1.21	14-Jun-2018	msaitoh	branches: 1.21.2; s/regisrer/register/. No functional change.
1.20	14-Jun-2018	msaitoh	The revision number of Yukon FE A2 is not 0x3 but 0x2. Same as other OSes.
1.19	13-Jun-2018	jdolecek	actually recognize the Yukon 2 variants up to SK_YUKON_OPTIMA2 as such
1.18	12-Jun-2018	jdolecek	move SK_HASH_BITS to if_skreg.h and reuse in if_msk.c (from OpenBSD)
1.17	12-Jun-2018	jdolecek	match three more variants - DGE550T_B1, 8040T, 8042 sync more sk_type conditionals from OpenBSD
1.16	09-Jun-2018	jdolecek	port basic match support for several more Yukon variants from OpenBSD, together with some basic conditionals I've found in that driver; only compile-tested, so very likely needs further work to actually work adresses also PR kern/43507
1.15	10-Oct-2014	christos	branches: 1.15.18; PR/49270: Hauke Fath: Samsung NC10 Marvell Yukon ethernet not supported
1.14	29-May-2011	phx	branches: 1.14.14; The Genesis interface provides a register in the device specific config space to automatically revert the byte order in all descriptors. Make sure this feature is inactive. Otherwise the driver is unable to receive or transmit any frame.
1.13	18-Nov-2008	chris	branches: 1.13.8; 1.13.10; Fix Yukon EC Ultra cold power up issue. For the EC Ultra it is necessary to update some extra registers during reset. Without doing so causes the system to hang at boot. The only workaround I found was to PXE boot before booting into NetBSD. This change is based on the code from FreeBSD's if_msk.c. Specifically the msk_phy_power function. Also add an splnet/splx across mii_tick. This matches most other network drivers. Change posted for review on 3rd Oct 2008 to tech-net. No feedback received.
1.12	28-Apr-2008	martin	branches: 1.12.6; 1.12.8; 1.12.10; Remove clause 3 and 4 from TNF licenses
1.11	31-Jan-2007	msaitoh	branches: 1.11.40; 1.11.42; 1.11.44; fix bit definitions for the RX FIFO Flush mode
1.10	30-Jan-2007	msaitoh	apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs
1.9	30-Jan-2007	msaitoh	sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41
1.8	09-Sep-2006	riz	branches: 1.8.4; 1.8.6; Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.7	31-May-2006	riz	branches: 1.7.4; 1.7.6; Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.6	14-Feb-2006	riz	branches: 1.6.2; 1.6.8; Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.5	11-Dec-2005	christos	branches: 1.5.2; 1.5.4; 1.5.6; merge ktrace-lwp.
1.4	30-May-2005	christos	branches: 1.4.2; - const poisoning - avoid variable shadowing.
1.3	26-Sep-2004	skd	branches: 1.3.10; Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.2	26-Aug-2003	jdolecek	branches: 1.2.2; 1.2.4; 1.2.6; retain the FreeBSD RCS Id for xmaciireg.h for reference
1.1	26-Aug-2003	jdolecek	Add driver for SysKonnect SK-9821 and 3COM 3C940 gigabit ethernet boards From FreeBSD via OpenBSD, port to NetBSD done by Stephen Degler Changes relative to submitted version: * yukonreg.h and xmaciireg.h merged into if_skreg.h * bhack[] constified + other small editing changes * use 'Ethernet address' rather than 'address' in attach message XXX completely untested by me, needs further cleanup Driver provided in PR kern/22511 by Stephen Degler
1.2.6.3	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #10636): sys/dev/pci/if_skreg.h: revision 1.7 sys/dev/pci/if_sk.c: revision 1.26 Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.2.6.2	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #10403): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.2.6.1	07-Jan-2005	jdc	Pull up revision 1.3 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.2.4.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.4.5	19-Oct-2004	skrll	Sync with HEAD
1.2.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.4.2	03-Aug-2004	skrll	Sync with HEAD
1.2.4.1	26-Aug-2003	skrll	file if_skreg.h was added on branch ktrace-lwp on 2004-08-03 10:49:09 +0000
1.2.2.1	07-Jan-2005	jdc	Pull up revision 1.3 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.3.10.2	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #1350): sys/dev/pci/if_skreg.h: revision 1.7 sys/dev/pci/if_sk.c: revision 1.26 Add support for the DLink DGE-530T revision B1 (Tested by Tillman Hodgson) and the DGE-560T (untested).
1.3.10.1	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #1230): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.4.2.3	26-Feb-2007	yamt	sync with head.
1.4.2.2	30-Dec-2006	yamt	sync with head.
1.4.2.1	21-Jun-2006	yamt	sync with head.
1.5.6.2	01-Jun-2006	kardel	Sync with head.
1.5.6.1	22-Apr-2006	simonb	Sync with head.
1.5.4.1	09-Sep-2006	rpaulo	sync with head
1.5.2.1	18-Feb-2006	yamt	sync with head.
1.6.8.1	19-Jun-2006	chap	Sync with head.
1.6.2.2	14-Sep-2006	yamt	sync with head.
1.6.2.1	26-Jun-2006	yamt	sync with head.
1.7.6.2	01-Feb-2007	ad	Sync with head.
1.7.6.1	18-Nov-2006	ad	Sync with head.
1.7.4.1	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/if_skreg.h: revision 1.8 sys/dev/pci/if_msk.c: revision 1.1 sys/dev/pci/files.pci: revision 1.268 sys/dev/pci/if_mskvar.h: revision 1.1 Add msk(4), a driver for Marvell Yukon 2 gigabit ethernet chips, from Mark Kettenis of OpenBSD. There are still some outstanding issues with this driver, namely: - Checksum offload is unsupported - There is a significant amount of code duplication from sk(4) - There remain some 'magic numbers' - Performance is not heavily tested, and likely to be lower than the chip is capable of in some cases. Syncing some of the aforementioned 'magic numbers' with the Marvell FreeBSD driver should help here. Tested on a motherboard with Marvell 88E8053 ethernet, under NetBSD/i386 and NetBSD/amd64.
1.8.6.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.8.4.1	26-Aug-2007	liamjfoy	Pull up following revision(s) (requested by riz in ticket #816): sys/dev/pci/if_skreg.h: revision 1.10 sys/dev/pci/if_skreg.h: revision 1.11 sys/dev/pci/pcidevs: revision 1.866 sys/dev/pci/if_skreg.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.36 sys/dev/pci/if_msk.c: revision 1.4 sys/dev/pci/if_msk.c: revision 1.5 sys/dev/pci/if_msk.c: revision 1.6 sys/dev/pci/if_msk.c: revision 1.7 sys/dev/pci/if_mskvar.h: revision 1.2 sys/dev/pci/if_mskvar.h: revision 1.3 Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set. add some msk devices sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41 sync with if_skreg.h rev. 1.41 need more work? apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs fix bit definitions for the RX FIFO Flush mode Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.11.44.2	04-May-2009	yamt	sync with head.
1.11.44.1	16-May-2008	yamt	sync with head.
1.11.42.1	18-May-2008	yamt	sync with head.
1.11.40.2	17-Jan-2009	mjf	Sync with HEAD.
1.11.40.1	02-Jun-2008	mjf	Sync with HEAD.
1.12.10.1	20-Nov-2008	snj	Pull up following revision(s) (requested by chris in ticket #84): sys/dev/pci/if_skreg.h: revision 1.13 sys/dev/pci/if_msk.c: revision 1.22 Fix Yukon EC Ultra cold power up issue. For the EC Ultra it is necessary to update some extra registers during reset. Without doing so causes the system to hang at boot. The only workaround I found was to PXE boot before booting into NetBSD. This change is based on the code from FreeBSD's if_msk.c. Specifically the msk_phy_power function. Also add an splnet/splx across mii_tick. This matches most other network drivers. Change posted for review on 3rd Oct 2008 to tech-net. No feedback received.
1.12.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.12.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.13.10.1	06-Jun-2011	jruoho	Sync with HEAD.
1.13.8.1	31-May-2011	rmind	sync with head
1.14.14.1	03-Dec-2017	jdolecek	update from HEAD
1.15.18.3	30-Sep-2018	pgoyette	Ssync with HEAD
1.15.18.2	28-Jul-2018	pgoyette	Sync with HEAD
1.15.18.1	25-Jun-2018	pgoyette	Sync with HEAD
1.21.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.21.2.1	10-Jun-2019	christos	Sync with HEAD
1.26.4.1	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #450): sys/dev/pci/if_msk.c: revision 1.94 sys/dev/pci/if_skreg.h: revision 1.27 Make Yukon EX, FE+, SUPR stable. The code is mainly taken from FreeBSD. At least, this change made my own Yukon EX machine (HP ProBook 4501s) much stable than before.
1.33.2.1	02-Aug-2025	perseant	Sync with HEAD
1.20	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.19	03-Jul-2018	jdolecek	attach the rnd source only once even with dual-port adapters
1.18	13-Apr-2015	riastradh	branches: 1.18.16; 1.18.18; Convert sys/dev to use <sys/rndsource.h>.
1.17	02-Feb-2012	tls	branches: 1.17.6; 1.17.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.16	19-Nov-2011	tls	branches: 1.16.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.15	08-Sep-2008	christos	branches: 1.15.28; cf_attach_decl_new
1.14	20-Jun-2008	cube	branches: 1.14.2; Use a mutex instead of splvm() to protect the list of jubo-ready mbufs, as done with nfe(4) a while ago. Issue reported by Gary Duzan, who kindly fixed the patch I had sent him.
1.13	28-Apr-2008	martin	branches: 1.13.2; 1.13.4; Remove clause 3 and 4 from TNF licenses
1.12	04-Mar-2007	christos	branches: 1.12.36; 1.12.38; 1.12.40; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.11	04-Sep-2006	riz	branches: 1.11.8; Remove unused members of sk_softc.
1.10	28-Mar-2006	riz	Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months.
1.9	14-Feb-2006	riz	branches: 1.9.2; 1.9.4; 1.9.6; Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.8	11-Feb-2006	rpaulo	PR 31304: Contribute to the random pool. From Rhialto.
1.7	11-Dec-2005	christos	branches: 1.7.2; 1.7.4; 1.7.6; merge ktrace-lwp.
1.6	30-May-2005	christos	branches: 1.6.2; - const poisoning - avoid variable shadowing.
1.5	27-Feb-2005	perry	branches: 1.5.2; nuke trailing whitespace
1.4	26-Sep-2004	skd	branches: 1.4.4; 1.4.6; Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.3	02-Nov-2003	wiz	branches: 1.3.2; 1.3.4; 1.3.6; boundary, not boundry. Inspired by Tom Cosgrove.
1.2	15-Oct-2003	briggs	Make sure descriptors get bus_dmamap_sync()ed. Restructure a little to make this a little easier. Also try to avoid allocating things at reset time.
1.1	26-Aug-2003	jdolecek	Add driver for SysKonnect SK-9821 and 3COM 3C940 gigabit ethernet boards From FreeBSD via OpenBSD, port to NetBSD done by Stephen Degler Changes relative to submitted version: * yukonreg.h and xmaciireg.h merged into if_skreg.h * bhack[] constified + other small editing changes * use 'Ethernet address' rather than 'address' in attach message XXX completely untested by me, needs further cleanup Driver provided in PR kern/22511 by Stephen Degler
1.3.6.4	07-Sep-2006	rpaulo	Pull up following revision(s) (requested by riz in ticket #10680): sys/dev/pci/if_skvar.h: revision 1.10 share/man/man4/sk.4: revision 1.9 sys/dev/pci/if_sk.c: revision 1.22 Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months. Jumbo frames now work under NetBSD, so note it.
1.3.6.3	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #10403): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.3.6.2	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #10402): sys/dev/pci/if_sk.c: revision 1.19 sys/dev/pci/if_skvar.h: revision 1.8 PR 31304: Contribute to the random pool. From Rhialto.
1.3.6.1	07-Jan-2005	jdc	Pull up revision 1.4 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.3.4.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.3.4.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.3.4.5	19-Oct-2004	skrll	Sync with HEAD
1.3.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.3.4.2	03-Aug-2004	skrll	Sync with HEAD
1.3.4.1	02-Nov-2003	skrll	file if_skvar.h was added on branch ktrace-lwp on 2004-08-03 10:49:09 +0000
1.3.2.1	07-Jan-2005	jdc	Pull up revision 1.4 (requested by abs in ticket #973). Fix Yukon ram sizing. Fix tx queue (slist can be corrupted when tx interrupts hit within tx_encap. Lower interrupt moderation timer to (improves performance). Improve chip identification. (from linux sk98lin driver). Keep tx queue running by kicking the tx bmu repeatedly.
1.4.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.4.1	29-Apr-2005	kent	sync with -current
1.5.2.3	03-Nov-2006	tron	Pull up following revision(s) (requested by riz in ticket #1480): sys/dev/pci/if_skvar.h: revision 1.10 sys/dev/pci/if_sk.c: revision 1.22 Add jumbo frames support, from OpenBSD (mcbride). As seen on tech-net for the last N months.
1.5.2.2	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #1230): sys/dev/pci/if_skreg.h: revision 1.6 sys/dev/pci/if_skvar.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.20 via patch Not all chips supported by this driver use the same interrupt moderation timer frequency; Take this into account. Information gathered from the sk98lin driver for Linux, from Marvell. Also add sysctl support for changing the interrupt moderation timer at runtime; each board is controlled independently. Discussed on tech-net beginning in November 2005.
1.5.2.1	29-Mar-2006	tron	Pull up following revision(s) (requested by riz in ticket #1229): sys/dev/pci/if_sk.c: revision 1.19 sys/dev/pci/if_skvar.h: revision 1.8 PR 31304: Contribute to the random pool. From Rhialto.
1.6.2.3	03-Sep-2007	yamt	sync with head.
1.6.2.2	30-Dec-2006	yamt	sync with head.
1.6.2.1	21-Jun-2006	yamt	sync with head.
1.7.6.1	22-Apr-2006	simonb	Sync with head.
1.7.4.1	09-Sep-2006	rpaulo	sync with head
1.7.2.1	18-Feb-2006	yamt	sync with head.
1.9.6.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.9.4.1	19-Apr-2006	elad	sync with head.
1.9.2.2	14-Sep-2006	yamt	sync with head.
1.9.2.1	01-Apr-2006	yamt	sync with head.
1.11.8.1	12-Mar-2007	rmind	Sync with HEAD.
1.12.40.2	04-May-2009	yamt	sync with head.
1.12.40.1	16-May-2008	yamt	sync with head.
1.12.38.1	18-May-2008	yamt	sync with head.
1.12.36.3	28-Sep-2008	mjf	Sync with HEAD.
1.12.36.2	29-Jun-2008	mjf	Sync with HEAD.
1.12.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.13.4.1	27-Jun-2008	simonb	Sync with head.
1.13.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.13.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.14.2.1	19-Oct-2008	haad	Sync with HEAD.
1.15.28.1	17-Apr-2012	yamt	sync with head
1.16.2.1	18-Feb-2012	mrg	merge to -current.
1.17.24.1	06-Jun-2015	skrll	Sync with HEAD
1.17.6.1	03-Dec-2017	jdolecek	update from HEAD
1.18.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.18.18.1	10-Jun-2019	christos	Sync with HEAD
1.18.16.1	28-Jul-2018	pgoyette	Sync with HEAD
1.66	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.65	10-Feb-2024	andvar	s/alloted/allotted/ in comments.
1.64	21-Sep-2022	thorpej	No need to include <sys/malloc.h>
1.63	21-Sep-2022	thorpej	Fix a memory leak in an extremely uncommon error path in ste_start().
1.62	15-Mar-2020	thorpej	Don't bother with IFF_OACTIVE.
1.61	13-Mar-2020	thorpej	- Add the D-Link products, identified by subsystem ID. - On some variations, the internal PHY is ghosted at #0 and #1. Work around this by ignoring PHY #0 accesses unless we don't find one, and then look for one there as a fall-back if we don't detect anything else. - Fix access width when setting the TxDMAUrgentThresh register. - Support MBUFTRACE.
1.60	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.59	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.58	02-Nov-2019	tnn	branches: 1.58.2; cast away unused return value
1.57	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.56	28-May-2019	msaitoh	branches: 1.56.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.55	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.54	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.53	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.52	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.51	26-Jun-2018	msaitoh	branches: 1.51.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.50	15-Dec-2016	ozaki-r	branches: 1.50.8; 1.50.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.49	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.48	07-Jul-2016	msaitoh	branches: 1.48.2; KNF. Remove extra spaces. No functional change.
1.47	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.46	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.45	29-Mar-2014	christos	branches: 1.45.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.44	27-Oct-2012	chs	branches: 1.44.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.43	22-Jul-2012	matt	branches: 1.43.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.42	13-Nov-2010	uebayasi	branches: 1.42.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.41	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.40	19-Jan-2010	pooka	branches: 1.40.2; 1.40.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.39	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.38	27-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Compile test only.
1.37	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.36	28-Apr-2008	martin	branches: 1.36.14; Remove clause 3 and 4 from TNF licenses
1.35	10-Apr-2008	cegger	branches: 1.35.2; 1.35.4; use aprint_*_dev and device_xname
1.34	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.33	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.32	19-Jan-2008	dyoung	branches: 1.32.2; 1.32.6; Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.31	19-Oct-2007	ad	branches: 1.31.2; 1.31.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.30	14-Oct-2007	xtraeme	branches: 1.30.2; IC Plus Corp IP100A support for ste(4), submitted by Pedro Alejandro Lopez-Valencia in PR kern/36605.
1.29	26-Aug-2007	dyoung	branches: 1.29.2; 1.29.4; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.28	09-Jul-2007	ad	branches: 1.28.2; 1.28.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.27	04-Mar-2007	christos	branches: 1.27.2; 1.27.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.26	07-Jan-2007	mlelstv	branches: 1.26.2; Make watchdog poll interrupts before re-initializing the chip. Fixes kern/34834.
1.25	16-Nov-2006	christos	branches: 1.25.2; __unused removal on arguments; approved by core.
1.24	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.23	17-Jun-2006	christos	branches: 1.23.4; 1.23.6; re-factor the pci powestate api. reviewed by gimpy
1.22	11-Dec-2005	christos	branches: 1.22.4; 1.22.8; 1.22.14; 1.22.16; merge ktrace-lwp.
1.21	27-Feb-2005	perry	branches: 1.21.4; nuke trailing whitespace
1.20	30-Oct-2004	thorpej	branches: 1.20.4; 1.20.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.19	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.18	08-Jul-2004	mycroft	Program the "MaxFrameSize" register so that it's correct for VLAN encapsulation, per PR 20195.
1.17	05-Jun-2003	tsutsui	branches: 1.17.2; 1.17.4; Add a missed htole32() for DMA txdescs. Problem reported by Todd Mortimer on port-macppc.
1.16	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.15	21-Oct-2002	fair	Change the "dontcare bits" argument to ifmedia_init() to IFM_IMASK so that it is possible select PHY instances other than the first one (instance zero), if there is more than one PHY attached.
1.14	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.13	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.12	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.11	05-Jun-2002	bouyer	STE_CarrierSenseErrors is 8bit wide, not 16bit. Reenable statistics interrupts, reading it as 16bit caused it to not be cleared,and thus the eventual interrupt to not be cleared (and this caused an infinite loop in interrupt routine). While I'm there count STE_CarrierSenseErrors as oerror (this is incremented each time we try to send a packet and the link is down).
1.10	05-Jun-2002	bouyer	Rework transmit underrun logic: - when a transmit underrun occurs, only reset the transmit logic. This prevents the link from going down at the MII level (the same logic can probably be used for other transmit errors) - set the urgent threshold to half the start threshold. From experiments this helps a bit decreasing the number of underruns Now that underruns don't cause the interface to stop for several seconds, make back out the default transmit threshold to 128.
1.9	01-Jun-2002	bouyer	When the chip gets a transmit underrun, the driver resets the chip, which cause the link to go down for several seconds on my Dlink DFE580-TX (quad-port DL1002). This behavior is unacceptable, especially as the driver usually gets several underrun at once, causing the link to be almost unaviable for several 10s of seconds. Bump the default transmit threshold to 1504 bytes to avoid transmit underrun.
1.8	01-Jun-2002	bouyer	On a Dlink DFE580-TX (DL-1002), attempting to transmit packets while the link is down cause the chip to create an IE_UpdateStats condition which can't be cleared, causing the driver to enter an interrupt loop. Workaround: mask IE_UpdateStats
1.7	13-Nov-2001	lukem	branches: 1.7.8; 1.7.10; add RCSID
1.6	23-Jul-2001	thorpej	branches: 1.6.2; Make sure to initialize the all Rx descriptors properly in *_init() even if mbufs for them are already allocated.
1.5	23-Jul-2001	thorpej	Adjust some reset delays.
1.4	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.3	30-Jun-2001	thorpej	branches: 1.3.2; Update for D-Link ST201 change.
1.2	29-Jun-2001	thorpej	In the `copy small' path, make sure to align the payload.
1.1	19-Jun-2001	thorpej	Device driver for the Sundance Tech. ST-201 10/100 Ethernet. This chip is found on the D-Link DFE-550TX.
1.3.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	03-Aug-2001	lukem	update to -current
1.6.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.6.2.6	11-Nov-2002	nathanw	Catch up to -current
1.6.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.6.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.6.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.6.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.6.2.1	23-Jul-2001	nathanw	file if_ste.c was added on branch nathanw_sa on 2001-08-24 00:10:07 +0000
1.7.10.2	17-Jun-2003	msaitoh	Pull up revisions 1.17 (requested by tsutsui in ticket #1315): Add a missed htole32() for DMA txdescs.
1.7.10.1	10-Jun-2002	tv	Pull up revision 1.8 (requested by 128 in ticket #0): - mask IE_UpdateStats interrupt: the driver fails to clear the interrupt and enter an interrupt loop (it may be a chip bug, I didn't have the time to investigate yet) if it attemps to transmit some packets while the link is down (like unplug/replug the cable, ifconfig down/ifconfig up or after a transmit underrun condition, see below) - bump the default transmit threshold to 1504 to avoid transmit underruns. The driver resets the chip after a transmit underrun condition, which makes the link go down for several seconds, which is unacceptable for a server usage (after bootup you'll get 10s of underruns before the threshold is high enouth and things stabilize, which means the link is almost unusable for minutes) - update documentation to mention the quad-port DFE580TX, and mention the driver in notes.
1.7.8.1	20-Jun-2002	gehenna	catch up with -current.
1.17.4.1	23-Jul-2004	he	branches: 1.17.4.1.2; Pull up revision 1.18 (requested by mycroft in ticket #698): Program the "MaxFrameSize" register so that it is correct for VLAN encapsulation. Fixes PR#20195.
1.17.4.1.2.1	24-Jan-2005	he	Pull up revision 1.20 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.17.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.17.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.17.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.17.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.17.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.17.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.20.4.1	29-Apr-2005	kent	sync with -current
1.21.4.7	24-Mar-2008	yamt	sync with head.
1.21.4.6	17-Mar-2008	yamt	sync with head.
1.21.4.5	21-Jan-2008	yamt	sync with head
1.21.4.4	27-Oct-2007	yamt	sync with head.
1.21.4.3	03-Sep-2007	yamt	sync with head.
1.21.4.2	26-Feb-2007	yamt	sync with head.
1.21.4.1	21-Jun-2006	yamt	sync with head.
1.22.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.22.14.1	19-Jun-2006	chap	Sync with head.
1.22.8.1	26-Jun-2006	yamt	sync with head.
1.22.4.1	09-Sep-2006	rpaulo	sync with head
1.23.6.2	10-Dec-2006	yamt	sync with head.
1.23.6.1	22-Oct-2006	yamt	sync with head
1.23.4.2	12-Jan-2007	ad	Sync with head.
1.23.4.1	18-Nov-2006	ad	Sync with head.
1.25.2.2	22-Oct-2007	pavel	Pull up following revisions (requested by xtraeme in ticket #931): sys/dev/pci/pcidevs: revision 1.905 sys/dev/pci/if_ste.c: revision 1.30 IC Plus Corp IP100A support for ste(4), submitted by Pedro Alejandro Lopez-Valencia in PR kern/36605.
1.25.2.1	01-Apr-2007	bouyer	branches: 1.25.2.1.2; Pull up following revision(s) (requested by mlelstv in ticket #552): sys/dev/pci/if_ste.c: revision 1.26 Make watchdog poll interrupts before re-initializing the chip. Fixes kern/34834.
1.25.2.1.2.1	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.26.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.27.4.1	11-Jul-2007	mjf	Sync with head.
1.27.2.3	23-Oct-2007	ad	Sync with head.
1.27.2.2	09-Oct-2007	ad	Sync with head.
1.27.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.28.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.28.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.28.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.29.4.1	14-Oct-2007	yamt	sync with head.
1.29.2.2	23-Mar-2008	matt	sync with HEAD
1.29.2.1	06-Nov-2007	matt	sync with HEAD
1.30.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.31.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.31.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.32.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.32.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.32.2.1	24-Mar-2008	keiichi	sync with head.
1.35.4.4	11-Aug-2010	yamt	sync with head.
1.35.4.3	11-Mar-2010	yamt	sync with head
1.35.4.2	16-May-2009	yamt	sync with head
1.35.4.1	16-May-2008	yamt	sync with head.
1.35.2.1	18-May-2008	yamt	sync with head.
1.36.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.40.4.2	05-Mar-2011	rmind	sync with head
1.40.4.1	30-May-2010	rmind	sync with head
1.40.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.42.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.42.8.1	30-Oct-2012	yamt	sync with head
1.43.2.3	03-Dec-2017	jdolecek	update from HEAD
1.43.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.43.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.44.2.1	18-May-2014	rmind	sync with head
1.45.6.3	05-Feb-2017	skrll	Sync with HEAD
1.45.6.2	09-Jul-2016	skrll	Sync with HEAD
1.45.6.1	19-Mar-2016	skrll	Sync with HEAD
1.48.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.50.14.3	26-Jan-2019	pgoyette	Sync with HEAD
1.50.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.50.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.50.8.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.51.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.51.2.1	10-Jun-2019	christos	Sync with HEAD
1.56.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.58.2.1	29-Feb-2020	ad	Sync with head.
1.5	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.4	25-Dec-2007	perry	branches: 1.4.6; 1.4.8; 1.4.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.3	24-Jun-2002	bouyer	branches: 1.3.22; 1.3.80; 1.3.86; 1.3.90; 1.3.94; Typo fix
1.2	03-May-2002	thorpej	branches: 1.2.2; Bump the number of Tx DMA segments from 8 to 16 (the zero-copy socket code sometimes sees more than 8).
1.1	19-Jun-2001	thorpej	branches: 1.1.2; 1.1.4; Device driver for the Sundance Tech. ST-201 10/100 Ethernet. This chip is found on the D-Link DFE-550TX.
1.1.4.4	01-Aug-2002	nathanw	Catch up to -current.
1.1.4.3	20-Jun-2002	nathanw	Catch up to -current.
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	19-Jun-2001	nathanw	file if_stereg.h was added on branch nathanw_sa on 2001-08-24 00:10:07 +0000
1.1.2.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.1.2.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.2.1	15-Jul-2002	gehenna	catch up with -current.
1.3.94.1	02-Jan-2008	bouyer	Sync with HEAD
1.3.90.1	26-Dec-2007	ad	Sync with head.
1.3.86.1	18-Feb-2008	mjf	Sync with HEAD.
1.3.80.1	09-Jan-2008	matt	sync with HEAD
1.3.22.1	21-Jan-2008	yamt	sync with head
1.4.10.1	16-May-2008	yamt	sync with head.
1.4.8.1	18-May-2008	yamt	sync with head.
1.4.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.94	04-Oct-2025	thorpej	Add a shared function to query the common properties used for configuring an Ethernet address.
1.93	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.92	29-Jun-2024	riastradh	branches: 1.92.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.91	10-Feb-2024	andvar	s/alloted/allotted/ in comments.
1.90	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.89	22-Jan-2022	martin	Explicitly cast a __MASK(40) to bus_addr_t (for 32bit bus_addr_t archs)
1.88	22-Jan-2022	skrll	Ensure bus_dmatag_subregion is called with an inclusive max_addr everywhere.
1.87	02-Jul-2020	msaitoh	prop_data_data_nocopy -> prop_data_value
1.86	15-Mar-2020	thorpej	Don't bother with IFF_OACTIVE.
1.85	13-Mar-2020	thorpej	Support MBUFTRACE.
1.84	07-Mar-2020	thorpej	If bus_dmatag_subregion(), actually fall back on the base 64-bit DMA tag.
1.83	02-Mar-2020	thorpej	Duh, acctually use all 64-bits when giving the rings to the chip.
1.82	01-Mar-2020	thorpej	Add explicit casts in bus_dmatag_subregion() calls to avoid overflow / truncation warnings on 32-bit platforms.
1.81	01-Mar-2020	thorpej	Harmonize bus_dmatag_subregion() handling with other drivers.
1.80	29-Feb-2020	thorpej	Use the 64-bit PCI DMA tag if it's available, but range-limit to 40-bit because that's what the hardware can do.
1.79	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.78	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.77	14-Jan-2020	msaitoh	No functional change: - Move some definitions from if_stgereg.h to if_stge.c again because those are not chip (registers or descriptors) definitions. - Use proplib to pass information that loading DSP code is required when PHY reset.
1.76	09-Jan-2020	msaitoh	Reduce diff against OpenBSD. No functional change. - USE CSR_{READ,WRITE}_*() macro. - Move some macros from if_stge.c to if_stgereg.h
1.75	26-Dec-2019	msaitoh	branches: 1.75.2; Improve flow control.
1.74	26-Dec-2019	msaitoh	Apply OpenBSD rev. 1.34: > Set Rx early threshold and Tx DMA thesholds. > > From Linux via FreeBSD
1.73	10-Nov-2019	msaitoh	Add missing initialization of sc_dev.
1.72	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.71	07-Oct-2019	msaitoh	Add support IC Plus IP1000* PHYs. PR/kern 42314 first reported by Tomokazu HARADA and patch provided by Andrius V.
1.70	09-Jul-2019	msaitoh	branches: 1.70.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.69	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.68	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.67	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.66	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.65	26-Jun-2018	msaitoh	branches: 1.65.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.64	28-Sep-2017	christos	branches: 1.64.2; only get vtag when we have vtag like the other drivers.
1.63	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.62	15-Dec-2016	ozaki-r	branches: 1.62.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.61	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.60	07-Jul-2016	msaitoh	branches: 1.60.2; KNF. Remove extra spaces. No functional change.
1.59	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.58	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.57	29-Mar-2014	christos	branches: 1.57.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	30-Mar-2013	christos	branches: 1.56.4; remove trailing whitespace
1.55	22-Jul-2012	matt	branches: 1.55.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.54	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.53	12-Mar-2011	phx	branches: 1.53.4; 1.53.8; When shutting down the network interface it is not enough to stop it. Some boards also need to reset the chip, otherwise bad things can happen.
1.52	08-Mar-2011	phx	On the ST1023 (and compatible, like IP1000A), the mac-address can also be passed via device properties, when the serial EEPROM is missing or empty.
1.51	13-Nov-2010	uebayasi	branches: 1.51.2; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.50	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.49	19-Jan-2010	pooka	branches: 1.49.2; 1.49.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.48	14-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested on D-Link DL-4000.
1.47	17-May-2009	tsutsui	Split device_t/softc. Tested on D-Link DL-4000.
1.46	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.45	28-Apr-2008	martin	branches: 1.45.14; Remove clause 3 and 4 from TNF licenses
1.44	10-Apr-2008	cegger	branches: 1.44.2; 1.44.4; use aprint_*_dev and device_xname
1.43	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.42	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.41	07-Feb-2008	dyoung	branches: 1.41.2; 1.41.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.40	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.39	19-Oct-2007	ad	branches: 1.39.2; 1.39.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.38	26-Aug-2007	dyoung	branches: 1.38.2; 1.38.6; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.37	09-Jul-2007	ad	branches: 1.37.2; 1.37.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.36	04-Mar-2007	christos	branches: 1.36.2; 1.36.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.35	16-Nov-2006	christos	branches: 1.35.4; __unused removal on arguments; approved by core.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	17-Jun-2006	christos	branches: 1.33.4; 1.33.6; re-factor the pci powestate api. reviewed by gimpy
1.32	11-Dec-2005	christos	branches: 1.32.4; 1.32.8; 1.32.14; 1.32.16; merge ktrace-lwp.
1.31	28-Jul-2005	christos	Don't use a global variable to indicate that we are stge1023. Use an sc member. Pointed out by thorpej.
1.30	25-Jul-2005	kleink	Add missing brace to previous.
1.29	25-Jul-2005	christos	PR/29925: Mihai CHELARU: Fix for stge 1023.
1.28	25-Jun-2005	bouyer	branches: 1.28.2; Write the station address using 8-bits writes instead of 16-bits. This is what the linux driver does, and makes the DGE-550T work without the STGE_CU_BUG hack. So remove the STGE_CU_BUG hack. Set bit 0x0020 in STGE_DebugCtrl too, the linux driver does it (the comments note this as a workaround, without more details. This doesn't seem to make things worse). Also initialize STGE_RxDMABurstThresh and STGE_RxDMAUrgentThresh, using values from the linux driver. Approved by Jason Thorpe.
1.27	16-May-2005	bouyer	Fix some big-endian issues. stge(4) now works on sparc64 with hardware checksums.
1.26	13-May-2005	wiz	Fix typo reported by Brad at comstyle com.
1.25	02-May-2005	yamt	split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.24	27-Feb-2005	perry	branches: 1.24.2; nuke trailing whitespace
1.23	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition
1.22	30-Oct-2004	thorpej	branches: 1.22.4; 1.22.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.21	02-Oct-2004	wiz	Fix typo reported by Alexander Yurchenko in private mail.
1.20	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.19	01-Mar-2003	mjacob	branches: 1.19.2; 1.19.6; Argh. Commited the wrong file.
1.18	01-Mar-2003	mjacob	Fix RFD_TCI macro.
1.17	10-Feb-2003	christos	Centralize the interrupt bits for the 3 registers where they are being used.
1.16	10-Feb-2003	christos	make this compile again.
1.15	10-Feb-2003	mjacob	Add in h/w VLAN tag support. Put in code that, if enabled, makes my copper Sundance TI cards work- promiscuous mode is needed for this for now. The default is to have this off until I or somebody figures out what is broken- I think it's stuff with the Marvell PHY.
1.14	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.13	21-Oct-2002	fair	Change the "dontcare bits" argument to ifmedia_init() to IFM_IMASK so that it is possible select PHY instances other than the first one (instance zero), if there is more than one PHY attached.
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.9	13-Nov-2001	lukem	add RCSID
1.8	10-Oct-2001	mjacob	The flow control registers are 16 bits wide, not 32.
1.7	06-Oct-2001	thorpej	Create the Tx DMA maps with a size large enough for a jumbo Ethernet frame.
1.6	30-Jul-2001	thorpej	branches: 1.6.2; 1.6.4; 1.6.6; Oops, make Rx interrupt deferral actually work.
1.5	27-Jul-2001	thorpej	Tell the MII layer to advertise PAUSE capability, and configure the chip's PAUSE threshold.
1.4	25-Jul-2001	thorpej	Jumbo frames do not currently work (chip goes catatonic -- silicon bug?), so don't set the capability.
1.3	25-Jul-2001	thorpej	Match the Antares Gigabit Ethernet.
1.2	25-Jul-2001	thorpej	Remove a debugging #define.
1.1	25-Jul-2001	thorpej	Driver for the Sundance Tech./Tamarack TC9021 Gigabit Ethernet controller.
1.6.6.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.6.4.7	29-Dec-2002	thorpej	Sync with HEAD.
1.6.4.6	11-Nov-2002	nathanw	Catch up to -current
1.6.4.5	18-Oct-2002	nathanw	Catch up to -current.
1.6.4.4	14-Nov-2001	nathanw	Catch up to -current.
1.6.4.3	22-Oct-2001	nathanw	Catch up to -current.
1.6.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.6.4.1	30-Jul-2001	nathanw	file if_stge.c was added on branch nathanw_sa on 2001-08-24 00:10:08 +0000
1.6.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.6.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.2	03-Aug-2001	lukem	update to -current
1.6.2.1	30-Jul-2001	lukem	file if_stge.c was added on branch kqueue on 2001-08-03 04:13:16 +0000
1.19.6.3	05-Jul-2005	riz	Pull up revision 1.28 (requested by bouyer in ticket #2062): Write the station address using 8-bits writes instead of 16-bits. This is what the linux driver does, and makes the DGE-550T work without the STGE_CU_BUG hack. So remove the STGE_CU_BUG hack. Set bit 0x0020 in STGE_DebugCtrl too, the linux driver does it (the comments note this as a workaround, without more details. This doesn't seem to make things worse). Also initialize STGE_RxDMABurstThresh and STGE_RxDMAUrgentThresh, using values from the linux driver. Approved by Jason Thorpe.
1.19.6.2	21-Jun-2005	riz	Pull up revision 1.27 (requested by bouyer in ticket #2009): Fix some big-endian issues. stge(4) now works on sparc64 with hardware checksums.
1.19.6.1	24-Jan-2005	he	Pull up revision 1.22 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.19.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.19.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.19.2.4	19-Oct-2004	skrll	Sync with HEAD
1.19.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.19.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.19.2.1	25-Aug-2004	skrll	Sync with HEAD.
1.22.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.22.4.1	29-Apr-2005	kent	sync with -current
1.24.2.2	05-Jul-2005	tron	Pull up revision 1.28 (requested by bouyer in ticket #548): Write the station address using 8-bits writes instead of 16-bits. This is what the linux driver does, and makes the DGE-550T work without the STGE_CU_BUG hack. So remove the STGE_CU_BUG hack. Set bit 0x0020 in STGE_DebugCtrl too, the linux driver does it (the comments note this as a workaround, without more details. This doesn't seem to make things worse). Also initialize STGE_RxDMABurstThresh and STGE_RxDMAUrgentThresh, using values from the linux driver. Approved by Jason Thorpe.
1.24.2.1	22-Jun-2005	tron	Pull up revision 1.27 (requested by bouyer in ticket #477): Fix some big-endian issues. stge(4) now works on sparc64 with hardware checksums.
1.28.2.8	24-Mar-2008	yamt	sync with head.
1.28.2.7	17-Mar-2008	yamt	sync with head.
1.28.2.6	11-Feb-2008	yamt	sync with head.
1.28.2.5	21-Jan-2008	yamt	sync with head
1.28.2.4	27-Oct-2007	yamt	sync with head.
1.28.2.3	03-Sep-2007	yamt	sync with head.
1.28.2.2	30-Dec-2006	yamt	sync with head.
1.28.2.1	21-Jun-2006	yamt	sync with head.
1.32.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.32.14.1	19-Jun-2006	chap	Sync with head.
1.32.8.1	26-Jun-2006	yamt	sync with head.
1.32.4.1	09-Sep-2006	rpaulo	sync with head
1.33.6.2	10-Dec-2006	yamt	sync with head.
1.33.6.1	22-Oct-2006	yamt	sync with head
1.33.4.1	18-Nov-2006	ad	Sync with head.
1.35.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.36.4.1	11-Jul-2007	mjf	Sync with head.
1.36.2.3	23-Oct-2007	ad	Sync with head.
1.36.2.2	09-Oct-2007	ad	Sync with head.
1.36.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.37.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.37.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.37.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.38.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.38.2.2	23-Mar-2008	matt	sync with HEAD
1.38.2.1	06-Nov-2007	matt	sync with HEAD
1.39.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.39.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.41.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.41.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.41.2.1	24-Mar-2008	keiichi	sync with head.
1.44.4.6	11-Aug-2010	yamt	sync with head.
1.44.4.5	11-Mar-2010	yamt	sync with head
1.44.4.4	16-Sep-2009	yamt	sync with head
1.44.4.3	20-Jun-2009	yamt	sync with head
1.44.4.2	16-May-2009	yamt	sync with head
1.44.4.1	16-May-2008	yamt	sync with head.
1.44.2.1	18-May-2008	yamt	sync with head.
1.45.14.2	23-Jul-2009	jym	Sync with HEAD.
1.45.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.49.4.3	21-Apr-2011	rmind	sync with head
1.49.4.2	05-Mar-2011	rmind	sync with head
1.49.4.1	30-May-2010	rmind	sync with head
1.49.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.51.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.53.8.1	18-Feb-2012	mrg	merge to -current.
1.53.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.4.2	30-Oct-2012	yamt	sync with head
1.53.4.1	17-Apr-2012	yamt	sync with head
1.55.2.3	03-Dec-2017	jdolecek	update from HEAD
1.55.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.2.1	23-Jun-2013	tls	resync from head
1.56.4.1	18-May-2014	rmind	sync with head
1.57.6.3	05-Feb-2017	skrll	Sync with HEAD
1.57.6.2	09-Jul-2016	skrll	Sync with HEAD
1.57.6.1	19-Mar-2016	skrll	Sync with HEAD
1.60.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.62.8.3	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1439): sys/dev/sbus/sio16.c: revision 1.25 sys/dev/gpib/mt.c: revision 1.33 sys/arch/mvme68k/dev/wdsc.c: revision 1.33 sys/dev/gpib/ct.c: revision 1.31 sys/dev/isa/mcd.c: revision 1.119 sys/dev/pci/if_stge.c: revision 1.73 sys/dev/gpib/ppi.c: revision 1.25 sys/dev/pcmcia/if_tr_pcmcia.c: revision 1.29 sys/arch/amiga/dev/zssc.c: revision 1.46 sys/arch/mac68k/obio/iwm_fd.c: revision 1.57 sys/arch/amiga/dev/if_qn.c: revision 1.48 sys/dev/isa/tcic2_isa.c: revision 1.28 sys/dev/isa/uha_isa.c: revision 1.42 sys/dev/pci/neo.c: revision 1.55 sys/arch/next68k/dev/nextdisplay.c: revision 1.22 sys/dev/isa/if_iy.c: revision 1.111 Add missing initialization of sc_dev. Initialize sc_dev correctly to avoid null pointer dereference when bus_space_map() failed. Add missing initialization of sc_dev.
1.62.8.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.62.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.64.2.3	26-Jan-2019	pgoyette	Sync with HEAD
1.64.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.64.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.65.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.65.2.1	10-Jun-2019	christos	Sync with HEAD
1.70.2.3	25-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #476): sys/dev/pci/if_stgereg.h: revision 1.6 sys/arch/amd64/conf/ALL: revision 1.124 sys/dev/mii/files.mii: revision 1.51 share/man/man4/vge.4: revision 1.8 share/man/man4/vge.4: revision 1.9 distrib/sets/lists/man/mi: revision 1.1656 sys/arch/i386/conf/ALL: revision 1.471 share/man/man4/mii.4: revision 1.28 sys/dev/pci/if_stge.c: revision 1.71 sys/dev/mii/ipgphy.c: revision 1.1 sys/dev/mii/ipgphy.c: revision 1.2 share/man/man4/Makefile: revision 1.684 sys/dev/mii/ipgphy.c: revision 1.3 sys/dev/mii/ipgphyreg.h: revision 1.1 sys/dev/mii/ipgphy.c: revision 1.4 sys/dev/mii/ipgphyreg.h: revision 1.2 sys/dev/mii/ipgphy.c: revision 1.5 sys/dev/mii/ipgphyreg.h: revision 1.3 sys/dev/DEVNAMES: revision 1.322 sys/arch/i386/conf/GENERIC: revision 1.1211 sys/arch/amd64/conf/GENERIC: revision 1.537 share/man/man4/ipgphy.4: revision 1.1 share/man/man4/ipgphy.4: revision 1.2 Add support IC Plus IP1000* PHYs. PR/kern 42314 first reported by Tomokazu HARADA and patch provided by Andrius V. New sentence, new line. Fix date. Add RCS Id. KNF. No functional change. STGE_PhyCtrl is not PHY register but MAC register, so use ukphy_status() for IP1000A device. - Use auto-negotiation when forcing 1000BASE-T. - Add XXX comment for strange pause setting code. I suspect this is wrong. - On my environments, 1000BASE-T half duplex doesn't work, so we might remove IFM_1000T_HDX from this device in future. - Define IPGPHY_SCR's address. - Whitespace fix. Add missing NetBSD RCS Id. Reported by Andrius V.
1.70.2.2	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #425): sys/dev/sbus/sio16.c: revision 1.25 sys/dev/gpib/mt.c: revision 1.33 sys/arch/mvme68k/dev/wdsc.c: revision 1.33 sys/dev/gpib/ct.c: revision 1.31 sys/dev/isa/mcd.c: revision 1.119 sys/dev/pci/if_stge.c: revision 1.73 sys/dev/gpib/ppi.c: revision 1.25 sys/dev/pcmcia/if_tr_pcmcia.c: revision 1.29 sys/arch/amiga/dev/zssc.c: revision 1.46 sys/arch/mac68k/obio/iwm_fd.c: revision 1.57 sys/arch/amiga/dev/if_qn.c: revision 1.48 sys/dev/isa/tcic2_isa.c: revision 1.28 sys/dev/isa/uha_isa.c: revision 1.42 sys/dev/pci/neo.c: revision 1.55 sys/arch/next68k/dev/nextdisplay.c: revision 1.22 sys/dev/isa/if_iy.c: revision 1.111 Add missing initialization of sc_dev. Initialize sc_dev correctly to avoid null pointer dereference when bus_space_map() failed. Add missing initialization of sc_dev.
1.70.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.75.2.2	29-Feb-2020	ad	Sync with head.
1.75.2.1	17-Jan-2020	ad	Sync with head.
1.92.2.1	02-Aug-2025	perseant	Sync with HEAD
1.9	14-Jan-2020	msaitoh	No functional change: - Move some definitions from if_stgereg.h to if_stge.c again because those are not chip (registers or descriptors) definitions. - Use proplib to pass information that loading DSP code is required when PHY reset.
1.8	09-Jan-2020	msaitoh	Reduce diff against OpenBSD. No functional change. - USE CSR_{READ,WRITE}_*() macro. - Move some macros from if_stge.c to if_stgereg.h
1.7	26-Dec-2019	msaitoh	branches: 1.7.2; Fix address of STGE_MaxFrameSize. Same as other OSes.
1.6	07-Oct-2019	msaitoh	Add support IC Plus IP1000* PHYs. PR/kern 42314 first reported by Tomokazu HARADA and patch provided by Andrius V.
1.5	28-Apr-2008	martin	branches: 1.5.80; 1.5.88; 1.5.94; Remove clause 3 and 4 from TNF licenses
1.4	25-Dec-2007	perry	branches: 1.4.6; 1.4.8; 1.4.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.3	10-Feb-2003	christos	branches: 1.3.18; 1.3.76; 1.3.82; 1.3.86; 1.3.90; Centralize the interrupt bits for the 3 registers where they are being used.
1.2	08-Nov-2002	enami	s/than than/than/.
1.1	25-Jul-2001	thorpej	branches: 1.1.2; 1.1.4; Driver for the Sundance Tech./Tamarack TC9021 Gigabit Ethernet controller.
1.1.4.3	11-Nov-2002	nathanw	Catch up to -current
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	25-Jul-2001	nathanw	file if_stgereg.h was added on branch nathanw_sa on 2001-08-24 00:10:08 +0000
1.1.2.2	03-Aug-2001	lukem	update to -current
1.1.2.1	25-Jul-2001	lukem	file if_stgereg.h was added on branch kqueue on 2001-08-03 04:13:16 +0000
1.3.90.1	02-Jan-2008	bouyer	Sync with HEAD
1.3.86.1	26-Dec-2007	ad	Sync with head.
1.3.82.1	18-Feb-2008	mjf	Sync with HEAD.
1.3.76.1	09-Jan-2008	matt	sync with HEAD
1.3.18.1	21-Jan-2008	yamt	sync with head
1.4.10.1	16-May-2008	yamt	sync with head.
1.4.8.1	18-May-2008	yamt	sync with head.
1.4.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.94.2	21-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #627): sys/dev/pci/if_stgereg.h: revision 1.7 Fix address of STGE_MaxFrameSize. Same as other OSes.
1.5.94.1	25-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #476): sys/dev/pci/if_stgereg.h: revision 1.6 sys/arch/amd64/conf/ALL: revision 1.124 sys/dev/mii/files.mii: revision 1.51 share/man/man4/vge.4: revision 1.8 share/man/man4/vge.4: revision 1.9 distrib/sets/lists/man/mi: revision 1.1656 sys/arch/i386/conf/ALL: revision 1.471 share/man/man4/mii.4: revision 1.28 sys/dev/pci/if_stge.c: revision 1.71 sys/dev/mii/ipgphy.c: revision 1.1 sys/dev/mii/ipgphy.c: revision 1.2 share/man/man4/Makefile: revision 1.684 sys/dev/mii/ipgphy.c: revision 1.3 sys/dev/mii/ipgphyreg.h: revision 1.1 sys/dev/mii/ipgphy.c: revision 1.4 sys/dev/mii/ipgphyreg.h: revision 1.2 sys/dev/mii/ipgphy.c: revision 1.5 sys/dev/mii/ipgphyreg.h: revision 1.3 sys/dev/DEVNAMES: revision 1.322 sys/arch/i386/conf/GENERIC: revision 1.1211 sys/arch/amd64/conf/GENERIC: revision 1.537 share/man/man4/ipgphy.4: revision 1.1 share/man/man4/ipgphy.4: revision 1.2 Add support IC Plus IP1000* PHYs. PR/kern 42314 first reported by Tomokazu HARADA and patch provided by Andrius V. New sentence, new line. Fix date. Add RCS Id. KNF. No functional change. STGE_PhyCtrl is not PHY register but MAC register, so use ukphy_status() for IP1000A device. - Use auto-negotiation when forcing 1000BASE-T. - Add XXX comment for strange pause setting code. I suspect this is wrong. - On my environments, 1000BASE-T half duplex doesn't work, so we might remove IFM_1000T_HDX from this device in future. - Define IPGPHY_SCR's address. - Whitespace fix. Add missing NetBSD RCS Id. Reported by Andrius V.
1.5.88.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.5.88.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.5.80.1	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1496): sys/dev/pci/if_stgereg.h: revision 1.7 Fix address of STGE_MaxFrameSize. Same as other OSes.
1.7.2.1	17-Jan-2020	ad	Sync with head.
1.125	05-Nov-2024	andvar	s/UCP/UDP/ in comments.
1.124	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.123	23-May-2022	rin	branches: 1.123.10; Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.122	10-Nov-2021	msaitoh	s/desciptor/descriptor/ in comment.
1.121	05-Jun-2021	thorpej	Back out the following changes: http://mail-index.netbsd.org/source-changes/2020/02/29/msg114581.html http://mail-index.netbsd.org/source-changes/2020/02/29/msg114583.html They have reportedly caused some issues with link detection, and will be revisited later.
1.120	05-Mar-2020	msaitoh	branches: 1.120.8; 1.120.12; Add support for Farallon PN9000SX from FreeBSD.
1.119	05-Mar-2020	msaitoh	No functional change: - Remove extra semicolon - Remove extra whitespace
1.118	05-Mar-2020	msaitoh	No functional change: - Fix typo in comment. - Whitespace
1.117	05-Mar-2020	msaitoh	s/Intialize/Initialize/ in comment.
1.116	03-Mar-2020	msaitoh	Don't clear CFI and priority bit to pass them to the upper layer.
1.115	29-Feb-2020	thorpej	- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit DMA addresses. Modeled after similar usage in the bge(4) driver (the chips supported by bge(4) are descendants of those supported by ti(4)). - Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and using) bounce buffers needlessly.
1.114	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.113	10-Nov-2019	chs	branches: 1.113.2; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.112	09-Jul-2019	msaitoh	branches: 1.112.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.111	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.110	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.109	23-May-2019	msaitoh	-No functional change: - Simplify struct ethercom's pointer near ETHER_FIRST_MULTI(). - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.108	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.107	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.106	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.105	18-Jul-2018	sevan	Be consistent among ethernet drivers on the convention for printing ethernet addresses. NFC
1.104	26-Jun-2018	msaitoh	branches: 1.104.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.103	08-Feb-2018	dholland	branches: 1.103.2; Typos.
1.102	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.101	15-Dec-2016	ozaki-r	branches: 1.101.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.100	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.99	14-Jul-2016	msaitoh	branches: 1.99.2; - Use aprint_normal_dev() instead of aprint_error_dev() when printing Ethernet address. - Use aprint() more. - KNF.
1.98	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.97	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.96	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.95	25-Jul-2015	maxv	Memory leak. Same as r1.93. I don't know why Brainy didn't detect it earlier; or perhaps I forgot to report it. Found by Brainy.
1.94	27-Apr-2015	christos	fix mbuf leak on failure (Brainy)
1.93	29-Mar-2014	christos	branches: 1.93.4; 1.93.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.92	27-Feb-2014	joerg	Remove impossible checks.
1.91	27-Oct-2012	chs	branches: 1.91.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.90	13-Nov-2010	uebayasi	branches: 1.90.8; 1.90.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.89	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.88	19-Jan-2010	pooka	branches: 1.88.2; 1.88.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.87	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.86	27-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Compile test only.
1.85	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.84	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.83	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.82	07-Nov-2008	dyoung	branches: 1.82.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.81	10-Apr-2008	cegger	branches: 1.81.4; 1.81.10; 1.81.12; use aprint_*_dev and device_xname
1.80	07-Feb-2008	dyoung	branches: 1.80.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.79	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.78	19-Oct-2007	ad	branches: 1.78.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.77	07-Sep-2007	tnn	branches: 1.77.4; First part of ti(4) overhaul: - Sanitise macros - Apply KNF Should have no functional changes.
1.76	01-Sep-2007	dyoung	Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.75	29-Aug-2007	dyoung	Constify: LLADDR() -> CLLADDR().
1.74	04-Mar-2007	christos	branches: 1.74.2; 1.74.10; 1.74.14; 1.74.16; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.73	16-Nov-2006	christos	branches: 1.73.4; __unused removal on arguments; approved by core.
1.72	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.71	07-Sep-2006	dogcow	branches: 1.71.2; 1.71.4; remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
1.70	28-Mar-2006	thorpej	Use device_unit().
1.69	11-Dec-2005	christos	branches: 1.69.4; 1.69.6; 1.69.8; 1.69.10; 1.69.12; merge ktrace-lwp.
1.68	30-May-2005	christos	branches: 1.68.2; - const poisoning - avoid variable shadowing.
1.67	02-May-2005	yamt	split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.66	27-Feb-2005	perry	nuke trailing whitespace
1.65	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition also restructure and simplify input VLAN ID code handling somewhat, to not use local have_vlan/vlan_tag variables, and use the appropriate cur_rx condition on the one place where it's needed
1.64	04-Feb-2005	perry	de-__P
1.63	29-Sep-2004	yamt	branches: 1.63.4; 1.63.6; - mbuf external storage is of (caddr_t), not (caddr_t *). - remove a needless cast.
1.62	21-Sep-2004	yamt	introduce M_EXT_RW to allow mbuf external storage R/W. discussed on tech-net@.
1.61	27-Jul-2004	he	Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable). Bump the default values for these to the values used by FreeBSD, and also adjust ti_init_rx_ring_jumbo() to use the same constant that FreeBSD uses. Yes, this consumes more kernel memory. The effect of this is that you can use jumbo frames in a back-to-back setup with TCP windows up to about 250KB and get ~930Mbit/s throughput, while we were earlier limited to around 3-400Mbit/s, and trying to push above that mark by widening the TCP window caused ti0: jumbo buffer allocation failed messages to be logged and a corresponding stall in the traffic.
1.60	18-Mar-2004	bouyer	branches: 1.60.2; Use the bus_space__stream_ methods to access the chip's registers. As we turn the chip to big-endian mode on big-endian systems, we should never byte-swap the data read/written from/to registers. Tested on sparc64. Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)
1.59	18-Mar-2004	bouyer	cur_rx->ti_vlan_tag holds the 802.1q VLAN tag and the 802.1p priority. Trimm the priority, as the upper layers won't do it and will drop the packet if priority is not 0. While there, print the revision in the "unsupported chip revision" printf.
1.58	24-Feb-2004	wiz	becuase -> because. From Peter Postma.
1.57	09-Feb-2003	mjacob	branches: 1.57.2; Add missing binding of VLAN tag to received packet.
1.56	31-Jan-2003	thorpej	Change ext_size to a size_t, and update the signature of ext_free.
1.55	21-Jan-2003	itojun	vlan tagging mistake during m_tag transition
1.54	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.53	17-Jan-2003	itojun	switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag implementation. it will simplify porting across *bsd (such as kame/altq), and make us more synchronized. from Joel Wilsson
1.52	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.51	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.50	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.49	16-Jul-2002	bouyer	Add missing BUS_DMA_NOWAIT to bus_dmamap_load_mbuf() calls. Should fix port-alpha/17615.
1.48	01-Jun-2002	lukem	SIMPLEQ rototill: - implement SIMPLEQ_REMOVE(head, elm, type, field). whilst it's O(n), this mirrors the functionality of SLIST_REMOVE() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE() - remove the unnecessary elm arg from SIMPLEQ_REMOVE_HEAD(). this mirrors the functionality of SLIST_REMOVE_HEAD() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE_HEAD() - remove notes about SIMPLEQ not supporting arbitrary element removal - use SIMPLEQ_FOREACH() instead of home-grown for loops - use SIMPLEQ_EMPTY() appropriately - use SIMPLEQ_*() instead of accessing sqh_first,sqh_last,sqe_next directly - reorder manual page; be consistent about how the types are listed - other minor cleanups
1.47	02-May-2002	thorpej	branches: 1.47.2; 1.47.4; Change the semantics of mbuf external storage "ext_free" routines so that they're more useful for arbitrary types of external storage: * Add an "mbuf " argument to (ext_free)(). If non-NULL, (ext_free)() is expected to free the mbuf itself. This allows (ext_free)() to use the mbuf for bookkeeping (e.g. deferring the work to a helper thread). If the "mbuf " argument is NULL, we are assumed to be in a context which is safe for performing the destructor operation now. Adjust MEXTREMOVE() and MFREE() routines for above change. * Update "ade" and "ti" drivers for new semantics.
1.46	28-Apr-2002	thorpej	Use MEXTADD() rather than open-coding it.
1.45	20-Mar-2002	eeh	Use bus_space_vaddr() correctly. (Why use bus_space_vaddr() at all?)
1.44	26-Nov-2001	itojun	make it compile without INET
1.43	13-Nov-2001	lukem	add RCSID
1.42	20-Aug-2001	wiz	precede, not preceed.
1.41	14-Aug-2001	mrg	add (u_long) casts for sparc64.
1.40	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.39	07-Jul-2001	thorpej	branches: 1.39.2; bzero -> memset
1.38	07-Jul-2001	thorpej	bcopy -> memcpy, strcpy
1.37	05-Jul-2001	thorpej	For product names that include the media in the name, don't bother saying "Gigabit".
1.36	30-Jun-2001	bjh21	IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.
1.35	30-Jun-2001	thorpej	Sync the Tx descriptors pre- and post- on the Tigon 2.
1.34	30-Jun-2001	thorpej	When we're done setting up the General Information Block, sync it. Sync the NIC-maintained stats structure as appropriate, as well.
1.33	30-Jun-2001	thorpej	Use macros (like those in every other network driver I have written) to compute the offset of a structure in the DMA'd control data, and the DMA address of that structure.
1.32	30-Jun-2001	thorpej	Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions, for the same reason we split the tx encap routine.
1.31	30-Jun-2001	thorpej	Split the transmit encap routine into two versions: one for the Tigon 1 and another for the Tigon 2. This eliminates a conditional in the loop around the DMA segments, and will make further cleanup easier.
1.30	30-Jun-2001	thorpej	Move the `ti_tx_ring_nic' pointer out of the ring-data structure and into the softc. Garbage-collect the `ti_cmd_ring' pointer, since bus_space is properly used to access the command ring.
1.29	30-Jun-2001	thorpej	When a transmit is complete, sync (and unload!!) the DMA map used for the packet. When setting up a packet for transmit, sync the DMA map being used.
1.28	30-Jun-2001	thorpej	Move the Tigon/Tigon2 firmware, version 12.4.13+wpaul, to a more appropriate location.
1.27	29-Jun-2001	thorpej	Correct spelling of media in the product names.
1.26	27-Jun-2001	bouyer	When setting/changing an address, no need to call ti_init() if the interface is already up. Especially this allow netatalk to start properly on a ti interface.
1.25	12-Jun-2001	wiz	receive, not recieve
1.24	07-Jun-2001	bouyer	Set TI_OPMODE_1_DMA_ACTIVE only if hardware assisted checksum isn't used (FreeBSD does it this way). My Tigon II works with hardware assisted checksum now.
1.23	03-Jun-2001	thorpej	In ti_init2(), use a simpler, more obvious, and correct MTU calculation to give the firmware.
1.22	03-Jun-2001	thorpej	Replace a bunch of home-grown Ethernet related constants with ones defined in <net/if_ether.h>
1.21	02-Jun-2001	thorpej	Implement support for IP/TCP/UDP checksum offloading provided by network interfaces. This works by pre-computing the pseudo-header checksum and caching it, delaying the actual checksum to ip_output() if the hardware cannot perform the sum for us. In-bound checksums can either be fully-checked by hardware, or summed up for final verification by software. This method was modeled after how this is done in FreeBSD, although the code is significantly different in most places. We don't delay checksums for IPv6/TCP, but we do take advantage of the cached pseudo-header checksum. Note: hardware-assisted checksumming defaults to "off". It is enabled with ifconfig(8). See the manual page for details. Implement hardware-assisted checksumming on the DP83820 Gigabit Ethernet, 3c90xB/3c90xC 10/100 Ethernet, and Alteon Tigon/Tigon2 Gigabit Ethernet.
1.20	29-Jan-2001	enami	branches: 1.20.2; Handle allmulti case correctly as a NetBSD network driver; if we are requested range of multicast address or too many multicast address, program multicast filter to receive all multicast address. And set/clear IFF_ALLMULTI flag properly.
1.19	18-Jan-2001	jdolecek	constify
1.18	14-Jan-2001	thorpej	Use splnet() to block network interrupts, not splimp().
1.17	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.16	14-Dec-2000	thorpej	ALTQ'ify.
1.15	17-Nov-2000	bouyer	Pull up FreeBSD changes: 1.39: Have if_ti stop "hiding" the softc pointer in the buffer region. Rather, use the available void * passed to the free routine and pass the softc pointer through there. 1.33: Add support for the Netgear GA620T copper gigabit card. 1.32: Tweak probe message so that 1000baseSX and 1000baseT cards are explicitly identified. 1.31: Update the Tigon driver to support 1000baseTX gigE over copper AceNIC cards. This basically involves switching to the 12.4.13 firmware, plus a couple of minor tweaks to the driver. NetBSD changes: get rid of ti_inuse, the mbuf ref counting code should call ti_free() when needed. Use hardware 802.1q support.
1.14	15-Nov-2000	thorpej	Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
1.13	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.12	12-Nov-2000	bouyer	- use jumbo rx if mtu is too big to fit in one mbuf, not when larger than ETHERMTU. - add support for ETHERCAP_VLAN_MTU.
1.11	01-Oct-2000	thorpej	Move the check for "promisc + unicast + not for us" into ether_input(), and change Ethernet drivers to always pass all received frames to ether_input() (with a few exceptions, which are documented in the code).
1.10	24-Sep-2000	martin	Add calls to ether_addmulti and ether_delmulti. This code looks a bit different than other drivers due to the way ti_setmulti manages the filters. Patch from Keijiro Ehara.
1.9	24-Sep-2000	jdolecek	use new generic bootverbose instead of local definition
1.8	30-Mar-2000	augustss	branches: 1.8.4; Remove register declarations.
1.7	17-Mar-2000	bouyer	Fix botched merge in previous.
1.6	17-Mar-2000	bouyer	- remove leftovers of FreeBSD code in #if 0/#endif - fix a bug in ti_encap() where a NULL pointer could be used if we run out of tx descriptors (return ENOMEM instead, and let the caller assert OACTIVE) - sync boot messages with NetBSD's pci standart (print ti_name and irq line) - add a shutdown hook to stop the board, so that the board will not DMA at random places while we're rebooting - convert all but TIGON 1 transmit descriptors to use bus_space (this last one isn't trivial to get rigth and I don't have the hardware to test) - If we can't map the mem space linar, retry to map it nolinear. Punt if we couldn't map the mem space linear and we found a Tigon 1 chip. With these changes a Tigon 2 board (netgear BA620 at last) can be used in my DS20.
1.5	06-Mar-2000	thorpej	Set ifp->if_baudrate as appropriate.
1.4	25-Jan-2000	thorpej	"i/o" -> "memory"
1.3	25-Jan-2000	thorpej	IFM_1000_FX -> IFM_1000_SX
1.2	06-Oct-1999	drochner	branches: 1.2.2; 1. include headers for NS support 2. pull in a fix from FreeBSD: revision 1.18 date: 1999/09/17 18:04:14; author: wpaul; state: Exp; lines: +3 -3 Remember to account for ETHER_ALIGN when setting the maxmimum packet length for mini receive ring. The max length was MHLEN, however the mbufs are actually shortened to MHLEN - ETHER_ALIGN to force payload alignment.
1.1	01-Sep-1999	drochner	branches: 1.1.2; 1.1.4; quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver. todo: -IPv6 -clean up jumbo buffer allocation - NetBSD provides an opaque argument to the free function, thus doesn't need the hack done here -deal correctly with the mapping of the shared memory
1.1.4.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.2.6	06-Jul-2000	he	Pull up revision 1.3 (requested by he): Make the ti(4) driver compile again after the wi(4) related pullups of if_media.h changed IFM_1000_FX to IFM_1000_SX.
1.1.2.5	01-Jun-2000	he	Apply patch (requested by bouyer): Fix compile error.
1.1.2.4	30-Apr-2000	he	Pull up revisions 1.6-1.7 (via patch, requested by bouyer): Fix bug which could result in a panic. Add a shutdown hook to stop the board to avoid memory corruption on reboot. Use bus_space for TIGON 2 chips, so that they will work even if we can't map the memory space linearly (e.g. on alpha).
1.1.2.3	06-Oct-1999	he	Pull up revision 1.2 (requested by drochner): Make it compile in GENERIC by including NS headers. Pull in fix from FreeBSD.
1.1.2.2	05-Oct-1999	he	Pull up revision 1.1+patch (new, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.1.2.1	01-Sep-1999	he	file if_ti.c was added on branch netbsd-1-4 on 1999-10-05 22:08:51 +0000
1.2.2.5	11-Feb-2001	bouyer	Sync with HEAD.
1.2.2.4	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.4.4	13-Nov-2001	he	Pull up revision 1.26 (requested by bouyer): When setting/changing an address, do not call ti_init() if the interface is already up. This allows netatalk to start when using this interface.
1.8.4.3	13-Mar-2001	he	Pull up revision 1.20 (requested by tsutsui): Handle allmulti case correctly as a NetBSD network driver; if we are requested range of multicast address or too many multicast address, program multicast filter to receive all multicast address. And set/clear IFF_ALLMULTI flag properly.
1.8.4.2	31-Dec-2000	jhawk	Pull up revision 1.12, 1.15 via patch (requested by bouyer): For the ti driver, add hardware 802.1Q support, update firmware to 12.4.13+wpaul, and support the Netgear GA620T copper gigabit card, and the Alteon ACEnic 100baseSX and 1000baseT gigabit cards. Various small changes from FreeBSD.
1.8.4.1	26-Sep-2000	martin	Pullup revision 1.10 (approved by thorpej). Fix multicast filters. Patch from Keijiro Ehara.
1.20.2.9	17-Jan-2003	thorpej	Sync with HEAD.
1.20.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.20.2.7	01-Aug-2002	nathanw	Catch up to -current.
1.20.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.20.2.5	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.20.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.20.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.20.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.20.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.39.2.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.39.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.39.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.39.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.39.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.39.2.1	03-Aug-2001	lukem	update to -current
1.47.4.3	26-May-2004	tron	Pull up revision 1.60 (requested by bouyer in ticket #1698): Use the bus_space__stream_ methods to access the chip's registers. As we turn the chip to big-endian mode on big-endian systems, we should never byte-swap the data read/written from/to registers. Tested on sparc64. Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)
1.47.4.2	26-May-2004	tron	Pull up revision 1.59 (requested by bouyer in ticket #1697): cur_rx->ti_vlan_tag holds the 802.1q VLAN tag and the 802.1p priority. Trimm the priority, as the upper layers won't do it and will drop the packet if priority is not 0. While there, print the revision in the "unsupported chip revision" printf.
1.47.4.1	18-Jul-2002	lukem	Pull up revision 1.49 (requested by thorpej in ticket #513): Add missing BUS_DMA_NOWAIT to bus_dmamap_load_mbuf() calls. Should fix port-alpha/17615.
1.47.2.2	20-Jul-2002	gehenna	catch up with -current.
1.47.2.1	20-Jun-2002	gehenna	catch up with -current.
1.57.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.57.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.57.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.57.2.5	19-Oct-2004	skrll	Sync with HEAD
1.57.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.57.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.57.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.57.2.1	03-Aug-2004	skrll	Sync with HEAD
1.60.2.1	28-Jul-2004	tron	Pull up revision 1.61 (requested by he in ticket #719): Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable). Bump the default values for these to the values used by FreeBSD, and also adjust ti_init_rx_ring_jumbo() to use the same constant that FreeBSD uses. Yes, this consumes more kernel memory. The effect of this is that you can use jumbo frames in a back-to-back setup with TCP windows up to about 250KB and get ~930Mbit/s throughput, while we were earlier limited to around 3-400Mbit/s, and trying to push above that mark by widening the TCP window caused ti0: jumbo buffer allocation failed messages to be logged and a corresponding stall in the traffic.
1.63.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.63.6.1	12-Feb-2005	yamt	sync with head.
1.63.4.1	29-Apr-2005	kent	sync with -current
1.68.2.6	11-Feb-2008	yamt	sync with head.
1.68.2.5	15-Nov-2007	yamt	sync with head.
1.68.2.4	27-Oct-2007	yamt	sync with head.
1.68.2.3	03-Sep-2007	yamt	sync with head.
1.68.2.2	30-Dec-2006	yamt	sync with head.
1.68.2.1	21-Jun-2006	yamt	sync with head.
1.69.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.69.10.1	19-Apr-2006	elad	sync with head.
1.69.8.2	14-Sep-2006	yamt	sync with head.
1.69.8.1	01-Apr-2006	yamt	sync with head.
1.69.6.1	22-Apr-2006	simonb	Sync with head.
1.69.4.1	09-Sep-2006	rpaulo	sync with head
1.71.4.2	10-Dec-2006	yamt	sync with head.
1.71.4.1	22-Oct-2006	yamt	sync with head
1.71.2.1	18-Nov-2006	ad	Sync with head.
1.73.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.74.16.3	23-Mar-2008	matt	sync with HEAD
1.74.16.2	08-Nov-2007	matt	sync with -HEAD
1.74.16.1	06-Nov-2007	matt	sync with HEAD
1.74.14.4	11-Nov-2007	joerg	Sync with HEAD.
1.74.14.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.74.14.2	02-Oct-2007	joerg	Sync with HEAD.
1.74.14.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.74.10.2	10-Sep-2007	skrll	Sync with HEAD.
1.74.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.74.2.3	23-Oct-2007	ad	Sync with head.
1.74.2.2	09-Oct-2007	ad	Sync with head.
1.74.2.1	01-Sep-2007	ad	Update for pool_cache API changes.
1.77.4.2	13-Nov-2007	bouyer	Sync with HEAD
1.77.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.78.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.78.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.80.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.80.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.81.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.81.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.81.4.4	11-Aug-2010	yamt	sync with head.
1.81.4.3	11-Mar-2010	yamt	sync with head
1.81.4.2	16-May-2009	yamt	sync with head
1.81.4.1	04-May-2009	yamt	sync with head.
1.82.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.88.4.2	05-Mar-2011	rmind	sync with head
1.88.4.1	30-May-2010	rmind	sync with head
1.88.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.90.18.3	03-Dec-2017	jdolecek	update from HEAD
1.90.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.90.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.90.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.90.8.1	30-Oct-2012	yamt	sync with head
1.91.2.1	18-May-2014	rmind	sync with head
1.93.6.6	05-Feb-2017	skrll	Sync with HEAD
1.93.6.5	05-Oct-2016	skrll	Sync with HEAD
1.93.6.4	09-Jul-2016	skrll	Sync with HEAD
1.93.6.3	19-Mar-2016	skrll	Sync with HEAD
1.93.6.2	22-Sep-2015	skrll	Sync with HEAD
1.93.6.1	06-Jun-2015	skrll	Sync with HEAD
1.93.4.1	04-Aug-2015	snj	Pull up following revision(s) (requested by maxv in ticket #913): sys/dev/pci/if_ti.c: revisions 1.94, 1.95 fix mbuf leak on failure (Brainy) -- Memory leak. Same as r1.93. I don't know why Brainy didn't detect it earlier; or perhaps I forgot to report it. Found by Brainy.
1.99.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.101.8.3	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1590: sys/dev/pci/if_ti.c 1.103, 1.106, 1.108, 1.116-1.120 via patch sys/dev/pci/if_tireg.h 1.26 - Use pci_intr_establish_xname() - Don't clear CFI and priority bit to pass them to the upper layer. - Add support for Farallon PN9000SX from FreeBSD. - u_int_{8,16,32}_t -> uint_{8,16,32}_t - Fix typo in comment. - Style fix.
1.101.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.101.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.103.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.103.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.104.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.104.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.104.2.1	10-Jun-2019	christos	Sync with HEAD
1.112.2.1	15-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1014: sys/dev/pci/if_ti.c 1.116-1.120 via patch sys/dev/pci/if_tireg.h 1.26 - Don't clear CFI and priority bit to pass them to the upper layer. - Add support for Farallon PN9000SX from FreeBSD. - Fix typo in comment. - Style fix.
1.113.2.1	29-Feb-2020	ad	Sync with head.
1.120.12.1	06-Jun-2021	cjep	sync with head
1.120.8.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.123.10.1	02-Aug-2025	perseant	Sync with HEAD
1.32	21-Feb-2023	andvar	s/jumo/jumbo/ in comments.
1.31	05-Jun-2021	thorpej	Back out the following changes: http://mail-index.netbsd.org/source-changes/2020/02/29/msg114581.html http://mail-index.netbsd.org/source-changes/2020/02/29/msg114583.html They have reportedly caused some issues with link detection, and will be revisited later.
1.30	02-Apr-2020	msaitoh	branches: 1.30.8; 1.30.12; No functional change: - s/u_int64_t/uint64_t/ - s/alignes/aligns/ - indent, tabify
1.29	02-Apr-2020	msaitoh	s/impliment/implement/ in comment.
1.28	05-Mar-2020	msaitoh	u_int_t -> unit_t. No funcitonal change.
1.27	05-Mar-2020	msaitoh	miscelaneous -> miscellaneous in comment.
1.26	05-Mar-2020	msaitoh	No functional change: - Fix typo in comment. - Whitespace
1.25	29-Feb-2020	thorpej	G/C the never-used version of ti_hostaddr that can't possibly be correct.
1.24	29-Feb-2020	thorpej	- Change the definition and usage of TI_HOSTADDR() to allow for 64-bit DMA addresses. Modeled after similar usage in the bge(4) driver (the chips supported by bge(4) are descendants of those supported by ti(4)). - Use the 64-bit PCI DMA tag if available, otherwise we're allocating (and using) bounce buffers needlessly.
1.23	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.22	13-Sep-2019	msaitoh	branches: 1.22.2; if_flags is neither int nor short. It's unsigned short.
1.21	05-Mar-2019	msaitoh	branches: 1.21.4; Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.20	27-Oct-2012	chs	branches: 1.20.30; 1.20.38; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.19	07-Sep-2007	tnn	branches: 1.19.58; 1.19.68; First part of ti(4) overhaul: - Sanitise macros - Apply KNF Should have no functional changes.
1.18	04-Mar-2007	christos	branches: 1.18.2; 1.18.10; 1.18.14; 1.18.16; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.17	11-Dec-2005	christos	branches: 1.17.26; merge ktrace-lwp.
1.16	30-May-2005	christos	branches: 1.16.2; - const poisoning - avoid variable shadowing.
1.15	27-Feb-2005	perry	nuke trailing whitespace
1.14	27-Jul-2004	he	branches: 1.14.4; 1.14.6; Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable). Bump the default values for these to the values used by FreeBSD, and also adjust ti_init_rx_ring_jumbo() to use the same constant that FreeBSD uses. Yes, this consumes more kernel memory. The effect of this is that you can use jumbo frames in a back-to-back setup with TCP windows up to about 250KB and get ~930Mbit/s throughput, while we were earlier limited to around 3-400Mbit/s, and trying to push above that mark by widening the TCP window caused ti0: jumbo buffer allocation failed messages to be logged and a corresponding stall in the traffic.
1.13	18-Mar-2004	bouyer	branches: 1.13.2; Use the bus_space__stream_ methods to access the chip's registers. As we turn the chip to big-endian mode on big-endian systems, we should never byte-swap the data read/written from/to registers. Tested on sparc64. Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)
1.12	14-May-2003	wiz	branches: 1.12.2; coaless -> coalesce.
1.11	30-Jun-2001	bjh21	branches: 1.11.16; IFM_1000_TX -> IFM_1000_T, as (breifly) discussed on tech-net.
1.10	30-Jun-2001	thorpej	Add macros to sync the various bits of the DMA'd control data.
1.9	30-Jun-2001	thorpej	Use macros (like those in every other network driver I have written) to compute the offset of a structure in the DMA'd control data, and the DMA address of that structure.
1.8	30-Jun-2001	thorpej	Split the tx-end-of-frame routine into Tigon 1 and Tigon 2 verions, for the same reason we split the tx encap routine.
1.7	30-Jun-2001	thorpej	Split the transmit encap routine into two versions: one for the Tigon 1 and another for the Tigon 2. This eliminates a conditional in the loop around the DMA segments, and will make further cleanup easier.
1.6	30-Jun-2001	thorpej	Move the `ti_tx_ring_nic' pointer out of the ring-data structure and into the softc. Garbage-collect the `ti_cmd_ring' pointer, since bus_space is properly used to access the command ring.
1.5	03-Jun-2001	thorpej	Replace a bunch of home-grown Ethernet related constants with ones defined in <net/if_ether.h>
1.4	15-May-2001	lukem	delint: () protect some macro args
1.3	17-Nov-2000	bouyer	branches: 1.3.2; Pull up FreeBSD changes: 1.39: Have if_ti stop "hiding" the softc pointer in the buffer region. Rather, use the available void * passed to the free routine and pass the softc pointer through there. 1.33: Add support for the Netgear GA620T copper gigabit card. 1.32: Tweak probe message so that 1000baseSX and 1000baseT cards are explicitly identified. 1.31: Update the Tigon driver to support 1000baseTX gigE over copper AceNIC cards. This basically involves switching to the 12.4.13 firmware, plus a couple of minor tweaks to the driver. NetBSD changes: get rid of ti_inuse, the mbuf ref counting code should call ti_free() when needed. Use hardware 802.1q support.
1.2	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.1	01-Sep-1999	drochner	branches: 1.1.2; 1.1.4; 1.1.14; quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver. todo: -IPv6 -clean up jumbo buffer allocation - NetBSD provides an opaque argument to the free function, thus doesn't need the hack done here -deal correctly with the mapping of the shared memory
1.1.14.1	31-Dec-2000	jhawk	Pull up revision 1.3 (requested by bouyer): For the ti driver, add hardware 802.1Q support, update firmware to 12.4.13+wpaul, and support the Netgear GA620T copper gigabit card, and the Alteon ACEnic 100baseSX and 1000baseT gigabit cards. Various small changes from FreeBSD.
1.1.4.1	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.2	05-Oct-1999	he	Pull up revision 1.1 (new, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.1.2.1	01-Sep-1999	he	file if_tireg.h was added on branch netbsd-1-4 on 1999-10-05 22:06:48 +0000
1.3.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.3.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.11.16.1	26-May-2004	tron	Pull up revision 1.13 (requested by bouyer in ticket #1698): Use the bus_space__stream_ methods to access the chip's registers. As we turn the chip to big-endian mode on big-endian systems, we should never byte-swap the data read/written from/to registers. Tested on sparc64. Finally fix kern/13341 by Jason R. Thorpe (really, the hard work of putting bus_dmamap_sync() calls at the right places has been done my Jason mid-2001 :)
1.12.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.12.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.12.2.1	03-Aug-2004	skrll	Sync with HEAD
1.13.2.1	28-Jul-2004	tron	Pull up revision 1.14 (requested by he in ticket #719): Make the SSLOTS, MSLOTS and JSLOTS tunable (redefineable). Bump the default values for these to the values used by FreeBSD, and also adjust ti_init_rx_ring_jumbo() to use the same constant that FreeBSD uses. Yes, this consumes more kernel memory. The effect of this is that you can use jumbo frames in a back-to-back setup with TCP windows up to about 250KB and get ~930Mbit/s throughput, while we were earlier limited to around 3-400Mbit/s, and trying to push above that mark by widening the TCP window caused ti0: jumbo buffer allocation failed messages to be logged and a corresponding stall in the traffic.
1.14.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.4.1	29-Apr-2005	kent	sync with -current
1.16.2.2	27-Oct-2007	yamt	sync with head.
1.16.2.1	03-Sep-2007	yamt	sync with head.
1.17.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.18.16.1	06-Nov-2007	matt	sync with HEAD
1.18.14.1	02-Oct-2007	joerg	Sync with HEAD.
1.18.10.1	10-Sep-2007	skrll	Sync with HEAD.
1.18.2.1	09-Oct-2007	ad	Sync with head.
1.19.68.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.19.58.1	30-Oct-2012	yamt	sync with head
1.20.38.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.20.38.1	10-Jun-2019	christos	Sync with HEAD
1.20.30.1	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1590: sys/dev/pci/if_ti.c 1.103, 1.106, 1.108, 1.116-1.120 via patch sys/dev/pci/if_tireg.h 1.26 - Use pci_intr_establish_xname() - Don't clear CFI and priority bit to pass them to the upper layer. - Add support for Farallon PN9000SX from FreeBSD. - u_int_{8,16,32}_t -> uint_{8,16,32}_t - Fix typo in comment. - Style fix.
1.21.4.1	15-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1014: sys/dev/pci/if_ti.c 1.116-1.120 via patch sys/dev/pci/if_tireg.h 1.26 - Don't clear CFI and priority bit to pass them to the upper layer. - Add support for Farallon PN9000SX from FreeBSD. - Fix typo in comment. - Style fix.
1.22.2.1	29-Feb-2020	ad	Sync with head.
1.30.12.1	06-Jun-2021	cjep	sync with head
1.30.8.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.126	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.125	02-Sep-2022	thorpej	Remove unnecessary inclusion of <net/netisr.h>.
1.124	23-Feb-2022	andvar	fix various typos in comments, mainly immediatly/immediately/, as well shared and recently fixed typos in OpenBSD code by Jonathan Grey.
1.123	06-Sep-2021	andvar	fix typos in word "segment" and two additional typos in if_tl.c.
1.122	07-Jul-2020	msaitoh	No functional change: - u_int32_t -> uint32_t - KNF.
1.121	11-Jun-2020	thorpej	Update for proplib(3) API changes.
1.120	28-Feb-2020	msaitoh	When a media change is requested, don't just all (*if_init)(). That's needlessly disruptive and incompatible with future locking changes. We can use ether_mediachange() instead.
1.119	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.118	22-Dec-2019	thorpej	branches: 1.118.2; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.117	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.116	28-May-2019	msaitoh	branches: 1.116.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.115	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.114	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.113	24-Apr-2019	msaitoh	Remove extra test of SIOCSIFMEDIA.
1.112	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.111	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.110	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.109	26-Jun-2018	msaitoh	branches: 1.109.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.108	22-Jun-2018	msaitoh	It's not required to include net/bpfdesc.h. Remove it.
1.107	23-May-2017	ozaki-r	branches: 1.107.2; 1.107.8; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.106	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.105	07-Jul-2016	msaitoh	branches: 1.105.2; KNF. Remove extra spaces. No functional change.
1.104	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.103	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.102	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.101	10-Aug-2014	tls	branches: 1.101.4; Merge tls-earlyentropy branch into HEAD.
1.100	29-Mar-2014	christos	branches: 1.100.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.99	30-Mar-2013	christos	branches: 1.99.4; remove trailing whitespace
1.98	22-Jul-2012	matt	branches: 1.98.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.97	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.96	13-Nov-2010	uebayasi	branches: 1.96.8; 1.96.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.95	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.94	19-Jan-2010	pooka	branches: 1.94.2; 1.94.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.93	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.92	05-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested Compaq Netelligent 10/100 TX.
1.91	16-Nov-2008	tsutsui	Fix tyop in #ifdef TLDEBUG part.
1.90	04-Oct-2008	bouyer	branches: 1.90.2; Add missing splx(). Funny that it didn't cause issues ...
1.89	01-Jun-2008	tsutsui	branches: 1.89.4; Split device_t/softc, with misc cosmetic changes.
1.88	31-May-2008	tsutsui	My Compaq Netelligent 10/100 TX has 24C02 (2kbit) serial EEPROM, so pass 256 (bytes) rather than 512 to seeprom_bootstrap_read().
1.87	22-May-2008	dyoung	KNF.
1.86	10-Apr-2008	cegger	branches: 1.86.2; 1.86.4; 1.86.6; use aprint_*_dev and device_xname
1.85	19-Jan-2008	dyoung	branches: 1.85.6; Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.84	19-Oct-2007	ad	branches: 1.84.2; 1.84.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.83	09-Jul-2007	ad	branches: 1.83.6; 1.83.8; 1.83.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.82	04-Mar-2007	christos	branches: 1.82.2; 1.82.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.81	25-Dec-2006	wiz	branches: 1.81.2; Spell "schedule" correctly. From Zafer Aydogan.
1.80	24-Dec-2006	rumble	Make this compile again when KASSERT is enabled.
1.79	24-Dec-2006	rumble	Revert the previous change and specify the appropriate page boundary to bus_dmamap_create(9) via a proplib number entry. Recommended by Izumi Tsutsui.
1.78	23-Dec-2006	rumble	The Set Engineering GIO board for sgimips (IP22) sits behind a special bridge that precludes the ThunderLAN's DMA engine from performing segment transfers across page boundaries. Add logic under #ifdef TL_SETENG_GFE to split up these segment transfers appropriately. It's unknown whether this issue could also affect the RX path, though no problems have been observed yet.
1.77	16-Nov-2006	christos	branches: 1.77.2; __unused removal on arguments; approved by core.
1.76	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.75	07-Sep-2006	dogcow	branches: 1.75.2; 1.75.4; remove more vestiges of CCITT, LLC, HDLC, NS, and NSIP.
1.74	07-Sep-2006	rumble	Avoid a panic on adapter check interrupts: the 'tl_restart' callout expects a pointer to struct ifnet, not tl_softc.
1.73	30-Aug-2006	rumble	Make this compile again when TLDEBUG is defined.
1.72	24-Dec-2005	perry	branches: 1.72.4; 1.72.8; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.71	11-Dec-2005	christos	merge ktrace-lwp.
1.70	30-May-2005	christos	branches: 1.70.2; - const poisoning - avoid variable shadowing.
1.69	20-Apr-2005	bouyer	Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.68	04-Feb-2005	perry	branches: 1.68.4; de-__P
1.67	23-Jan-2005	dan	branches: 1.67.2; attach tl(4) as an event source to rnd(4) patches from Sean Davis in PR/28556 (slightly modified to actually catch our own interrupts)
1.66	30-Oct-2004	thorpej	branches: 1.66.4; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.65	15-Jun-2004	bouyer	Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.64	10-Nov-2003	wiz	branches: 1.64.2; Spell address with two d's. Inspired by similar changes in OpenBSD, originating from Jonathon Gray and forwarded by jmc@openbsd.
1.63	02-Nov-2003	wiz	deferred, not deffered. Inspired by Tom Cosgrove.
1.62	05-Oct-2003	tsutsui	Increase number of transmit/receive buffers from 10 to 32 to avoid RX buffer overrun.
1.61	05-Oct-2003	tsutsui	Pass proper dma mapsize of TX/RX mbufs to bus_dmamap_sync(). Now tl(4) works on sgimips.
1.60	05-Oct-2003	tsutsui	Oops, more white space nit.
1.59	05-Oct-2003	tsutsui	TAB/space nits.
1.58	30-Sep-2003	thorpej	New generic I2C framework. Supports bit-bang and "intelligent" I2C interface controllers (of varying intelligence levels). Contributed by Wasabi Systems, Inc. Primarily written by Steve Woodford, with some modification by me.
1.57	19-Mar-2003	bouyer	branches: 1.57.2; tl(4) supports ETHERCAP_VLAN_MTU.
1.56	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.55	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.54	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.53	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.52	17-Apr-2002	wiz	branches: 1.52.4; Fix two typos in comments. kern/16386 by Sean Davis.
1.51	07-Mar-2002	christos	stupid void pointers... tl_init() needs struct ifnet *...
1.50	05-Mar-2002	itojun	bring in latest ALTQ from kjc. ALTQify some of the drivers.
1.49	28-Feb-2002	christos	PR/15754: Sean Davis: Initialize if_softc before calling the mii init routines to avoid NULL pointer de-referencing in the callbacks.
1.48	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.47	13-Nov-2001	lukem	add RCSID
1.46	20-Sep-2001	bouyer	- Allocate control structures at attach time rather than at init time. Avoids using bus_dmamem_alloc/bus_dmamem_map at interrupt time. Should fix PRs kern/13924 and kern/13979 from dive@endersgame.net - while I'm there convert to use ether_ioctl.
1.45	07-Aug-2001	bouyer	branches: 1.45.2; Properly set/clear IFF_OACTIVE.
1.44	06-Aug-2001	bouyer	Improvement: allocate one page of dmamem for Rx, Tx and nullbuff instead of 3 different ones. Also pass the rigth seg/nseg to bus_dmamem_free(). Correct types in debug printf's
1.43	03-Aug-2001	bouyer	- kill the carrier detect code. It's not usefull any more and can give false informations on a loaded ethernet segement. - convert to bus_dma(9) (better late than never :) - add proper le32toh/htole32 so that it works on big-endian system (tested on macppc). Close PR kern/10327.
1.42	07-Jul-2001	thorpej	branches: 1.42.2; bzero -> memset
1.41	07-Jul-2001	thorpej	bcopy -> memcpy, strcpy
1.40	21-Jun-2001	bouyer	Proper media support for Compaq Deskpro 4000. PR kern/13244 from Andrew Gillham.
1.39	28-Dec-2000	sommerfeld	branches: 1.39.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.38	14-Dec-2000	thorpej	ALTQ'ify.
1.37	15-Nov-2000	thorpej	Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
1.36	01-Oct-2000	thorpej	Move the check for "promisc + unicast + not for us" into ether_input(), and change Ethernet drivers to always pass all received frames to ether_input() (with a few exceptions, which are documented in the code).
1.35	05-Sep-2000	thorpej	Cast arg to vtophys() to vaddr_t.
1.34	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.33	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.32	23-Mar-2000	thorpej	branches: 1.32.4; New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.31	06-Mar-2000	thorpej	No longer necessary to futz with ifp->if_baudrate here.
1.30	02-Feb-2000	thorpej	Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass flags down from the parent to child vi mii_attach().
1.29	02-Feb-2000	thorpej	Bring some order to the chaos which was the MII code function naming "conventions".
1.28	12-Dec-1999	tron	As suggested by Jason Thorpe back out revision 1.27 because it broke PHY probing for the ThunderLan driver.
1.27	19-Nov-1999	thorpej	Make the ThunderLAN and VIA Rhine drivers use the common MII bit-bang module.
1.26	12-Nov-1999	thorpej	Call mii_down() as appropriate.
1.25	04-Nov-1999	thorpej	Adapt to mii_phy_probe() change.
1.24	18-May-1999	thorpej	branches: 1.24.2; 1.24.4; 1.24.8; Rework layer 2 protocol input routines. Instead of calling e.g. ether_input() directly, call the function pointer (if_input)(ifp, m). The input routine expects the packet header to be at the head of the packet, and will adjust as necessary. Privatize the layer 2 input and output routines, allowing _ifattach() to set them up as appropriate.
1.23	25-Mar-1999	bouyer	branches: 1.23.4; 1.23.6; First pass at support of Compaq NetFlex 3/P PCI: this one has IO and Mem base address register inverted. Still need some MII hack to get it fully fonctionnal.
1.22	11-Jan-1999	tron	Because memory mapped access doesn't fail only with the docking station for the TI TravelMate 5000 but also with at least one Compaq NIC we prefer I/O mapped access again. Fixes PR kern/6690.
1.21	30-Oct-1998	thorpej	branches: 1.21.4; Manuel confirms that defaulting to AUTO is ok, to Do It.
1.20	18-Aug-1998	thorpej	vm_offset_t -> vaddr_t
1.19	15-Aug-1998	thorpej	Oops, don't forget to tick the MII once a second.
1.18	15-Aug-1998	bouyer	Fix typos in last commit.
1.17	15-Aug-1998	bouyer	correct indent.
1.16	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.15	11-Aug-1998	thorpej	Adapt to the new MII code. This required splitting some structures into if_tlvar.h, since the ThunderLAN PHY driver needs to know some things about it's parent (and since it can only attach to a ThunderLAN chip, this is the easiest way).
1.14	08-Aug-1998	mycroft	Use splnet, not splimp.
1.13	05-Jul-1998	jonathan	branches: 1.13.2; defopt NS, NSIP.
1.12	05-Jul-1998	jonathan	defopt INET, NETATALK.
1.11	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.10	05-May-1998	thorpej	Slighly less brutal hack to deal with broken memory-mapped access of the ThunderLAN on the TI TravelMate 5000 docking station: single it out, and allow memory-mapped access on the Compaq products.
1.9	11-Feb-1998	bouyer	Correct a bogosity in the adapter->mii attach code pointed out by cgd: in pci/if_tl.c, call config_found() with a print function, instead of printing ourself a message in if_tl.c if no miibus was found. The print function is in mii/mii.c (mii_adapter_print()) so that it can be used by any adapter (idea from the scsi system).
1.8	12-Jan-1998	thorpej	Update for config changes.
1.7	30-Nov-1997	drochner	make it compile without "__BROKEN_INDIRECT_CONFIG"
1.6	18-Nov-1997	bouyer	Fix for systems with multiple tl interfaces: nullbuf (used to pad packets < ETHER_MIN_SIZE) is used for all tl interfaces. Allocates only once, and never deallocate it (as we can't say if another instance of the driver is interface is using it).
1.5	17-Nov-1997	thorpej	Sigh, now that we've fixed a bug in PCI space management (the bug caused i/o-mapped space to always be used), we discover that at least one ThunderLAN interface can't read the EEPROM properly if memory-mapped access is used. Kludge around this for now by "prefering" i/o space.
1.4	17-Nov-1997	thorpej	Fix a couple of bugs/annoyances in match/attach: - Do PCI space configuration like the other drivers. In particular, don't _disable_ the space we're not using because some lame firmware implementations might not reenable it on warm boot. Also, prefer memory space always. - Make match and info-gathering in attach table-driven. - Rearrange things a bit to be a bit more visually pleasing during boot. Also, fixup some #include problems.
1.3	16-Nov-1997	christos	PR/4510: William Coldwell: Add TI thunderlan support.
1.2	21-Oct-1997	bouyer	Correct 'ThunderLAN' spelling
1.1	17-Oct-1997	bouyer	branches: 1.1.2; Driver for Texas Instruments' thunderland network controller (present in some Compaq products).
1.1.2.5	10-Nov-1998	cgd	patch from bouyer to address sysinst and media selection issues: set the default media to UTP, and don't include 'none' in the media list, so that sysinst will be able to set media properly. Done differently in -current, because there the code uses MII. (bouyer)
1.1.2.4	18-Nov-1997	mellon	Pull rev 1.6 up from trunk (bouyer)
1.1.2.3	17-Nov-1997	thorpej	Sync w/ trunk.
1.1.2.2	17-Nov-1997	thorpej	Sync w/ trunk.
1.1.2.1	21-Oct-1997	mrg	pull up from trunk: Correct ThunderLAN spelling
1.13.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.21.4.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.23.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.23.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.24.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.24.4.1	15-Nov-1999	fvdl	Sync with -current
1.24.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.24.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.24.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.32.4.3	11-Mar-2002	he	Revert pull-up of revision 1.51 (requested by christos): tl_init() changed interface between the netbsd-1-5 branch and -current. Therefore, application of 1.51 is wrong, so undo it here.
1.32.4.2	09-Mar-2002	he	Pull up revisions 1.49,1.51 (via patch, requested by christos): Make sure ifp->sc_softc is initialized before we set the media callbacks, and correct argument error in invocation of tl_init().
1.32.4.1	13-Nov-2001	he	Pull up revision 1.40 (requested by bouyer): Fix media support for Compaq Deskpro 4000. Fixes PR#13244.
1.39.2.7	18-Oct-2002	nathanw	Catch up to -current.
1.39.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.39.2.5	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.39.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.39.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.39.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.39.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.42.2.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.42.2.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.42.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.42.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.42.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.42.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.45.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.52.4.3	01-May-2005	tron	Pull up revision 1.69 (requested by bouyer in ticket #5756): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.52.4.2	22-Jun-2004	tron	Pull up revision 1.65 (requested by bouyer in ticket #1718): Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.52.4.1	30-Jun-2003	grant	Pull up revision 1.57 (requested by bouyer in ticket #1350): tl(4) supports ETHERCAP_VLAN_MTU.
1.57.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.57.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.57.2.5	24-Jan-2005	skrll	Sync with HEAD.
1.57.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.57.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.57.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.57.2.1	03-Aug-2004	skrll	Sync with HEAD
1.64.2.1	22-Jun-2004	tron	branches: 1.64.2.1.2; Pull up revision 1.65 (requested by bouyer in ticket #537): Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.64.2.1.2.3	01-May-2005	tron	Pull up revision 1.69 (requested by bouyer in ticket #1438): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.64.2.1.2.2	16-Apr-2005	tron	Pull up revision 1.67 (requested by dan in ticket #1117): attach tl(4) as an event source to rnd(4) patches from Sean Davis in PR/28556 (slightly modified to actually catch our own interrupts)
1.64.2.1.2.1	24-Jan-2005	he	Pull up revision 1.66 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.66.4.1	29-Apr-2005	kent	sync with -current
1.67.2.1	12-Feb-2005	yamt	sync with head.
1.68.4.1	21-Apr-2005	tron	Pull up revision 1.69 (requested by bouyer in ticket #180): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.70.2.5	21-Jan-2008	yamt	sync with head
1.70.2.4	27-Oct-2007	yamt	sync with head.
1.70.2.3	03-Sep-2007	yamt	sync with head.
1.70.2.2	30-Dec-2006	yamt	sync with head.
1.70.2.1	21-Jun-2006	yamt	sync with head.
1.72.8.2	14-Sep-2006	yamt	sync with head.
1.72.8.1	03-Sep-2006	yamt	sync with head.
1.72.4.1	09-Sep-2006	rpaulo	sync with head
1.75.4.2	10-Dec-2006	yamt	sync with head.
1.75.4.1	22-Oct-2006	yamt	sync with head
1.75.2.2	12-Jan-2007	ad	Sync with head.
1.75.2.1	18-Nov-2006	ad	Sync with head.
1.77.2.1	04-Mar-2007	bouyer	Pull up following revision(s) (requested by rumble in ticket #478): sys/dev/pci/if_tl.c: revision 1.78 - 1.80 sys/dev/ic/tulip.c: revision 1.148 Make the ThunderLAN part of the Set Engineering GIO board for sgimips (IP22) work. Extract mac addresses on Phobos G100 boards.
1.81.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.82.4.1	11-Jul-2007	mjf	Sync with head.
1.82.2.2	23-Oct-2007	ad	Sync with head.
1.82.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.83.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.83.8.2	23-Mar-2008	matt	sync with HEAD
1.83.8.1	06-Nov-2007	matt	sync with HEAD
1.83.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.84.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.84.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.85.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.85.6.2	05-Oct-2008	mjf	Sync with HEAD.
1.85.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.86.6.2	10-Oct-2008	skrll	Sync with HEAD.
1.86.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.86.4.4	11-Aug-2010	yamt	sync with head.
1.86.4.3	11-Mar-2010	yamt	sync with head
1.86.4.2	16-Sep-2009	yamt	sync with head
1.86.4.1	04-May-2009	yamt	sync with head.
1.86.2.1	04-Jun-2008	yamt	sync with head
1.89.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.89.4.1	19-Oct-2008	haad	Sync with HEAD.
1.90.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.94.4.2	05-Mar-2011	rmind	sync with head
1.94.4.1	30-May-2010	rmind	sync with head
1.94.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.96.12.1	18-Feb-2012	mrg	merge to -current.
1.96.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.96.8.2	30-Oct-2012	yamt	sync with head
1.96.8.1	17-Apr-2012	yamt	sync with head
1.98.2.3	03-Dec-2017	jdolecek	update from HEAD
1.98.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.98.2.1	23-Jun-2013	tls	resync from head
1.99.4.1	18-May-2014	rmind	sync with head
1.100.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.101.4.5	28-Aug-2017	skrll	Sync with HEAD
1.101.4.4	05-Feb-2017	skrll	Sync with HEAD
1.101.4.3	09-Jul-2016	skrll	Sync with HEAD
1.101.4.2	19-Mar-2016	skrll	Sync with HEAD
1.101.4.1	06-Jun-2015	skrll	Sync with HEAD
1.105.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.107.8.4	26-Jan-2019	pgoyette	Sync with HEAD
1.107.8.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.107.8.2	28-Jul-2018	pgoyette	Sync with HEAD
1.107.8.1	25-Jun-2018	pgoyette	Sync with HEAD
1.107.2.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.107.2.1	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.109.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.109.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.109.2.1	10-Jun-2019	christos	Sync with HEAD
1.116.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.118.2.1	29-Feb-2020	ad	Sync with head.
1.131	20-Dec-2023	thorpej	malloc(9) -> kmem(9), and plug a memory leak on detach while I'm here.
1.130	08-May-2021	thorpej	Use pci_compatible_match().
1.129	07-Jul-2020	msaitoh	branches: 1.129.6; No functional change: - u_int32_t -> uint32_t - KNF.
1.128	02-Jul-2020	msaitoh	prop_data_data_nocopy -> prop_data_value
1.127	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.126	29-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - KNF
1.125	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.124	07-Jul-2016	msaitoh	branches: 1.124.16; 1.124.18; KNF. Remove extra spaces. No functional change.
1.123	29-Mar-2014	christos	branches: 1.123.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.122	23-Sep-2012	chs	branches: 1.122.2; match some more devices.
1.121	11-Nov-2011	jakllsch	branches: 1.121.6; 1.121.10; Use GPP_GPC in the last place it isn't already used.
1.120	11-Nov-2011	jakllsch	Improve punctuation and consistency in some comments.
1.119	10-Nov-2011	jakllsch	When applying the board-specific bits for the SMC 9332, the mediasw has already been set to non-NULL, don't require it to be NULL.
1.118	26-Jul-2011	dyoung	branches: 1.118.2; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.117	09-Jul-2011	christos	- add a routine to get the name of the card.
1.116	21-Jan-2010	martin	Remove sparc specific code, rely on device properties instead
1.115	01-Sep-2009	jmcneill	aprint-ify
1.114	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.113	17-Apr-2009	cegger	backout rev. 1.112. I was misleaded by the wrong pci(9) manpage.
1.112	17-Apr-2009	cegger	pci_activate() wants a softc and not a device
1.111	17-Apr-2009	cegger	check return code of tlp_attach()
1.110	17-Apr-2009	cegger	Free resources in error path. While here, fix copy&paste error I introduced in rev. 1.108.
1.109	17-Apr-2009	cegger	use aprint_*
1.108	17-Apr-2009	cegger	Detach tlp(4) during shutdown. Free unused resources on attachment.
1.107	17-Apr-2009	cegger	device_t/softc split. Tested with tlp at pci
1.106	08-Jun-2008	tsutsui	branches: 1.106.6; 1.106.12; Use device_lookup_private() rather than using cd_devs[] directly to get softc. XXX maybe we should change a type of cd_devs[] in struct cfdriver from (void *) to device_t.
1.105	28-Apr-2008	martin	branches: 1.105.2; Remove clause 3 and 4 from TNF licenses
1.104	10-Apr-2008	cegger	branches: 1.104.2; 1.104.4; use aprint_*_dev and device_xname
1.103	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.102	19-Oct-2007	ad	branches: 1.102.12; 1.102.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.101	18-Feb-2007	tsutsui	branches: 1.101.4; 1.101.16; 1.101.18; 1.101.22; - KNF, ANSIfy, TAB/space cleanup - unwrap some lines - u_intNN_t -> uintNN_t
1.100	10-Feb-2007	dsl	branches: 1.100.2; Remove some ancient debugging leftovers left in by Michael Lorenz.
1.99	09-Feb-2007	macallan	add a quirks entry for the DEC21140 found on UMAX E100 cards
1.98	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.97	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.96	30-Aug-2006	rumble	branches: 1.96.2; 1.96.4; Add quirks for the Phobos GIO Fast Ethernet boards.
1.95	26-Jul-2006	itohy	branches: 1.95.2; Remove ADM9511's "signature". I don't think it is used anywhere.
1.94	21-Jul-2006	gdamore	Update ADMtek pci device ids to match data sheets. While here, add the 9513 device (although I've never seen one, I have a datasheet for it. :-)
1.93	18-Jul-2006	itohy	Make new device (vendor 0x1317, product 0x9511) work as AN983.
1.92	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.91	20-May-2006	rpaulo	branches: 1.91.2; The check for LMC cards should return NULL, not 0. No functional change.
1.90	05-May-2006	thorpej	branches: 1.90.2; Remove the devprop API and switch everthing over to the new proplib. Add a new device_properties() accessor for device_t that returns the device's property dictionary.
1.89	20-Apr-2006	christos	Put back the code that avoids matching the lmc card. As pointed out by David Boggs, it does not matter if the lmc driver routine returns 100 for the match, because if the lmc driver is loaded via an LKM after boot, the other drivers would have claimed the card already.
1.88	25-Mar-2006	rpaulo	Add support for the Conexant RS7112. From Frank Wille <frank@phoenix.owl.de>.
1.87	25-Mar-2006	thorpej	Use device_parent().
1.86	18-Feb-2006	thorpej	branches: 1.86.2; 1.86.4; 1.86.6; - Don't expose dev_propdb directly -- provide devprop_*() wrappers instead. - Rework the ARMADILLO / epe device properties interaction so that it actually associates the MAC address property with the epe device instance.
1.85	16-Feb-2006	thorpej	Remove the ugly algor_get_ethaddr() hack. We have a properties infrastructure now, so use that to provide the Ethernet address of the built-in Ethernet. XXX Still a little ugly in the tlp driver, but fixing that would require XXX reworking the SROM usage in a broader sense. While here, do the P4032 DMA window hack in device_register() as well.
1.84	06-Dec-2005	christos	branches: 1.84.2; 1.84.4; 1.84.6; Remove the "don't match lmc cards" test, because the lmc driver returns a match of 100 which will supersede the de and tlp match if present. If not present, then these two drivers will fail to match an mii. Thanks to thorpej for the explanation.
1.83	06-Dec-2005	thorpej	Don't use magic constants in previous. (Not that we should be doing this here in the first place...)
1.82	06-Dec-2005	christos	Don't match lmc cards by accident.
1.81	17-Jul-2005	rpaulo	If the DEC board name is not known, remove the leading whitespace in sc_name. This closes PR#30458. ok'ed Jason Thorpe.
1.80	23-Jun-2005	rpaulo	branches: 1.80.2; Add support for ASIX AX88140A/AX88141 chipsets using only the 10/100 MII with inspiration from FreeBSD dc(4) driver. Tested on i386 and macppc. Reviewed by Jason Thorpe and ok'ed by Hubert Feyrer.
1.79	13-Jun-2005	tron	Also use prom_getether() under NetBSD-sparc(64) if the MAC address retrieved from the PROM starts with ff:ff:ff. The driver now works with the onboard interface in the Tadpole SPARCle.
1.78	27-Feb-2005	perry	branches: 1.78.2; nuke trailing whitespace
1.77	21-Aug-2004	thorpej	branches: 1.77.4; 1.77.6; Use ANSI function decls and make use of static.
1.76	17-Mar-2004	martin	branches: 1.76.4; For sparc*: myetheraddr is no more
1.75	28-Nov-2003	cube	Add support for the 3com OfficeConnect 10/100B card, which has an ADMtek AN985 controller, handled by tlp(4). From Mike M. Volokhov <mishka AT terabyte DOT com DOT ua> in current-users.
1.74	03-Nov-2003	ichiro	Fix uninitialized variable warnings
1.73	27-Jun-2003	itohy	branches: 1.73.2; Properly reset PHY on ZNYX boards which are composed of 21143 and Quality Semiconductor QS6611 PHY. QS6611 requires explicit reset on some hardware configuration. See ``Using the Intel 21143 with the Quality Semiconductor QS6611 Revision D4'' <http://www.intel.com/design/network/applnots/278239.htm> for details. Models which should be supported by this change are ZX345Q, ZX346Q, ZX348Q, (non-Q versions use different PHYs) ZX414, ZX412, ZX444, ZX442, ZX212, ZX214, ZX374, ZX372, ZX244, ZX424, ZX422. Tested on ZX346Q. Some information is from the ZNYX driver source code available from the ZNYX web site <http://www.znyx.com/>.
1.72	10-Mar-2003	thorpej	Fix an error message.
1.71	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.70	09-Oct-2002	jdolecek	tlp_pci_vpc_21140_quirks(): make agnostic to signed/unsigned char; fixes compile on cats
1.69	08-Oct-2002	minoura	Add quirk for Connectix VirtualPC emulated 21140, just to print the name from the SROM.
1.68	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.67	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.66	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.65	17-Apr-2002	mycroft	Remove TLP_MATCH_* options.
1.64	04-Apr-2002	chs	add support for Cogent EM100 (aka Adaptec ANA-6910) TX and FX.
1.63	03-Apr-2002	thorpej	Add support for the Netwinder on-board 21143 (but don't try using it yet; there are still some problems).
1.62	26-Mar-2002	chs	add support for SMC 9332DST.
1.61	16-Mar-2002	chs	decode the sub-system IDs on adaptec cards so we can differentiate between multi-port cards (which need the shared-rom/intr stuff) and single-port cards (which must not have it). previously we enabled sharing for all adaptec cards, which caused problems if you had multiple single-port cards on the same PCI bus.
1.60	13-Mar-2002	chs	support the 4-port boards with the Cogent OUI as well.
1.59	13-Mar-2002	chs	support ANA-6944A boards (in addition to ANA-6922).
1.58	14-Feb-2002	chs	add support for Adaptec ANA-69xx multi-port cards.
1.57	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.56	07-Dec-2001	matt	Fix sense of if to grab address from alternation if srom is not in isv format on Davicom chips. Add sparc specific case in this case to call myetheraddr if oui is all 0s (for Netra X1).
1.55	13-Nov-2001	lukem	add RCSID
1.54	08-Jul-2001	thorpej	branches: 1.54.2; Remove unnecessary include files. From Onno van der Linden.
1.53	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.52	27-May-2001	thorpej	Add quirks for the Algorithmics P-5064 on-board 21143, and a way to get the Ethernet address from it (they don't have an SROM). XXX The latter should be replaced with devprops when they exist.
1.51	24-Feb-2001	cgd	branches: 1.51.2; C requires that labels be followed by statements.
1.50	08-Jan-2001	thorpej	Add Accton EN2242 10/100 MiniPCI Ethernet.
1.49	08-Jan-2001	thorpej	Use the AN985 media switch on the AN985.
1.48	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.47	03-Oct-2000	thorpej	Add support for the ADMtek AN983 and AN985.
1.46	03-Aug-2000	castor	Support some differences in the Macronix 98715AEC-C and E chips from the other 98715* series. The MX98715AEC-[C,E] use a different location in the serial eerom for LED control, and programming it with the original location's values caused unpredictable behavior. Also, start integrating fixes where media changes on an adapter under load may fail. There's more work to be done here, but I need to sort out our internal changes a little more carefully.
1.45	17-Jul-2000	tron	Don't make assumptions of the actual value of PCI_PMCSR_STATE_D0. Problem noted by John Hawkinson.
1.44	17-Jul-2000	tron	Don't clobber bits from 2 to 31 when trying to change the power state. Problem noted by John Hawkinson.
1.43	16-Jul-2000	tron	If card is in power state D3 put it into power state D0 so that it will at least work after the next reboot.
1.42	15-Jul-2000	tron	Avoid printing duplicate colon when card is waked up.
1.41	15-Jul-2000	tron	Improve kernel message when card is in power state D3. Instead of... epic0 at pci0 dev 11 function 0 epic0: unable to wake up from ... ... print this: epic0 at pci0 dev 11 function 0: unable to wake up from power state D3
1.40	25-Jun-2000	sommerfeld	Defopt TLP_MATCH_*
1.39	27-May-2000	thorpej	branches: 1.39.2; While DM9102A boards tend to have ISV-format SROMs (likely to describe the HomePNA PHY typically connected to the external MII interface), DM9102 implementations (often found on motherboards) do not. Handle this.
1.38	26-May-2000	thorpej	branches: 1.38.2; Add support for the Davicom DM9102 and DM9102A 10/100 Ethernet chips. Partially based on diffs submitted by Matthew Orgass <darkstar@pgh.net> and IWAMOTO Toshihiro <iwamoto@sat.t.u-tokyo.ac.jp>.
1.37	04-Apr-2000	thorpej	- if_tlp_cardbus.c: Use the `pmreg' pointer passed by by cardbus_get_capability() [mirror change already made to if_tlp_pci.c] - if_tlp_cardbus.c: If we don't find an ISV SROM, try to grab the Ethernet address from the CIS. - if_tlp_cardbus.c: set CardBus cards to store-and-forward mode from the get-go. - Put the TxThresh tables in tulipvar.h, and use them in the CardBus and PCI front-ends to go to store-and-forward mode. - Document the Xircom X3201-3 clone a little more.
1.36	26-Mar-2000	soren	For Cobalt Networks 21143's, always use store-and-forward and toggle a GPIO line to fully reset the PHY.
1.35	23-Mar-2000	mycroft	Rather than guessing at the location of the PMCSR, use the pointer returned by pci_get_capability(). (This is, after all, a standardized interface...)
1.34	15-Mar-2000	thorpej	Add power management support to the `tlp' driver. The battery on my laptop lasts a whole lot longer now.
1.33	07-Mar-2000	mycroft	Determine the size of the SROM by probing the chip using the standard Microwire protocol. Eliminates the need for srom_quirks and some other trash.
1.32	26-Jan-2000	thorpej	Give the ISV filter a chance to verify 21143 pass 4.1 SROMs before running them through the quirk table.
1.31	26-Jan-2000	thorpej	Additional hackery for the Cobalt interfaces. Even though these are pass 4.1 21143s, they have a 128-byte SROM, and thus use a different SROM address size than all other pass 4.1 21143s. A plea to hardware designers -- please read application notes for the components you're using before putting your hardware together. Thanks.
1.30	25-Jan-2000	thorpej	Clean up the SROM addrbits code a little.
1.29	25-Jan-2000	thorpej	Add a chip type for the Xircom X3201-3, and centralize the declaration of the chip names array.
1.28	25-Jan-2000	thorpej	Handle Cobalt Networks interfaces, which are 21143 pass 4.1 devices with a simple (MAC address only) SROM format. From Soren S. Jorvang <soren@wheel.dk>
1.27	11-Dec-1999	thorpej	- Don't try to run the chip in big-endian mode. Not all chips support this consistently, and it doesn't always work even when the chip supports it. - Make sure things DMA'd to the chip that the chip interprets are in little-endian mode.
1.26	07-Dec-1999	thorpej	After reading the 21143 manual again, make sure SNOOZE is cleared in the CFDA register, too.
1.25	07-Dec-1999	thorpej	Add the PMCSR offset for the 21143, and make sure to bring the 21142 and 21143 out of sleep mode. From Masanori Kanaoka.
1.24	07-Dec-1999	thorpej	If, on the 21143, we find that we don't have an ISV SROM, try reading it again, with 8 address bits. 21143 Pass 4.1 has a larger SROM, for storing the CIS when used in a CardBus application.
1.23	19-Nov-1999	thorpej	- Use the common MII bit-bang module. - Add support other than 128-byte SROMs. - Add support for the 21142/21143 MII.
1.22	04-Nov-1999	thorpej	Use Memory Read Line and Memory Read Multiple on the 21140, 21140A, 21142, and 21143, if the PCI bus says its okay.
1.21	28-Oct-1999	thorpej	Add some 21142/21143 goo.
1.20	30-Sep-1999	thorpej	branches: 1.20.2; 1.20.4; 1.20.6; Changes to get the Lite-On/Macronix 82C115 working: - 82C115 has a 128-bit multicast hash table, not 512-bit. - Correct the way the MAC address is read from the SROM, after re-reading the MX98715A Application Note. Other semi-related changes: - Differentiate between MX98715 and MX98715A. - Improve the Macronix link-up/link-down detection.
1.19	29-Sep-1999	thorpej	- Get a little closer on the Macronix chips. Still some more work to do. - Add some basic support for the ADMtek AL981 clone (which, unfortunately, I can't test until my sample boards arrive).
1.18	28-Sep-1999	thorpej	Add code to take a device out of Power-Down state if it's in such a state, and attempt to attach Macronix chips.
1.17	26-Sep-1999	thorpej	Add a quirk entry for the Asante 21140[A] boards that don't use the ISV SROM format. For these boards, we provide the GPIO pin direction info, a separate reset hook, and hard-wire them to MII-on-SIO. Based on a patch submitted by Luoqi Chen <luoqi@chekov.watermarkgroup.com>.
1.16	26-Sep-1999	thorpej	Some 21140[A] boards don't use the ISV SROM format (e.g. Asante boards), but instead use the older DEC Address ROM format. From a patch supplied by Luoqi Chen <luoqi@chekov.watermarkgroup.com>.
1.15	26-Sep-1999	thorpej	Add quirk entry support for the 21041 and 21140 chips. Initial quirks for these: extract the spiffy DEC board names (e.g. DE500-AA).
1.14	26-Sep-1999	thorpej	Make tlp_read_srom() work on big-endian systems. From a patch supplied by Luoqi Chen <luoqi@chekov.watermarkgroup.com>.
1.13	25-Sep-1999	thorpej	Add support for the DECchip 21140[A]. This currently only works with boards which use MII for media attachment. ISV SROM format information lifted from Matt Thomas's `de' driver. Thanks to Dave Sainty for experimenting w/ his 21140A MII boards, and for supplying a fix to the MII bit-bang code (PR #8382).
1.12	20-Sep-1999	thorpej	Add support for the DECchip 21041. This also adds some preliminary support for the ISV SROM format used in the 2114{1,1A,2,3}. Note, like the 21040, auto-sense is not yet supported for the 21041. Add a "pre-init" hook, which will be used for the 2114x and PNIC (currently) which allows chip-specific code to set up and write OPMODE before the chip is reset. This is necessary in order for the chip's internal pathways to get initialized properly for MII/SYM/PCS/SIA media attachments. Thanks to Dave Sainty for the hint from the `de' driver that inspired the pre-init hook.
1.11	17-Sep-1999	thorpej	Fix a few buglets in the multi-port 21040 code. ZNYX ZX314 confirmed working now.
1.10	14-Sep-1999	thorpej	Make sure 21040 chips are at least pass 2.0 and that 21140 chips are at least pass 1.1.
1.9	14-Sep-1999	thorpej	Move the PCI device number to the base Tulip softc; it will be needed to decode the SROM on 21041 and 2114x multi-port boards.
1.8	14-Sep-1999	thorpej	First crack at shared ROM/shared interrupts on 21040 multi-port boards, plus quirk identification for non-multiple-port boards.
1.7	14-Sep-1999	thorpej	Oops, fix Winbond regshift.
1.6	14-Sep-1999	thorpej	Change the semantics of `sc_regshift'; it is now the shift required to make a CSR index number into a CSR offset. Introduce a couple of macros to handle CSR name -> offset conversion.
1.5	14-Sep-1999	thorpej	- More PNIC NWAY stuff. Still more work to do, here. - Fix the SROM checksum routine. - Add code to parse the old DEC Address ROM SROM format. - Rearrange the statchg routines a bit, to make them consistent with one another. - Add support for the DECchip 21040. XXX No support for media autosense yet, and no support for any of the multi-port boards yet.
1.4	08-Sep-1999	thorpej	Add chip types for several more Tulip clones (ASIX, ADMtex, PNIC II, Davicom), and match the Compex RL100-ATX; it uses the Winbond 89C840F (RL100-TX is a Macronix 98713).
1.3	08-Sep-1999	thorpej	Update for Lite-On product ID change.
1.2	03-Sep-1999	thorpej	Byte swap the EEPROM words on the PNIC. From Stephen Ma.
1.1	01-Sep-1999	thorpej	This is the long-awaited "new Tulip driver", a complete, from-scratch rewrite of the driver for the DECchip 21x4x Ethernet chips, and a variety of clones. Currently, the driver supports the Winbond 89C840F (this works pretty well), and the Lite-On PNIC (e.g. NetGear PCI boards), however Lite-On support may be broken [I may simply have a busted test board]. Eventually, support for the Macronix and ASIX chips will filter into this driver, and then, slowly, support for the genuine DEC chips, and maybe even the DE-425 EISA model.
1.20.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.20.4.1	15-Nov-1999	fvdl	Sync with -current
1.20.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.20.2.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.20.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.20.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.38.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.39.2.7	24-Apr-2001	he	Apply patch (requested by he): Fix typo in last patch, so that this compiles again.
1.39.2.6	23-Apr-2001	he	Pull up revisions 1.47-1.50 (via patch, requested by hubertf): Support Accton EN2242 and other AmdTek AN985 compatible cards with the tlp(4) driver.
1.39.2.5	09-Aug-2000	castor	Pulling up ic/tulipvar.h diffs between 1.36 and 1.37 ic/tulipreg.h diffs between 1.21 and 1.20 ic/ tulip.c diffs between 1.70 and 1.71 pci/if_tlp_pci.c diffs between 1.43 and 1.46 requested by tls and thorpej, approved by thorpej > date: 2000/08/03 03:07:31; author: castor; state: Exp; lines: +5 -1 Support some differences in the Macronix 98715AEC-C and E chips from the other 98715* series. The MX98715AEC-[C,E] use a different location in the serial eerom for LED control, and programming it with the original location's values caused unpredictable behavior. Also, start integrating fixes where media changes on an adapter under load may fail. There's more work to be done here, but I need to sort out our internal changes a little more carefully.
1.39.2.4	16-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.12 syssrc/sys/dev/pci/if_tlp_pci.c, 1.43 If card is in power state D3 put it into power state D0 so that it will at least work after the next reboot.
1.39.2.3	15-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.11 syssrc/sys/dev/pci/if_tlp_pci.c, 1.42 Avoid printing duplicate colon when card is waked up.
1.39.2.2	15-Jul-2000	tron	Sync with NetBSD-current (pull up approved by thorpej): syssrc/sys/dev/pci/if_epic_pci.c, 1.9 syssrc/sys/dev/pci/if_tlp_pci.c, 1.41 Improve kernel message when card is in power state D3. Instead of... epic0 at pci0 dev 11 function 0 epic0: unable to wake up from ... ... print this: epic0 at pci0 dev 11 function 0: unable to wake up from power state D3
1.39.2.1	25-Jun-2000	sommerfeld	defopt TLP_MATCH_*. pullup approved by thorpej
1.51.2.10	29-Dec-2002	thorpej	Sync with HEAD.
1.51.2.9	18-Oct-2002	nathanw	Catch up to -current.
1.51.2.8	20-Jun-2002	nathanw	Catch up to -current.
1.51.2.7	17-Apr-2002	nathanw	Catch up to -current.
1.51.2.6	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.51.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.51.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.51.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.51.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.51.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.54.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.54.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.54.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.54.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.54.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.73.2.7	11-Dec-2005	christos	Sync with head.
1.73.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.73.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.73.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.73.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.73.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.73.2.1	03-Aug-2004	skrll	Sync with HEAD
1.76.4.1	24-Jul-2005	riz	Pull up revision 1.81 (requested by rpaulo in ticket #5506): If the DEC board name is not known, remove the leading whitespace in sc_name. This closes PR#30458. ok'ed Jason Thorpe.
1.77.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.77.4.1	29-Apr-2005	kent	sync with -current
1.78.2.3	19-Jul-2005	riz	Pull up revision 1.81 (requested by rpaulo in ticket #600): If the DEC board name is not known, remove the leading whitespace in sc_name. This closes PR#30458. ok'ed Jason Thorpe.
1.78.2.2	01-Jul-2005	tron	Pull up revision 1.80 (requested by rpaulo in ticket #493): Add support for ASIX AX88140A/AX88141 chipsets using only the 10/100 MII with inspiration from FreeBSD dc(4) driver. Tested on i386 and macppc. Reviewed by Jason Thorpe and ok'ed by Hubert Feyrer.
1.78.2.1	18-Jun-2005	riz	Pull up revision 1.79 (requested by tron in ticket #451): Also use prom_getether() under NetBSD-sparc(64) if the MAC address retrieved from the PROM starts with ff:ff:ff. The driver now works with the onboard interface in the Tadpole SPARCle.
1.80.2.5	24-Mar-2008	yamt	sync with head.
1.80.2.4	27-Oct-2007	yamt	sync with head.
1.80.2.3	26-Feb-2007	yamt	sync with head.
1.80.2.2	30-Dec-2006	yamt	sync with head.
1.80.2.1	21-Jun-2006	yamt	sync with head.
1.84.6.2	01-Jun-2006	kardel	Sync with head.
1.84.6.1	22-Apr-2006	simonb	Sync with head.
1.84.4.1	09-Sep-2006	rpaulo	sync with head
1.84.2.1	18-Feb-2006	yamt	sync with head.
1.86.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.86.6.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.86.4.2	11-May-2006	elad	sync with head
1.86.4.1	19-Apr-2006	elad	sync with head.
1.86.2.5	03-Sep-2006	yamt	sync with head.
1.86.2.4	11-Aug-2006	yamt	sync with head
1.86.2.3	26-Jun-2006	yamt	sync with head.
1.86.2.2	24-May-2006	yamt	sync with head.
1.86.2.1	01-Apr-2006	yamt	sync with head.
1.90.2.1	19-Jun-2006	chap	Sync with head.
1.91.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.95.2.1	10-Sep-2006	tron	Pull up following revision(s) (requested by rumble in ticket #137): sys/dev/pci/if_tlp_pci.c: revision 1.96 Add quirks for the Phobos GIO Fast Ethernet boards.
1.96.4.2	10-Dec-2006	yamt	sync with head.
1.96.4.1	22-Oct-2006	yamt	sync with head
1.96.2.1	18-Nov-2006	ad	Sync with head.
1.100.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.101.22.1	25-Oct-2007	bouyer	Sync with HEAD.
1.101.18.2	23-Mar-2008	matt	sync with HEAD
1.101.18.1	06-Nov-2007	matt	sync with HEAD
1.101.16.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.101.4.1	23-Oct-2007	ad	Sync with head.
1.102.16.3	29-Jun-2008	mjf	Sync with HEAD.
1.102.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.102.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.102.12.1	24-Mar-2008	keiichi	sync with head.
1.104.4.5	11-Mar-2010	yamt	sync with head
1.104.4.4	16-Sep-2009	yamt	sync with head
1.104.4.3	16-May-2009	yamt	sync with head
1.104.4.2	04-May-2009	yamt	sync with head.
1.104.4.1	16-May-2008	yamt	sync with head.
1.104.2.2	17-Jun-2008	yamt	sync with head.
1.104.2.1	18-May-2008	yamt	sync with head.
1.105.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.106.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.106.6.1	28-Apr-2009	skrll	Sync with HEAD.
1.118.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.118.2.2	30-Oct-2012	yamt	sync with head
1.118.2.1	17-Apr-2012	yamt	sync with head
1.121.10.3	03-Dec-2017	jdolecek	update from HEAD
1.121.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.121.10.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.121.6.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #691): sys/dev/pci/aac_pci.c: revision 1.34 sys/dev/pci/if_sip.c: revision 1.155 sys/dev/pci/if_tlp_pci.c: revision 1.122 sys/dev/pci/if_an_pci.c: revision 1.34 match some more devices.
1.122.2.1	18-May-2014	rmind	sync with head
1.123.6.1	09-Jul-2016	skrll	Sync with HEAD
1.124.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.124.18.1	10-Jun-2019	christos	Sync with HEAD
1.124.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.129.6.1	13-May-2021	thorpej	Sync with HEAD.
1.13	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.12	07-Jul-2020	msaitoh	No functional change: - u_int32_t -> uint32_t - KNF.
1.11	18-Oct-2014	snj	src is too big these days to tolerate superfluous apostrophes. It's "its", people!
1.10	19-Oct-2009	bouyer	branches: 1.10.22; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.9	25-Dec-2007	perry	branches: 1.9.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.8	11-Dec-2005	christos	branches: 1.8.46; 1.8.52; 1.8.56; 1.8.60; merge ktrace-lwp.
1.7	27-Feb-2005	perry	branches: 1.7.4; nuke trailing whitespace
1.6	06-Aug-2001	bouyer	branches: 1.6.20; 1.6.28; 1.6.30; Improvement: allocate one page of dmamem for Rx, Tx and nullbuff instead of 3 different ones. Also pass the rigth seg/nseg to bus_dmamem_free(). Correct types in debug printf's
1.5	03-Aug-2001	bouyer	- kill the carrier detect code. It's not usefull any more and can give false informations on a loaded ethernet segement. - convert to bus_dma(9) (better late than never :) - add proper le32toh/htole32 so that it works on big-endian system (tested on macppc). Close PR kern/10327.
1.4	15-Aug-1998	bouyer	branches: 1.4.24; 1.4.26; Fix typos in last commit.
1.3	15-Aug-1998	bouyer	correct indent.
1.2	17-Nov-1997	thorpej	Cosmetic change.
1.1	17-Oct-1997	bouyer	branches: 1.1.2; Driver for Texas Instruments' thunderland network controller (present in some Compaq products).
1.1.2.1	17-Nov-1997	thorpej	Sync w/ trunk.
1.4.26.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.4.24.1	24-Aug-2001	nathanw	Catch up with -current.
1.6.30.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.6.28.1	29-Apr-2005	kent	sync with -current
1.6.20.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.4.1	21-Jan-2008	yamt	sync with head
1.8.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.8.56.1	26-Dec-2007	ad	Sync with head.
1.8.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.8.46.1	09-Jan-2008	matt	sync with HEAD
1.9.10.1	11-Mar-2010	yamt	sync with head
1.10.22.1	03-Dec-2017	jdolecek	update from HEAD
1.18	07-Jul-2020	msaitoh	No functional change: - u_int32_t -> uint32_t - KNF.
1.17	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.16	02-Feb-2012	tls	branches: 1.16.6; 1.16.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.15	19-Nov-2011	tls	branches: 1.15.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.14	19-Oct-2009	bouyer	branches: 1.14.12; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.13	01-Jun-2008	tsutsui	Split device_t/softc, with misc cosmetic changes.
1.12	11-Dec-2005	christos	branches: 1.12.70; 1.12.72; 1.12.74; 1.12.76; merge ktrace-lwp.
1.11	23-Jan-2005	dan	attach tl(4) as an event source to rnd(4) patches from Sean Davis in PR/28556 (slightly modified to actually catch our own interrupts)
1.10	10-Nov-2003	wiz	branches: 1.10.6; 1.10.8; Spell address with two d's. Inspired by similar changes in OpenBSD, originating from Jonathon Gray and forwarded by jmc@openbsd.
1.9	02-Nov-2003	wiz	deferred, not deffered. Inspired by Tom Cosgrove.
1.8	30-Sep-2003	thorpej	New generic I2C framework. Supports bit-bang and "intelligent" I2C interface controllers (of varying intelligence levels). Contributed by Wasabi Systems, Inc. Primarily written by Steve Woodford, with some modification by me.
1.7	20-Sep-2001	bouyer	branches: 1.7.18; - Allocate control structures at attach time rather than at init time. Avoids using bus_dmamem_alloc/bus_dmamem_map at interrupt time. Should fix PRs kern/13924 and kern/13979 from dive@endersgame.net - while I'm there convert to use ether_ioctl.
1.6	06-Aug-2001	bouyer	branches: 1.6.2; Improvement: allocate one page of dmamem for Rx, Tx and nullbuff instead of 3 different ones. Also pass the rigth seg/nseg to bus_dmamem_free(). Correct types in debug printf's
1.5	03-Aug-2001	bouyer	- kill the carrier detect code. It's not usefull any more and can give false informations on a loaded ethernet segement. - convert to bus_dma(9) (better late than never :) - add proper le32toh/htole32 so that it works on big-endian system (tested on macppc). Close PR kern/10327.
1.4	23-Mar-2000	thorpej	branches: 1.4.6; 1.4.8; New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.3	11-Jan-1999	tron	branches: 1.3.8; Because memory mapped access doesn't fail only with the docking station for the TI TravelMate 5000 but also with at least one Compaq NIC we prefer I/O mapped access again. Fixes PR kern/6690.
1.2	15-Aug-1998	bouyer	branches: 1.2.4; correct indent.
1.1	11-Aug-1998	thorpej	Adapt to the new MII code. This required splitting some structures into if_tlvar.h, since the ThunderLAN PHY driver needs to know some things about it's parent (and since it can only attach to a ThunderLAN chip, this is the easiest way).
1.2.4.1	11-Dec-1998	kenh	The beginnings of interface detach support. Still some bugs, but mostly works for me. This work was originally by Bill Studenmund, and cleaned up by me.
1.3.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.8.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.8.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.4.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.4.6.1	24-Aug-2001	nathanw	Catch up with -current.
1.6.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.7.18.4	24-Jan-2005	skrll	Sync with HEAD.
1.7.18.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.18.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.18.1	03-Aug-2004	skrll	Sync with HEAD
1.10.8.1	29-Apr-2005	kent	sync with -current
1.10.6.1	16-Apr-2005	tron	Pull up revision 1.11 (requested by dan in ticket #1117): attach tl(4) as an event source to rnd(4) patches from Sean Davis in PR/28556 (slightly modified to actually catch our own interrupts)
1.12.76.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.12.74.2	11-Mar-2010	yamt	sync with head
1.12.74.1	04-May-2009	yamt	sync with head.
1.12.72.1	04-Jun-2008	yamt	sync with head
1.12.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.14.12.1	17-Apr-2012	yamt	sync with head
1.15.2.1	18-Feb-2012	mrg	merge to -current.
1.16.24.1	06-Jun-2015	skrll	Sync with HEAD
1.16.6.1	03-Dec-2017	jdolecek	update from HEAD
1.76	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.75	10-Feb-2024	andvar	s/alloted/allotted/ in comments.
1.74	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.73	10-Mar-2020	thorpej	On NetBSD it's spelled "__NO_STRICT_ALIGNMENT". Adjust txp_rx_reclaim() accordingly and structure it like other NetBSD drivers so as to re-use the original Rx buffer rather than doing a needless free/alloc cycle. Note this happened to work previously on my Qube2 because IP, TCP, etc. perform their own alignment checks and react accordingly. However, it's not clear that ALL protocols do this yet, so it's better to just do the safe thing for now.
1.72	10-Mar-2020	thorpej	- Add support for MBUFTRACE. - Fix an mbuf leak I introduced in rev 1.68 (failed to increment "cnt" variable in txp_start(), which subsequently caused txp_tx_reclaim() to never run). - Use the context address pointer in the Tx descriptor to stash the pointer to the Tx job software state in a manner like that which is used for Rx software state.
1.71	09-Mar-2020	thorpej	Add an assertion.
1.70	09-Mar-2020	thorpej	Now that we're allocating rx software state from a fixed pool, we don't have to round-trip pointers in a sketchy way -- we can round- trip indices.
1.69	09-Mar-2020	thorpej	Allocate a static set of receive software state descriptors at attach time, and create a simple allocator for them for use by the rx reclaim handlers. Don't create and destroy DMA maps for receoive buffers in the interrupt handlers -- just use the maps that are created at attach time.
1.68	08-Mar-2020	thorpej	Give each Tx DMA map 16 DMA segments rather than the previous absurdly large number, and structure the loop in txp_start() similarly to other drivers (e.g. ste_start() in the ste(4) driver). Similar in spirit to OpenBSD's rev 1.126, but implemented a bit differently.
1.67	08-Mar-2020	thorpej	Upgrade to newer firmware needed by some txp(4) variants. From FreeBSD via OpenBSD.
1.66	08-Mar-2020	thorpej	Follow the example of the MII code and don't register the half-duplex versions of each media type.
1.65	08-Mar-2020	thorpej	- Use BUS_ADDR_{LO,HI}32(), and rather than forcing the descriptor DMA addresses into a 64-bit value (and thus using 64-bit shifts when unnecessary), just reference the address from the DMA map directly. - Add some missing byte-swaps.
1.64	29-Feb-2020	thorpej	Use the 64-bit PCI DMA tag, if available.
1.63	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.62	06-Dec-2019	maxv	branches: 1.62.2; localify
1.61	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.60	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.59	09-Jul-2019	msaitoh	branches: 1.59.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.58	20-Jun-2019	msaitoh	Fix panic on attach. This bug was added in rev. 1.56. I think it's not required to call txp_set_filter() in the attach function.
1.57	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.56	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.55	23-May-2019	msaitoh	No functional change: - Change ac(was arpcom) to ec(ethercom) - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.54	07-May-2019	msaitoh	- Use bus_dmamem_unmap() correctly in txp_dma_free() to prevent panic. The code was wrong since rev. 1.1. This panic was found by adding KASSERT in uvm_map.c rev. 1.351. This bug may be related to PR kern/26204. - Use aprint_normal() for non-error message.
1.53	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.52	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.51	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.50	25-Jul-2018	msaitoh	Initialize some members in a mbuf which is on stack.
1.49	26-Jun-2018	msaitoh	branches: 1.49.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.48	26-Sep-2017	knakahara	branches: 1.48.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.47	15-Dec-2016	ozaki-r	branches: 1.47.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.46	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.45	14-Jul-2016	msaitoh	branches: 1.45.2; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.44	14-Jul-2016	msaitoh	- Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.43	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.42	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.41	16-Jun-2014	msaitoh	branches: 1.41.4; IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.40	29-Mar-2014	christos	branches: 1.40.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.39	27-Oct-2012	chs	branches: 1.39.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.38	13-Nov-2010	uebayasi	branches: 1.38.8; 1.38.14; 1.38.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.37	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.36	19-Jan-2010	pooka	branches: 1.36.2; 1.36.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.35	27-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Compile test only.
1.34	18-Apr-2009	tsutsui	Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.33	18-Mar-2009	cegger	bcopy -> memcpy
1.32	18-Mar-2009	cegger	bzero -> memset
1.31	18-Mar-2009	cegger	bcmp -> memcmp
1.30	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.29	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.28	16-Feb-2009	cegger	fix media priorities: IEEE 802.3 Annex 28B.3 specifies the following relative priorities of the technologies supported by 802.3 Selector Field value: 1000BASE-T full duplex 1000BASE-T 100BASE-T2 full duplex 100BASE-TX full duplex 100BASE-T2 100BASE-T4 100BASE-TX 10BASE-T full duplex 10BAST-T Our drivers give 100BASE-T4 a higher priority than 100BASE-TX full duplex. Fix this. This patch is based on changes in FreeBSD and OpenBSD. Patch presented on tech-kern and tech-net: http://mail-index.netbsd.org/tech-kern/2009/02/15/msg004397.html http://mail-index.netbsd.org/tech-net/2009/02/15/msg001064.html got no comments, no objections.
1.27	07-Nov-2008	dyoung	branches: 1.27.4; * Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.26	10-Apr-2008	cegger	branches: 1.26.4; 1.26.10; 1.26.12; 1.26.14; 1.26.18; use aprint_*_dev and device_xname
1.25	09-Apr-2008	dyoung	Use device_t and accessors. Use cfdata_t and __arraycount(). Miscellaneous cosmetic changes.
1.24	09-Apr-2008	dyoung	s/pcimatch/match/ to avoid conflict with a global name.
1.23	07-Feb-2008	dyoung	branches: 1.23.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.22	19-Oct-2007	ad	branches: 1.22.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.21	01-Sep-2007	dyoung	branches: 1.21.4; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.20	09-Jul-2007	ad	branches: 1.20.2; 1.20.6; 1.20.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.19	04-Mar-2007	christos	branches: 1.19.2; 1.19.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.18	16-Nov-2006	christos	branches: 1.18.4; __unused removal on arguments; approved by core.
1.17	04-Nov-2006	tsutsui	Remove vtophys (in comment).
1.16	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.15	14-Apr-2006	christos	branches: 1.15.8; 1.15.10; Coverity CID 2314: Free memory if we fail to allocate rings.
1.14	28-Dec-2005	christos	branches: 1.14.4; 1.14.6; 1.14.8; 1.14.10; 1.14.12; We don't have access to in_cksum() if no INET.
1.13	11-Dec-2005	christos	merge ktrace-lwp.
1.12	18-Jun-2005	chs	branches: 1.12.2; adapt to const/volatile changes.
1.11	02-May-2005	yamt	split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.10	27-Feb-2005	perry	branches: 1.10.2; nuke trailing whitespace
1.9	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition
1.8	30-Oct-2004	thorpej	branches: 1.8.4; 1.8.6; When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.7	01-Jul-2004	is	disestablish the interupt if we fail to (among other things) download the firmware successfully. This prevents the kernel from crashing later at interupt enable time, if firmware download fails like with my 3c990B card.
1.6	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.5	27-Oct-2003	thorpej	branches: 1.5.2; - Change callout_setfunc() to require that the callout handle is already initialized. Update the txp(4) to compensate. - Statically initialize the TCP timer callout handles in the tcpcb template. We still use callout_setfunc(), but that call is now much less expensive. Add a comment that the compiler is likely to unroll the loop (so don't sweat that it's there).
1.4	20-Aug-2003	drochner	make the vlan stuff work
1.3	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.2	07-Jul-2003	drochner	some cleanup/fixes: -get boot messages to NetBSD's standards (ie at least 3 lines) -get some card information from PCI ID and subsystem ID -fix media handling and link reporting for fiber cards -port cksum offloading to NetBSD and enable as far as appropriate (TCP/UDP transmit offloading still disabled, as in OpenBSD)
1.1	01-Jul-2003	drochner	branches: 1.1.2; VERY quick and VERY dirty port of the OpenBSD "txp" driver for 3Com 3C(R)990 cards. No TCP/UDP nor IPSEC offloading for now. The ifmedia part needs some work, but I only have a 100-FX card; hopefully someone will clean this up.
1.1.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.1.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.1	03-Aug-2004	skrll	Sync with HEAD
1.5.2.1	10-Jul-2004	tron	branches: 1.5.2.1.2; Pull up revision 1.7 (requested by is in ticket #626): disestablish the interupt if we fail to (among other things) download the firmware successfully. This prevents the kernel from crashing later at interupt enable time, if firmware download fails like with my 3c990B card.
1.5.2.1.2.1	24-Jan-2005	he	Pull up revision 1.8 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.8.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.2.1	28-Oct-2005	jmc	Pullup (requested by chs in ticket #912) Fix txp cards so this compiles and enable in GENERIC.
1.12.2.5	11-Feb-2008	yamt	sync with head.
1.12.2.4	27-Oct-2007	yamt	sync with head.
1.12.2.3	03-Sep-2007	yamt	sync with head.
1.12.2.2	30-Dec-2006	yamt	sync with head.
1.12.2.1	21-Jun-2006	yamt	sync with head.
1.14.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.14.10.1	19-Apr-2006	elad	sync with head.
1.14.8.1	24-May-2006	yamt	sync with head.
1.14.6.1	22-Apr-2006	simonb	Sync with head.
1.14.4.1	09-Sep-2006	rpaulo	sync with head
1.15.10.2	10-Dec-2006	yamt	sync with head.
1.15.10.1	22-Oct-2006	yamt	sync with head
1.15.8.1	18-Nov-2006	ad	Sync with head.
1.18.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.19.4.1	11-Jul-2007	mjf	Sync with head.
1.19.2.3	23-Oct-2007	ad	Sync with head.
1.19.2.2	09-Oct-2007	ad	Sync with head.
1.19.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.20.8.2	23-Mar-2008	matt	sync with HEAD
1.20.8.1	06-Nov-2007	matt	sync with HEAD
1.20.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.20.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.20.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.21.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.22.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.23.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.23.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.26.18.1	21-Apr-2010	matt	sync to netbsd-5
1.26.14.1	01-May-2009	snj	Pull up following revision(s) (requested by cegger in ticket #474): sys/dev/mii/inphy.c: revision 1.51 sys/dev/mii/iophy.c: revision 1.35 sys/dev/mii/nsphy.c: revision 1.56 sys/dev/mii/rlphy.c: revision 1.25 sys/dev/mii/ukphy_subr.c: revision 1.11 sys/dev/pci/if_txp.c: revision 1.28 fix media priorities: IEEE 802.3 Annex 28B.3 specifies the following relative priorities of the technologies supported by 802.3 Selector Field value: 1000BASE-T full duplex 1000BASE-T 100BASE-T2 full duplex 100BASE-TX full duplex 100BASE-T2 100BASE-T4 100BASE-TX 10BASE-T full duplex 10BAST-T Our drivers give 100BASE-T4 a higher priority than 100BASE-TX full duplex. Fix this. This patch is based on changes in FreeBSD and OpenBSD. Patch presented on tech-kern and tech-net: http://mail-index.netbsd.org/tech-kern/2009/02/15/msg004397.html http://mail-index.netbsd.org/tech-net/2009/02/15/msg001064.html got no comments, no objections.
1.26.12.3	28-Apr-2009	skrll	Sync with HEAD.
1.26.12.2	03-Mar-2009	skrll	Sync with HEAD.
1.26.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.26.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.26.4.3	11-Aug-2010	yamt	sync with head.
1.26.4.2	11-Mar-2010	yamt	sync with head
1.26.4.1	04-May-2009	yamt	sync with head.
1.27.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.36.4.2	05-Mar-2011	rmind	sync with head
1.36.4.1	30-May-2010	rmind	sync with head
1.36.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.38.18.3	03-Dec-2017	jdolecek	update from HEAD
1.38.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.38.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.38.14.1	09-Nov-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #1189): sys/dev/pci/if_wm.c: revision 1.270 sys/dev/mii/inphy.c: revision 1.53 sys/dev/mii/glxtphy.c: revision 1.24 sys/dev/mii/tlphy.c: revision 1.62 sys/dev/mii/iophy.c: revision 1.37 sys/dev/mii/brgphy.c: revision 1.70 sys/dev/mii/ihphy.c: revision 1.8 sys/dev/mii/bmtphy.c: revision 1.31 sys/dev/mii/urlphy.c: revision 1.30 sys/dev/mii/makphy.c: revision 1.40 sys/dev/mii/qsphy.c: revision 1.48 sys/dev/mii/igphy.c: revision 1.23 sys/dev/mii/nsphy.c: revision 1.58 sys/dev/mii/mvphy.c: revision 1.10 sys/dev/pci/if_txp.c: revision 1.41 sys/dev/mii/nsphy.c: revision 1.59 sys/dev/mii/rlphy.c: revision 1.28 sys/dev/mii/icsphy.c: revision 1.49 sys/dev/mii/rlphy.c: revision 1.29 sys/dev/mii/lxtphy.c: revision 1.49 sys/dev/mii/ciphyreg.h: revision 1.5 sys/dev/mii/nsphyter.c: revision 1.38 sys/dev/mii/sqphy.c: revision 1.50 sys/dev/mii/gentbi.c: revision 1.26 sys/dev/mii/gentbi.c: revision 1.27 sys/dev/mii/tqphy.c: revision 1.39 sys/dev/mii/ikphy.c: revision 1.10 sys/dev/mii/dmphy.c: revision 1.35 sys/dev/mii/amhphy.c: revision 1.20 sys/dev/mii/acphy.c: revision 1.24 sys/dev/mii/ciphy.c: revision 1.25 sys/dev/mii/brgphyreg.h: revision 1.8 sys/dev/mii/ukphy_subr.c: revision 1.12 sys/dev/ic/rtl80x9.c: revision 1.16 sys/arch/mips/adm5120/dev/if_admsw.c: revision 1.12 sys/dev/pci/if_kse.c: revision 1.28 sys/dev/mii/ukphy_subr.c: revision 1.13 sys/dev/mii/mii.h: revision 1.18 sys/dev/mii/gphyter.c: revision 1.29 No functional change: - Fix typo. - Remove trailing white spaces. - Capitalize comments. - Tabify. - KNF. IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.38.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.38.8.1	30-Oct-2012	yamt	sync with head
1.39.2.1	18-May-2014	rmind	sync with head
1.40.2.1	10-Aug-2014	tls	Rebase.
1.41.4.4	05-Feb-2017	skrll	Sync with HEAD
1.41.4.3	05-Oct-2016	skrll	Sync with HEAD
1.41.4.2	09-Jul-2016	skrll	Sync with HEAD
1.41.4.1	19-Mar-2016	skrll	Sync with HEAD
1.45.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.47.8.3	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.47.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.47.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.48.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.48.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.49.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.49.2.1	10-Jun-2019	christos	Sync with HEAD
1.59.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.62.2.1	29-Feb-2020	ad	Sync with head.
1.11	10-Mar-2020	thorpej	Protect against multiple inclusion.
1.10	09-Mar-2020	thorpej	Allocate a static set of receive software state descriptors at attach time, and create a simple allocator for them for use by the rx reclaim handlers. Don't create and destroy DMA maps for receoive buffers in the interrupt handlers -- just use the maps that are created at attach time.
1.9	08-Mar-2020	thorpej	Give each Tx DMA map 16 DMA segments rather than the previous absurdly large number, and structure the loop in txp_start() similarly to other drivers (e.g. ste_start() in the ste(4) driver). Similar in spirit to OpenBSD's rev 1.126, but implemented a bit differently.
1.8	08-Mar-2020	thorpej	Upgrade to newer firmware needed by some txp(4) variants. From FreeBSD via OpenBSD.
1.7	08-Mar-2020	thorpej	- Use BUS_ADDR_{LO,HI}32(), and rather than forcing the descriptor DMA addresses into a 64-bit value (and thus using 64-bit shifts when unnecessary), just reference the address from the DMA map directly. - Add some missing byte-swaps.
1.6	27-Oct-2012	chs	branches: 1.6.38; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.5	10-Apr-2008	cegger	branches: 1.5.38; 1.5.48; use aprint_*_dev and device_xname
1.4	04-Mar-2007	christos	branches: 1.4.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	11-Dec-2005	christos	branches: 1.3.26; merge ktrace-lwp.
1.2	07-Jul-2003	drochner	branches: 1.2.16; some cleanup/fixes: -get boot messages to NetBSD's standards (ie at least 3 lines) -get some card information from PCI ID and subsystem ID -fix media handling and link reporting for fiber cards -port cksum offloading to NetBSD and enable as far as appropriate (TCP/UDP transmit offloading still disabled, as in OpenBSD)
1.1	01-Jul-2003	drochner	branches: 1.1.2; VERY quick and VERY dirty port of the OpenBSD "txp" driver for 3Com 3C(R)990 cards. No TCP/UDP nor IPSEC offloading for now. The ifmedia part needs some work, but I only have a 100-FX card; hopefully someone will clean this up.
1.1.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.1	03-Aug-2004	skrll	Sync with HEAD
1.2.16.1	03-Sep-2007	yamt	sync with head.
1.3.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.4.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.48.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.5.38.1	30-Oct-2012	yamt	sync with head
1.6.38.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.89	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.88	29-Jun-2024	riastradh	branches: 1.88.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.87	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.86	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.85	17-Sep-2022	thorpej	Eliminate use of IFF_OACTIVE.
1.84	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.83	22-Jan-2022	martin	Explicitly cast a __MASK(48) to bus_addr_t (for 32bit bus_addr_t archs)
1.82	22-Jan-2022	skrll	Ensure bus_dmatag_subregion is called with an inclusive max_addr everywhere.
1.81	24-Jul-2021	andvar	Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889. Also fixed some additional typos in comments, found on review of same files or typos.
1.80	21-Mar-2020	thorpej	branches: 1.80.8; Use the 64-bit DMA tag, if available, with some constraints: - The data buffers can come from anywhere in the 64-bit region, but the upper 16 bits of the data buffer address is stored in a single register, so just treat that as 0 and create a 48-bit restricted DMA tag. - The descriptor address registers share a single register for the upper 32-bits, so enforce a 4G boundary when allocating memory for the descriptors.
1.79	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.78	27-Dec-2019	msaitoh	branches: 1.78.2; s/opration/operation/
1.77	21-Nov-2019	msaitoh	- Set duplex correctly when user setting is not IFM_AUTO. - When the link is up, set VGE_DIAGCTL not from user media setting but from the current active link status.
1.76	19-Nov-2019	msaitoh	Fixes a bug that "ifmedia vge0 media 1000baseT-FDX" causes device timeout. If the interface's media is NOT in the "best" mode (i.e. other than IFM_AUTO), control VGE_DIAGCTL_GMII bit.
1.75	08-Oct-2019	msaitoh	Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.74	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.73	09-Jul-2019	msaitoh	branches: 1.73.2; Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.72	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.71	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.70	23-May-2019	msaitoh	-No functional change: - Simplify struct ethercom's pointer near ETHER_FIRST_MULTI(). - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.69	11-Apr-2019	msaitoh	Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.68	05-Mar-2019	msaitoh	Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.67	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.66	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.65	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.64	18-Jul-2018	sevan	Be consistent among ethernet drivers on the convention for printing ethernet addresses. NFC
1.63	26-Jun-2018	msaitoh	branches: 1.63.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.62	14-Jan-2018	maxv	branches: 1.62.2; Dedup. m_defrag is already a common function, no need to reimplement it there. Meanwhile this should fix two bugs (that I couldn't investigate more than that since I don't have this hardware): the mbuf passed to vge_m_defrag was leaked, and the tags were not copied in the returned mbuf.
1.61	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.60	15-Dec-2016	ozaki-r	branches: 1.60.8; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.59	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.58	10-Jun-2016	ozaki-r	branches: 1.58.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.57	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.56	29-Mar-2014	christos	branches: 1.56.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.55	13-Sep-2013	martin	Remove unused variable
1.54	30-Mar-2013	christos	branches: 1.54.4; remove trailing whitespace
1.53	22-Jul-2012	matt	branches: 1.53.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.52	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.51	05-Apr-2010	joerg	branches: 1.51.8; 1.51.12; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.50	19-Jan-2010	pooka	branches: 1.50.2; 1.50.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.49	07-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested on VT6122.
1.48	16-May-2009	tsutsui	Split device_t/softc split, tested on VT6212L.
1.47	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.46	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.45	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.44	09-Feb-2009	nonaka	fix comment.
1.43	26-Nov-2008	joerg	branches: 1.43.4; - ifconfig down support - ALTQ support
1.42	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.41	10-Apr-2008	cegger	branches: 1.41.4; 1.41.10; 1.41.12; 1.41.14; use aprint_*_dev and device_xname
1.40	07-Feb-2008	dyoung	branches: 1.40.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.39	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.38	19-Oct-2007	ad	branches: 1.38.2; 1.38.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.37	01-Sep-2007	dyoung	branches: 1.37.4; Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.36	09-Jul-2007	ad	branches: 1.36.2; 1.36.6; 1.36.8; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.35	12-Jun-2007	tsutsui	Fix one more bug I introduced in rev 1.21: Clear sc->sc_rx_consumed before allocating RX mbufs by vge_newbuf() in vge_intr() otherwise vge_newbuf() might get KASSERT() failure on DIAGNOSTIC kernel when the interface is reinitialized during RX.
1.34	04-Mar-2007	christos	branches: 1.34.2; 1.34.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.33	01-Dec-2006	tsutsui	branches: 1.33.2; 1.33.4; 1.33.8; Free resouces on failure in vge_allocmem().
1.32	01-Dec-2006	tsutsui	- vge_probe() -> vge_match() - KNF, unwrap lines
1.31	26-Nov-2006	tsutsui	Fix stupid bugs found on sparc64: - checking if dmamap != NULL is not valid because dmamap is not cleared in bus_dmamap_unload(9) - no need check RX mbufs and call m_freem() and bus_dmamap_unload() in vge_init() since it's done in prior vge_stop() Now vge(4) works fine on Ultra5.
1.30	26-Nov-2006	tsutsui	Include <sys/device.h> for struct device.
1.29	26-Nov-2006	tsutsui	- prepare inline functions to set 48 bit DMA address which avoid 64 bit arithmetic on ILP32 (sizeof(bus_addr_t) == 4, exactly) hosts - prepare and use VGE_PREV_TXDESC() macro - use VGE_[TR]XDESCSYNC() calls more efficiently - wrap a sanity check against VGE_RDSTS_OWN in vge_newbuf() with #ifdef DIAGNOSTIC since it should not happen - use sc->sc_tx_free (number of free TX descs) to check if TX packets are queued or sent - call vge_start() only if the interupt is actually handled by this driver - some misc cosmetics
1.28	26-Nov-2006	tsutsui	Introduce a new VGE_RX_BUFSIZE macro which represent default RX mbuf length and define VGE_RX_PAD (which is 4 bytes) for ETHER_ALIGN (2 bytes) padding only in !__NO_STRICT_ALIGNMENT case to avoid confusion. As per comments from Murata Shuuichirou in PR kern/31323. Tested on i386 and macppc.
1.27	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.26	04-Nov-2006	tsutsui	Rather than copying m_pkthdr.csum_flags in m_defrag() (which was committed in rev 1.22), save it before calling m_defrag(). I haven't confirmed whether the m_pkthdr.csum_flags is preserved during m_defrag(), but the previous way sometimes makes vge(4) chip mad...
1.25	04-Nov-2006	tsutsui	Return a proper error from vge_encap() if m_defrag() fails.
1.24	01-Nov-2006	tsutsui	Fix another botch in rev 1.14: vge_encap() should bail out if there is not enough free TX descriptor _OR_ TX descriptor is still owned by the chip. Anyway, we already check if the TX descriptor already has an mbuf to be sent in vge_start() so no need to bother to check sc_tx_free and VGE_TDSTS_OWN in the descriptor in normal case, then make it use KASSERT(9) or wrap with #ifdef DIAGNOSTIC. XXX1: I don't know why original FreeBSD driver checks if a number of free TX descriptors is more than two, not zero. XXX2: Is it better to check a number of free descriptors in vge_start() like other our drivers rather than mbuf chain for each descriptor?
1.23	01-Nov-2006	tsutsui	Remove a debug statement which was slipped in rev 1.21.
1.22	01-Nov-2006	tsutsui	Also copy csum_flags and csum_data in vge_m_defrag(). Fix checksum error problem on sending fragmented large packets, which was introduced in rev 1.14. BTW, should we have m_defrag() in MI for other drivers?
1.21	31-Oct-2006	tsutsui	Make vge(4) look a NetBSD native driver like pcn(4) or epic(4) etc: - merge if_vgevar.h into if_vge.c since no other file refers it - rename some macro (VGE_TX_DESC_CNT -> VGE_NTXDESC etc.) and structs - change prefixes of structure members to represent parents - put TX and RX descriptors into single struct vge_control_data and allocate DMA memory at once - no need to specify BUS_DMA_ALLOCNOW - define appropriate macro for offsets and addresses of DMA descriptors - define struct vge_txsoft and vge_rxsoft, and put common data for each descriptor into them - remove struct vge_list_data and move its members into struct vge_softc - remove #ifdef DEVICE_POLLING code we don't support - merge vge_[tr]x_list_init() functions into vge_init() - use aprint_error() to print errors - put splnet(9) where appropriate and remove unused VGE_{LOCK,UNLOCK}() macro - clear TX timeout only if there is no pending descriptor - check dm_nsegs properly otherwise padding short packets could fail - prepare zero'ed DMA memory to pad short packets rather than putting random data - fix a wrong VGE_TXDESCSYNC() usage which should be VGE_TXFRAGSYNC() Tested on my i386 which is my development machine. (more bus_dmamap_sync(9) cleanup will be done later)
1.20	21-Oct-2006	tsutsui	Fix hardware VLAN tagging support on vge(4): - On TX, vge(4) seems to assume that tags are written in little endian. We already use htole32() to write values into descriptors, so extra byteswap by htons() is not needed there. - On RX, vge(4) seems to store tags in network byteorder. We have to swap byteorder regardless of host's byteorder (i.e. we have to use bswap16() rather than ntohs()) because we already use le32toh() to read values from descriptors. Anyway, no need to use htons()/ntohs() because there is no stream data. Tested on both i386 and macppc, and OK'ed by Pavel Cahyna.
1.19	17-Oct-2006	tsutsui	Fix a typo in comment.
1.18	15-Oct-2006	tsutsui	In vge_newbuf(), don't adjust packet length twice by m_adj(9) in case of __NO_STRICT_ALIGNMENT. We can adjust m_data directly since it's allocated by MCLGET(9) just above. (though the length is not used anyway)
1.17	15-Oct-2006	tsutsui	Umm, remove a DELAY() added in previous. It seems to cause another problem.
1.16	15-Oct-2006	tsutsui	On sgimips (R5000 O2), it seems that the first EEPROM read data right after vge_reset() could be corrupted. For workaround, add a dummy EEPROM read in vge_reset() so that MAC address is properly set on the machine. While here, add a DELAY() in busy loop in vge_read_eeprom().
1.15	14-Oct-2006	tsutsui	- use MGETHDR() and MCLGET() rather than m_gethdr() and m_clget() - KNF, ANSIfy, remove register decls - some more cosmetics
1.14	14-Oct-2006	tsutsui	- call bus_dmamap_sync() properly (sigh) - merge vge_dma_map_rx_desc() and vge_dma_map_tx_desc() into callers - declare DMA descs volatile - remove unused macro/declarations - bzero -> memset - some more cosmetics
1.13	14-Oct-2006	tsutsui	- use correct dmamap size for TX/RX DMA descriptors - use __NO_STRICT_ALIGNMENT directly rather than local VGE_FIXUP_RX - no need to use BUS_DMA_ALLOCNOW - remove unneeded members from softc XXX: Is vge_fixup_rx() really more efficient than memmove(9), XXX: or allocating a new buffer and memcpy(9) into it? XXX: Anyway, vge(4) is not recommended for non-x86 hosts at all XXX: because it requires copying buffers by CPU on RX.
1.12	14-Oct-2006	tsutsui	Umm, set Ethernet address into softc properly.
1.11	14-Oct-2006	tsutsui	- fix eeprom read function so that MAC address can be read correctly on big endian machines - add some missed byteswap functions for DMA descriptors
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	08-Mar-2006	lukem	branches: 1.9.10; 1.9.12; Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.8	24-Dec-2005	perry	branches: 1.8.4; 1.8.6; 1.8.8; 1.8.10; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.7	21-Nov-2005	martin	This driver does only access the bus_space_tag via bus_space_* functions, so there is no need to map it linear. From Garrett D'Amore in PR kern/32137.
1.6	08-Sep-2005	christos	branches: 1.6.6; PR/31130: FUKAUMI Naoki: Fix vge(4) SIOC{ADD,DEL}MULTI handling
1.5	02-May-2005	yamt	branches: 1.5.2; split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.4	05-Mar-2005	jdolecek	branches: 1.4.2; 1.4.4; 1.4.6; g/c debugging printf
1.3	05-Mar-2005	jdolecek	modify vge_m_defrag() to support > MCLBYTES chains, by allocating appropriate number of mbuf clusters; also don't free the original mbuf chain simplify vge_encap() and make error reporting more consistent normalize vge_start() (including use of IFQ_* macros), and handle the case when the input mbuf chain is defragmented JUMBO packets seem to work fine now
1.2	27-Feb-2005	perry	branches: 1.2.2; nuke trailing whitespace
1.1	20-Feb-2005	jdolecek	Add device driver support for the VIA Networking Technologies VT6122 gigabit ethernet chip and integrated 10/100/1000 copper PHY. Obtained from: FreeBSD
1.2.2.5	11-Dec-2005	christos	Sync with head.
1.2.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.3	08-Mar-2005	skrll	Sync with HEAD.
1.2.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.1	27-Feb-2005	skrll	file if_vge.c was added on branch ktrace-lwp on 2005-03-04 16:45:19 +0000
1.4.6.2	29-Apr-2005	kent	sync with -current
1.4.6.1	05-Mar-2005	kent	file if_vge.c was added on branch kent-audio2 on 2005-04-29 11:29:06 +0000
1.4.4.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.4.1	05-Mar-2005	yamt	file if_vge.c was added on branch yamt-km on 2005-03-19 08:35:11 +0000
1.4.2.1	11-Sep-2005	tron	Pull up following revision(s) (requested by christos in ticket #756): sys/dev/pci/if_vge.c: revision 1.6 PR/31130: FUKAUMI Naoki: Fix vge(4) SIOC{ADD,DEL}MULTI handling
1.5.2.6	11-Feb-2008	yamt	sync with head.
1.5.2.5	21-Jan-2008	yamt	sync with head
1.5.2.4	27-Oct-2007	yamt	sync with head.
1.5.2.3	03-Sep-2007	yamt	sync with head.
1.5.2.2	30-Dec-2006	yamt	sync with head.
1.5.2.1	21-Jun-2006	yamt	sync with head.
1.6.6.1	22-Nov-2005	yamt	sync with head.
1.8.10.1	19-Apr-2006	elad	sync with head.
1.8.8.1	13-Mar-2006	yamt	sync with head.
1.8.6.1	22-Apr-2006	simonb	Sync with head.
1.8.4.1	09-Sep-2006	rpaulo	sync with head
1.9.12.2	10-Dec-2006	yamt	sync with head.
1.9.12.1	22-Oct-2006	yamt	sync with head
1.9.10.2	12-Jan-2007	ad	Sync with head.
1.9.10.1	18-Nov-2006	ad	Sync with head.
1.33.8.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.33.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.33.2.1	18-Jun-2007	liamjfoy	Pull up following revision(s) (requested by tsutsui in ticket #731): sys/dev/pci/if_vge.c: revision 1.35 Fix one more bug I introduced in rev 1.21: Clear sc->sc_rx_consumed before allocating RX mbufs by vge_newbuf() in vge_intr() otherwise vge_newbuf() might get KASSERT() failure on DIAGNOSTIC kernel when the interface is reinitialized during RX.
1.34.4.1	11-Jul-2007	mjf	Sync with head.
1.34.2.4	23-Oct-2007	ad	Sync with head.
1.34.2.3	09-Oct-2007	ad	Sync with head.
1.34.2.2	15-Jul-2007	ad	Sync with head.
1.34.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.36.8.2	23-Mar-2008	matt	sync with HEAD
1.36.8.1	06-Nov-2007	matt	sync with HEAD
1.36.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.36.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.36.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.37.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.38.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.38.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.40.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.40.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.41.14.4	07-Dec-2008	bouyer	Pull up following revision(s) (requested by joerg in ticket #143): sys/dev/pci/if_vge.c: revision 1.43 via patch - ifconfig down support - ALTQ support
1.41.14.3	30-Nov-2008	bouyer	Back out ticket 143, it doesn't build.
1.41.14.2	30-Nov-2008	bouyer	Apply patch, requested by joerg in ticket 143: Fix build failure, vge_stop() needs 2 arguments in netbsd-5.
1.41.14.1	29-Nov-2008	bouyer	Pull up following revision(s) (requested by joerg in ticket #143): sys/dev/pci/if_vge.c: revision 1.43 - ifconfig down support - ALTQ support
1.41.12.2	03-Mar-2009	skrll	Sync with HEAD.
1.41.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.41.10.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.41.4.5	11-Aug-2010	yamt	sync with head.
1.41.4.4	11-Mar-2010	yamt	sync with head
1.41.4.3	16-Sep-2009	yamt	sync with head
1.41.4.2	16-May-2009	yamt	sync with head
1.41.4.1	04-May-2009	yamt	sync with head.
1.43.4.2	23-Jul-2009	jym	Sync with HEAD.
1.43.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.4.1	30-May-2010	rmind	sync with head
1.50.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.51.12.1	18-Feb-2012	mrg	merge to -current.
1.51.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.51.8.2	30-Oct-2012	yamt	sync with head
1.51.8.1	17-Apr-2012	yamt	sync with head
1.53.2.3	03-Dec-2017	jdolecek	update from HEAD
1.53.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.53.2.1	23-Jun-2013	tls	resync from head
1.54.4.1	18-May-2014	rmind	sync with head
1.56.6.3	05-Feb-2017	skrll	Sync with HEAD
1.56.6.2	09-Jul-2016	skrll	Sync with HEAD
1.56.6.1	19-Mar-2016	skrll	Sync with HEAD
1.58.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.60.8.3	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1406): sys/dev/pci/if_vgereg.h: revision 1.5 sys/dev/pci/if_vge.c: revision 1.75 Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.60.8.2	13-May-2019	martin	Pull up the following, via patch, requested by msaitoh in ticket #1263: sys/dev/mii/brgphy.c 1.84 sys/dev/mii/ciphy.c 1.33 via patch sys/dev/mii/rgephy.c 1.53 sys/arch/arm/imx/if_enet.c 1.18 sys/arch/mips/adm5120/dev/if_admsw.c 1.19-1.20 sys/dev/pci/if_bge.c 1.329 sys/dev/pci/if_bnx.c 1.81 sys/dev/pci/if_et.c 1.21 sys/dev/pci/if_lii.c 1.22 sys/dev/pci/if_msk.c 1.87 sys/dev/pci/if_nfe.c 1.68 sys/dev/pci/if_sk.c 1.95 sys/dev/pci/if_ti.c 1.107 sys/dev/pci/if_txp.c 1.52 sys/dev/pci/if_vge.c 1.69 sys/dev/usb/if_axen.c 1.38 sys/dev/usb/if_aue.c 1.149 Fix a bug that the duplex of manual media setting may be wrong when the IFM_GMASK bit other than IFM_[FH]DX is set.
1.60.8.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.62.2.4	26-Jan-2019	pgoyette	Sync with HEAD
1.62.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.62.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.62.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.63.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.63.2.1	10-Jun-2019	christos	Sync with HEAD
1.73.2.2	25-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #477): sys/dev/pci/if_vge.c: revision 1.76 sys/dev/pci/if_vge.c: revision 1.77 Fixes a bug that "ifmedia vge0 media 1000baseT-FDX" causes device timeout. If the interface's media is NOT in the "best" mode (i.e. other than IFM_AUTO), control VGE_DIAGCTL_GMII bit. - Set duplex correctly when user setting is not IFM_AUTO. - When the link is up, set VGE_DIAGCTL not from user media setting but from the current active link status.
1.73.2.1	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #345): sys/dev/pci/if_vgereg.h: revision 1.5 sys/dev/pci/if_vge.c: revision 1.75 Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.78.2.1	29-Feb-2020	ad	Sync with head.
1.80.8.1	01-Aug-2021	thorpej	Sync with HEAD.
1.88.2.1	02-Aug-2025	perseant	Sync with HEAD
1.8	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.7	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.6	24-Jul-2021	andvar	Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889. Also fixed some additional typos in comments, found on review of same files or typos.
1.5	08-Oct-2019	msaitoh	branches: 1.5.12; Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.4	11-Jul-2019	msaitoh	branches: 1.4.2; Fix typo (s/supress/suppress/).
1.3	31-Oct-2006	tsutsui	branches: 1.3.134; 1.3.142; Make vge(4) look a NetBSD native driver like pcn(4) or epic(4) etc: - merge if_vgevar.h into if_vge.c since no other file refers it - rename some macro (VGE_TX_DESC_CNT -> VGE_NTXDESC etc.) and structs - change prefixes of structure members to represent parents - put TX and RX descriptors into single struct vge_control_data and allocate DMA memory at once - no need to specify BUS_DMA_ALLOCNOW - define appropriate macro for offsets and addresses of DMA descriptors - define struct vge_txsoft and vge_rxsoft, and put common data for each descriptor into them - remove struct vge_list_data and move its members into struct vge_softc - remove #ifdef DEVICE_POLLING code we don't support - merge vge_[tr]x_list_init() functions into vge_init() - use aprint_error() to print errors - put splnet(9) where appropriate and remove unused VGE_{LOCK,UNLOCK}() macro - clear TX timeout only if there is no pending descriptor - check dm_nsegs properly otherwise padding short packets could fail - prepare zero'ed DMA memory to pad short packets rather than putting random data - fix a wrong VGE_TXDESCSYNC() usage which should be VGE_TXFRAGSYNC() Tested on my i386 which is my development machine. (more bus_dmamap_sync(9) cleanup will be done later)
1.2	14-Oct-2006	tsutsui	- call bus_dmamap_sync() properly (sigh) - merge vge_dma_map_rx_desc() and vge_dma_map_tx_desc() into callers - declare DMA descs volatile - remove unused macro/declarations - bzero -> memset - some more cosmetics
1.1	20-Feb-2005	jdolecek	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.10; 1.1.38; 1.1.40; Add device driver support for the VIA Networking Technologies VT6122 gigabit ethernet chip and integrated 10/100/1000 copper PHY. Obtained from: FreeBSD
1.1.40.2	10-Dec-2006	yamt	sync with head.
1.1.40.1	22-Oct-2006	yamt	sync with head
1.1.38.1	18-Nov-2006	ad	Sync with head.
1.1.10.1	30-Dec-2006	yamt	sync with head.
1.1.8.2	29-Apr-2005	kent	sync with -current
1.1.8.1	20-Feb-2005	kent	file if_vgereg.h was added on branch kent-audio2 on 2005-04-29 11:29:06 +0000
1.1.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	20-Feb-2005	yamt	file if_vgereg.h was added on branch yamt-km on 2005-03-19 08:35:11 +0000
1.1.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.1	20-Feb-2005	skrll	file if_vgereg.h was added on branch ktrace-lwp on 2005-03-04 16:45:19 +0000
1.3.142.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.3.134.1	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1406): sys/dev/pci/if_vgereg.h: revision 1.5 sys/dev/pci/if_vge.c: revision 1.75 Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.4.2.1	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #345): sys/dev/pci/if_vgereg.h: revision 1.5 sys/dev/pci/if_vge.c: revision 1.75 Copy vge_clrwol() from FreeBSD and call it in vge_attach() to recover from powerdown mode. Fixes PR kern/41525 reported by Aran Clauson.
1.5.12.1	01-Aug-2021	thorpej	Sync with HEAD.
1.7	31-Oct-2006	tsutsui	This has been megered into if_vge.c.
1.6	15-Oct-2006	tsutsui	Oops, use correct dmamap to sync DMA descriptors. Found by DIAGNOSTIC code in sgimips/bus.c:_bus_dmamap_sync_mips3().
1.5	14-Oct-2006	tsutsui	- use MGETHDR() and MCLGET() rather than m_gethdr() and m_clget() - KNF, ANSIfy, remove register decls - some more cosmetics
1.4	14-Oct-2006	tsutsui	- call bus_dmamap_sync() properly (sigh) - merge vge_dma_map_rx_desc() and vge_dma_map_tx_desc() into callers - declare DMA descs volatile - remove unused macro/declarations - bzero -> memset - some more cosmetics
1.3	14-Oct-2006	tsutsui	- use correct dmamap size for TX/RX DMA descriptors - use __NO_STRICT_ALIGNMENT directly rather than local VGE_FIXUP_RX - no need to use BUS_DMA_ALLOCNOW - remove unneeded members from softc XXX: Is vge_fixup_rx() really more efficient than memmove(9), XXX: or allocating a new buffer and memcpy(9) into it? XXX: Anyway, vge(4) is not recommended for non-x86 hosts at all XXX: because it requires copying buffers by CPU on RX.
1.2	01-Oct-2005	jdolecek	branches: 1.2.26; 1.2.28; VGE_FIXUP_RX is not needed on amd64, disable it for all non-strict alignment architectures fix ETHER_ALIGN to 2 (same value as on FreeBSD) - appears VGE_FIXUP_RX code cuts part of packet otherwise; also add comment about it's purpose PR: 31323 by Murata Shuuichirou
1.1	20-Feb-2005	jdolecek	branches: 1.1.2; 1.1.6; 1.1.8; 1.1.10; Add device driver support for the VIA Networking Technologies VT6122 gigabit ethernet chip and integrated 10/100/1000 copper PHY. Obtained from: FreeBSD
1.1.10.2	30-Dec-2006	yamt	sync with head.
1.1.10.1	21-Jun-2006	yamt	sync with head.
1.1.8.2	29-Apr-2005	kent	sync with -current
1.1.8.1	20-Feb-2005	kent	file if_vgevar.h was added on branch kent-audio2 on 2005-04-29 11:29:06 +0000
1.1.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	20-Feb-2005	yamt	file if_vgevar.h was added on branch yamt-km on 2005-03-19 08:35:11 +0000
1.1.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.1	20-Feb-2005	skrll	file if_vgevar.h was added on branch ktrace-lwp on 2005-03-04 16:45:19 +0000
1.2.28.2	10-Dec-2006	yamt	sync with head.
1.2.28.1	22-Oct-2006	yamt	sync with head
1.2.26.1	18-Nov-2006	ad	Sync with head.
1.115	15-Jul-2025	ozaki-r	vioif: count tx pcq full
1.114	14-Feb-2025	joe	use right byte size of data resource to be freed instead of sizeof pointer use of sizeof pointer frees only 8 bytes of the large vioif_tx_context struct which leaks the rest of the memory
1.113	10-Nov-2024	mlelstv	Add MBUFTRACE
1.112	29-Jun-2024	riastradh	branches: 1.112.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.111	21-Mar-2024	isaki	Ensure that the number of bus_dma segments doesn't exceed VirtIO queue size. This fixes reproducible panics when the host's VirtIO queue size is too small, less than or equal to VIRTIO_NET_TX_MAXNSEGS(=16). PR kern/58049.
1.110	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.109	13-May-2023	andvar	fix typos in comments.
1.108	11-May-2023	yamaguchi	Fix missing check for netq->netq_stopping in vioif_rx_intr() Reported-by: syzbot+5120b7a1f97a3f5ca052@syzkaller.appspotmail.com https://syzkaller.appspot.com/bug?id=243cf4115808e49774a49294f63200770399660b
1.107	27-Mar-2023	nakayama	Use PRIuBUSSIZE to print bus_size_t variables.
1.106	24-Mar-2023	yamaguchi	vioif(4): fix wrong memory allocation size
1.105	23-Mar-2023	yamaguchi	vioif(4): clear flags when configure is failed
1.104	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.103	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.102	23-Mar-2023	yamaguchi	vioif(4): divide IFF_OACTIVE into per-queue
1.101	23-Mar-2023	yamaguchi	vioif(4): reorganize functions iThis change is move of function and rename, and this is no functional change.
1.100	23-Mar-2023	yamaguchi	vioif(4): rename sc_hdr_segs to sc_segs
1.99	23-Mar-2023	yamaguchi	vioif(4): added functions to manipulate network queues
1.98	23-Mar-2023	yamaguchi	vioif(4): added new data structure for network queues and moved the same parameters in vioif_txqueue and vioif_rxqueue into the new structure
1.97	23-Mar-2023	yamaguchi	vioif(4): added __predct_false to error check
1.96	23-Mar-2023	yamaguchi	vioif(4): prepare slot before dequeuing
1.95	23-Mar-2023	yamaguchi	vioif(4): added a structure to manage variables for packet processings
1.94	23-Mar-2023	yamaguchi	vioif(4): increase output error counter
1.93	23-Mar-2023	yamaguchi	vioif(4): merge drain into clear of queue
1.92	23-Mar-2023	yamaguchi	vioif(4): divide interrupt handler for receiving into dequeuing and preparing of buffers
1.91	23-Mar-2023	yamaguchi	vioif(4): drain receive buffer on stopping the device to remove branch in vioif_populate_rx_mbufs_locked()
1.90	23-Mar-2023	yamaguchi	vioif(4): fix missing virtio_enqueue_abort for error handling
1.89	23-Mar-2023	yamaguchi	vioif(4): added event counters related to receive processing
1.88	23-Mar-2023	yamaguchi	vioif(4): adjust receive buffer to ETHER_ALIGN
1.87	23-Mar-2023	yamaguchi	vioif(4): stop interrupt before schedule handler
1.86	23-Mar-2023	yamaguchi	vioif(4): rename {txq,rxq}_active to {txq,rxq}_running_handle
1.85	23-Mar-2023	yamaguchi	vioif(4): use device reset to stop interrupt completely
1.84	23-Mar-2023	yamaguchi	vioif(4): access to txq_active and rxq_active with lock held
1.83	23-Mar-2023	yamaguchi	vioif(4): remove unnecessary lock release if_percpuq_enqueue() can call with rxq->rxq_lock held because of per-cpu.
1.82	12-Sep-2022	knakahara	branches: 1.82.4; Uniform vioif's link status to if_link_state. Implemented by yamaguchi@n.o. Let vioif(4) know LINK_STATE_UNKNOWN.
1.81	04-May-2022	simonb	White space KNF nits.
1.80	16-Apr-2022	andvar	fix various typos in comments and log messages.
1.79	13-Apr-2022	uwe	virtio: use the new syntax for snprintb(3) format strings. The old syntax is limited to 32 bits only (and has 1-based bit numbers which is rather incovenient too).
1.78	13-Apr-2022	yamaguchi	vioif(4): issue VIRTIO_NET_CTRL_MAC_ADDR_SET command only when VIRTIO_NET_F_CTRL_MAC_ADDR is negotiated
1.77	31-Mar-2022	yamaguchi	vioif(4): remove unnecessary lock acquirement The lock was hold to wait for completion of interrupt handlers. But, they are already stopped by rxq_stopping and txq_stopping flags. pointed out by riastradh@n.o, thanks.
1.76	29-Mar-2022	yamaguchi	vioif(4): Added a comment about stopping packet processing
1.75	24-Mar-2022	yamaguchi	vioif(4): adopt ether_set_ifflags_cb
1.74	24-Mar-2022	yamaguchi	vioif(4): register MAC address to a device
1.73	24-Mar-2022	yamaguchi	vioif(4): fix missing error handling
1.72	24-Mar-2022	yamaguchi	vioif(4): do not schedule packet processing while stopping the device
1.71	28-Oct-2021	yamaguchi	virtio: stop reinit for safety when a device resetting is failed
1.70	08-Feb-2021	skrll	Trailing whitespace
1.69	03-Feb-2021	reinoud	Oops, made a mistake in my last commit
1.68	03-Feb-2021	reinoud	Allocate enough space for the bus_dmamap_t arrays for rxq_hdr_dmamaps[] and txq_hdr_maps[]
1.67	31-Jan-2021	reinoud	Although the header structure can be smaller, the headers are indexed as if they are full sized so allocate enough memory so the indexing works as expected and we are not scribbling outside bounds.
1.66	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.65	28-May-2020	riastradh	branches: 1.65.2; Allocate proper storage for the event counter group names. Can't use a stack buffer for these because the evcnt remembers the pointer!
1.64	25-May-2020	yamaguchi	Stop all processing related to rx before that related to tx for safety
1.63	25-May-2020	yamaguchi	Use evcnt(9) to record error status in vioif(4)
1.62	25-May-2020	yamaguchi	Introduce the lock for vioif_softc to avoid a race condition in vioif_update_link_status() The function is called in both vioif_init() and softint.
1.61	25-May-2020	yamaguchi	Populate mbufs in the packet receiving process, not in a softint
1.60	25-May-2020	yamaguchi	Always hold tx lock in deferred transmit to send all packets There may be packets that enqueued before another transmission releases the lock after finish of its transmission. When using mutex_try_enter(), vioif_deferred_transmit() can not sends them. pointed out by knakahara@n.o
1.59	25-May-2020	yamaguchi	Remove redundant checks. There is the same check in vioif_send_common_locked()
1.58	25-May-2020	yamaguchi	Replace macros with static functions for refactoring
1.57	25-May-2020	yamaguchi	Fix typo in comments
1.56	25-May-2020	yamaguchi	Fix the wrong segment size in vioif(4)
1.55	25-May-2020	yamaguchi	Introduce packet handling in softint or kthread for vioif(4)
1.54	25-May-2020	yamaguchi	Set handlers implemented in child device of virtio(4) to virtqueue instead of the commonized function
1.53	25-May-2020	yamaguchi	Obsolete VIOIF_SOFTINT_INTR The kernel option is introduced to realize softint-based if_input. Since the same scheme has been implemented in if_percpuq_enqueue(), the option is no longer needed. pointed out by ozaki-r@n.o.
1.52	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.51	01-Oct-2019	chs	branches: 1.51.2; in many device attach paths, allocate memory with KM_SLEEP instead of KM_NOSLEEP and remove code to handle failures that can no longer happen.
1.50	14-Sep-2019	christos	- KNF - fix typo in error message - use aprint* everywhere - use loops to initialize mac - remove unused variables
1.49	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.48	23-May-2019	msaitoh	-No functional change: - Simplify struct ethercom's pointer near ETHER_FIRST_MULTI(). - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.47	04-Feb-2019	yamaguchi	Do not call virtio_start_vq_intr() for ctrlq unless the iface has a control queue
1.46	14-Jan-2019	yamaguchi	Add multiqueue support, vioif(4)
1.45	14-Jan-2019	yamaguchi	Set IFEF_MPSAFE flag
1.44	14-Jan-2019	yamaguchi	Functionize the same code related to ctrl vq in vioif(4)
1.43	14-Jan-2019	yamaguchi	Divide some elements of vioif_softc into txq, rxq, and ctrlq
1.42	14-Jan-2019	yamaguchi	Make macros not depend on vioif_softc
1.41	26-Jun-2018	msaitoh	branches: 1.41.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.40	10-Jun-2018	jakllsch	remove irrelevant pci(9) #includes from virtio child drivers
1.39	08-Feb-2018	dholland	branches: 1.39.2; Typos.
1.38	01-Jun-2017	chs	remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.37	17-May-2017	jdolecek	more precise m_freem() on error paths, and update m after the m_defrag() call
1.36	17-May-2017	jdolecek	simplify vioif_start() - remove the delivery attempts on failure and retries, leave that for the dedicated thread if dma map load fails, retry after m_defrag(), but continue processing other queue items regardless set interface queue length according to the length of virtio queue, so that higher layer won't queue more than interface can manage to keep in flight use the mutexes always, not just with NET_MPSAFE, so they continue being exercised and hence working; they also enforce proper IPL level inspired by discussion around PR kern/52211, thanks to Masanobu SAITOH for the m_defrag() idea and code
1.35	17-May-2017	jdolecek	do not set IFF_OACTIVE if dma map load or the virtio reserve fails; this causes interface to ignore any further TX requests if this happens when there are no other TX requests in progress fixes kern/52211 by Juergen Hannken-Illjes
1.34	28-Mar-2017	ozaki-r	branches: 1.34.4; Handle config change interrupts to inhibit sending packets while link down PR kern/52103 by s-yamaguchi@IIJ
1.33	28-Mar-2017	ozaki-r	Don't write to read-only VIRTIO_NET_S_LINK_UP bit The bit is defined as read-only in the Virtio PCI Card Specification. The fix is inspired by FreeBSD. PR kern/52103 by s-yamaguchi@IIJ
1.32	25-Mar-2017	jdolecek	reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.31	17-Jan-2017	ozaki-r	Fix unlocking in vioif_rx_filter
1.30	28-Dec-2016	ozaki-r	branches: 1.30.2; Protect ec_multi* with mutex The data can be accessed from sysctl, ioctl, interface watchdog (if_slowtimo) and interrupt handlers. We need to protect the data against parallel accesses from them. Currently the mutex is applied to some drivers, we need to apply it to all drivers in the future. Note that the mutex is adaptive one for ease of implementation but some drivers access the data in interrupt context so we cannot apply the mutex to every drivers as is. We have two options: one is to replace the mutex with a spin one, which requires some additional works (see ether_multicast_sysctl), and the other is to modify the drivers to access the data not in interrupt context somehow.
1.29	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.28	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.27	29-Nov-2016	uwe	vioif_start() - do not call virtio_enqueue_abort() after error from virtio_enqueue_reserve(), as it's already done by the latter, so we ended up with a kind of "double free" that messed up out free list of vq_entry's. This is even documented in a "typical usage" comment in virtio.c (and those quotes are not intended to be sarcastic). PR 51132 - virtio net device stuck for UDP burst transmission
1.26	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.25	29-Aug-2016	ozaki-r	Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.24	10-Jun-2016	ozaki-r	branches: 1.24.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.23	17-May-2016	pooka	Try to get more packets going if the transmit interrupt indicates some were sent. Doing so avoids a situation where vioif_start never gets called in case the sendqueue fills up and therefore the interface perpetually drops all packets due to the queue being full. (not sure why all drivers need to do this themselves; just keeping up with the joneses) Problem reported and patch tested by jmmlmendes and yasukata at repo.rumpkernel.org/rumprun
1.22	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.21	10-Jan-2016	christos	PR/50636: Ryo ONODERA: Reduce memory use
1.20	29-Oct-2015	christos	simplify
1.19	29-Oct-2015	ozaki-r	Name virtqueue index
1.18	27-Oct-2015	christos	- Print the negotiated feature bits. - Use aprint_error_dev on error, instead of printf - Add missing abort call.
1.17	26-Oct-2015	ozaki-r	Support MSI-X in virtio Currently only vioif(4) uses the feature. knakahara@ helped to migrate to pci_intr_alloc(9). Thanks!
1.16	05-May-2015	ozaki-r	Use NULL for initialization of sc_config_change
1.15	16-Jan-2015	ozaki-r	Introduce defflag for NET_MPSAFE
1.14	25-Dec-2014	ozaki-r	Reuse mbuf when retrying in vioif_start Otherwise, the old mbuf will leak.
1.13	24-Dec-2014	ozaki-r	Take TX/RX locks when sc_stopping = true in if_stop Taking the locks is needed to ensure ongoing TX/RX operations finish. Otherwise, if_stop may run during TX/RX operations.
1.12	19-Dec-2014	ozaki-r	Implement softint-based interrupt handling in if_vioif Softint-based interrupt handling is considered as a future direction of the (network) device driver architecture in NetBSD. pq3etsec of ppc is already implemented based on the architecture (unlike pq3etsec, this change doesn't include softint-based if_start). In this architecture, a hardware interrupt handler just schedules a softint and the softint performs actual interrupt processing. It reduces processing in hardware interrupt context and allows Layer 2 network stack (e.g., bridge, vlan and even bpf) run in softint context, which makes it easy to implement fine-grain locking in the layer. This is an experimental implementation of the architecture in if_viof. virtio introduces a new flag VIRTIO_F_PCI_INTR_SOFTINT. If a driver of virtio sets it to sc_flags of virtio_softc, virtio calls softint_schedule in virtio_intr instead of directly calling the interrupt handler of the driver. When VIOIF_SOFTINT_INTR is on, vioif doesn't use the existing softint (vioif_rx_softint) that is called from vioif_rx_vq_done. Because vioif_rx_softint already runs in softint context and another softint isn't needed. This change actually improves performance in some cases. The feature is disabled by default and enabled when SOFTINT_INTR is set somewhere (normally in a kernel configuration).
1.11	09-Oct-2014	ozaki-r	branches: 1.11.2; Add ETHERCAP_VLAN_MTU capability to vioif
1.10	08-Oct-2014	ozaki-r	Add missing semicolon
1.9	08-Oct-2014	ozaki-r	Don't turn promisc off in vioif_deferred_init if already configured as promisc
1.8	13-Aug-2014	pooka	Don't use config_deferred_interrupts() for vioif_deferred_init(), just run it once as part of if_init(). The problem with the former is that it will execute the deferred init routine in-place when !cold, and since vioif_deferred_init() finishing depends on virtio interrupts which are established only after config_deferred_interrupts() is called, the vioif attach method would deadlock when !cold.
1.7	22-Jul-2014	ozaki-r	branches: 1.7.2; Make if_vioif MPSAFE - Introduce VIOIF_MPSAFE - It's enabled only when NET_MPSAFE is defined in if.h or the kernel config - Add tx and rx mutex locks - Locking them is performance sensitive, so it's not used when !VIOIF_MPSAFE - Set SOFTINT_MPSAFE to vioif_rx_softint only when VIOIF_MPSAFE
1.6	22-Jul-2014	ozaki-r	Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio It is set by a child driver, e.g., if_vioif. If set, virtio sets PCI_INTR_MPSAFE for pci_intr_establish.
1.5	18-Jul-2014	ozaki-r	Don't set SOFTINT_MPSAFE to vioif_rx_softint vioif_rx_softint calls vioif_populate_rx_mbufs that is not MPSAFE. vioif_populate_rx_mbufs is also called via vioif_ioctl and so can be called by two LWPs simultaneously, resulting in kernel panic. PR kern/49007
1.4	09-May-2013	minoura	branches: 1.4.6; Fix a typo, and remove an unused member. This should fix the problem that recent Qemu dies during configuring a vioif.
1.3	30-Mar-2013	christos	remove trailing whitespace
1.2	19-Nov-2011	jmcneill	branches: 1.2.6; 1.2.8; 1.2.12; 1.2.14; fix build when ALTQ is defined
1.1	30-Oct-2011	hannken	branches: 1.1.2; Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.1.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.2.1	17-Apr-2012	yamt	sync with head
1.2.14.3	24-Sep-2016	bouyer	Pull up following revision(s) (requested by ozaki-r in ticket #1401): sys/dev/pci/if_vioif.c: revision 1.25 Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.2.14.2	07-Aug-2014	msaitoh	Pull up following revision(s) (requested by ozaki-r in ticket #1102): sys/dev/pci/if_vioif.c: revision 1.5 Don't set SOFTINT_MPSAFE to vioif_rx_softint vioif_rx_softint calls vioif_populate_rx_mbufs that is not MPSAFE. vioif_populate_rx_mbufs is also called via vioif_ioctl and so can be called by two LWPs simultaneously, resulting in kernel panic. PR kern/49007
1.2.14.1	12-May-2013	riz	Pull up following revision(s) (requested by minoura in ticket #890): sys/dev/pci/if_vioif.c: revision 1.4 Fix a typo, and remove an unused member. This should fix the problem that recent Qemu dies during configuring a vioif.
1.2.12.3	03-Dec-2017	jdolecek	update from HEAD
1.2.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.12.1	23-Jun-2013	tls	resync from head
1.2.8.3	24-Sep-2016	bouyer	Pull up following revision(s) (requested by ozaki-r in ticket #1401): sys/dev/pci/if_vioif.c: revision 1.25 Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.2.8.2	07-Aug-2014	msaitoh	Pull up following revision(s) (requested by ozaki-r in ticket #1102): sys/dev/pci/if_vioif.c: revision 1.5 Don't set SOFTINT_MPSAFE to vioif_rx_softint vioif_rx_softint calls vioif_populate_rx_mbufs that is not MPSAFE. vioif_populate_rx_mbufs is also called via vioif_ioctl and so can be called by two LWPs simultaneously, resulting in kernel panic. PR kern/49007
1.2.8.1	12-May-2013	riz	branches: 1.2.8.1.2; Pull up following revision(s) (requested by minoura in ticket #890): sys/dev/pci/if_vioif.c: revision 1.4 Fix a typo, and remove an unused member. This should fix the problem that recent Qemu dies during configuring a vioif.
1.2.8.1.2.2	24-Sep-2016	bouyer	Pull up following revision(s) (requested by ozaki-r in ticket #1401): sys/dev/pci/if_vioif.c: revision 1.25 Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.2.8.1.2.1	07-Aug-2014	msaitoh	Pull up following revision(s) (requested by ozaki-r in ticket #1102): sys/dev/pci/if_vioif.c: revision 1.5 Don't set SOFTINT_MPSAFE to vioif_rx_softint vioif_rx_softint calls vioif_populate_rx_mbufs that is not MPSAFE. vioif_populate_rx_mbufs is also called via vioif_ioctl and so can be called by two LWPs simultaneously, resulting in kernel panic. PR kern/49007
1.2.6.3	09-Jun-2013	msaitoh	Pull up following revision(s) (requested by minoura in ticket #1861): sys/dev/pci/if_vioif.c: revision 1.4 Fix a typo, and remove an unused member. This should fix the problem that recent Qemu dies during configuring a vioif. Fixes PR#47780.
1.2.6.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.2.6.1	19-Nov-2011	riz	file if_vioif.c was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.4.6.1	10-Aug-2014	tls	Rebase.
1.7.2.3	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1337): sys/dev/pci/if_vioif.c: revision 1.21 sys/dev/pci/virtio.c: revision 1.14 PR/50636: Ryo ONODERA: Fix incorrect kmem_free() -- PR/50636: Ryo ONODERA: Reduce memory use
1.7.2.2	18-Sep-2016	snj	Pull up following revision(s) (requested by ozaki-r in ticket #1244): sys/dev/pci/if_vioif.c: revision 1.25 Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.7.2.1	29-Dec-2014	martin	branches: 1.7.2.1.2; 1.7.2.1.4; Pull up following revision(s) (requested by ozaki-r in ticket #354): sys/dev/pci/if_vioif.c: revision 1.14 Reuse mbuf when retrying in vioif_start Otherwise, the old mbuf will leak.
1.7.2.1.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.7.2.1.2.1	18-Sep-2016	snj	Pull up following revision(s) (requested by ozaki-r in ticket #1244): sys/dev/pci/if_vioif.c: revision 1.25 Fix initializing wrong queues Pointed out by Mike Larkin. PR kern/51448
1.11.2.10	28-Aug-2017	skrll	Sync with HEAD
1.11.2.9	05-Feb-2017	skrll	Sync with HEAD
1.11.2.8	05-Dec-2016	skrll	Sync with HEAD
1.11.2.7	05-Oct-2016	skrll	Sync with HEAD
1.11.2.6	09-Jul-2016	skrll	Sync with HEAD
1.11.2.5	29-May-2016	skrll	Sync with HEAD
1.11.2.4	19-Mar-2016	skrll	Sync with HEAD
1.11.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.11.2.2	06-Jun-2015	skrll	Sync with HEAD
1.11.2.1	06-Apr-2015	skrll	Sync with HEAD
1.24.2.4	26-Apr-2017	pgoyette	Sync with HEAD
1.24.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.24.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.24.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.30.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.34.4.1	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.39.2.3	18-Jan-2019	pgoyette	Synch with HEAD
1.39.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.39.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.41.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.41.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.41.2.1	10-Jun-2019	christos	Sync with HEAD
1.51.2.1	29-Feb-2020	ad	Sync with head.
1.65.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.82.4.6	22-Feb-2025	martin	Pull up following revision(s) (requested by joe in ticket #1055): sys/dev/pci/if_vioif.c: revision 1.114 PR kern/59086: use right byte size of data resource to be freed instead of sizeof pointer use of sizeof pointer frees only 8 bytes of the large vioif_tx_context struct which leaks the rest of the memory
1.82.4.5	07-Aug-2024	martin	Pull up following revision(s) (requested by isaki in ticket #772): sys/dev/pci/if_vioif.c: revision 1.111 Ensure that the number of bus_dma segments doesn't exceed VirtIO queue size. This fixes reproducible panics when the host's VirtIO queue size is too small, less than or equal to VIRTIO_NET_TX_MAXNSEGS(=16). PR kern/58049.
1.82.4.4	07-Aug-2024	martin	Pull up following revision(s) (requested by rin in ticket #770): sys/dev/pci/if_vioif.c: revision 1.108 Fix missing check for netq->netq_stopping in vioif_rx_intr()
1.82.4.3	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.82.4.2	01-Apr-2023	martin	Additionally pull up the following for ticket #128 to fix the build: sys/dev/pci/if_vioif.c: revision 1.107 sys/dev/pci/virtio.c: revision 1.71 Use PRIuBUSSIZE to print bus_size_t variables.
1.82.4.1	30-Mar-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #128): sys/dev/pci/if_vioif.c: revision 1.83-1.102,1.105,1.106 vioif(4): remove unnecessary lock release if_percpuq_enqueue() can call with rxq->rxq_lock held because of per-cpu. vioif(4): access to txq_active and rxq_active with lock held vioif(4): use device reset to stop interrupt completely vioif(4): rename {txq,rxq}_active to {txq,rxq}_running_handle vioif(4): stop interrupt before schedule handler vioif(4): adjust receive buffer to ETHER_ALIGN vioif(4): added event counters related to receive processing vioif(4): fix missing virtio_enqueue_abort for error handling vioif(4): drain receive buffer on stopping the device to remove branch in vioif_populate_rx_mbufs_locked() vioif(4): divide interrupt handler for receiving into dequeuing and preparing of buffers vioif(4): merge drain into clear of queue vioif(4): increase output error counter vioif(4): added a structure to manage variables for packet processings vioif(4): prepare slot before dequeuing vioif(4): added __predct_false to error check vioif(4): added new data structure for network queues and moved the same parameters in vioif_txqueue and vioif_rxqueue into the new structure vioif(4): added functions to manipulate network queues vioif(4): rename sc_hdr_segs to sc_segs vioif(4): reorganize functions This change is move of function and rename, and this is no functional change. vioif(4): divide IFF_OACTIVE into per-queue vioif(4): clear flags when configure is failed vioif(4): fix wrong memory allocation size
1.112.2.1	02-Aug-2025	perseant	Sync with HEAD
1.17	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.16	29-Jun-2024	riastradh	branches: 1.16.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.15	06-Mar-2024	andvar	Fix few typos in comments.
1.14	09-Feb-2024	andvar	s/incompatiable/incompatible/ in error messages.
1.13	28-Aug-2023	msaitoh	Set IFM_FDX flag if the link is up.
1.12	10-Aug-2023	riastradh	vmxnet(4): Fix various MP bugs. - Defer reset to workqueue. => vmxnet3_stop_locked is forbidden in softint. => XXX Problem: We still take the core lock in softint, and we still take the core lock around vmxnet3_stop_locked. TBD. - Touch if_flags only under IFNET_LOCK. => Cache ifp->if_flags & IFF_PROMISC in vmxnet3_ifflags_cb. => Don't call vmxnet3_set_rxfilter unless up and running; cache this as vmx_mcastactive. Use ENETRESET in vmxnet3_ifflags_cb instead of calling vmxnet3_set_rxfilter directly. . (The cache is currently serialized by the core lock, but it might reasonably be serialized by an independent lock like in usbnet(9).) - Fix vmxnet3_stop_rendezvous so it actually does something. => New vxtxq_stopping, vxrxq_stopping variables synchronize with Rx/Tx interrupt handlers. - Sprinkle IFNET_LOCK and core lock assertions.
1.11	16-Sep-2022	knakahara	branches: 1.11.4; It is correct to update ifp statistics per packet, just remove commented out old statistics code. The code is commented out by if_vmx.c:r1.57.
1.10	16-Sep-2022	knakahara	Add ALWAYS_TXDEFER option to vmx(4), too.
1.9	06-Jul-2022	msaitoh	Call txeof first, then rxeof for the consistency. There are three functions where the txeof and rxeof are called. The MSI-X interrupt function and the workqueue function call rxeof first, then rxeof. For legacy interrupt. rxeof is called first. Modify it to match with other two.
1.8	24-May-2022	msaitoh	Reduce diff among vmxnet3_{legacy_intr,txrxq_intr,handle_que}(). NFCI.
1.7	30-Mar-2022	knakahara	vmxnet3_stop_rendezvous() requires workqueue_wait() to wait workqueue not scheduled yet.
1.6	13-Feb-2022	riastradh	vmx(4): Omit needless membar_sync wrapper. Nothing uses this, and it is unlikely that any part of this driver will need store-before-load ordering to synchronize with the host.
1.5	13-Oct-2021	knakahara	Fix vmx(4) receive workqueue problem on qemu, reported by ozaki-r@n.o, thanks. ESXi stops interrupt by vmxnet3_disable_intr() immediately. In contrast, qemu does not stop interrupt soon. Hmm, qemu's behavior seems not to be correct, however, there is no explicit specification of vmxnet3. So, we have to accept that behavior.
1.4	15-Oct-2020	ryo	Reduces useless "vmxnet3_read_cmd(sc, VMXNET3_CMD_GET_LINK)" calls
1.3	15-Oct-2020	ryo	Fixed a bug where 'ifconfig vmx0 down up' caused a failure to send. This procedure was leaving sc->vmx_link_active = 0. ok nakahara@
1.2	14-Oct-2020	ryo	modularize, and fix build error
1.1	14-Oct-2020	ryo	vmx(4) should be MI. moved to sys/dev/pci from sys/arch/x86/pci
1.11.4.2	21-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #904): sys/dev/pci/if_vmx.c: revision 1.12 vmxnet(4): Fix various MP bugs. - Defer reset to workqueue. => vmxnet3_stop_locked is forbidden in softint. => XXX Problem: We still take the core lock in softint, and we still take the core lock around vmxnet3_stop_locked. TBD. - Touch if_flags only under IFNET_LOCK. => Cache ifp->if_flags & IFF_PROMISC in vmxnet3_ifflags_cb. => Don't call vmxnet3_set_rxfilter unless up and running; cache this as vmx_mcastactive. Use ENETRESET in vmxnet3_ifflags_cb instead of calling vmxnet3_set_rxfilter directly. . (The cache is currently serialized by the core lock, but it might reasonably be serialized by an independent lock like in usbnet(9).) - Fix vmxnet3_stop_rendezvous so it actually does something. => New vxtxq_stopping, vxrxq_stopping variables synchronize with Rx/Tx interrupt handlers. - Sprinkle IFNET_LOCK and core lock assertions.
1.11.4.1	11-May-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #687): sys/dev/pci/if_vmx.c: revision 1.13 Set IFM_FDX flag if the link is up.
1.16.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1	14-Oct-2020	ryo	vmx(4) should be MI. moved to sys/dev/pci from sys/arch/x86/pci
1.138	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.137	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.136	17-Sep-2022	thorpej	Eliminate use of IFF_OACTIVE.
1.135	24-Jul-2021	andvar	s/structres/structures + s/whenver/whenever
1.134	08-May-2021	thorpej	Use pci_compatible_match().
1.133	07-Feb-2020	thorpej	branches: 1.133.10; Use callout_setfunc() / callout_schedule().
1.132	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.131	13-Sep-2019	msaitoh	branches: 1.131.2; if_flags is neither int nor short. It's unsigned short.
1.130	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.129	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.128	23-May-2019	msaitoh	-No functional change: - Simplify struct ethercom's pointer near ETHER_FIRST_MULTI(). - Simplify MII structure initialization. - u_int_t -> uint_t. - KNF
1.127	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.126	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.125	18-Jul-2018	sevan	Be consistent among ethernet drivers on the convention for printing ethernet addresses. NFC
1.124	26-Jun-2018	msaitoh	branches: 1.124.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.123	28-Feb-2018	flxd	branches: 1.123.2; Cosmetic printf/aprint changes. Add missing return statement after failed pci_intr_establish().
1.122	28-Feb-2018	flxd	Cosmetic fix for the Physical Address Registers PAR0 to PAR5.
1.121	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.120	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.119	14-Jul-2016	msaitoh	branches: 1.119.2; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.118	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.117	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.116	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.115	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.114	10-Aug-2014	tls	branches: 1.114.4; Merge tls-earlyentropy branch into HEAD.
1.113	29-Mar-2014	christos	branches: 1.113.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.112	27-Dec-2012	jmcneill	branches: 1.112.2; - reset the chip if the tx engine gets stuck after a link state change, from OpenBSD - no need to do a full reset of the chip when enabling or disabling promiscuous mode
1.111	22-Jul-2012	matt	branches: 1.111.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.110	02-Feb-2012	tls	branches: 1.110.2; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.109	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.108	19-Nov-2011	tls	branches: 1.108.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.107	16-Dec-2010	jakllsch	branches: 1.107.8; Enable ETHERCAP_VLAN_MTU in vr(4). Tested on VT6105M and VT8237R+. May or may not work on earlier chips, for now just assume it will. Provoked by PR#44243.
1.106	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.105	28-Apr-2010	hubertf	Silence for boot -z
1.104	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.103	24-Feb-2010	dyoung	branches: 1.103.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.102	19-Jan-2010	pooka	branches: 1.102.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.101	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.100	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.99	26-Sep-2009	jmcneill	add suspend/resume support
1.98	06-Sep-2009	tsutsui	Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested VIA VT86C100A (which is probed as VT3043).
1.97	23-Aug-2009	jmcneill	Don't maintain a list of descriptions for each supported device, that's what PCIVERBOSE is for; use pci_devinfo instead.
1.96	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.95	09-Jul-2008	joerg	branches: 1.95.4; 1.95.8; 1.95.12; - device/softc split
1.94	09-Jul-2008	joerg	Constify device table.
1.93	28-Apr-2008	martin	branches: 1.93.2; 1.93.4; 1.93.6; Remove clause 3 and 4 from TNF licenses
1.92	10-Apr-2008	cegger	branches: 1.92.2; 1.92.4; use aprint_*_dev and device_xname
1.91	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.90	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.89	19-Jan-2008	dyoung	branches: 1.89.2; 1.89.6; Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.88	19-Oct-2007	ad	branches: 1.88.2; 1.88.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.87	09-Jul-2007	ad	branches: 1.87.6; 1.87.8; 1.87.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.86	15-Jun-2007	jmcneill	Add VT6105M (Rhine III) to the list of supported devices, as found in the Soekris net5501.
1.85	04-Mar-2007	christos	branches: 1.85.2; 1.85.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.84	16-Nov-2006	christos	branches: 1.84.2; 1.84.4; 1.84.8; __unused removal on arguments; approved by core.
1.83	05-Nov-2006	tsutsui	Use uintNN_t.
1.82	03-Nov-2006	tsutsui	Pull several fixes which improve TX error handling from FreeBSD's if_vr.c rev 1.52: - check more error status in TX descriptor and restart TX module appropriately in vr_txeof() - check more error interrupt status in vr_intr() I can't confirm whether these changes actually fix TX stalls because I can't reproduce the problem I had about seven years ago (I guess it might be caused by excessive collisions on a dumb hub), but at least they don't seem to have bad side effects on normal operations on my macppc.
1.81	02-Nov-2006	tsutsui	Don't touch the VR_STICKHW register on VT3043. The commit log in FreeBSD's if_vr.c rev 1.43 says "This is really only for the VT6102, but it doesn't hurt the older chips," but at least it hurts my VT86C100A (which returns a product ID of VT3043) on macppc and causes kernel MCHK trap while the same board on i386 and VT6102 on macppc have no problem with it.
1.80	02-Nov-2006	tsutsui	Pull some RX fixes from FreeBSD's if_vr.c rev 1.59: - in vr_rxeoc() (i.e. on RX error interrupts), disable RX before calling vr_rxeof() and check it actually stopped - no recovery is needed for VR_ISR_DROPPED, so just account ierrors - also account ierrors in vr_rxeoc()
1.79	02-Nov-2006	tsutsui	- make DMA descriptor members volatile - set VR_RXSTAT_OWN after all other descriptor data is set With these changes, vr_init() seems to initialize the vr chip stuck on RX properly.
1.78	20-Oct-2006	scw	Force a reload of the EEPROM if the MAC address is all zeroes. Fixes PR kern/34812.
1.77	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.76	17-Jun-2006	christos	branches: 1.76.4; 1.76.6; re-factor the pci powestate api. reviewed by gimpy
1.75	11-Dec-2005	christos	branches: 1.75.4; 1.75.8; 1.75.14; 1.75.16; merge ktrace-lwp.
1.74	30-Jan-2005	thorpej	branches: 1.74.6; Eliminate use of M_HASFCS.
1.73	27-Jan-2005	scw	In vr_start(), don't forget to free the new mbuf if the second bus_dmamap_load_mbuf() fails.
1.72	25-Jan-2005	jmmv	Capture incomplete packets coming from the card, as this driver doesn't support this feature. This avoids multiple crashes that I've had in the past. Also ensure that packets are not empty when DIAGNOSTIC is set. However, this is just another sanity check of the received packets, but does not address the real problem. The issue seems to be the following: if the card receives data while doing a reset (vr_init), it later finds a bunch of empty packets in the receive ring. This explains the crashes I've hit: running a program which needs promiscuous mode (dhclient) while the card was already running in that mode (tcpdump). In this situation, it's easy that the second reset receives stuff from the network. Unfortunately, I don't know why the card is producing these packets... While here, fix a typo in a comment.
1.71	13-Jan-2005	jmmv	branches: 1.71.2; 1.71.4; Fix two typos.
1.70	30-Oct-2004	thorpej	When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.69	21-Aug-2004	thorpej	Use ANSI function decls and make use of static.
1.68	15-Feb-2004	jdolecek	branches: 1.68.4; add rnd hooks
1.67	14-Nov-2003	jmcneill	Don't set VR_EECSR_LOAD before reading the MAC address on the VT6102 either.
1.66	13-Nov-2003	scw	On Rhine III, don't set VR_EECSR_LOAD before reading the MAC address. This reloads the entire EEPROM, not just the MAC address, which can cause problems for the host PCI bus under certain circumstances. The chip already loads the EEPROM at powerup/reset anyway. XXX: This probably applies to the other Rhine variants too, but I don't have a data sheet to confirm this behaviour.
1.65	17-Oct-2003	tsutsui	- No need to set VR_TXCTL_TLINK in vr_start() because all TX buffers have only single segment. - No need to set VR_CMD_TX_ON in vr_start(). - Initialize BCR0 and BCR1 registers. - Change RX DMA threshold 128bytes. Mostly from FreeBSD via OpenBSD.
1.64	17-Oct-2003	tsutsui	Misc cosmetics.
1.63	17-Oct-2003	tsutsui	Use common PCI macro/functions for PWRMGMT.
1.62	23-Aug-2003	dogcow	add VT6105 support to vr(4)
1.61	10-Apr-2003	christos	branches: 1.61.2; Add MIIF_FORCEANEG from Martin Husemann
1.60	15-Jan-2003	bouyer	Zero out the buffer when padding packet to ETHER_MIN_LEN-ETHER_CRC_LEN
1.59	03-Jan-2003	lha	From freebsd if_vr.c:1.43,1.56 and if_vrreg.h:1.11,1.14 - Windows put the chip in suspended mode, make sure we unsuspend it. 1.43, by Takefumi SAYO <stake@po.shiojiri.ne.jp> - Detect the revision of the Rhine chip we're using, and force reset when the chip supports it. 1.65, by silby@freebsd.org
1.58	21-Oct-2002	fair	Change the "dontcare bits" argument to ifmedia_init() to IFM_IMASK so that it is possible select PHY instances other than the first one (instance zero), if there is more than one PHY attached.
1.57	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.56	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.55	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.54	28-Jul-2002	christos	my card prints transmit aborted and then hangs. Let's try to reset it.
1.53	13-Nov-2001	lukem	branches: 1.53.8; 1.53.10; add RCSID
1.52	14-Aug-2001	mrg	mtod() with uintptr_t, not bus_addr_t.
1.51	23-Jul-2001	thorpej	Make sure to initialize the all Rx descriptors properly in *_init() even if mbufs for them are already allocated.
1.50	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.49	07-Jul-2001	thorpej	branches: 1.49.2; bcopy -> memcpy, strcpy
1.48	12-Jun-2001	thorpej	Don't need INET or NS includes here.
1.47	12-Jun-2001	wiz	receive, not recieve
1.46	29-Jan-2001	tsutsui	branches: 1.46.2; Use ether_crc32_be() for multicast hash calculation.
1.45	29-Jan-2001	enami	Handle allmulti case correctly as a NetBSD network driver; if we are requested range of multicast address or too many multicast address, program multicast filter to receive all multicast address. And set/clear IFF_ALLMULTI flag properly.
1.44	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.43	19-Dec-2000	thorpej	Fix a problem with the ALTQ changes that can cause bogus memory refernces. Problem reported by Luke Mewburn.
1.42	14-Dec-2000	thorpej	ALTQ'ify.
1.41	15-Nov-2000	thorpej	Move bpfattach()/bpfdetach() calls into ether_ifattach()/ether_ifdetach().
1.40	15-Oct-2000	thorpej	Don't trim off the FCS, pass it up.
1.39	11-Oct-2000	thorpej	Use ether_ioctl().
1.38	01-Oct-2000	thorpej	Move the check for "promisc + unicast + not for us" into ether_input(), and change Ethernet drivers to always pass all received frames to ether_input() (with a few exceptions, which are documented in the code).
1.37	13-Sep-2000	tron	Use correct product id for VIA VT6102 chip. Problem pointed out by Jaromir Doleeek in private e-mail.
1.36	13-Sep-2000	tron	Add support VIA VT6102 chip found e.g on the D-Link DFE530TX NIC. Fixes PR kern/11001 by Petter Lindquist.
1.35	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.34	23-Mar-2000	thorpej	branches: 1.34.4; New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.33	06-Mar-2000	thorpej	No longer necessary to futz with ifp->if_baudrate here.
1.32	02-Feb-2000	thorpej	Don't dry to diving MIIF_NOISOLATE in the PHY drivers. Instead, pass flags down from the parent to child vi mii_attach().
1.31	02-Feb-2000	thorpej	Bring some order to the chaos which was the MII code function naming "conventions".
1.30	12-Dec-1999	thorpej	Use htole32() and le32toh().
1.29	19-Nov-1999	thorpej	Make the ThunderLAN and VIA Rhine drivers use the common MII bit-bang module.
1.28	12-Nov-1999	thorpej	Call mii_down() as appropriate.
1.27	04-Nov-1999	thorpej	Adapt to mii_phy_probe() change.
1.26	20-Sep-1999	thorpej	branches: 1.26.2; 1.26.4; 1.26.8; Fix packet reception in promiscuous mode: compare the Ethernet destination address against our station address if it's not a multicast packet. Either the Rhine manual lies about the phys/broadcast/multicast rxstatus bits, or the Rhine chip is just broken. Fixes a redirect storm problem reported by Laine Stump on current-users.
1.25	14-Aug-1999	hwr	Initialize error variable. Should fix spurious init errors. From Dave Huang in PR 8205.
1.24	14-Aug-1999	hwr	Remove dulicate "Ethernet" output. From Dave Huang in PR 8206
1.23	03-Aug-1999	thorpej	Be a little nicer about memory usage: * Don't allocate receive buffers until the interface is actually brought up, and release all of them if the interface is taken down. * Add a knob (defaults to off) which will copy an incoming packet to a single header mbuf if it is small enough to fit in one, rather than burning an entire cluster on it. Note that this change will be mostly moot if/when sbcompress() it changed to handle compressing clusters.
1.22	18-May-1999	thorpej	Rework layer 2 protocol input routines. Instead of calling e.g. ether_input() directly, call the function pointer (if_input)(ifp, m). The input routine expects the packet header to be at the head of the packet, and will adjust as necessary. Privatize the layer 2 input and output routines, allowing _ifattach() to set them up as appropriate.
1.21	26-Apr-1999	thorpej	Make the VIA Rhine driver work on big-endian systems. From Izumi Tsutsui, PR #7305.
1.20	24-Apr-1999	thorpej	Fix an mbuf leak in the strict alignment case of packet reception.
1.19	24-Mar-1999	thorpej	branches: 1.19.2; Don't enable I/O or memory access here! These bits in the PCI CSR are set up by the firmware, and indicate which mapping types the system supports.
1.18	12-Feb-1999	thorpej	Major overhaul of the Rhine/Rhine-II driver, structuring it a little more like the SMC83C100 EPIC/100 driver: * Rather than using pointers to the head and tail of the transmit and receive rings, use wrapping indexes into arrays. This is a little more obvious when reading the code. * More cleanly separate the hardware descriptor from the software descriptor. * bus_dma it everywhere. * Implement interrupt pacing and avoid a potential race in the transmit loop. Now this looks more or less like the Rhine driver I was working on when this driver was committed :-) Update copyright notice to reflect that.
1.17	05-Feb-1999	thorpej	Make this driver work on the Alpha. Apparently the receive DMA mechanism has the same 4-byte alignment requirement that the transmit side does. This causes the packet payload to be misaligned. So, on systems which require strict alignment, we must copy the incoming frame to a new packet buffer, suitably aligned.
1.16	05-Feb-1999	thorpej	Fix the interrupt handler to actually acknowledge the interupt to the dispatch code (!!!). How did this work even on the i386 before?
1.15	05-Feb-1999	thorpej	Fix some formatting nits.
1.14	05-Feb-1999	thorpej	Remove some duplicate softc members, and comment some of the softc fields.
1.13	05-Feb-1999	thorpej	Remove some unnecessary splnet's.
1.12	05-Feb-1999	thorpej	splimp -> splnet
1.11	05-Feb-1999	thorpej	Use the generic MII support code, not our own.
1.10	05-Feb-1999	thorpej	Use MII command opcodes from <sys/dev/mii.h>
1.9	05-Feb-1999	thorpej	Rename VR_MIICMD_DATAIN and VR_MIICMD_DATAOUT to actually match what the data sheet says. (The names and their usage were exactly reversed in the code.)
1.8	02-Feb-1999	thorpej	Use pcidevs.h to get vendor and product IDs.
1.7	02-Feb-1999	thorpej	Take software-only structure definitions out of the register description header.
1.6	02-Feb-1999	thorpej	Remove support for FreeBSD from the driver; so many changes are going to be required to make it properly portable to all of NetBSD's platforms, that it just wouldn't be worth the effort.
1.5	01-Feb-1999	thorpej	Minimal changes to make this compile on the Alpha. Much more work is needed to make this a truly portable driver.
1.4	26-Jan-1999	sakamoto	NetBSD RCS id.
1.3	22-Jan-1999	sakamoto	Change bootverbose 1 to 0. Add vr_lookup.
1.2	21-Jan-1999	sakamoto	Add NetBSD autoconfiguration support. KNF.
1.1	21-Jan-1999	sakamoto	branches: 1.1.1; Initial revision
1.1.1.1	21-Jan-1999	sakamoto	Import FreeBSD's VIA Rhine fast ethernet PCI NIC driver written by Bill Paul <wpaul@ctr.columbia.edu>.
1.19.2.3	22-Sep-1999	cgd	pull up rev 1.26 from trunk (requested by thorpej): Fix packet reception in promiscuous mode: compare the Ethernet destination address against our station address if it's not a multicast packet. Fixes a redirect storm problem.
1.19.2.2	26-Apr-1999	perry	branches: 1.19.2.2.2; 1.19.2.2.4; pullup 1.20->1.21 (thorpej): make the driver work on big-endian systems
1.19.2.1	26-Apr-1999	perry	pullup 1.19->1.20 (thorpej): fix an mbuf leak
1.19.2.2.4.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.19.2.2.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.26.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.26.4.1	15-Nov-1999	fvdl	Sync with -current
1.26.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.26.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.26.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.26.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.34.4.2	13-Mar-2001	he	Pull up revision 1.45 (requested by tsutsui): Handle allmulti case correctly as a NetBSD network driver; if we are requested range of multicast address or too many multicast address, program multicast filter to receive all multicast address. And set/clear IFF_ALLMULTI flag properly.
1.34.4.1	13-Sep-2000	tron	Pull up from current (approved by thorpej): Add support VIA VT6102 chip found e.g on the D-Link DFE530TX NIC. syssrc/sys/dev/pci/if_vr.c 1.36-1.37 syssrc/sys/dev/pci/pcidevs 1.256-1.258
1.46.2.8	17-Jan-2003	thorpej	Sync with HEAD.
1.46.2.7	07-Jan-2003	thorpej	Sync with HEAD.
1.46.2.6	11-Nov-2002	nathanw	Catch up to -current
1.46.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.46.2.4	01-Aug-2002	nathanw	Catch up to -current.
1.46.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.46.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.46.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.49.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.49.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.49.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.49.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.49.2.1	03-Aug-2001	lukem	update to -current
1.53.10.1	27-Jan-2003	jmc	Pullup revisions 1.59-1.60 (requested by bouyer in ticket #1092) Zero out the buffer when padding packet to ETHER_MIN_LEN-ETHER_CRC_LEN
1.53.8.1	29-Aug-2002	gehenna	catch up with -current.
1.61.2.7	04-Feb-2005	skrll	Sync with HEAD.
1.61.2.6	17-Jan-2005	skrll	Sync with HEAD.
1.61.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.61.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.61.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.61.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.61.2.1	03-Aug-2004	skrll	Sync with HEAD
1.68.4.1	24-Jan-2005	he	Pull up revision 1.70 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.71.4.1	12-Feb-2005	yamt	sync with head.
1.71.2.1	29-Apr-2005	kent	sync with -current
1.74.6.7	24-Mar-2008	yamt	sync with head.
1.74.6.6	17-Mar-2008	yamt	sync with head.
1.74.6.5	21-Jan-2008	yamt	sync with head
1.74.6.4	27-Oct-2007	yamt	sync with head.
1.74.6.3	03-Sep-2007	yamt	sync with head.
1.74.6.2	30-Dec-2006	yamt	sync with head.
1.74.6.1	21-Jun-2006	yamt	sync with head.
1.75.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.75.14.1	19-Jun-2006	chap	Sync with head.
1.75.8.1	26-Jun-2006	yamt	sync with head.
1.75.4.1	09-Sep-2006	rpaulo	sync with head
1.76.6.2	10-Dec-2006	yamt	sync with head.
1.76.6.1	22-Oct-2006	yamt	sync with head
1.76.4.1	18-Nov-2006	ad	Sync with head.
1.84.8.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.84.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.84.2.1	17-Jun-2007	riz	Pull up following revision(s) (requested by jmcneill in ticket #725): sys/dev/pci/if_vr.c: revision 1.86 Add VT6105M (Rhine III) to the list of supported devices, as found in the Soekris net5501.
1.85.4.1	11-Jul-2007	mjf	Sync with head.
1.85.2.3	23-Oct-2007	ad	Sync with head.
1.85.2.2	15-Jul-2007	ad	Sync with head.
1.85.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.87.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.87.8.2	23-Mar-2008	matt	sync with HEAD
1.87.8.1	06-Nov-2007	matt	sync with HEAD
1.87.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.88.8.1	20-Jan-2008	bouyer	Sync with HEAD
1.88.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.89.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.89.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.89.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.89.2.1	24-Mar-2008	keiichi	sync with head.
1.92.4.6	11-Aug-2010	yamt	sync with head.
1.92.4.5	11-Mar-2010	yamt	sync with head
1.92.4.4	16-Sep-2009	yamt	sync with head
1.92.4.3	16-May-2009	yamt	sync with head
1.92.4.2	04-May-2009	yamt	sync with head.
1.92.4.1	16-May-2008	yamt	sync with head.
1.92.2.1	18-May-2008	yamt	sync with head.
1.93.6.1	19-Oct-2008	haad	Sync with HEAD.
1.93.4.1	18-Jul-2008	simonb	Sync with head.
1.93.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.95.12.1	21-Apr-2010	matt	sync to netbsd-5
1.95.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.95.4.2	03-Oct-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #1022): sys/dev/pci/if_vr.c: revision 1.99 add suspend/resume support
1.95.4.1	03-Oct-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #1022): sys/dev/pci/if_vr.c: revision 1.98 Replace shutdownhook_establish(9) with pmf_device_register1(9). Tested VIA VT86C100A (which is probed as VT3043).
1.102.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.103.2.2	05-Mar-2011	rmind	sync with head
1.103.2.1	30-May-2010	rmind	sync with head
1.107.8.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.107.8.3	23-Jan-2013	yamt	sync with head
1.107.8.2	30-Oct-2012	yamt	sync with head
1.107.8.1	17-Apr-2012	yamt	sync with head
1.108.2.1	18-Feb-2012	mrg	merge to -current.
1.110.2.1	08-Feb-2013	riz	Pull up following revision(s) (requested by taca in ticket #783): sys/dev/pci/if_vr.c: revision 1.112 - reset the chip if the tx engine gets stuck after a link state change, from OpenBSD - no need to do a full reset of the chip when enabling or disabling promiscuous mode
1.111.2.3	03-Dec-2017	jdolecek	update from HEAD
1.111.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.111.2.1	25-Feb-2013	tls	resync with head
1.112.2.1	18-May-2014	rmind	sync with head
1.113.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.114.4.5	05-Feb-2017	skrll	Sync with HEAD
1.114.4.4	05-Oct-2016	skrll	Sync with HEAD
1.114.4.3	09-Jul-2016	skrll	Sync with HEAD
1.114.4.2	19-Mar-2016	skrll	Sync with HEAD
1.114.4.1	06-Jun-2015	skrll	Sync with HEAD
1.119.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.123.2.3	26-Jan-2019	pgoyette	Sync with HEAD
1.123.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.123.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.124.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.124.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.124.2.1	10-Jun-2019	christos	Sync with HEAD
1.131.2.1	29-Feb-2020	ad	Sync with head.
1.133.10.2	01-Aug-2021	thorpej	Sync with HEAD.
1.133.10.1	13-May-2021	thorpej	Sync with HEAD.
1.17	28-Feb-2018	flxd	Cosmetic fix for the Physical Address Registers PAR0 to PAR5.
1.16	05-Nov-2006	tsutsui	Use uintNN_t.
1.15	03-Nov-2006	tsutsui	Pull several fixes which improve TX error handling from FreeBSD's if_vr.c rev 1.52: - check more error status in TX descriptor and restart TX module appropriately in vr_txeof() - check more error interrupt status in vr_intr() I can't confirm whether these changes actually fix TX stalls because I can't reproduce the problem I had about seven years ago (I guess it might be caused by excessive collisions on a dumb hub), but at least they don't seem to have bad side effects on normal operations on my macppc.
1.14	02-Nov-2006	tsutsui	- make DMA descriptor members volatile - set VR_RXSTAT_OWN after all other descriptor data is set With these changes, vr_init() seems to initialize the vr chip stuck on RX properly.
1.13	11-Dec-2005	christos	branches: 1.13.20; 1.13.22; merge ktrace-lwp.
1.12	17-Oct-2003	tsutsui	branches: 1.12.16; - No need to set VR_TXCTL_TLINK in vr_start() because all TX buffers have only single segment. - No need to set VR_CMD_TX_ON in vr_start(). - Initialize BCR0 and BCR1 registers. - Change RX DMA threshold 128bytes. Mostly from FreeBSD via OpenBSD.
1.11	17-Oct-2003	tsutsui	Use common PCI macro/functions for PWRMGMT.
1.10	03-Jan-2003	lha	branches: 1.10.2; From freebsd if_vr.c:1.43,1.56 and if_vrreg.h:1.11,1.14 - Windows put the chip in suspended mode, make sure we unsuspend it. 1.43, by Takefumi SAYO <stake@po.shiojiri.ne.jp> - Detect the revision of the Rhine chip we're using, and force reset when the chip supports it. 1.65, by silby@freebsd.org
1.9	12-Feb-1999	thorpej	branches: 1.9.20; Garbage-collect some unneeded definitions.
1.8	05-Feb-1999	thorpej	Rename VR_MIICMD_DATAIN and VR_MIICMD_DATAOUT to actually match what the data sheet says. (The names and their usage were exactly reversed in the code.)
1.7	02-Feb-1999	thorpej	Use pcidevs.h to get vendor and product IDs.
1.6	02-Feb-1999	thorpej	Take software-only structure definitions out of the register description header.
1.5	02-Feb-1999	thorpej	Remove support for FreeBSD from the driver; so many changes are going to be required to make it properly portable to all of NetBSD's platforms, that it just wouldn't be worth the effort.
1.4	26-Jan-1999	sakamoto	NetBSD RCS id.
1.3	22-Jan-1999	sakamoto	Change bootverbose 1 to 0. Add vr_lookup.
1.2	21-Jan-1999	sakamoto	Add NetBSD autoconfiguration support. KNF.
1.1	21-Jan-1999	sakamoto	branches: 1.1.1; Initial revision
1.1.1.1	21-Jan-1999	sakamoto	Import FreeBSD's VIA Rhine fast ethernet PCI NIC driver written by Bill Paul <wpaul@ctr.columbia.edu>.
1.9.20.1	07-Jan-2003	thorpej	Sync with HEAD.
1.10.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.12.16.1	30-Dec-2006	yamt	sync with head.
1.13.22.1	10-Dec-2006	yamt	sync with head.
1.13.20.1	18-Nov-2006	ad	Sync with head.
1.39	24-Feb-2025	andvar	fix various typos in comments.
1.38	05-Sep-2024	andvar	s/vte_rxoef/vte_rxeof/ in debug message.
1.37	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.36	29-Jun-2024	riastradh	branches: 1.36.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.35	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.34	17-Sep-2022	thorpej	vte_ifstart(): Make the condition if the transmit loop the number of free descriptors, not "are there packets in the queue"; the latter is redundant with the IFQ_POLL() that happens within the loop after the number of available descriptors is checked. Eliminate use of IFF_OACTIVE.
1.33	03-Sep-2022	thorpej	Garbage-collect the remaining vestiges of netisr.
1.32	30-Aug-2021	andvar	Restore original MDC speed control register value after MAC reset, if it wasn't default. Fixes PR port-i386/53494. ok riastradh
1.31	07-Feb-2020	thorpej	Use callout_setfunc() / callout_schedule().
1.30	04-Feb-2020	thorpej	Use ifmedia_fini().
1.29	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.28	12-Nov-2019	maya	branches: 1.28.2; Fix whitespace. From vezhlys
1.27	30-Oct-2019	msaitoh	if_percpuq(9) automatically increments if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.26	28-May-2019	msaitoh	branches: 1.26.2; Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.25	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.24	23-May-2019	msaitoh	No functional change: - Simplify MII structure initialization and reference. - u_int_t -> uint_t. - KNF
1.23	05-Feb-2019	msaitoh	Remove very old IFF_NOTRAILERS flag.
1.22	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.21	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.20	26-Jun-2018	msaitoh	branches: 1.20.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.19	22-Jun-2018	msaitoh	It's not required to include net/bpfdesc.h. Remove it.
1.18	26-Jun-2017	christos	branches: 1.18.4; PR/52333: Jia-Ju Bai: explicitly pass BUS_DMA_NOWAIT to bus_dmamap_load_mbuf() called from an interrupt context. It should not matter since the maps are constructed with BUS_DMA_ALLOCNOW, but :-)
1.17	23-May-2017	ozaki-r	branches: 1.17.2; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.16	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.15	11-Jul-2016	msaitoh	branches: 1.15.2; KNF. No functional change.
1.14	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.13	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.12	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.11	10-Aug-2014	tls	branches: 1.11.4; Merge tls-earlyentropy branch into HEAD.
1.10	29-Mar-2014	christos	branches: 1.10.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.9	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.8	30-Mar-2013	christos	branches: 1.8.4; remove trailing whitespace
1.7	22-Jul-2012	matt	branches: 1.7.2; Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.6	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.5	02-Feb-2012	tls	Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.4	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.3	28-Apr-2011	bouyer	branches: 1.3.4; 1.3.6; 1.3.10; - Properly count collisions - call vte_rxfilter() after starting transmit/receive; this seems to fix some IPv6-related multicast issues for me - when the MCR0_BROADCAST bit is set broadcast receive is disabled. Rename it to MCR0_BROADCAST_DIS to reflect its function, and fix usage. - Fix loop over multicast addresses to not set the same address in the 3 perfect filter slots.
1.2	02-Apr-2011	mbalmer	Fix misplaced parenthesis. From henning.petersen@t-online.de, thanks.
1.1	26-Jan-2011	bouyer	branches: 1.1.2; 1.1.4; Add a driver for the RDC Semiconductor RDC R6040 10/100 Ethernet controller, as found in the vortex86 SoCs (http://www.vortex86dx.com). Ported from freebsd. Not added to amd64's GENERIC because this CPU is 32bit only. thanks to DM&P Electronics, Inc for providing documentation and sample devices for this work.
1.1.4.4	31-May-2011	rmind	sync with head
1.1.4.3	21-Apr-2011	rmind	sync with head
1.1.4.2	05-Mar-2011	rmind	sync with head
1.1.4.1	26-Jan-2011	rmind	file if_vte.c was added on branch rmind-uvmplock on 2011-03-05 20:53:45 +0000
1.1.2.2	08-Feb-2011	bouyer	Sync with HEAD
1.1.2.1	26-Jan-2011	bouyer	file if_vte.c was added on branch bouyer-quota2 on 2011-02-08 16:19:50 +0000
1.3.10.1	18-Feb-2012	mrg	merge to -current.
1.3.6.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.6.2	30-Oct-2012	yamt	sync with head
1.3.6.1	17-Apr-2012	yamt	sync with head
1.3.4.2	06-Jun-2011	jruoho	Sync with HEAD.
1.3.4.1	28-Apr-2011	jruoho	file if_vte.c was added on branch jruoho-x86intr on 2011-06-06 09:08:15 +0000
1.7.2.3	03-Dec-2017	jdolecek	update from HEAD
1.7.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.7.2.1	23-Jun-2013	tls	resync from head
1.8.4.1	18-May-2014	rmind	sync with head
1.10.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.11.4.6	28-Aug-2017	skrll	Sync with HEAD
1.11.4.5	05-Feb-2017	skrll	Sync with HEAD
1.11.4.4	05-Oct-2016	skrll	Sync with HEAD
1.11.4.3	09-Jul-2016	skrll	Sync with HEAD
1.11.4.2	19-Mar-2016	skrll	Sync with HEAD
1.11.4.1	06-Jun-2015	skrll	Sync with HEAD
1.15.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.17.2.3	03-Sep-2021	martin	Pull up following revision(s) (requested by andvar in ticket #1693): sys/dev/pci/if_vte.c: revision 1.32 Restore original MDC speed control register value after MAC reset, if it wasn't default. Fixes PR port-i386/53494. ok riastradh
1.17.2.2	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1427): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.17.2.1	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.18.4.4	26-Jan-2019	pgoyette	Sync with HEAD
1.18.4.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.18.4.2	28-Jul-2018	pgoyette	Sync with HEAD
1.18.4.1	25-Jun-2018	pgoyette	Sync with HEAD
1.20.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.20.2.1	10-Jun-2019	christos	Sync with HEAD
1.26.2.2	03-Sep-2021	martin	Pull up following revision(s) (requested by andvar in ticket #1339): sys/dev/pci/if_vte.c: revision 1.32 Restore original MDC speed control register value after MAC reset, if it wasn't default. Fixes PR port-i386/53494. ok riastradh
1.26.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #403): sys/arch/arm/broadcom/bcm53xx_eth.c: revision 1.39 sys/dev/pcmcia/if_xi.c: revision 1.91 sys/dev/ic/aic6915.c: revision 1.40 sys/dev/pci/if_tl.c: revision 1.117 sys/arch/arm/gemini/gemini_gmac.c: revision 1.18 sys/dev/ic/elinkxl.c: revision 1.133 sys/dev/pci/if_ste.c: revision 1.57 sys/dev/pci/if_alc.c: revision 1.43 sys/dev/pci/if_stge.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.34 sys/dev/pci/if_age.c: revision 1.62 sys/dev/pci/if_txp.c: revision 1.60 sys/dev/ic/i82557.c: revision 1.156 sys/dev/pci/if_vte.c: revision 1.27 sys/arch/powerpc/booke/dev/pq3etsec.c: revision 1.47 sys/arch/arm/gemini/if_gpn.c: revision 1.13 if_percpuq(9) and ether_input() automatically increment if_ipackets, so don't add number of RX frames from device's statistics counter to if_ipackets to avoid double count.
1.28.2.1	29-Feb-2020	ad	Sync with head.
1.36.2.1	02-Aug-2025	perseant	Sync with HEAD
1.2	28-Apr-2011	bouyer	branches: 1.2.4; - Properly count collisions - call vte_rxfilter() after starting transmit/receive; this seems to fix some IPv6-related multicast issues for me - when the MCR0_BROADCAST bit is set broadcast receive is disabled. Rename it to MCR0_BROADCAST_DIS to reflect its function, and fix usage. - Fix loop over multicast addresses to not set the same address in the 3 perfect filter slots.
1.1	26-Jan-2011	bouyer	branches: 1.1.2; 1.1.4; Add a driver for the RDC Semiconductor RDC R6040 10/100 Ethernet controller, as found in the vortex86 SoCs (http://www.vortex86dx.com). Ported from freebsd. Not added to amd64's GENERIC because this CPU is 32bit only. thanks to DM&P Electronics, Inc for providing documentation and sample devices for this work.
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	05-Mar-2011	rmind	sync with head
1.1.4.1	26-Jan-2011	rmind	file if_vtereg.h was added on branch rmind-uvmplock on 2011-03-05 20:53:45 +0000
1.1.2.2	08-Feb-2011	bouyer	Sync with HEAD
1.1.2.1	26-Jan-2011	bouyer	file if_vtereg.h was added on branch bouyer-quota2 on 2011-02-08 16:19:50 +0000
1.2.4.2	06-Jun-2011	jruoho	Sync with HEAD.
1.2.4.1	28-Apr-2011	jruoho	file if_vtereg.h was added on branch jruoho-x86intr on 2011-06-06 09:08:15 +0000
1.5	01-Feb-2020	thorpej	Adopt <net/if_stats.h>.
1.4	14-Apr-2015	riastradh	branches: 1.4.18; 1.4.26; Include <sys/rndsource.h> where it is actually used. I had removed <sys/rnd.h> from files that didn't mention anything of the rnd(9) API. But they included other files which assumed <sys/rnd.h> had already been included.
1.3	02-Feb-2012	tls	branches: 1.3.6; 1.3.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.2	19-Nov-2011	tls	branches: 1.2.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.1	26-Jan-2011	bouyer	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; Add a driver for the RDC Semiconductor RDC R6040 10/100 Ethernet controller, as found in the vortex86 SoCs (http://www.vortex86dx.com). Ported from freebsd. Not added to amd64's GENERIC because this CPU is 32bit only. thanks to DM&P Electronics, Inc for providing documentation and sample devices for this work.
1.1.10.1	17-Apr-2012	yamt	sync with head
1.1.8.2	06-Jun-2011	jruoho	Sync with HEAD.
1.1.8.1	26-Jan-2011	jruoho	file if_vtevar.h was added on branch jruoho-x86intr on 2011-06-06 09:08:15 +0000
1.1.4.2	05-Mar-2011	rmind	sync with head
1.1.4.1	26-Jan-2011	rmind	file if_vtevar.h was added on branch rmind-uvmplock on 2011-03-05 20:53:45 +0000
1.1.2.2	08-Feb-2011	bouyer	Sync with HEAD
1.1.2.1	26-Jan-2011	bouyer	file if_vtevar.h was added on branch bouyer-quota2 on 2011-02-08 16:19:50 +0000
1.2.2.1	18-Feb-2012	mrg	merge to -current.
1.3.24.1	06-Jun-2015	skrll	Sync with HEAD
1.3.6.1	03-Dec-2017	jdolecek	update from HEAD
1.4.26.1	29-Feb-2020	ad	Sync with head.
1.4.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.57	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.56	14-Jul-2016	msaitoh	branches: 1.56.16; 1.56.18; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.55	29-Mar-2014	christos	branches: 1.55.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.54	30-Jan-2012	drochner	branches: 1.54.6; 1.54.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.53	26-Jul-2011	dyoung	branches: 1.53.2; 1.53.6; Replace anonymous constants, 0x10, 0x14, ..., with PCI_BAR(0), PCI_BAR(1), .... There was no change in the generated assembly. I used this semantic patch: @ mapsit @ identifier bar; expression pact; @@ ( pci_mapreg_map \| Cardbus_mapreg_map )(pact, bar, ...) @ depends on mapsit @ identifier mapsit.bar; @@ ( - #define bar 0x10 + #define bar PCI_BAR(0) \| - #define bar 0x14 + #define bar PCI_BAR(1) \| - #define bar 0x18 + #define bar PCI_BAR(2) \| - #define bar 0x1C + #define bar PCI_BAR(3) \| - #define bar 0x20 + #define bar PCI_BAR(4) )
1.52	23-Nov-2010	christos	fix typos
1.51	23-Nov-2010	christos	split if_wi
1.50	26-Nov-2009	njoly	branches: 1.50.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.49	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability. XXX: should these pmf(9) calls be moved into MI attach functions XXX: using function pointers for suspend and resume passed via softc?
1.48	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.47	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.46	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.45	28-Apr-2008	martin	branches: 1.45.14; Remove clause 3 and 4 from TNF licenses
1.44	10-Apr-2008	cegger	branches: 1.44.2; 1.44.4; use aprint_*_dev and device_xname
1.43	09-Dec-2007	jmcneill	branches: 1.43.10; Merge jmcneill-pm branch.
1.42	19-Oct-2007	ad	branches: 1.42.4; 1.42.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.41	16-Nov-2006	christos	branches: 1.41.8; 1.41.22; 1.41.24; 1.41.28; __unused removal on arguments; approved by core.
1.40	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.39	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.38	11-Dec-2005	christos	branches: 1.38.20; 1.38.22; merge ktrace-lwp.
1.37	22-Jun-2005	dyoung	branches: 1.37.2; Resolve conflicts in importation of 18-May-2005 ath(4) / net80211(9) from FreeBSD. Introduce compatibility shims (sys/dev/ic/ath_netbsd.[ch], sys/net80211/ieee80211_netbsd.[ch]). Update drivers (an, atu, atw, awi, ipw, iwi, rtw, wi) for the new net80211(9) API.
1.36	27-Feb-2005	perry	nuke trailing whitespace
1.35	21-Aug-2004	thorpej	branches: 1.35.4; 1.35.6; Use ANSI function decls and make use of static.
1.34	07-Aug-2004	mycroft	Don't pass a MAC address to wi_attach() -- yet.
1.33	17-Jul-2004	mycroft	Remove special-casing of Intersil product name.
1.32	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.31	28-Mar-2004	nakayama	Avoid use of wi_pci_reset when it uses other chip like PLX bridge chip. This solves PCI bus error on sparc64 with Eumitcom WL11000P.
1.30	14-Dec-2003	dyoung	PR 23428: support National Datacomm Corporation NCP130A2 board. Thanks T. M. Pederson for sending the patches.
1.29	07-Dec-2003	dyoung	Enable RSSI-based rate-adaptation for wi(4). This enables wi(4)-based APs to run at "full speed" where before they ran at just 1 or 2Mb/s. The AP will adapt the data rate for each client based on packet losses and the received signal strength. I have also enabled rate adaptation for STA mode and for IBSS mode, but the hardware gives us less control over the data rate in those modes.
1.28	16-Nov-2003	dyoung	Add data-link type DLT_IEEE802_11_RADIO to wi and atw. DLT_IEEE802_11_RADIO lets you monitor radio stats like received signal strength, which diversity antenna was used, channel/frequency, modulation, and data rate.
1.27	13-Oct-2003	dyoung	Switch wi(4) to the new 802.11 layer.
1.26	29-Sep-2003	scw	Copy our pci_chipset_tag_t into wi_pci_softc instead of a pointer to pci_attach_args. The latter is allocated on the stack during auto- configuration and so will not be valid after that time. It's amazing how the old code worked for so long. I guess pci_attach_args is allocated deep in pid#0's kernel stack on most platforms.
1.25	29-Jun-2003	fvdl	branches: 1.25.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.24	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.23	12-Apr-2003	christos	PR/21108: Christopher SEKIYA: if_wi_pci.c rev 1.21 is incorrect for PLX-9052
1.22	29-Mar-2003	jdc	Add support for USRobotics Wireless Access PCI adapter (model 132415). From Sergio Jimenez in PR kern/20642.
1.21	29-Mar-2003	dyoung	Support so-called "PLX" bridge for wi(4). From FreeBSD via Simon Burge. Simon writes, "Here's the patch that Grant and I have been using on a combinations of architectures and both NetBSD 1.6 and -current. "This is based on FreeBSD - look for /* We have to do a magic PLX poke to enable interrupts */ in http://www.freebsd.org/cgi/cvsweb.cgi/src/sys/dev/wi/if_wi_pci.c
1.20	28-Mar-2003	perry	Fix PR/20903 for real. Identical patches supplied by: Masuda Hideo <h-masuda@ootani.nagata.kobe.jp> Christopher SEKIYA <wileyc@rezrov.net>
1.19	27-Mar-2003	perry	ifdef out something dependent on PCI_VENDOR_COREGA for now.
1.18	27-Mar-2003	perry	Tell driver to accept the Corega CGWLPCIA11. Submitted by h-masuda@ootani.nagata.kobe.jp in PR #20903
1.17	27-Mar-2003	dyoung	Introduced sc_reset to wi_softc. The bus-specific front-end gets to fill this.
1.16	02-Jan-2003	dyoung	Since the I/O and memory enable bits are written to the PCI CSR by the PCI bus code, and because wi(4) does not use bus-mastering, we may entirely skip writing the PCI CSR.
1.15	01-Jan-2003	dyoung	For confidence's sake, enable I/O and memory accesses (which were previously NOT enabled), but do not enable bus-mastering (which was previously enabled, but which neither Lucent nor Prism support).
1.14	31-Dec-2002	dyoung	Per discussion in PR 18794, fix PCI resets. Alas, this does not seem to fix PR 18794.
1.13	21-Oct-2002	soren	WI_PCI_CBMA belongs in dev/pci.
1.12	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.11	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.10	30-Sep-2002	onoe	Nuke unneeded member sc_ifp and use #define sc_if instead, to prepare jumbo commit for wi.c
1.9	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.8	23-Sep-2002	thorpej	Make the PCI reset more robust. From David Young <dyoung@ojctech.com>.
1.7	04-Apr-2002	jdolecek	branches: 1.7.4; move wi_pci_reset() from ic/wi.c to pci/if_wi_pci.c, where is belongs
1.6	04-Apr-2002	jdolecek	fix sligh confusion on whether or not to supply dv_xname to printf() in error cases of attach Noticed in kern/16197 by George E Sollish.
1.5	15-Jan-2002	augustss	Note to self: set variable before using it.
1.4	13-Jan-2002	augustss	Cosmetic error message fix.
1.3	13-Jan-2002	augustss	Add code to handle PCMCIA cards placed on a PCI card with the PLX chip. From OpenBSD and FreeBSD.
1.2	13-Nov-2001	lukem	branches: 1.2.2; add RCSID
1.1	13-Oct-2001	ichiro	branches: 1.1.2; Add Intersil Prism2.5 Mini-PCI wavelan.
1.1.2.10	03-Jan-2003	thorpej	Sync with HEAD.
1.1.2.9	11-Nov-2002	nathanw	Catch up to -current
1.1.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.7	05-Oct-2002	gmcgarry	Remove CVS conflicts.
1.1.2.6	23-Sep-2002	thorpej	Sync with HEAD.
1.1.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.1.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.1.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.1.2.1	13-Oct-2001	nathanw	file if_wi_pci.c was added on branch nathanw_sa on 2001-10-22 20:41:24 +0000
1.2.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.2.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.1	13-Nov-2001	thorpej	file if_wi_pci.c was added on branch kqueue on 2002-01-10 19:56:47 +0000
1.7.4.1	28-Jul-2003	he	Pull up revisions 1.21,1.23 (via patch, requested by grant in ticket #1281): Support the so-called ``PLX'' bridge for wi(4).
1.25.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.25.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.25.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.25.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.25.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.25.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.25.2.2	03-Aug-2004	skrll	Sync with HEAD
1.25.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.35.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.35.4.1	29-Apr-2005	kent	sync with -current
1.37.2.3	21-Jan-2008	yamt	sync with head
1.37.2.2	27-Oct-2007	yamt	sync with head.
1.37.2.1	30-Dec-2006	yamt	sync with head.
1.38.22.2	10-Dec-2006	yamt	sync with head.
1.38.22.1	22-Oct-2006	yamt	sync with head
1.38.20.1	18-Nov-2006	ad	Sync with head.
1.41.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.41.24.2	09-Jan-2008	matt	sync with HEAD
1.41.24.1	06-Nov-2007	matt	sync with HEAD
1.41.22.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.41.22.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.41.22.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.41.22.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.41.22.1	04-Sep-2007	joerg	Convert wi(4) to jmcneill-pm style power management.
1.41.8.1	23-Oct-2007	ad	Sync with head.
1.42.6.1	11-Dec-2007	yamt	sync with head.
1.42.4.1	26-Dec-2007	ad	Sync with head.
1.43.10.1	02-Jun-2008	mjf	Sync with HEAD.
1.44.4.4	11-Mar-2010	yamt	sync with head
1.44.4.3	16-Sep-2009	yamt	sync with head
1.44.4.2	16-May-2009	yamt	sync with head
1.44.4.1	16-May-2008	yamt	sync with head.
1.44.2.1	18-May-2008	yamt	sync with head.
1.45.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.50.4.1	05-Mar-2011	rmind	sync with head
1.53.6.1	18-Feb-2012	mrg	merge to -current.
1.53.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.2.1	17-Apr-2012	yamt	sync with head
1.54.10.1	18-May-2014	rmind	sync with head
1.54.6.2	03-Dec-2017	jdolecek	update from HEAD
1.54.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.6.1	05-Oct-2016	skrll	Sync with HEAD
1.56.18.1	10-Jun-2019	christos	Sync with HEAD
1.56.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.801	10-Nov-2024	mlelstv	Add MBUFTRACE
1.800	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.799	29-Jun-2024	riastradh	branches: 1.799.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.798	21-Feb-2024	msaitoh	wm(4): Fix upper 16bit of Image Unique ID(EtrackID). Don't override uid1 variable while reading option ROM version to print Image Unique ID correctly.
1.797	29-Jan-2024	msaitoh	wm(4): Fix compile error without WM_EVENT_COUNTERS.
1.796	29-Jan-2024	msaitoh	wm(4): Drop frames if the RX descriptor ring has no room on multiqueue system. Drop frames if the RX descriptor ring has no room. This is enabled only on multiqueue system to avoid bad influence to other queues. The drop count can be seen by the RQDPC counter (wmN rxqXXdrop evcnt).
1.795	29-Jan-2024	msaitoh	wm(4): Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. The iqdrops counter should include not only MPC(Missed Packet Count) but also RQDPC(Receive Queue Drop Packet Count). Same as ixgbe(4) and igc(4). Note that the RQDPC is not currently counted because SRRCTL_DROP_EN is not set.
1.794	26-Jan-2024	msaitoh	wm(4): Print RX packet buffer size.
1.793	18-Jan-2024	msaitoh	Add Intel I219-{LM,V}(20,21) support.
1.792	21-Nov-2023	gutteridge	if_wm.c: fix spelling and grammar in a comment
1.791	11-Oct-2023	msaitoh	wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.790	11-Oct-2023	msaitoh	wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss.
1.789	25-Sep-2023	rin	wm(4): Replace /* nothing */ to __nothing for evcnt macros Prevent empty if bodies for !WM_EVENT_COUNTERS.
1.788	25-Sep-2023	msaitoh	wm(4): Modify descriptions of flow control related event counters.
1.787	25-Sep-2023	msaitoh	wm(4): the IAC(Interrupt Assertion Count) register is for 82571 and newer.
1.786	25-Sep-2023	msaitoh	wm(4): Add some info to some event counters. - The difference between the RUC(Rx Undersize) and RFC(Rx Fragment) is whether the CRC is valid or not. - The difference between the ROC(Rx Oversize) and RJC(Rx Jabber) is whether the CRC is valid or not. - LENERRS(Length Errors) is for the length/type field <= 1500.
1.785	25-Sep-2023	msaitoh	wm(4): Use SCVPC and HRMPC for 82575 and newer. The SCVPC(SerDes/SGMII Code Violation Packet Count) register and the HRMPC (Header Redirection Missed Packet Count) register were added in if_wm.c rev. 1.776 but the location in the code were incorrect. Fix them.
1.784	25-Aug-2023	msaitoh	Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.783	25-Aug-2023	msaitoh	Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@.
1.782	23-Jun-2023	msaitoh	Add missing read to count Circuit Breaker Rx Dropped Packet correctly.
1.781	11-May-2023	msaitoh	Fix typo. s/ictxact/ictxatc/. No functional change.
1.780	11-May-2023	msaitoh	Use WM_IS_ICHPCH(). No functional change.
1.779	11-May-2023	msaitoh	Fix prc511's comment and description.
1.778	11-May-2023	msaitoh	Add SOICZIFDATA (ifconfig -z) support for evcnt(9). First update the statistics data, then clear the event counters, and finally copy and clear if_data via ether_ioctl().
1.777	11-May-2023	msaitoh	Move statistics updating code from wm_tick() to new wm_update_stats(). - To reuse. - No functional change.
1.776	11-May-2023	msaitoh	Add some new event counters. Add the following counters for 82575 and newer except 80003, ICHs and PCHs: - Only 82576 document describes about the circuit breaker, so the following two might be only for 82575: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - 82575's document doesn't describe the following two, but we can see the same value as GO{T,R}C have: - Host Good Octets RX - Host Good Octets TX - 82575's document doesn't describe the LENERRS (Length Errors) counter. I don't know if it has. - Perhaps Non-SerDes/SGMII devices don't have SCVPC (SerDes/SGMII Code Violation Packet) register. We don't care if it's SerDes/SGMII or not for now. - HRMPC (Header Redirection Missed Packet) appears only once in 8257[56]'s datasheet. FreeBSD's igb counts it, so we do, too. - Count the following two for I350 and newer. I don't know if PCHs have: - EEE TX LPI - EEE RX LPI
1.775	11-May-2023	msaitoh	Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero.
1.774	11-May-2023	msaitoh	Don't add "Count" for event counter's description.
1.773	11-May-2023	msaitoh	Fix a bug that the transmit underrun counter is incorrectly counted. The transmit underrun bit in the transmit status filed is only for 82544 (and older?), so don't use the counter for newer chips. The bit is reserved for newer chips, but the bit sometimes set on 82575 at least.
1.772	11-May-2023	msaitoh	IC{TX,RX}*C registers are for older than 82575.
1.771	11-May-2023	msaitoh	Sort lines. No functional change. Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match.
1.770	11-May-2023	msaitoh	Sort lines. No functional change. Rearrange the order of the registers so that they are roughly in ascending order.
1.769	11-May-2023	msaitoh	Add note for the TORL register. The TOR register includes error, flow control and broadcast rejected.
1.768	11-May-2023	msaitoh	Count some 64bit counters correctly. - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Found by knakahara.
1.767	08-Dec-2022	knakahara	branches: 1.767.2; Workaround for some hypervisor environments. The environments cannot stop e1000 interrupt immediately.
1.766	26-Oct-2022	msaitoh	Add comment to ICH/PCH's lock stuff in wm_reset().
1.765	19-Oct-2022	msaitoh	Fix typo. No functional change. OK'd by knakahara.
1.764	08-Sep-2022	msaitoh	KNF. Modify comment. No fucntional change.
1.763	12-Aug-2022	riastradh	wm(4): Remove the non-MP-safe scaffolding. - Where we had #ifndef WM_MPSAFE splnet(), we also had WM_CORE_LOCK, which implies splnet, so just remove the conditional splnet. - Make the core lock unconditional and remove macro indirections. - Pass CALLOUT_MPSAFE, SOFTINT_MPSAFE, and WQ_MPSAFE directly without macro indirections.
1.762	12-Aug-2022	riastradh	wm(4): Audit sc->phy.acquire and sc->nvm.acquire. The return value must be used, because some of the acquire/release pairs hold a mutex from acquire to release, and failing to call release, or calling release twice, leads to an inconsistent state.
1.761	12-Aug-2022	riastradh	wm: use a workqueue to reset the interface when wm_watchdog determines it needs a reset. Author: Nick Hudson <skrll@netbsd.org> Committer: Taylor R Campbell <riastradh@NetBSD.org>
1.760	12-Aug-2022	riastradh	wm(4): if_flags and IFNET_LOCK audit Don't touch if_flags without IFNET_LOCK: - If only core lock is held, use sc_if_flags. - If only txq lock is held, use txq_stopping. => Verified all paths guarantee !txq_stopping, so assert. - Make sure sc_if_flags is updated on stop. - Make wm_init fail once we enter wm_detach. - Sprinkle assertions. Author: Taylor R Campbell <riastradh@NetBSD.org>
1.759	12-Aug-2022	riastradh	wm(4): Revert previous -- mistakenly committed unsquashed change part.
1.758	12-Aug-2022	riastradh	wm(4): if_flags and IFNET_LOCK audit Don't touch if_flags without IFNET_LOCK: - If only core lock is held, use sc_if_flags. - If only txq lock is held, use txq_stopping. => Verified all paths guarantee !txq_stopping, so assert. - Make sure sc_if_flags is updated on stop. - Sprinkle assertions.
1.757	08-Aug-2022	msaitoh	To avoid releasing mutex temporally, use new wm_set_mdio_slow_mode_hv_locked().
1.756	08-Aug-2022	msaitoh	Consistency use -1 instead of 1 for some error code. Advised by knakahara.
1.755	08-Aug-2022	msaitoh	Pass an error code correctly if phy.acquire() failed. This is not a real bug because the return value is not used.
1.754	08-Aug-2022	msaitoh	Pass an error code to the upper layer instead of -1. Advised by knakahara.
1.753	08-Aug-2022	msaitoh	Modify debug messages so that we can determine where it happened.
1.752	05-Aug-2022	skrll	Sprinkle const on splfoo call results.
1.751	03-Aug-2022	skrll	Add a KASSERT for the locking protocol in wm_ioctl. Read the interface up/down status from sc_if_flags (under WM_CORE_LOCK) when deciding if the multicast filter needs to be updated. Discussed with msaitoh@, knakahara@ and riastradh@
1.750	03-Aug-2022	skrll	Add some KASSERTs around the locking protocol. Discussed with msaitoh@, knakahara@ and riastradh@
1.749	29-Jul-2022	skrll	Improve a couple of comments
1.748	29-Jul-2022	skrll	Trailing whitespace
1.747	28-Jul-2022	skrll	KNF the KNF
1.746	22-Jul-2022	msaitoh	Add more statistics countes. - Add many statics counters that the chip has. - Attach event counters only if available.
1.745	19-Jul-2022	riastradh	wm(4): Use CSR_WRITE_FLUSH, not bus_space_barrier. The bus space is mapped non-prefetchable and non-cacheable, so there is no need for bus_space_barrier and it might not even have any effect -- bus_space_read/write are totally ordered for non- prefetchable, non-cacheable mappings anyway. Presumably the intent here was to wait for acknowledgment of the write from the device -- which bus_space_barrier does not guarantee, and which I expect requires a read transaction as CSR_WRITE_FLUSH performs. ok msaitoh@, knakahara@
1.744	19-Jul-2022	riastradh	wm(4): Use bus_dmamap_sync, not bus_space_barrier, for tx desc. The newly written tx descriptor needs to be visible to the device before the TDT register is set. There's no bus space mappings that need any barriers, though -- nothing mapped prefetchable or cacheable here. So bus_dmamap_sync (via wm_cdtxsync) is necessary, not bus_space_barrier. By accident, bus_space_barrier implies the same barrier instructions as bus_dmamap_sync on some architectures, but it wouldn't work at all if we were bouncing. ok msaitoh@, knakahara@
1.743	19-Jul-2022	skrll	KNF
1.742	19-Jul-2022	skrll	Don't need sys/malloc.h so remove the #include
1.741	19-Jul-2022	skrll	Turn a locking botch (shouldn't drop lock on error) into a KASSERT
1.740	11-Jul-2022	msaitoh	KNF. No functional change.
1.739	11-Jul-2022	msaitoh	Consistently use semicolon. No functional change.
1.738	06-Jul-2022	msaitoh	Call txeof first, then rxeof for the consistency. There are three functions where the txeof and rxeof are called. The MSI-X interrupt function and the workqueue function call rxeof first, then rxeof. For legacy interrupt. rxeof is called first. Modify it to match with other two.
1.737	06-Jul-2022	msaitoh	Fix compile error if WM_DEBUG is set.
1.736	04-Jul-2022	skrll	Be less tautological and more explanatory in a comment.
1.735	02-Jul-2022	skrll	Use PRIxBIT instead of casts
1.734	02-Jul-2022	skrll	wm_tick: Add missing splx(s) when not WM_MPSAFE
1.733	19-May-2022	msaitoh	Don't set the more flag when there is no packet to process. Move the location of the loop limit check next to the DD bit test. This change prevent from scheduling the next txrx_work that does nothing.
1.732	19-May-2022	msaitoh	Call wm_xxeof() only when limit > 0. It's not required to call wm_xxeof() when limit == 0. It means that all xxeof processing will be done by softint or workqueue. Currently, wm_xxeof() returns quickly before checking the descriptor head when limit == 0 and return with ture to set the more flag.
1.731	17-May-2022	msaitoh	Call m_freem(m) only if m0 == NULL.
1.730	03-May-2022	gutteridge	if_wm.c: various spelling and grammar fixes
1.729	26-Feb-2022	rillig	if_wm.c: fix value return from void function lint complained: if_wm.c(10028): error: void function wm_txrxintr_disable cannot return value [213] if_wm.c(10049): error: void function wm_txrxintr_enable cannot return value [213] No functional change.
1.728	26-Feb-2022	rillig	if_wm.c: remove stray semicolons from struct declaration C99 does not allow 'struct { int member;; }'. No binary change.
1.727	16-Feb-2022	msaitoh	Remove duplicated break. No functional change.
1.726	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.725	23-Dec-2021	hannken	Keep constants 32 bit, why does __BIT() return uintmax_t?
1.724	20-Dec-2021	skrll	aprint_verbose the DMA range used.
1.723	20-Dec-2021	skrll	whitespace in a comment
1.722	11-Dec-2021	skrll	Typo in comment
1.721	10-Dec-2021	skrll	Simplify. Same code before and after.
1.720	10-Dec-2021	skrll	Trailing whitespace
1.719	05-Dec-2021	msaitoh	s/checkum/checksum/ in comment.
1.718	05-Nov-2021	msaitoh	Fix a bug that device timeout still happens on ICH/PCH. Fixes PR kern/56478. - if_wm.c rev. 1.695 for PR kern/40981 still had a problem. The workaround flag was unexpectedly cleared when a cable is not connected, SIOCINITIFADDR was called, or if_init was called. Fix it. - Add debug printf()s.
1.717	05-Nov-2021	msaitoh	Use macro. No functional change.
1.716	04-Nov-2021	msaitoh	Add some sysctl info for debugging.
1.715	20-Oct-2021	msaitoh	Use wm_flush_desc_rings() more. - Use wm_flush_desc_rings() for newer than PCH_SPT, too. Same as other OSes. It seems that some devices has no this errata, but we don't know how to identify it. So just use >= PCH_SPT. - Use wm_flush_desc_rings() before chip reset in wm_resume(), too.
1.714	20-Oct-2021	msaitoh	Disable printf()s in wm_flush_desc_rings() because the code is verified. The problem was very rare, so I added those printf()s as we can see. Last week, knakahara and I found a procedure to reproduce the problem and verified it worked correctly. To reproduce the problem: 0) Use I219 V1 to V5. Not all of them have the problem and newer than V5 may have the problem. I used V2 for the test. 1) Define WM_DEBUG or add printf in wm_flush_desc_rings() to see DESCRING_STATUS_FLUSH_REQ bit. 2) Run "iperf -s" on the DUT. 3) Run "iperf3 -R -u -c 192.168.1.110 -w 8m -b 1g" repeatedly on another machine. Note that kern.sbmax should be increased on both machines. 4) run while true; do ifconfig wm0 down up sleep 30 done 5) After DESCRING_STATUS_FLUSH_REQ is set, without wm_flush_desc_rings(), TX stalls and reboot is required to recover from it. With wm_flush_desc_rings(), no device hang. 6) If you can't see the DESCRING_STATUS_FLUSH_REQ is set, change the media to 100BASE-T. Note that if_wm.c rev. 1.710 originally written by chuq is required for the test. Without the change, the device becomes unrecoverable state before wm_flush_desc_rings() and the function doesn't recover from the problem.
1.713	20-Oct-2021	msaitoh	Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Even without this change, the workaround works with my machines. - Add comment.
1.712	20-Oct-2021	knakahara	Fix return value of interrupt handler, pointed out by msaitoh@n.o. Even if stopping flag is set (and handler do nothing), the interrupt itself should be assume to be processed when ICR is not zero. If not, interrupt count is not incremented correctly.
1.711	20-Oct-2021	knakahara	Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. Because the difference has caused INTx/MSI own bugs.
1.710	20-Oct-2021	knakahara	Fix Tx stall, reported and implemented original patch by chs@n.o, thanks. Check if_snd queue because if.c::if_transmit() doesn't call ifp->if_start() when IFQ_ENQUEUE failed. Reviewed and tested by chs@n.o and msaitoh@n.o.
1.709	18-Oct-2021	jmcneill	Only print an error about missing I/O BARs for chips that need it.
1.708	13-Oct-2021	msaitoh	Add support for I219V 15-19 and I219LM 16-19.
1.707	13-Oct-2021	msaitoh	Print I219's version number.
1.706	03-Aug-2021	knakahara	Fix locking against myself reported by rin@n.o. In some machine, INTx interrupts occurs for some reason while holding txq_lock in softint context. e.g. https://dmesgd.nycbug.org/index.cgi?do=view&id=6139 The panic message is the following.
1.705	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.704	12-May-2021	knakahara	Fix i386 build failure with options WM_EVENT_COUNTERS. pointed out by msaitoh@n.o.
1.703	03-May-2021	rillig	branches: 1.703.2; if_wm: do not return a void value from a void function This is a GCC extension.
1.702	11-Mar-2021	msaitoh	branches: 1.702.4; Whitespace fixes. No functional change.
1.701	01-Mar-2021	knakahara	Refactor rxq->rxq_ptr updating.
1.700	01-Mar-2021	knakahara	Remove extra unlock/lock processing around if_percpuq_enqueue(). This temporary unlock/lock processing was required for direct calling ifp->if_input. After r1.391, wm(4) uses if_percpuq_enqueue() and it does not hold any mutexes.
1.699	17-Feb-2021	knakahara	In 64 bit architectures, WM_EVENT_COUNTER is enabled by default. No objection from tech-kern@n.o and tech-net@n.o. ok'ed by msaitoh@n.o.
1.698	17-Feb-2021	knakahara	wm(4) use atomic_{load,store}_relaxed for evcnt 64 bit counter. ok'ed by msaitoh@n.o.
1.697	19-Nov-2020	msaitoh	Add some sysctl values for debugging TX/RX queues.
1.696	16-Nov-2020	rin	Fix little-endian dependence in wm_rxeof(), by which packets cannot be received by 82574 and successors on big-endian machines. Tested by aarch64eb with I210-T1 on ROCKPro64. Thanks msaitoh for discussion! XXX pullup to netbsd-9 and netbsd-8
1.695	02-Nov-2020	knakahara	Workaround for ihphy and atphy(ICH/PCH, 82580 and I350). These phys stop DMA while link is down which causes device timeout. Fix PR/kern 40981 Reviewed and tested by msaitoh@n.o, thanks. XXX pullup-[89]
1.694	30-Oct-2020	msaitoh	branches: 1.694.2; Add WMPHY_I350. Not used yet.
1.693	30-Oct-2020	msaitoh	Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. OK'd by knakahara. % sysctl -d hw.wm0.debug_flags hw.wm0.debug_flags: Debug flags: 0x01 LINK 0x02 TX 0x04 RX 0x08 GMII 0x10 MANAGE 0x20 NVM 0x40 INIT 0x80 LOCK
1.692	28-Oct-2020	msaitoh	Add missing drain for pcq in wm_stop_locked(). OK'd by knakahara. This change fixes two problems: 1. If the pcq is full and watchdog timer is fired, the full state of the pcq is kept and wm_transmit() returns with ENOBUFS. 2. ifconfig down doesn't free mbufs in the pcq.
1.691	16-Oct-2020	msaitoh	Fixes a problem that the attach function reported "wm_gmii_setup_phytype: Unknown PHY model. OUI=000000, model=0000" and "PHY type is still unknown." Don't call wm_gmii_setup_phytype() three times if the interface uses SGMII with internal MDIO. Tested with I354(Rangeley(SGMII(MDIO))) and I350(SERDES(SFP), SGMII(SFP)).
1.690	24-Sep-2020	msaitoh	Small fixes from knakahara: - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Fix comment.
1.689	16-Sep-2020	msaitoh	Use 12K for packet buffer for jumbo frame on PCH2 and newer. XXX Note that Linux Use 14K.
1.688	16-Sep-2020	msaitoh	Add a workaround for jumbo frame on PCH2 and newer. Tested by chs@. - Add wm_lv_jumbo_workaround_ich8lan() and use it. From FreeBSD. XXX For KUMCTRLSTA_OFFSET_HD_CTRL register modification, it's doubtful. FreeBSD and Linux do the same thing that they set the same value on both jumbo frame's enable case and the disable case. It seems the default value is 0x0b0c and it's not changed on the enable case, so it might be a bug on the enable case or the modification is not required. - Rename I219_UNKNOWN1 to I82579_UNKNOWN1.
1.687	15-Sep-2020	msaitoh	Add new flag named WM_F_CRC_STRIP and use it. No functional change. This change also sets the RCTL_SECRC bit on I211 but it doesn't change the behavior because I211 always strips CRC like I35[04] and I210.
1.686	24-Aug-2020	msaitoh	s/ressource/resource/. Found by knakahara.
1.685	05-Aug-2020	msaitoh	Set if_baudrate for non-MII device. Before this commit, it was 0.
1.684	05-Aug-2020	msaitoh	Explicitly cast from uint16_t to uint32_t before shifting 16bit left when printing Image Unique ID. Found by kUBSan.
1.683	03-Aug-2020	msaitoh	Setup PCS and SGMII for SFP correctly. It still doesn't support SFP insertion/removal. Copper: wm2: SGMII(SFP) wm2: 0x1043c440<SPI,IOH_VALID,PCIE,SGMII,NEWQUEUE,ASF_FIRM,EEE,SFP> makphy0 at wm2 phy 6: Marvell 88E1111 Gigabit PHY, rev. 1 Fiber: wm3: SERDES(SFP) wm3: 0x10034440<SPI,IOH_VALID,PCIE,NEWQUEUE,ASF_FIRM,SFP> wm3: 1000baseSX, 1000baseSX-FDX, auto
1.682	03-Aug-2020	msaitoh	s/MII_IGPHY_/IGPHY_/. No functional change.
1.681	09-Jul-2020	msaitoh	"wmX: 0" on 82542 is difficult to understand, so don't print it.
1.680	06-Jul-2020	msaitoh	Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM.
1.679	27-Jun-2020	jmcneill	prop_data_data_nocopy -> prop_data_value
1.678	11-Jun-2020	msaitoh	No functional change: - Rename some macros and function. - Add newline.
1.677	11-Jun-2020	thorpej	Update for proplib(3) API changes.
1.676	30-Apr-2020	riastradh	Make WM_NRXDESC unsigned.
1.675	30-Apr-2020	riastradh	Revert "- rnd_add_*() call must be serialized. Unserialized accesses broke the" It is no longer necessary to serialize calls to rnd_add_data, so we can use the same rndsource for every queue. This also reverts some unrelated changes that were mixed in: 1. WM_NRXDESC was made unsigned 2. The specific samples were changed. (1) will be redone in a separate commit. (2) was buggy (wrote `=' where `\|=' was meant) and is not needed; it's fine to do rnd_add_data in a thread or softint too, not just in a hard interrupt handler.
1.674	09-Apr-2020	jdolecek	adjust the condition for reusing checksum context, wm_tx_offload() can be used even for multiqueue case enable only when the chip supports exactly one queue (same as FreeBSD) - anything multiqueue-capable later than 82574 should work, but I don't have the hw to test
1.673	08-Apr-2020	jdolecek	wm_tx_offload() and wm_nq_tx_offload() actually always return 0, make them return void and remove the check for their return value
1.672	08-Apr-2020	jdolecek	fixup field names in previous
1.671	08-Apr-2020	jdolecek	actually writing always the checksum offload context descriptor makes the hw do extra processing, avoid doing that if possible - on my computer with I219 this results in about 2% speedup in Tx performance change adoped from FreeBSD XXX should be also done for multiqueue case, but I don't have hw to test it
1.670	21-Mar-2020	thorpej	branches: 1.670.2; Don't bother with IFF_OACTIVE.
1.669	15-Mar-2020	thorpej	Define and implement a locking protocol for the ifmedia / mii layers: - MP-safe drivers provide a mutex to ifmedia that is used to serialize access to media-related structures / hardware regsiters. Converted drivers use the new ifmedia_init_with_lock() function for this. The new name is provided to ease the transition. - Un-converted drivers continue to call ifmedia_init(), which will supply a compatibility lock to be used instead. Several media-related entry points must be aware of this compatibility lock, and are able to acquire it recursively a limited number of times, if needed. This is a SPIN mutex with priority IPL_NET. - This same lock is used to serialize access to PHY registers and other MII-related data structures. The PHY drivers are modified to acquire and release the lock, as needed, and assert the lock is held as a diagnostic aid. The "usbnet" framework has had an overhaul of its internal locking protocols to fit in with the media / mii changes, and the drivers adapted. USB wifi drivers have been changed to provide their own adaptive mutex to the ifmedia later via a new ieee80211_media_init_with_lock() function. This is required because the USB drivers need an adaptive mutex. Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv. mcx also now calls ifmedia_init_with_lock() because it needs to also use an adaptive mutex. The mcx driver still needs to be fully converted to NET_MPSAFE.
1.668	18-Feb-2020	msaitoh	It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set.
1.667	18-Feb-2020	msaitoh	Whitespace fix. No functional change.
1.666	31-Jan-2020	knakahara	wm(4) unset RSS UDP flags like ixg(4) and other OS's one. To handle IP fragmented UDP first packet and second packet by the same Rx queue. I tested on I354.
1.665	31-Jan-2020	knakahara	Fix missing workqueue_destroy(). ok by msaitoh@n.o
1.664	31-Jan-2020	knakahara	Fix wm(4) create too many workqueue. Pointed out by msaitoh@n.o
1.663	29-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.662	24-Jan-2020	knakahara	wm(4) can use workqueue as deferred Rx/Tx handler like ixg(4). Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0, that is, use softint as before. ok by msaitoh@n.o.
1.661	23-Jan-2020	knakahara	fix softint_disestablish() in wm_detach(). ok by msaitoh@n.o XXX pullup-8, pullup-9
1.660	21-Jan-2020	msaitoh	Add Intel I219 LM10-LM15 and V10-V14 from OpenBSD.
1.659	20-Jan-2020	thorpej	No longer need to go to splnet() before calling ifmedia_ioctl() in the NET_MPSAFE; ifmedia_ioctl() will do this for us, if needed.
1.658	13-Dec-2019	msaitoh	branches: 1.658.2; KNF. No functional change.
1.657	12-Dec-2019	knakahara	fix build failure. ok by msaitoh@n.o.
1.656	12-Dec-2019	msaitoh	Add SFP support part 2. - Add code for 1000BASE-T SFP. It uses SGMII. - Add extra delay in wm_serdes_power_up_link_82575(). Same as other OSes. TODO: - Module insertion/removal support. Currently, SFP detection is only done in the driver's attach phase.
1.655	11-Dec-2019	msaitoh	Add SFP support part 1. The code for SerDes and SGMII setting is not include in this commit. SFP module removal interrupt detects but not used yet: - Detect SFP's 1000BASE-SX, 1000BASE-LX, 100BASE-FX and 1000BASE-T correctly. - Detect the Media Auto Sense feature. Not supported yet. - Add comment.
1.654	11-Dec-2019	msaitoh	- Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. XXX We should check for other NVM reads if they have the same problem. One of the solution to avoid this type of bug is to check the register layout in wm_nvm_read(). - Use __BIT() - Add comment
1.653	11-Dec-2019	msaitoh	- MSI-X doesn't use sc->sc_icr variable, so move the code into non-MSI-X part. No functional change intended. OK'd by knakahara. - Modify debug printfs a bit.
1.652	11-Dec-2019	msaitoh	No functional change: - Add defintion of CONNSW register and PCS_NPTX (not used yet). - Add some bit definitions of PCS_LCTL. - Renae macro. - Fix comment. Add comment. - KNF.
1.651	09-Dec-2019	msaitoh	Fix a bug that wm_sgmii_readreg_locked() may return error even if error isn't occured.
1.650	04-Dec-2019	msaitoh	Simplify. No functional change.
1.649	26-Sep-2019	msaitoh	Fix typo in comment.
1.648	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.647	04-Sep-2019	msaitoh	printf -> device_printf
1.646	01-Aug-2019	msaitoh	Use unsigned to avoid undefined behavior in wm_i82543_mii_sendbits(). Found by kUBSan.
1.645	30-Jul-2019	msaitoh	branches: 1.645.2; - rnd_add_*() call must be serialized. Unserialized accesses broke the sc->rnd_source structure and caused off by one in kern_rndq.c's rnd_add_data_ts(). This problem was found by kUBSan. Make krndsource_t's data per queue-pair and call rnd_add_uint32() in wm_txrxintr_msix(). One of the reason why rnd_add_uint32() calls were moved from wm_{tx,rx}eof() is that rnd(9) manual page says "To allow the system to gather the timing information accurately, this call should be placed within the actual hardware interrupt service routine." - Pass a better value to rnd_add_uint32() in wm_intr_legacy(). XXX pullup-8.
1.644	26-Jul-2019	msaitoh	Use aprint_debug_dev() instead of printf() in wm_ulp_disable().
1.643	25-Jul-2019	tnn	wm: change one bus_dmamap_load_mbuf() to bus_dmamap_load() equivalent This makes wm work on aarch64, where the bus_dma implementation requires the mbuf data pointer be immutable between bus_dmamap_load_mbuf and bus_dmamap_sync. Fix suggested by thorpej@
1.642	09-Jul-2019	msaitoh	Don't automatically set ec_capenable's ETHERCAP_VLAN_HWTAGGING bit in vlan_config() to make it user-controllable. Instead, set the bit in xxx_attach().
1.641	12-Jun-2019	msaitoh	Print ICH/PCH's NVM version.
1.640	07-Jun-2019	msaitoh	Use unsigned. Found by UBSan. OK'd by kamil.
1.639	28-May-2019	msaitoh	s/recieve/receive/
1.638	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.637	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.636	15-May-2019	ozaki-r	Store IFF_ALLMULTI in ec_flags instead of if_flags to avoid data races IFF_ALLMULTI is set/unset to if_flags via if_mcast_op. To avoid data races on if_flags, IFNET_LOCK was added for if_mcast_op. Unfortunately it produces a deadlock so we want to remove added IFNET_LOCK by avoiding the data races by another approach. This fix introduces ec_flags to struct ethercom and stores IFF_ALLMULTI to it. ec_flags is protected by ETHER_LOCK and thus IFNET_LOCK is no longer necessary for if_mcast_op. Note that the fix is applied only to MP-safe drivers that the data races matter. In the kernel, IFF_ALLMULTI is set by a driver and used by the driver itself. So changing the storing place doesn't break anything. One exception is ioctl(SIOCGIFFLAGS); we have to include IFF_ALLMULTI in a result if needed to export the flag as well as before. A upcoming commit will remove IFNET_LOCK. PR kern/54189
1.635	14-May-2019	ozaki-r	Remove unnecessary checks of IFF_ALLMULTI IFF_ALLMULTI is changed by only driver itself, so we don't need to check its change on ec_ifflags_cb. This is part of PR kern/54189. NFCI.
1.634	24-Apr-2019	msaitoh	SIOCS is 'S'et function and the ioctl argument is ifreq. SIOCG is 'G'et function and the ioctl argument is ifmediareq. Before this change, SIOCG modify sc->sc_flowflags unexpectedly. Don't hook SIOCGIFMEDIA because this driver uses MII(4) and ether_ioctl has the hook(if_ethersubr.c rev. 1.160). This driver might require some additional fixes for SIOCSIFMTU and other ioctl()s. XXX pullup-[78].
1.633	24-Apr-2019	msaitoh	KNF. No functional change.
1.632	25-Mar-2019	msaitoh	Pass error status to the upper layer in wm_gmii_{i82544,i80003}*().
1.631	05-Mar-2019	msaitoh	Add Intel I219 variations for Cannon Lake.
1.630	04-Mar-2019	msaitoh	Use do { ... } whille (0) for DPRINTF(x, y).
1.629	28-Feb-2019	khorben	Typo (s/vaule/value/)
1.628	23-Feb-2019	kamil	Fix GCC7 "avoid ambiguous 'else'" build error in if_wm.c Detected with kUBSan NetBSD/i386 build.
1.627	21-Feb-2019	knakahara	Fix panic when wm(4) is reattached. ok by msaitoh@n.o. This panic occured after kern_rndq.c:r1.90.
1.626	08-Feb-2019	msaitoh	- Add support for I210 SGMII Flash-less device. - Add KASSERT. - Update TODO list.
1.625	07-Feb-2019	msaitoh	Fix a bug that WOL didn't work on some chips since if_wm.c rev. 1.610. Set WUC_APME bit older than PCH. Will fixes PR kern/53945 reported by kardel@. Tested with my own 82574 card.
1.624	06-Feb-2019	msaitoh	I21[01]'s RAL size is not 15 but 16(== 82575's value). Same as other OSes.
1.623	31-Jan-2019	msaitoh	Add yet another PHY workaround. Disable generation of early preamble on 82577 PHY rev. 1 or 2 and 82578 PHY rev. 1.
1.622	31-Jan-2019	msaitoh	We now have IFM_2500_KX and IFM_1000_KX, so use them.
1.621	31-Jan-2019	msaitoh	Add some workarounds which use EMI register or EEE related: - PCH only: Add workaround for link disconnects on a busy hub in half duplex. - PCH and PCH2 only: Set MSE higher to enable link to stay up when noise is high. - PCH2 only: Drop link after 5 times MSE threshold was reached. - PCH2 only: Set EEE LPI Update Timer to 200usec. - For PCH2 and newer: When connected at 10Mbps half-duplex, some parts are excessively aggressive resulting in many collisions. To avoid this, increase the IPG and reduce Rx latency in the PHY. - For I21[789] and if EEE is enabled: Disable LPLU if both link partners support 100BaseT EEE and 100Full is advertised on both ends of the link, and enable Auto Enable LPI since there will be no driver to enable LPI while in Sx.
1.620	25-Jan-2019	msaitoh	80003's SERDES is not the same as 82575's but the same as legacy devices. Use the old methods on 80003. XXX The reason why this bug existed is that our order of WM_T_* was little different from FreeBSD's enum e1000_mac_type. From 80003 to PCH_CNP and from 82575 to I211 are swapped. I carefully read through if_wm.c and knew only serdes related test were wrong. When we change WM_T_* to match FreeBSD's, additional changes also will be required.
1.619	25-Jan-2019	msaitoh	Use __nothing (== ((void)0)) instead of "do {} while (/CONSTCOND/false)".
1.618	24-Jan-2019	msaitoh	No functional change intended: - Use "do {} while (/CONSTCOND/false)" for null DPRINTF(). - Reduce indent level of wm_linkintr_gmii().
1.617	22-Jan-2019	msaitoh	Change MII PHY read/write API from: int (mii_readreg_t)(device_t, int, int); void (mii_writereg_t)(device_t, int, int, int); to: int (mii_readreg_t)(device_t, int, int, uint16_t ); int (*mii_writereg_t)(device_t, int, int, uint16_t); Now we can test if a read/write operation failed or not by the return value. In 802.3 spec says that the PHY shall not respond to read/write transaction to the unimplemented register(22.2.4.3). Detecting timeout can be used to check whether a register is implemented or not (if the register conforms to the spec). ukphy(4) can be used this for MII_MMDACR and MII_MMDAADR. Note that I noticed that the following code do infinite loop in the read/wirte function. If it accesses unimplemented PHY register, it will hang. It should be fixed: arm/at91/at91emac.c arm/ep93xx/epe.c arm/omap/omapl1x_emac.c mips/ralink/ralink_eth.c arch/powerpc/booke/dev/pq3etsec.c(read) dev/cadence/if_cemac.c <- hkenken dev/ic/lan9118.c Tested with the following device: axe+ukphy axe+rgephy axen+rgephy (tested by Andrius V) wm+atphy wm+ukphy wm+igphy wm+ihphy wm+makphy sk+makphy sk+brgphy sk+gentbi msk+makphy sip+icsphy sip+ukphy re+rgephy bge+brgphy bnx+brgphy gsip+gphyter rtk+rlphy fxp+inphy (tested by Andrius V) tlp+acphy ex+exphy epic+qsphy vge+ciphy (tested by Andrius V) vr+ukphy (tested by Andrius V) vte+ukphy (tested by Andrius V) Not tested (MAC): arm:at91emac arm:cemac arm:epe arm:geminigmac arm:enet arm:cpsw arm:emac(omac) arm:emac(sunxi) arm:npe evbppc:temac macppc:bm macppc:gm mips:aumac mips:ae mips:cnmac mips:reth mips:sbmac playstation2:smap powerpc:tsec powerpc:emac(ibm4xx) sgimips:mec sparc:be sf ne(ax88190, dl10019) awge ep gem hme smsh mtd sm age alc ale bce cas et jme lii nfe pcn ste stge tl xi aue mue smsc udav url Not tested (PHY): amhphy bmtphy dmphy etphy glxtphy ikphy iophy lxtphy nsphyter pnaphy rdcphy sqphy tlphy tqphy urlphy
1.616	11-Jan-2019	msaitoh	Set I82579_LPI_CTRL register correctly in wm_set_eee_pchlan().
1.615	09-Jan-2019	msaitoh	- Remove an extra register read in wm_kmrn_lock_loss_workaround_ich8lan(). - Don't leave the MDICNFG register modified when the Power Management capability offset can't get.
1.614	07-Jan-2019	msaitoh	Add EEE(802.3az) support for I350, I210, I211, PCH2 and newer. Not yet for I354(C2000). It'll be supported after implementing MI MII clause 45 register read/write API.
1.613	03-Jan-2019	msaitoh	Fix error check. reported by David Binderman in kern/53821.
1.612	30-Dec-2018	msaitoh	Accept IFM_NONE.
1.611	21-Dec-2018	msaitoh	Fix WoL detection once again.
1.610	20-Dec-2018	msaitoh	- Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220.
1.609	20-Dec-2018	msaitoh	Fix WoL capability detection on ICH and PCH.
1.608	20-Dec-2018	msaitoh	Rename functions. Add comment. No functional change.
1.607	17-Dec-2018	msaitoh	When sc->phy.acquire() failed, it's not required to call sc->phy.release().
1.606	17-Dec-2018	knakahara	Fix txqueue assignment. Pointed out by yamaguchi@n.o, thanks. E.g. When ncpu is six and nqueue is four, the sequence error occurs. XXX pullup-8
1.605	15-Dec-2018	msaitoh	Print and clear the wakeup status only when the WUS register != 0.
1.604	14-Dec-2018	msaitoh	- Fix availability detection of WoL on some chips. This change has no effect other than dmesg because WM_F_WOL is currently not used to change the behavior: - For ICH/PCH, check the capability not from NVM but from the WUC register. Check the value before clearing the register. - 82580 and newer have per-port NVM block, so read the area correctly. Note that 82580, I350 and I354 may have PCI function 2 and 3. - Some devices can't detect WoL capability neither from NVM nor from WUC. Use PCI device ID and the function number. - Print the WUS (WakeUp Status) register bits when resume.
1.603	12-Dec-2018	msaitoh	Add some code for suspend/resume: - Rename wm_smbustopci() to wm_init_phy_workarounds_pchlan(). It will also called when resume. - Call wm_phy_resetisblocked() after PHY reset in wm_init_phy_workarounds_pchlan() to wait for the PHY to quiesce to an accessible state. - Add new wm_resume_workarounds_pchlan() function and use it in wm_resume(). This workaround is only for PCH2 and newer. - Don't call wm_disable_aspm() neither in wm_attach() nor in wm_resume() but in wm_reset(). - Do some initialization in wm_resume() when IFF_UP is NOT set. - Don't continue when it failed to acquire semaphore in wm_ulp_disable(). - Add comment.
1.602	28-Nov-2018	msaitoh	Fix comment. No functional change.
1.601	22-Nov-2018	msaitoh	- Add wm_k1_workaround_lv() from FreeBSD. It's PCH2 specifc: Workaround to set the K1 beacon duration for 82579 parts in 10Mbps. Disable K1 for 1000 and 100 speeds. - Make wm_link_stall_workaround_hv() and move an 82578 specific code into it. Don't apply the workaround if BMCR_LOOP bit is set. Same as FreeBSD. - Add comment. Modify comment.
1.600	20-Nov-2018	msaitoh	- Add wm_oem_bits_config_ich8lan() to control LPLU and GbE setting base on the NVM's info. - Modify wm_enable_wakeup() to reduce difference against FreeBSD and Linux. This modification affects to ICH8 and newer devices. I217 Rapid Start Technology support have not written yet (it's TODO).
1.599	20-Nov-2018	msaitoh	- Fix a bug that i82578 PHY can't detect correctly. - Print PHY OUI and model number if a PHY's can't be identified. XXX pullup-[78]
1.598	19-Nov-2018	msaitoh	- Control TX/RX descriptor snooping control bits on ICH8 and newer. Only on ICH8, No-snoop bits are opposite polarity. On my Thinkpas X61, the default value of this bits are all zero, so this commit changes the snoop function enable on the machine. I tested with some other PCH machines and those bits are all zero (enable snoop by default), so this commit won't affect to some machines. - Disable relax ordering on 82546GB(Device ID 0x1099 and 0x10b5) or >= ICH8. Same as other OSes.
1.597	14-Nov-2018	msaitoh	- Add new wm_gmii_{hv,i82544}_{read,write}reg_locked() and use them in wm_gmii_{hv,i82544}_{read,write}reg(). _locked() functions are not mii(4) API functions, so it's not required to keep the mii API. Change the PHY register type from int to uint16_t. It also change the usage of return value. It returns zero on success and non-zero on error. - Check the return value of _locked() function and treat it. - Use *writereg_locked() function to reduce race condition in wm_init_lcd_from_nvm(). - Add comment.
1.596	03-Nov-2018	christos	make this compile again (rv was uninitialized)
1.595	02-Nov-2018	msaitoh	KNF a bit. No functional change.
1.594	02-Nov-2018	msaitoh	- Add missing wm_gate_hw_phy_config_ich8lan(false) in wm_phy_post_reset() on PCH2. wm_gate_hw_phy_config_ich8lan(true) is called in wm_reset(), so wm_phy_post_reset(false) should be called after reset. - On PCH2, set the phy config counter to 50msec after (PHY) reset.
1.593	02-Nov-2018	msaitoh	After writing MDIC register, don't read quickly the same register but do delay(50). Same as otehr OSes.
1.592	02-Nov-2018	msaitoh	Fix a PCH2 specific bug that wrong register value can be read when boot. When a wrong value is read when boot, the read device ID was incorrect and ukphy(3) is attached instead of ihphy(4). The bug might also result in MDIC read/write error. How to reproduce: 0) Boot Windows. 1) Leave some minutes. 2) Reboot to NetBSD. To fix this problem, adding extra 100us delay at the end of wm_gmii_mdic_{read,write}reg() on PCH2. Same as FreeBSD and linux. Reported by David Brownlee a few days ago and also reported by jmcneill a half year ago. Tested with my own Thinkpad X220. XXX pullup-[78]
1.591	02-Nov-2018	msaitoh	Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This woraround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. I observed this problem on my Shuttle DS57U's I218 port six times in two months.
1.590	31-Oct-2018	msaitoh	- 82574 and newer's document says the status field has neither EC (Excessive Collision) bit nor LC (Late Collision) bit (reserved). Refer "PCIe GbE Controller Open Source Software Developer's Manual", 82574 datasheet and newer. XXX I saw the LC bit was set on I218 even though the media was full duplex, so the bit might be used for other meaning ...(I have no document). - Use macro.
1.589	05-Oct-2018	msaitoh	Fix KASSERT to prevent panic on CNP(Intel 300 series + I219) with INTx interrupt (Xen dom0 or pre netbsd-8). XXX pullup-[78]
1.588	12-Sep-2018	msaitoh	Fix comment.
1.587	12-Sep-2018	msaitoh	- Try m_defrag() to reduce the number of DMA segment if bus_dmamap_load_mbuf() returned EFBIG. When m_defrag() is called, txqNNdefrag event counter is incremented. If the 2nd try of bus_dmamap_load_mbuf() failed, txqNNtoomanyseg event counter is incremented. - Reduce the max number of DMA segemts from 256 to 64 (it's the same valuse as other BSD's (EM_MAX_SCATTER) and more than before if_wm.c rev. 1.75's value (40)) because we do m_defrag() now.
1.586	12-Sep-2018	msaitoh	- Split txdrop evcnt into pcqdrop and descdrop. - Simplify evcnt name by removing duplicated "tx" or "rx". - s/tu/underrun/
1.585	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.584	09-Aug-2018	msaitoh	Fix two bugs for non-SERDES fiber. Tested on 82544EI(0x1009), 82545GM(0x1027), 82546GB(0x107a), 82571GB(0x105f), 82572EI(0x107e). Tested on SERDES 82580(0x150e (Winyao SERDES SFP)) and 82580(0x150e) as regression test: - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. When the signal is detected & STATUS_LU == 0 & RXCW == 0, it's required to set CTRL_SLU and CTRL_FD bit. wm_check_for_link() is the function to do it. If a link status change interrupt occured, call wm_check_for_link() before checking current link staus. - Fix a bug that 82542 misunderstand fiber's signal detection. CTRL_SWDPIN(1) is used for fiber's signal and it's value is reversed only on 82543 and 82544. Simplify the code and move it into a new function wm_tbi_havesignal() and use it.
1.583	26-Jun-2018	msaitoh	branches: 1.583.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.582	20-Jun-2018	msaitoh	Style fix. No functional change.
1.581	01-Jun-2018	maxv	Rename M_CSUM_DATA_IPv6_HL -> M_CSUM_DATA_IPv6_IPHL M_CSUM_DATA_IPv6_HL_SET -> M_CSUM_DATA_IPv6_SET Reduces the diff against IPv4. Also, clarify the definitions.
1.580	25-May-2018	ozaki-r	Ensure to call if_register after interface initializations finish
1.579	10-May-2018	msaitoh	KNF. No binary change.
1.578	08-May-2018	msaitoh	Fix a bug that TX might stall because WM_TXQ_NO_SPACE is not cleared in if_init() (though I've never seen this problem). Clear txq->txq_flags in wm_init_tx_queue(). OK'd by knakahara.
1.577	08-May-2018	msaitoh	rxipsum and rxtusum are not interrupt counter, so use EVCNT_TYPE_MISC instead of EVCNT_TYPE_INTR.
1.576	23-Apr-2018	msaitoh	- Backout if_wm.c rev.1.574 and print "device timeout (lost interrupt)" when all descriptors in a queue are free. When all descriptors are free after wm_txeof(), it's caused by lost interrupt (though I've never seen it). One possibility is chip bug and another possibility is software bug. We should reset in any cases. If we don't reset and don't print error message, TX processing is done intermittently and user might not noticed the problem. - Rename txq_watchdog to txq_sending to make the meaning clear.
1.575	23-Apr-2018	msaitoh	Count timeout correctly. This change reduce timeout value for 80003 as expected. Reported by mouse@.
1.574	20-Apr-2018	msaitoh	Fix watchdog timer. Without this change, watchdog timer is unnecessary fired and device is initialized without any error message. OK'd by knakahara.
1.573	16-Apr-2018	msaitoh	KNF. No functional change.
1.572	13-Apr-2018	msaitoh	I354 uses an external PHY, so don't use wm_set_eee_i350().
1.571	13-Apr-2018	msaitoh	Enable I219.
1.570	12-Apr-2018	msaitoh	Add PCH_CNP support (I219 with Intel 300 series chipset). It's required more test, so it's disabled by default.
1.569	12-Apr-2018	msaitoh	If the extended configration size in the EXTCNFSIZE register is 0, don't continue.
1.568	12-Apr-2018	msaitoh	Our PCH_SPT part of wm_nvm_valid_bank_detect_ich8lan() was based on FreeBSD r287467. After that, they reverted it and committed the different code in r287762. r287762's bank detect code didn't work for us because our wm dirver had a problem in flash access. The problem was fixed in if_wm.c rev. 1.567, so we can use the new way now.
1.567	12-Apr-2018	msaitoh	On PCH_SPT (and newer), FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it should be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask.
1.566	01-Mar-2018	msaitoh	branches: 1.566.2; Do wm_reset_mdicnfg_82580() on 82580 only.
1.565	01-Mar-2018	msaitoh	- Fix a bug that wm_pll_workaround_i210() is not called on some cases. As I wrote in if_wm.c rev 1.329's commit message, There are three cases: a) Chip is I211 b) Chip is I210 and it uses INVM (not FLASH) c) Chip is I210 (it uses FLASH) and NVM image version < 3.25 Before this commit, the workaround function was not called on a) and b). - Rename function.
1.564	16-Feb-2018	knakahara	Apply RSS utility to wm(4). ok by msaitoh@n.o.
1.563	14-Feb-2018	knakahara	Fix a bug that RX and TX may stall on heavy load on wm(4) like ixgbe.c:r1.121. wm_rxeof() and wm_txeof() have loop limit and the function returns true if a packet still exists. XXX need pullup-8
1.562	30-Jan-2018	knakahara	Make wm(4) watchdog MP-safe. There is almost no influence on performance. wm(4) does not use ifp->if_watchdog now, that is, it does not touch ifp->if_timer. It also uses own callout(wm_tick) as watchdog now. The watchdog uses per-queue counter to check timeout. So, global lock is not required.
1.561	29-Jan-2018	knakahara	Fix unmatched return type. The return value of wm_txeof() is not useded yet.
1.560	29-Jan-2018	knakahara	Fix if_wm.c:r1.557 merge miss, sorry.
1.559	26-Jan-2018	knakahara	Fix 82574 MSI-X mode cannot receive packets after 82574 receives high rate traffic. In short, 82574 MSI-X mode does not cause RXQ MSI-X vector when 82574's phys FIFO overflows. I don't know why but 82574 causes not RXQ MSI-X vector but OTHER MSI-X vector at the situation. see: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v4.15-rc9&id=4aea7a5c5e940c1723add439f4088844cd26196d advised by msaitoh@n.o, thanks.
1.558	21-Jan-2018	christos	CID-1427779: Fix uninitialized variables
1.557	18-Jan-2018	knakahara	wm_txeof() can limit the loop count the same as wm_rxeof() now.
1.556	17-Jan-2018	knakahara	Fix duplicated "rxintr" evcnt counting. Pointed out by ozaki-r@n.o, thanks.
1.555	16-Jan-2018	knakahara	Fix wm_watchdog_txq() lock region. Not only wm_txeof() but also wm_watchdog_txq() itself requires txq_lock as it reads Tx descriptor management variables such as "txq_free". There is almost no influence on performance.
1.554	15-Jan-2018	knakahara	improve comments
1.553	15-Jan-2018	knakahara	Fix legacy Tx descriptors printing when WM_DEBUG is enabled.
1.552	04-Jan-2018	msaitoh	Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki.
1.551	01-Jan-2018	jnemeth	PR/52885 - Shinichi Doyashiki -- typo in comment
1.550	28-Dec-2017	msaitoh	Don't use MSI-X if we can use only one queue to save interrupt resource. Written by knakahara and tested by me.
1.549	08-Dec-2017	ozaki-r	Revert "Make if_timer MP-safe if IFEF_MPSAFE" Because it has decreased the performance of wm. And also I found that wm_watchdog doesn't work well with if_watchdog framework at all. Sharing one counter (if_timer) with multiple instances (hardware multi-queues) can't detect a single (or some) stall of them because other instances reset the counter even if the stalled one want the watchdog to fire. Interfaces without IFEF_MPSAFE works safely with the original if_watchdog thanks to KENREL_LOCK. OTOH, interfaces with IFEF_MPSAFE shouldn't use if_watchdog and should implement their own watchdog timer that works with multiple instances.
1.548	07-Dec-2017	msaitoh	Fix a bug that 8257[56] can't receive packet reported by Bert Kiers in PR#52717. For 82575 and 82576, the RX descriptors must be initialized after the setting of RCTL.EN in wm_set_filter(). This bug was added in if_wm.c rev. 1.515.
1.547	06-Dec-2017	ozaki-r	Make if_timer MP-safe if IFEF_MPSAFE if_timer, a counter used by if_watchdog (if_slowtimo), can be modified in if_watchdog and if_start and/or interrupt handlers of some device drivers. All such accesses were serialized by KERNEL_LOCK. If IFEF_MPSAFE is enabled, KERNEL_LOCK of if_start (and perhaps interrupt handlers) is omitted and if_timer becomes racy. Fix the race condition by protecting if_timer by a spin mutex. if_watchdog_reset and if_watchdog_stop are introduced to ensure to take the mutex on accessing if_timer. Interface with IFEF_MPSAFE enabled must use the functions. In addition, if_watchdog callout is now set CALLOUT_MPSAFE if IFEF_MPSAFE. It means that if_watchdog implemented by a driver must be MP-safe if the driver is set IFEF_MPSAFE. Currenlty interfaces with IFEF_MPSAFE implementing if_watchdog and accessing if_timer in if_start and interrupt handlers are only wm(4). wm is changed to use the functions. (Its watchdog handler (wm_watchdog) is already MP-safe. These contracts will be written somewhere in a further commit. Note that the spin mutex is now ifp->if_snd.ifq_lock to avoid adding another spin mutex to each interface. For now reusing it isn't problematic (see the comment to know why) thought if that does matter in the future, feel free to replace it with a new spin mutex. It's easy to do.
1.546	30-Nov-2017	msaitoh	- 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Cleanup comment.
1.545	30-Nov-2017	msaitoh	Don't allocate MSI-X interrupt on 82583. 82583 chip has a MSI-X capability in the PCI configuration space but it doesn't support it. At least the document doesn't say anything about MSI-X. Fixes PR#52767 reported by Shinichi Doyashiki. XXX pullup-8.
1.544	22-Nov-2017	msaitoh	Revert if_wmreg.h 1.104 and if_wm.c 1.542. It's not required to mask other than VLAN ID bits in VLAN tag.
1.543	16-Nov-2017	ozaki-r	Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@
1.542	23-Oct-2017	knakahara	fix wm(4) vlan panic. Reported and tested by Tom Ivar Helbekkmo, thanks. wm(4) used PRI bits and CFI bit as vlantag by mistake. It is found out by if_ether.h:r1.67. XXX need pullup-8
1.541	23-Oct-2017	msaitoh	If if_initialize() failed in the attach function, return.
1.540	20-Oct-2017	msaitoh	On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux).
1.539	26-Sep-2017	msaitoh	- Correct NVM word size check in wm_nvm_version() to prevent "EERD polling failed" error on attach. if_wm.c rev. 1.534 was not perfect. Reported by ozaki-r@. - Print NVM offset and word count when EERD polling failed. XXX pullup-8.
1.538	26-Sep-2017	knakahara	VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.537	31-Jul-2017	knakahara	rename unclear functions. pointed out by msaitoh@n.o.
1.536	28-Jul-2017	knakahara	add comment about confusing wm_rxeof() calling. pointed out by msaitoh@n.o.
1.535	28-Jul-2017	msaitoh	Check some wm_nvm_read()'s return vale.
1.534	28-Jul-2017	msaitoh	Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higer address words to prevent timeout.
1.533	27-Jul-2017	msaitoh	Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally." on I21[01]. PXE boot is one of the case. XXX pullup-[678].
1.532	26-Jul-2017	msaitoh	On 8254[17] with SPI ROM, wm_nvm_set_addrbits_size_eecd() calls wm_nvm_read(), so it's required to set flags and functions before calling it. It was broken again in if_wm.c rev. 1.530 (see also 1.313). Tested with Dell Power Edge 1850.
1.531	26-Jul-2017	msaitoh	- Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register vaule from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg.
1.530	25-Jul-2017	msaitoh	Use new nvm.{acquire,release}() for semaphore. Almost the same except the following: - 8257[12]: Don't directly access SPI but use EERD register. - 82575-I354: If the size of SPI ROM >= 32K words, use direct SPI access instead of EERD register access. - Add wm_nvm_eec_clock_raise() and wm_nvm_eec_clock_lower() and use them for Microwire/SPI bus control. Same as Linux and FreeBSD. - Redude timeout value for 80003 in wm_get_swfw_semaphore(). Same as Linux and FreeBSD.
1.529	20-Jul-2017	msaitoh	Fix a bug that sc->sc_flags can't be passed to igphy(4). This bug was added in if_wm.c rev 1.527.
1.528	18-Jul-2017	msaitoh	- Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Enable wm_init_lcd_from_nvm() again. Tested by Thinkpad X220.
1.527	18-Jul-2017	msaitoh	Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function is required to set correctly before wm_reset() because wm_reset() and some helper functions refer them.
1.526	18-Jul-2017	msaitoh	Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hungup. A semaphore is get/put in wm_nvm_{read,write}.
1.525	14-Jul-2017	msaitoh	Calling wm_init_lcd_from_nvm() makes hang on some machines reported by Brad Harder. Until fix this problem correctly, Disable whole of wm_init_lcd_from_nvm().
1.524	13-Jul-2017	msaitoh	IF TX is not required to flush, RX is also not required to flush. Same as other OSes.
1.523	13-Jul-2017	msaitoh	- Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - Add debug printf.
1.522	13-Jul-2017	msaitoh	- Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Add diagnostic code.
1.521	13-Jul-2017	msaitoh	- Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Add debug printf. - Rename variables.
1.520	12-Jul-2017	msaitoh	Remove debug printf().
1.519	12-Jul-2017	msaitoh	Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register.
1.518	06-Jul-2017	msaitoh	Move the location of wm_get_hw_control() in wm_init_locked() again. Doing it "after" wm_reset() is required for some AMT based machines to linkup 1Gbps. Tested with HP dc7700.
1.517	26-Jun-2017	msaitoh	- Make new wm_phy_post_reset() and use this function at all location after reseting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer.
1.516	26-Jun-2017	msaitoh	Add missing "else" in wm_nvm_release().
1.515	26-Jun-2017	msaitoh	- Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange().
1.514	26-Jun-2017	msaitoh	- Fix wm_set_ral(): - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. - Remove wrong comment.
1.513	26-Jun-2017	msaitoh	Print sc_flags with snprintb().
1.512	23-Jun-2017	msaitoh	Don't print the Image Unique ID if an NVM is iNVM (i210 and I211).
1.511	20-Jun-2017	msaitoh	- Add debug printf()s. - If a legacy interrupt is a spurious interrupt, don't print debug message. - Add comment.
1.510	19-Jun-2017	msaitoh	Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes.
1.509	12-Jun-2017	msaitoh	Use IFM_1000_KX and IFM_2500_KX.
1.508	13-Apr-2017	knakahara	branches: 1.508.4; wm(4) can disable msi/msix by build option and ddb command. suggested by nonaka@n.o. reviewed by msaitoh@n.o and nonaka@n.o.
1.507	12-Apr-2017	knakahara	improve comments about the difference between legacy/msi and msix. pointed out by msaitoh@n.o.
1.506	05-Apr-2017	msaitoh	Fix a bug that non-GMII devices don't send a routing message when the link status is changed. Call if_link_state_change(). Tested with 82572EI(TBI) and I350(SERDES). XXX pullup-[67].
1.505	24-Mar-2017	knakahara	fix typo pointed out by ozaki-r@n.o.
1.504	24-Mar-2017	knakahara	remove extra "+"
1.503	24-Mar-2017	knakahara	add comments about the situation wm(4) use MSI-X but does not use multiqueue.
1.502	24-Mar-2017	knakahara	make self-describing function the code reading sc->sc_nintrs or sc->sc_nqueues directly.
1.501	24-Mar-2017	knakahara	add prototype. I have forgotten the fix by wm_softint_establish(). Thank you christos@n.o!
1.500	22-Mar-2017	knakahara	add comments about transmit context descriptor
1.499	22-Mar-2017	knakahara	missing M_CSUM_IPv4 checking. It does not cause problems, but it is wrong. The implementations of M_CSUM_DATA_IPv4_IPHL() and M_CSUM_DATA_IPv6_HL() is the same, so that does not cause problems.
1.498	21-Mar-2017	knakahara	fix PR kern/52068 (82574 Tx csum offload bug in not only i386 but also amd64). The packets sent from txq[1] could not be enabled any tx offload. This has been enbugged since r1.454.
1.497	08-Mar-2017	kardel	PR kern/52039: use same safeguard as for the 82578
1.496	03-Mar-2017	knakahara	fix r1.492 bug, sorry
1.495	03-Mar-2017	knakahara	implement AIM(Adaptive Interrupt Moderation), but not enable yet. Currently, it causes performance degration if enabled. More ITR calcuration tuning is required...
1.494	03-Mar-2017	knakahara	make kernel config option WM_RX_PROCESS_LIMIT_DEFAULT and WM_RX_INTR_PROCESS_LIMIT_DEFAULT e.g. if - WM_RX_PROCESS_LIMIT_DEFAULT is set 0 - WM_RX_INTR_PROCESS_LIMIT_DEFAULT is set -1 (means almost infinite) that means wm(4) does not use polling mode.
1.493	03-Mar-2017	knakahara	implement polling mode for multiqueue. It can suppress performance degration at high load. e.g. I354 IP forwarding throughput performance + before - 133Mbps forwarding performance at 200Mbps input load - 41Mbps forwarding performance at 800Mbps input load + after - 150Mbps forwarding performance at 200Mbps input load - 150Mbps forwarding performance at 800Mbps input load
1.492	03-Mar-2017	knakahara	IFEF_START_MPSAFE flag should be set only if WM_MPSAFE defined. some refactor. pointed out and suggested by msaitoh@n.o, thanks.
1.491	01-Mar-2017	knakahara	refactor: functionalize ITR/EITR writing processing. tested I354 and 82545(VMware e1000).
1.490	01-Mar-2017	knakahara	divide ITR/EITR value from wm_softc to wm_queue as EITR register is bound to each H/W queue.
1.489	28-Feb-2017	knakahara	fix EITR setting. + 82574 - add EITR setting, that is MSI-X mode interrupt interval + 82575 - fix EITR value. 82575's EITR usage is the same as legacy (not NEWQUEUE) controllers - apply workaround which 82575's EITR does not have CNT_INGR bit + other NEWQUEUE controllers - fix interrupt interval field. NEWQUEUE (include 82575) controllers' interrupt interval field is 2:14 - use CNT_INGR bit which avoid to overwrite counter tested 82574, 82575 and I354. ok by msaitoh@n.o.
1.488	27-Feb-2017	knakahara	fix Tx queue assignment when the number of H/W queues is less than ncpu.
1.487	24-Feb-2017	knakahara	apply 82574 RFCTL workaround the same as FreeBSD and linux. In fact, this workaround reduces interrupt count.
1.486	24-Feb-2017	knakahara	82574 should follow MSI-X mode IMS manner. Howerver some legacy interrupts occur as MSI-X other interrupt yet...
1.485	18-Feb-2017	christos	Establish a softint for legacy, otherwise BOOM!
1.484	17-Feb-2017	knakahara	unify Tx softint and Rx softint to uniform ixg(4) design. Furthermore, that can remove wm_deferred_start() which holds all txq's txq_lock, so it would reduce lock contention between Tx interrupt handler and Tx softint handler.
1.483	17-Feb-2017	knakahara	rxq_ptr must be updated holding rxq_lock consistent with the Rx processing. In previous implementation, wm_rxeof() temporarily unlock rxq->rxq_lock to call if_percpuq_enqueue(), and then re-lock rxq->rxq_lock to update rxq->rxq_ptr. So, if multiple CPUs share the same rxq, there is race e.g. - CPU A: lock rxq->rxq_lock - CPU A: Rx processing include increment local variable "i" - CPU A: unlock rxq->rxq_lock - CPU A: call if_percpuq_enqueue() - CPU B: lock rxq->rxq_lock - CPU B: Rx processing include increment local variable "i" - CPU B: unlock rxq->rxq_lock - CPU B: call if_percpuq_enqueue() - CPU B: lock rxq->rxq_lock - CPU B: update rxq->rxq_ptr, that is, set CPU B's local "i" to rxq->rxq_ptr - CPU B: unlock rxq->rxq_lock - CPU A: lock rxq->rxq_lock - CPU A: update rxq->rxq_ptr, that is, set CPU A's local "i" to rxq->rxq_ptr - CPU A: unlock rxq->rxq_lock
1.482	17-Feb-2017	knakahara	more fix about IFF_OACTIVE. fix TODO of r1.479 and reduce IFF_OACTIVE manipulation.
1.481	17-Feb-2017	knakahara	refactor: functionize deferred_start code holding locks.
1.480	17-Feb-2017	knakahara	refactor: functionize txrxintr disable/enable code.
1.479	16-Feb-2017	knakahara	fix about IFF_OACTIVE. It is required manipulate IFF_OACTIVE for old ethernet controllers and ALTQ to use if_start. To avoid race, txq_lock of 0th txq is needed to write IFF_OACTIVE. TODO: read side refactor
1.478	13-Feb-2017	knakahara	fix typo
1.477	09-Feb-2017	knakahara	detach wm(4) evcnt. pointed out by msaitoh@n.o, thanks.
1.476	02-Feb-2017	knakahara	remove WM_CORE_LOCK in deferred start routine. It is enough to check each txq's txq_stopping with mutex_enter(not mutex_tryenter). reviewed by ozaki-r@n.o.
1.475	01-Feb-2017	msaitoh	Use new wm_gmii_setup_phytype() function to setup sc_phytype and mii_{read\|write}reg. This function is called twie. To identify PHY type, correct read/write function should be selected. To select correct read/write function, PCI ID or MAC type are required without accessing PHY registers. On the first call of this function, PHY ID is not known yet. Check PCI ID or MAC type. The list of the PCI ID may not be perfect, so the result might be incorrect. In the second call, PHY OUI and model are used to identify PHY type. It might not be perfpect because of the lack of compared entry, but it would be better than the first call. If the detected new result and previous assumption is different, diagnous message will be printed.
1.474	01-Feb-2017	msaitoh	Linux and FreeBSD defines 0x10f5 as E1000_DEV_ID_ICH9_IGP_M_AMT. In reality, This is not IGP but BM. Add new case to identify PHY type correctly on this device. Fixes PR#51924 reported byJarle Greipsland. This change fix only for 0x10f5. Need more work to be the best.
1.473	01-Feb-2017	knakahara	fix: wm(4) Rx descriptor is corrupted under high rate. don't write descriptor which is not set DD bit.
1.472	31-Jan-2017	knakahara	remove obsolete comment. wm(4) vlan have already tested on I354.
1.471	30-Jan-2017	knakahara	fix endianness of "rsshash"(currently, used for debug message only) and tabify.
1.470	27-Jan-2017	knakahara	add comment about txq->txq_lock contentions in wm_nq_transmit().
1.469	26-Jan-2017	knakahara	fix memory leak at wm_detach. pointed out and reviewed by msaitoh@n.o.
1.468	16-Jan-2017	maya	move declaration to #ifdef WM_DEBUG block as opposed to using __debugused fixes build with DEBUG but without WM_DEBUG ok martin
1.467	16-Jan-2017	knakahara	update todo comment
1.466	16-Jan-2017	knakahara	support {extended, advanced} receive descriptor for wm(4). support below two receive descriptors. - extended descriptor (used by 82574 only) - advanced descriptor (used by 82575 and newer) SPH(split header buffer) is not supported yet. ok by msaitoh@n.o.
1.465	10-Jan-2017	msaitoh	branches: 1.465.2; remove debug printf().
1.464	10-Jan-2017	msaitoh	Fix previous. The last code wasn't called.
1.463	10-Jan-2017	knakahara	sc->sc_core_stopping must be checked with holding sc_core_lock. pointed out by ozaki-r@n.o, thanks.
1.462	06-Jan-2017	msaitoh	- Set IPV6EXDIS bit in RFCTL register because of an Errata on 82575 and newer devices. - Update comment in wm_init_rss().
1.461	04-Jan-2017	knakahara	fix typo. pointed out by msaitoh@n.o.
1.460	28-Dec-2016	ozaki-r	Protect ec_multi* with mutex The data can be accessed from sysctl, ioctl, interface watchdog (if_slowtimo) and interrupt handlers. We need to protect the data against parallel accesses from them. Currently the mutex is applied to some drivers, we need to apply it to all drivers in the future. Note that the mutex is adaptive one for ease of implementation but some drivers access the data in interrupt context so we cannot apply the mutex to every drivers as is. We have two options: one is to replace the mutex with a spin one, which requires some additional works (see ether_multicast_sysctl), and the other is to modify the drivers to access the data not in interrupt context somehow.
1.459	26-Dec-2016	msaitoh	0x10cb is not 82801H but 82801I. No functional change (The macro name was changed but the vaule is the same).
1.458	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.457	13-Dec-2016	msaitoh	- Change to use 2500Base-KX correctly on C2000(I354). It worked, but the output of ifconfig and if_baudrate was not good. Tested by nils@: - The STATUS_TBIMODE bit in the STATUS register is deleted since 82575, so check for 82575 and newer first and then check for old devices. - Check the 2P5_SKU and 2P5_SKU_OVER bit for KX. - Set IFM_2500_SX instead of IFM_1000_SX for 2.5G. - Check SERDES's speed directly from the PCS layer (PCS_LSTS register) for old devices. - Style fix.
1.456	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.455	02-Dec-2016	knakahara	fix: ifp->if_obytes and ifp->if_omcasts are not counted up for packets which is sent later. pointed out by ozaki-r@n.o, thanks.
1.454	01-Dec-2016	knakahara	support Tx multiqueue for 82574(aka e1000e for VMware and QEMU).
1.453	21-Nov-2016	msaitoh	Add note about kumeran registers. No functional change.
1.452	18-Nov-2016	joerg	Remove extra pair of braces.
1.451	18-Nov-2016	msaitoh	Fix WOL related setting of the WUC register for other than PCH* in wm_enable_wakeup(). Tested with 82567V(ICH8) and 82583V.
1.450	17-Nov-2016	msaitoh	8257[124] should not set WM_F_ARC_SUBSYS_VALID.
1.449	17-Nov-2016	msaitoh	Fix flag check in wm_get_wakeup() - 8254[17]* should not set WM_F_ARC_SUBSYS_VALID. - Add missing WM_T_82541_2 and WM_T_82547_2 - Note that the setting of both Linux's FLAG_HAS_AMT and FreeBSD's has_amt flag is doubtful on 8257[34] and 82583.
1.448	16-Nov-2016	msaitoh	Move the location of wm_smbustopci() call.
1.447	16-Nov-2016	msaitoh	Sync wm_smbustopci() with Linux and FreeBSD. This change effects PCH and newer devices.
1.446	16-Nov-2016	msaitoh	Fix a bug that 8257[56], 82580, I35[04] and I21[01] didn't use wm_{get,release}_hw_control() correctly.
1.445	16-Nov-2016	msaitoh	Add three workarounds for PCH_{LPT,SPT}.
1.444	14-Nov-2016	msaitoh	Set CTRL_MEHE correctly (PCH_{LPT,SPT} only).
1.443	10-Nov-2016	msaitoh	- Move I219 DMA workaround into wm_flush_desc_rings() and call it before wm_reset(). - Rewite I219 TX DMA workaround based on OpenBSD's one. - Add I219 RX DMA workaroud from OpenBSD.
1.442	10-Nov-2016	msaitoh	Call wm_enable_phy_wakeup() on PCH2 and newer, too. Now these devices can do WOL. Tested with Thinkpad X220(PCH2).
1.441	08-Nov-2016	msaitoh	Fix wm_access_phy_wakeup_reg_bm(). This change has no effect because this function is used for WUC register and our driver currenlty doesn't access to it.
1.440	08-Nov-2016	msaitoh	Remove debug printf which was added in rev. 1.437.
1.439	08-Nov-2016	msaitoh	Call wm_enable_wakeup() in wm_detach() and wm_suspend(). Now wake on lan works on Thinkpad X61(ICH8). Some workarounds will be required on PCH*.
1.438	06-Nov-2016	msaitoh	Add "10/100" into non-gigabit devices' name.
1.437	02-Nov-2016	msaitoh	- Fix workaround which did dummy read BM_WUC register. This code was changed to drop BM_WUC_HOST_WU_BIT of BM_PROT_GEN_CFG register in FreeBSD r228386. The code was added rev. 1.149, but the location was not the best. Now I219 doesn't hang quickly after "ifconfig up". - wm_gmii_hv_{read/write}reg*(): USE PHY address 1 for some special registers. - Add check code for an 82578 workaround. Not completed yet(check only). - wm_release_hw_control(): Remove extra line. No any effect.
1.436	31-Oct-2016	knakahara	Fix locking against myself at wm_turn{on,off} when NET_MPSAFE is defined. Pointed out by ozaki-r@n.o, thanks.
1.435	28-Oct-2016	msaitoh	Fix PHY access on 82567(ICH8 or ICH10), 82574 and 82583: - Use wm_gmii_bm_{read,write}reg() on 82574 and 82573. - Issue page select correctly on BM PHYs.
1.434	28-Oct-2016	msaitoh	Change debug flags to be better than before.
1.433	28-Oct-2016	msaitoh	- Define WMPHY_I217, WMPHY_VF and WMPHY_210. - Use BME1000_PHY_PAGE_SELECT in wm_gmii_bm_{read,write}reg(). This change has no effect because GG82563_PHY_PAGE_SELECT and BME1000_PHY_PAGE_SELECT have the same value.
1.432	28-Oct-2016	msaitoh	- Use MII_ADDRMASK. - Remove debug printf().
1.431	28-Oct-2016	knakahara	Fix wm(4) input drop packet counter. WMREG_RNBC is incremented when there is no available buffers in host memory. However, ethernet controller can receive packets in such case if there is space in phy's FIFO. That is, ethernet controller drops packet only if there is no available buffers and there is no space in phy's FIFO. So, the number of dropped packets should be added WMREG_MPC only. ok by msaitoh@n.o
1.430	28-Oct-2016	msaitoh	- Remove an 82578 workaround which was for PCH rev < 3. FreeBSD removed this workaround in r228386. - Add an 82578 workaround which is for PHY rev < 2. From FreeBSD and Linux. - Add some DPRINTF()s.
1.429	28-Oct-2016	knakahara	Fix sc_stopping race. To scale, separate sc_stopping flag to wm_softc and each tx,rx queues. Pointed out by skrll@n.o, thanks. ok by msaitoh@n.o
1.428	26-Oct-2016	msaitoh	Use wm_gmii_82544_{read,write}reg() on non-82567 ICH8, 9 and 10.
1.427	26-Oct-2016	msaitoh	82567V_3 is BME1000_E_2(bm). Tested with Advantech AIMB-212 1st Ethernet port.
1.426	26-Oct-2016	msaitoh	82567V-3 is not ICH9 but ICH8.
1.425	21-Oct-2016	msaitoh	- Fix previous commit in wm_kmrn_readreg(). - Use wm_gii_mdic_readreg/writereg() in wm_access_phy_wakeup_reg_bm() because these functions are called with taking lock. - Add some DPRINTF()s.
1.424	21-Oct-2016	msaitoh	- Rewrite PHY related lock stuff. Almost the same as FreeBSD. This change will fix a bug that PHY read/write fail on some cases. - Increase delay in wm_phy_resetisblocked(). Same as FreeBSD. - Use semaphore in wm_hv_phy_workaround_ich8lan() and wm_k1_gig_workaround_hv()
1.423	20-Oct-2016	msaitoh	- Move call of wm_reset() in wm_attach() after setting PHY and NVM related flags because those flags are used in wm_reset(). - Use mutex for NVM access on ICH8 and newer devices. Same as FreeBSD.
1.422	20-Oct-2016	msaitoh	Add newer I219 devices (not enabled).
1.421	20-Oct-2016	msaitoh	Add more DPRINTF()s.
1.420	19-Oct-2016	msaitoh	No, functional change: - Add WM_DEBUG_LOCK - Add some KASSERT - Modify comment.
1.419	19-Oct-2016	msaitoh	- Drop the host wakeup bit after resetting PHY on PCH and newer devices. - Increase delay while toggling LANPHYPC
1.418	11-Oct-2016	skrll	Read sc_if_flags after taking core lock
1.417	10-Aug-2016	knakahara	restructure wm(4) evcnt to support multiqueue evcnt. ok by msaitou@n.o.
1.416	11-Jul-2016	knakahara	branches: 1.416.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.415	20-Jun-2016	knakahara	wm(4) can enable IFEF_START_MPSAFE now.
1.414	14-Jun-2016	knakahara	fix: rxq->rxq_lock is uninitialized if NET_MPSAFE is not defined.
1.413	14-Jun-2016	skrll	Simplify the code. WM_[RT]X_* macros are removed and replaced with mutex_* calls directly. OK from nakahara@ and msaitoh@
1.412	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.411	30-May-2016	knakahara	fix r1.409 mistake
1.410	23-May-2016	knakahara	Update TODO: evcnt
1.409	23-May-2016	knakahara	fix: ethernet controllers which use wiseman_txdesc_t (earlier than 82575) hang possibly.
1.408	23-May-2016	knakahara	GENERIC kernel (disabled NET_MPSAFE kernel) also needs txq_lock to avoid race between tx processing and tx interrupt handler.
1.407	20-May-2016	knakahara	Update TODO - "TX Multiqueue" -> "TX Multiqueue improvement" - add "Advanced Recieve Descriptor" (RX NEWQUEUE)
1.406	19-May-2016	knakahara	fix build failure by gcc 4.8.5
1.405	19-May-2016	knakahara	unify TX and RX interrupt handler to use MSI-X vector efficiently
1.404	19-May-2016	knakahara	user-friendly interrupt affinity offset
1.403	19-May-2016	knakahara	initial TX multiqueue support
1.402	18-May-2016	knakahara	Whether Tx interrupt handler kicks if_start() should not depend on wm_txeof() return value.
1.401	18-May-2016	knakahara	rename WM_TXQ_WORKING to WM_TXQ_NO_SPACE
1.400	18-May-2016	knakahara	device drivers should not divert ifp->if_flags to manage H/W satatus.
1.399	18-May-2016	knakahara	refactor: eliminate "txq->txq_descs_size", use WM_TXDESCS_SIZE instead.
1.398	18-May-2016	knakahara	fix unmatched dma sync size for NEWQUEUE. This bug would have caused wm_nq_tx_offload() not to work.
1.397	11-May-2016	msaitoh	Fix compile error on some archs (e.g. i386).
1.396	11-May-2016	knakahara	Interrupt handlers read wm_rxqueue or wm_txqueue, so they must be freed after disestablishing interrupt handlers.
1.395	11-May-2016	msaitoh	Do I219 TX DMA workaround only when TDLEN(0) != 0
1.394	06-May-2016	msaitoh	- Modify wm_smbustopci() to reduce the diff against Linux and FreeBSD. - wm_gate_hw_phy_config_ich8lan() is for younger than PCH2. - Add debug code.
1.393	06-May-2016	msaitoh	Add some workaround code for I219. It still doesn't work.
1.392	06-May-2016	msaitoh	Basic support for I219. It doesn't work on I219, so it's disabled.
1.391	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.390	05-Feb-2016	msaitoh	s/max/ralmax/. No functional change.
1.389	29-Jan-2016	msaitoh	Remove debug message.
1.388	07-Jan-2016	msaitoh	- printf->DPRINTF in wm_serdes_mediastatus(). - KNF.
1.387	25-Dec-2015	msaitoh	Modify comment for 82544 Errata 9 "Certain register cannot be written with particular alignments in PCI-X bus operation".
1.386	25-Dec-2015	msaitoh	- Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Fix typo in comment and modify comment by tnn@. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change.
1.385	22-Dec-2015	knakahara	fix potentially bug
1.384	22-Dec-2015	knakahara	fix incorrect KASSERT
1.383	18-Dec-2015	knakahara	set Tx/Rx interrupts affinity to other than CPU#0 as much as possible
1.382	13-Dec-2015	christos	PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment.
1.381	30-Oct-2015	msaitoh	- On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being.
1.380	30-Oct-2015	msaitoh	- Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - KNF.
1.379	30-Oct-2015	msaitoh	Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT.
1.378	30-Oct-2015	msaitoh	Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done to wm(4) and igphy(4). Now PR#44893 should be fixed.
1.377	30-Oct-2015	msaitoh	Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH* too.
1.376	30-Oct-2015	msaitoh	igphy(4),ukphy(4): Set mii_mpd_* entries. wm(4): check PHY type correctly.
1.375	29-Oct-2015	msaitoh	Fix a bug that the multiqueue setting is done in a multiqueue capabile chip but can't use MSI-X on a machine. In that case, only one queue must be used. Written by knakahara@ and tested by me. This change should be fix a problem which was reported by Robert Swindells.
1.374	22-Oct-2015	knakahara	if_wm uses pci_intr_alloc stubs now.
1.373	22-Oct-2015	knakahara	add CTASSERT and KNF.
1.372	22-Oct-2015	knakahara	initial port rss_getkey() from freebsd, and use it instead of random().
1.371	14-Oct-2015	msaitoh	- Update TODO: - "Multiqueue" -> "TX Multiqueue" because RX Multiqueue was added. - Sort entries. - Grouping functions.
1.370	13-Oct-2015	christos	sprinkle __diagused
1.369	13-Oct-2015	knakahara	wm_linkintr_msix() should use WM_DEBUG_LINK instead of WM_DEBUG_TX.
1.368	13-Oct-2015	knakahara	recover if_wm.c:r1.348
1.367	13-Oct-2015	knakahara	fix: XEN3_DOM0(not defined __HAVE_PCI_MSI_MSIX) build
1.366	13-Oct-2015	knakahara	fix: WM_DEBUG build
1.365	13-Oct-2015	knakahara	support RX multiqueue. ok by msaitoh@n.o
1.364	13-Oct-2015	knakahara	ready to support RX multiqueue. ok by msaitoh@n.o
1.363	13-Oct-2015	knakahara	refactor: rearrange interrupt handlers argument
1.362	13-Oct-2015	knakahara	refactor: rearrange function arguments without interrupt handlers
1.361	13-Oct-2015	knakahara	change const value register macros to macro expressions to support multiqueue.
1.360	13-Oct-2015	knakahara	refactor: separate interrupts setup processing from wm_attach
1.359	13-Oct-2015	knakahara	allocating queues should be finished before establishing interrupts.
1.358	13-Oct-2015	knakahara	refactor: separate discriptors initialization and registers initialization
1.357	13-Oct-2015	knakahara	separete TX, RX queue mutex from wm_softc and rearrange WM_BOTH_LOCK. ok by msaitoh@n.o
1.356	13-Oct-2015	knakahara	refactor: separate TX, RX queue variables from wm_softc (without mutex)
1.355	13-Oct-2015	knakahara	refactor: separate busdma initialize processing as functions
1.354	13-Oct-2015	knakahara	separate TX dma control data and RX dma control data. ok by msaitoh@n.o
1.353	13-Oct-2015	knakahara	refactor: separate busdma allocate processing as functions.
1.352	13-Oct-2015	knakahara	refactor: change some macros to functions change below macros to functions - WM_CDTXSYNC - WM_CDRXSYNC - WM_INIT_RXDESC
1.351	08-Oct-2015	msaitoh	Fix a bug that LSC's interrupt storm occured when MSI-X is used. It was observed only on 82575.
1.350	30-Sep-2015	msaitoh	- Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Print the NVM image version on 82583, too. Pointed out by joerg@.
1.349	28-Sep-2015	knakahara	When the driver uses MSI-X, it tunes RX's EITR the same as TX's. ok by msaitoh@n.o
1.348	28-Sep-2015	knakahara	use sc->sc_itr instead of hard-coded number. ok by msaitoh@n.o
1.347	07-Sep-2015	msaitoh	- Check iNVM's image version and print it. - Update TODO.
1.346	17-Aug-2015	knakahara	Add kernel code to support intrctl(8).
1.345	28-Jul-2015	msaitoh	Print interrupt type correctly.
1.344	23-Jul-2015	msaitoh	Fix a bug that I21[01] NVM revision >= 3.25 can't be attached. Reported by MATSUI Yoshihiro.
1.343	23-Jul-2015	msaitoh	The WM_F_EEPROM_EERDEEWR is not required on iNVM device.
1.342	22-Jul-2015	knakahara	fix typo
1.341	22-Jul-2015	knakahara	Sorry, if_wm turns MSI/MSI-X default off by my mistake.
1.340	21-Jul-2015	knakahara	mofity for if_wm and if_bge to use pci_intr_alloc()
1.339	15-Jul-2015	msaitoh	Cleanup.
1.338	15-Jul-2015	msaitoh	Enable MSI/MSI-X feature by default.
1.337	15-Jul-2015	msaitoh	Don't use MSI on 8257[12] because of errata 63 "Byte Enables 2 and 3 are not set on MSI writes".
1.336	26-Jun-2015	msaitoh	Fix a bug that wm_detach() didn't unmap the FLASH's area. Now "drvctl -d wm0" -> "drvctl -r pci0" works on ICH* and PCH*.
1.335	13-Jun-2015	msaitoh	Add MSI/MSI-X support written by Kengo Nakahara. Some old devices' support is written by me. It's disabled by default. If you'd like to use, define WM_MSI_MSIX. Tested with: 8254[3405617] (INTx even if it has MSI CAP because of a errata) 8257[12], 82583 ICH8, ICH10, PCH2, PCH_LPT(I21[78]) (MSI) 8257[456], 82580, I35[04], I21[01] (MSI-X) Not tested: 82542, 82573, 80003, ICH9, PCH,
1.334	12-Jun-2015	msaitoh	- Print NVM version on 8257[124], too. - Modify NVM version string a bit.
1.333	10-Jun-2015	msaitoh	Remove extra if_start() in wm_ioctl().
1.332	08-Jun-2015	msaitoh	Call wm_pll_workaround_i210() at the end of the wm_reset() function, too.
1.331	06-Jun-2015	msaitoh	Fix the output of the option ROM version.
1.330	06-Jun-2015	msaitoh	Cleanup.
1.329	06-Jun-2015	msaitoh	- Add workaround for I210 Errata 25 and I211 Errata 10. - Add wm_gmii_gs40g_{read\|write}reg() and use it to access non-standatrd page. - Add wm_pll_workaround_i210() and call it when chip is i211 chip is i210 and it use INVM chip is i210 and NVM image version < 3.25 - Add comment - Rename macros.
1.328	06-Jun-2015	msaitoh	Print NVM image version.
1.327	06-Jun-2015	msaitoh	Revert previos. Sorry, I committed in another working directory...
1.326	06-Jun-2015	msaitoh	Print NVM image version.
1.325	02-Jun-2015	msaitoh	Fix a lot of bugs to make 82575 and newer's SERDES based systems work. - Add SERDES specific functions. - Fix IO pin configuration. - Reset autonego timer when link becomes up. TODO: - Fix a bug that SFP ROM can't read. - Perhaps some work is required for 8257[12] serdes systems. - Remove duplicated code in TBI's link related functions.
1.324	02-Jun-2015	msaitoh	No functional change: - Reorder functions. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick().
1.323	02-Jun-2015	msaitoh	Fix a bug that flags related to semaphore were incorrectly checked in wm_kmrn_{readreg,writereg}. i80003, ICH* and PCH* had this problem.
1.322	22-May-2015	msaitoh	- Currently, WM_F_EEE bit is not set on all chips. it's intended to not to use all of EEE fuctions but wm_set_eee_i350() leaves IPCNFG_10BASE_TE bit and it causes link negotiation problem on some old switches. Disable 10BASE-Te function, too. - Call wm_set_eee_i350() on some chips, too.
1.321	16-May-2015	msaitoh	Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines and NICs. Tested with Shuttle DS57U(iNVM based) and other non iNVM based I210 machines.
1.320	04-May-2015	msaitoh	For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way.
1.319	04-May-2015	msaitoh	Modify (E)ITR, TIDV and TADV related code: - ITR regiser are not documented in 82575 and newer devices' manual. The documets say "E"ITR(0) has no alias (to old ITR). But in reality, the alias really exists. When EITR(0) is written, the old ITR is changed. Before this commit, ITR was written after EITR was written. It causes that EITR's value (450) was overwritten with old ITR(1500). Set sc_itr first and use the value and don't set ITR if a device >= 82575 (which has MSI-X multi queue function). - Older than 82540 devices have no TADV register. - 82575 and newer devices have no TIDV and TADV registers.
1.318	04-May-2015	msaitoh	Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the Same as other OSes
1.317	04-May-2015	msaitoh	Remove WMREG_TQSA_LO and WMREG_TQSA_HIGH. Those registers are not described in documents and other OS's drivers don't access it. (I have no the first chip(82542)'s document. Those registers might be described in the document).
1.316	17-Apr-2015	msaitoh	Fix a bug that newer revision of I218-{LM,V} use wrong PHY access functions. The problem only occured on devices that the PCI device ID was 0x15a[0123].
1.315	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.314	28-Mar-2015	msaitoh	Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789.
1.313	16-Feb-2015	msaitoh	Fix a bug that the first access to NVM is failed on 8254[17] which use SPI EEPROM. Observed on Dell PowerEdge [12]850. Thanks Tom Ivar Helbekkmo for debugging.
1.312	15-Feb-2015	msaitoh	- Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment.
1.311	13-Feb-2015	msaitoh	Fix a bug that sc_mediatype is evaluated incorrectly. The real problem might be occured only on fiber and serdes cases.
1.310	29-Jan-2015	msaitoh	Fix a bug that a workaround for SWSM.SMBI bit doesn't work correctly. This problem was only occured with old boot ROM on 8257[12]. XXX pullup to netbsd-[67]
1.309	16-Jan-2015	ozaki-r	Introduce defflag for NET_MPSAFE
1.308	16-Nov-2014	msaitoh	branches: 1.308.2; Add C2000 KX and 2.5G.
1.307	24-Oct-2014	msaitoh	Simplify. No functional change.
1.306	24-Oct-2014	msaitoh	Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location.
1.305	17-Oct-2014	snj	There's an r in "determine"
1.304	10-Oct-2014	msaitoh	- Support DH89XXCC devices. - Add extra delay for 82580 and newer devices except DH89XXCC SGMII device. Same as FreeBSD.
1.303	07-Oct-2014	ozaki-r	Tidy up locking in wm_ioctl
1.302	07-Oct-2014	ozaki-r	Don't call ifmedia_ioctl with holding locks
1.301	06-Oct-2014	msaitoh	Fix debug message.
1.300	06-Oct-2014	msaitoh	Delete 82580ER related code. It was from FreeBSD and was removed in r203049.
1.299	06-Oct-2014	msaitoh	Add some devices. Not tested because I have no any of these devices/cards.
1.298	16-Sep-2014	msaitoh	Add some new I218 devices.
1.297	11-Sep-2014	msaitoh	It's not required to print "failed to detect NVM bank" with aprint_error_dev(). Use DPRINTF(). Same as {Free,Open}BSD.
1.296	04-Sep-2014	msaitoh	Remove one obsolete TODO entry.
1.295	03-Sep-2014	msaitoh	Read SFF SFP ROM and configure driver from the value.
1.294	01-Sep-2014	msaitoh	Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Calculate NVM word size correctly. - Determine timeout value based on the NVM word size.
1.293	29-Aug-2014	msaitoh	No binary change: - Move some NVM related macros from if_wm.c to if_wmreg.h. - Rename some macros for consistency.
1.292	28-Aug-2014	msaitoh	Add Internal SERDES mode support newer than or equal to 82575. Currently, it supports fiber device only. If you use 1000BaseT SFP module, it won't work. We have to write code to read SFP ROM and set SGMII mode if the module is copper.
1.291	26-Aug-2014	msaitoh	Modify WMP_F_1000{X,T} to WMP_F_{FIBER,COPPER}. No binary change.
1.290	24-Aug-2014	msaitoh	Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102.
1.289	10-Aug-2014	tls	branches: 1.289.2; Merge tls-earlyentropy branch into HEAD.
1.288	07-Aug-2014	msaitoh	Add one more TODO: - We should check XXX'ed comments.
1.287	06-Aug-2014	msaitoh	Don't check SWSM_SMBI bit if WM_F_LOCK_SWSM isn't set. Fix a problem when using vmware with e1000"e". With e1000e which is regarded as 82574L, wm_gmii_init() fails with "could not acquire SWSM SMBI" message without this change. This problem doesn't occur with real 82574L card. TODO: - Change the function name. It's difficult to understand. The name was from old Intel original e1000 driver. - Read Intel original driver again and check ALL location where semaphore is used.
1.286	05-Aug-2014	msaitoh	Add some TODOs in comment.
1.285	31-Jul-2014	msaitoh	Fix fiber link problem (PR#44776 and PR#30880). Tested with 82543GC, 82544EI, 82545EM, 82546GB 82571EB and 82572EI fiber cards. - Don't use the RXCFG interrupt. It's not required and the interrupt is very heavy (a lot of interrupts). Same as {Free,Open}BSD. - Modify wm_tbi_mediachange() to be close to em_setup_fiber_serdes_link() of {Free,Open}BSD. At least, don't forget to set duplex setting. - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. Tested with my own 82545EM card.
1.284	31-Jul-2014	msaitoh	Fix a bug that I added in rev. 1.273. Some old card stopped at the attach function with ""Please update the Bootagent" message. The code to clear the SWSM_SMBI bit is only for 8257[1234] and 82583.
1.283	28-Jul-2014	ozaki-r	Use two different mutexes for tx and rx This change splits the mutex of wm into two: one for tx and the other for rx. By doing so, lock contentions can be reduced. We lock both for other operations that need locking, e.g., init, stop and ioctl. The modification doesn't change the behavior of the driver.
1.282	25-Jul-2014	msaitoh	- Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD.
1.281	23-Jul-2014	msaitoh	No functional change: - Grouping functions. - KNF.
1.280	23-Jul-2014	msaitoh	No functional change: - Add missing prototypes. - Rename functions for consistency and clarify. - KNF.
1.279	22-Jul-2014	msaitoh	Add SERDES support from Matthias Drochner. Not all devices were tested but all SERDES entries are enabled.
1.278	14-Jul-2014	msaitoh	Cleanup comments. No functional change.
1.277	14-Jul-2014	ozaki-r	Remove a duplicated error message
1.276	11-Jul-2014	msaitoh	Fix compile error.
1.275	11-Jul-2014	msaitoh	Rename some WM_F_* flags to clarify. Renumbering flags. No functional change. WM_F_EEPROM_HANDSHAKE -> WM_F_LOCK_EECD WM_F_EEPROM_SEMAPHORE -> WM_F_LOCK_SWSM WM_F_SWFW_SYNC -> WM_F_LOCK_SWFW WM_F_SWFWHW_SYNC -> WM_F_LOCK_EXTCNF
1.274	11-Jul-2014	msaitoh	Restore the WM_F_EEPROM_SEMAPHORE flags for 82575-I354 and 80003 cases. The flag was removed in last commit.
1.273	11-Jul-2014	msaitoh	Fix some bugs realted to semaphore. This change fixes a problem which was exposed in if_wm.c rev. 1.271. Tested by riastradh@. - Clear the SMBI bit in SWSM register before accessing NVM and PHY in wm_attach(). Same as FreeBSD. - Fix a bug that 82573 doesn't put the hardware semaphore. Same as FreeBSD r256200.
1.272	01-Jul-2014	ozaki-r	Make if_wm MPSAFE - Make it MPSAFE only when NET_MPSAFE - otherwise, its instructions are almost same as before - the only change is IFQ_POLL/IFQ_DEQUEUE which is now single IFQ_DEQUEUE - Protect driver operations with a lock - further work would make it separate - Apply MPSAFE flag to - callout_init - pci_intr_establish - Stop proceeding packets when the driver is likely to stop for graceful ifconfig down Tested on Rangeley (I354) and KVM (e1000) Reviewed by msaitoh@
1.271	30-Jun-2014	ozaki-r	Acquire SW semaphore in wm_get_swsm_semaphore It is required to serialize device drivers running on different CPUs. The basic instructions are same as the official Intel driver. ok msaitoh@
1.270	16-Jun-2014	msaitoh	IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.269	27-May-2014	tls	From dyoung@ -- bump max TX DMA size to avoid pathological condition with TSO.
1.268	29-Mar-2014	christos	branches: 1.268.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.267	25-Mar-2014	christos	kill sprintf
1.266	07-Jan-2014	msaitoh	Insert completion barrier between register write and delay(). One exception is chip reset (to avoid hangup).
1.265	29-Dec-2013	msaitoh	Add support For I354(C2000 interna Ethernet controller): - Add I354 support. - Fix SGMII+MDIO case. SGMII+I2C is not supported yet. - Not tested well. - Sprinkle "XXX" to check later.
1.264	13-Sep-2013	martin	Remove unused variables
1.263	08-Sep-2013	msaitoh	Remove extra debug message.
1.262	16-Jul-2013	msaitoh	Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth.
1.261	27-Jun-2013	msaitoh	branches: 1.261.2; Move the location of wm_check_mng_mode() and wm_get_wakeup() in wm_attach(). Those functions access EEPROM, so they have to call after identifying EEPROM access type. This modification may change the behavior of BMC(IPMI).
1.260	25-Jun-2013	msaitoh	Sync the wm_enable_mng_pass_thru() function with FreeBSD. Don't check MANC_EN_MAC_ADDR_FILTER bit. Add 82574 and 82583 specific check. This modification may change the setting of WM_F_HAS_MANAGE flag on some machines. Sync the wm_release_manageablilty() fucntion with FreeBSD. Set MANC_ARP_EN. This change enables HW ARP function when entering suspend. When the chip is 82580(ER) or I350, set WM_F_ASF_FIRMWARE_PRES flag and check for the WM_F_ARC_SUBSYS_VALID flag. Same as FreeBSD.
1.259	25-Jun-2013	msaitoh	Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies.
1.258	19-Jun-2013	msaitoh	Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg.
1.257	19-Jun-2013	msaitoh	Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC.
1.256	19-Jun-2013	msaitoh	Fix some bugs on I21[78] to make those devices stable.
1.255	11-Jun-2013	msaitoh	Disable LPLU when the device is WM_T_PCH_LPT series (I21[78]). This change fixes a bug that wm interface don't negotiate to 1000BaseT.
1.254	11-Jun-2013	msaitoh	The wm_linkintr_gmii() function is called from interrupt, so call mii_pollstat() instead of mii_tick().
1.253	04-Jun-2013	msaitoh	Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2 but I've never seen the problem.
1.252	03-Jun-2013	msaitoh	Use wm_gmii_hv_{read,write}reg() for I21[78].
1.251	03-Jun-2013	msaitoh	Cleanup. No functional change.
1.250	02-Jun-2013	msaitoh	- Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878 - Remove a quick hack for NVM checksum that I added in last commit (rev. 1.249)
1.249	02-Jun-2013	msaitoh	Add I217 (and I218) support. - The name of I217 is similar to I210, but the function is rather similar to PCH2. - Not tested well. Tested with my own Intel DQ87PG which has I217LM onboard. - It seems that PCH2 and PCH_LPT specific function for RAL should be written. - Quick hack for the NVM checksum mismatch. if_wm.c currently has no wm_write_eeprom(), so it cannot update NVM's "updated bit". To avoid this problem, check only the last 12bits of the checksum. My own DQ87PG's updated bit is not set, and I could avoid the problem using with this hack.
1.248	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.247	18-Apr-2013	msaitoh	Add support I21[01]. Only tested with my own I210-T1 Ethernet Server Adapter.
1.246	30-Mar-2013	christos	remove trailing whitespace
1.245	13-Feb-2013	msaitoh	No functional change. - Fix comment. - Use macro.
1.244	13-Feb-2013	msaitoh	No functional change. - "struct mii_data *mii = &sc->sc_mii" and use it. - Add comments. - Fix some comments.
1.243	12-Feb-2013	msaitoh	Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542.
1.242	07-Feb-2013	msaitoh	Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350.
1.241	07-Feb-2013	msaitoh	Skip 64bit BAR correctly. I don't know if this bug causes a real problem.
1.240	30-Jan-2013	msaitoh	- Add WM_DEBUG_NVM. - If WM_DEBUG_NVM is enabled, dump the FLASH ROM data.
1.239	12-Dec-2012	msaitoh	- 82578DC is not PCH2 but PCH. 82579V is not PCH but PCH2. This bug was introduced in rev. 1.221. Reported by FUKAUMI Naoki. - Fix comment.
1.238	15-Nov-2012	msaitoh	Change style a bit.
1.237	07-Nov-2012	msaitoh	Fix a bug that PHY isn't set to low-power mode on PCH and PCH2.
1.236	12-Oct-2012	msaitoh	Explicitly cast to uint64_t instead of using PRIxPADDR because sizeof(paddr_t) != sizeof(bus_addr_t) on some architecture. See: http://mail-index.netbsd.org/source-changes-d/2012/10/09/msg005353.html
1.235	09-Oct-2012	msaitoh	Use PRIxPADDR to print a DMA address. This fix a problem that if_wm.c can't compile with WM_DEBUG on non-64bit platforms.
1.234	01-Sep-2012	matt	branches: 1.234.2; Shut up gcc about some uninitialized variables.
1.233	30-Aug-2012	msaitoh	Enable VLAN hardware tagging on all chips that have the new queue mechanism. Tested with 82575{EB,GB}, 82576, 82580, I350 and ICH9.
1.232	29-Aug-2012	bouyer	Make vlan and all ip/ip6 checksum offload work for the I350. On newer devices, when using the legacy TX descriptors, vlan-related flags that were set on the last descriptor of a packet have to be set on the first one. For tso/checksum offloads, a new "advanced" descriptor format has to be used. Change wcd_txdescs to a union defining all types of descriptors (they are all 16-bytes wide). Define a new tx function wm_nq_start(), which handle newer devices. There is some code duplication with wm_start(), but adding support to the existing wm_start() would make it a if () {} else {} maze. This also allows to get rid of some workaround for older chips that are not needed here. Use wm_nq_start() instead of wm_start() for the I350 (this should probably be for all WM_F_NEWQUEUE devices, but I have no hardware but the I350 to test). Call ifp->if_start() instead of wm_start() where is matters. Tested on a I350, and a i80003 (which use the old format), both with and without vlans, with and without checksum offloads.
1.231	09-Aug-2012	msaitoh	Fix the check of the device type in last commit. Reported by Thomas Klausner.
1.230	09-Aug-2012	msaitoh	Add workaround for QEMU and the variants that fail on EEPROM access. This problem was discovered a few years ago, but some variants and cloud services still have the bug. This problem is not NetBSD's bug but qemus' bug. For NetBSD users, existence of buggy virtual machines is sad thing, so we add a workaroud.
1.229	22-Jul-2012	matt	Fix mii_statchg to take a 'struct ifnet *' instead of device_t. This fixes problem with a common MDIO bus used for multiple interfaces. Some drivers converted to CFATTACL_DECL_NEW.
1.228	25-May-2012	msaitoh	Add support Intel I350 Ethernet.
1.227	02-Feb-2012	tls	branches: 1.227.2; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.226	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.225	28-Nov-2011	bouyer	branches: 1.225.2; There is problem with 82576 chips (and probably 82575 too) with hardware vlan tagging: some packets are sent untagged on the wire. Follow OpenBSD and disable hardware vlan tagging for these chips (I couldn't find a hint in other open-source drivers at what could be wrong ...)
1.224	19-Nov-2011	tls	First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.223	01-Jul-2011	matt	branches: 1.223.2; Fix uninitialized data warning found by gcc 4.5
1.222	24-May-2011	msaitoh	Fix invalid PBA setting on ICH{9,10}. This bug was introduced in the last commit. I noticed this problem via mail from spz.
1.221	20-May-2011	msaitoh	- Add PCH2 support. - Add 82579 support. - Change PBA size for PCH from 10K to 26K as FreeBSD's em-7.1.7 - Add yet another 82567V support. - Add ICH10+HANKSVILL support. - Add 82580 quad-1000BaseX support.
1.220	22-Feb-2011	dyoung	Use /FALLTHROUGH/.
1.219	06-Feb-2011	bouyer	wm_gmii_reset(): restore generic reset delays to what they were before rev 1.186. This makes the following hardware find its PHY again, and I can't see how these larger delays could be a problem for other hardware: wm0 at pci6 dev 7 function 0: Intel i82541GI 1000BASE-T Ethernet, rev. 5 wm0: interrupting at ioapic2 pin 0, event channel 5 wm0: 32-bit 66MHz PCI bus wm0: 65536 word (16 address bits) SPI EEPROM wm0: Ethernet address 00:13:72:54:ee:13 igphy0 at wm0 phy 1: Intel IGP01E1000 Gigabit PHY, rev. 0 igphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto
1.218	26-Jan-2011	msaitoh	Fix MAC address check on 8257[156] and 80003 case. Some cards have non 0xffff pointer but those don't use alternative MAC address in reality. So we check whether the broadcast bit is set or not like Intel's e1000 driver. Fixes PR kern/44072 reported by Jean-Yves Moulin. Remove extra Warning for newer cards.
1.217	14-Dec-2010	dyoung	branches: 1.217.2; 1.217.4; Stop wm(4) from needlessly resetting when you add or delete a vlan(4): ifconfig vlan0 create vlan 2 vlanif wm0 ifconfig vlan0 destroy It pays to avoid a reset because after a reset it can take about 30 seconds before wm0 is back on the net. The patch garnered no objections on tech-net@. Sponsored by CoyotePoint Systems Inc.
1.216	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.215	16-Oct-2010	taca	- Change "else" + space + tab "if" to "else" + space + "if". - A word "mechanism" was accidently separated in to "me" and "chanism" in comment. No functional chanage and I found it while reading if_wm.c yesterday.
1.214	26-Jul-2010	jym	Add PAE to ALL kernel, so that most paddr_t format string errors get caught during compilation. While here, fix the compilation for ALL.
1.213	21-Jul-2010	msaitoh	Use ether_set_ifflags_cb()
1.212	19-Jul-2010	jakllsch	Omit U+00AE "REGISTERED SIGN" in a product name due to its non-ASCII nature. wm_release_hw_control() in wm_detach() before we unmap the registers we need. Unmap I/O space during detach.
1.211	14-Jul-2010	msaitoh	- s/TBDA/TDBA/. It stands for Transmit Descriptor Base Address. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. TODO: - ip4csum doesn't work on 82575 and newer devices (reported by Paul Goyette), so we have to fix it.
1.210	28-Jun-2010	msaitoh	Fix a bug that wm_attach() fails after reading MAC address. Fixes PR#43533. Fix an uninitialized value on some cases.
1.209	25-Jun-2010	msaitoh	Only 82571 shares port 0 of EEMNGCTL_CFGDONE.
1.208	25-Jun-2010	msaitoh	Add some code to support 82580[ER]. Tested on my own I340-T4. - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add some code to support 82580. TODO: - ukphy -> somephy
1.207	25-May-2010	msaitoh	Fix the names of 82577L[MC] LAN controllers (for mobile). Fix typo.
1.206	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.205	10-Mar-2010	msaitoh	branches: 1.205.2; 82576 is dual port, so check the FUNCID and increment the MAC address for the 2nd port.
1.204	07-Mar-2010	msaitoh	Fix the type of sc_ss. That is not bus_space_handle_t but bus_size_t. Reported by jdc.
1.203	07-Mar-2010	msaitoh	- Add code for WOL, ASF, IPMI and Intel AMT. - wm_enable_wakeup() is disabled by default. If you want to use WOL with the Magic Packet, define WM_WOL. - Add the following flags: WM_F_ASF_FIRMWARE_PRESENT WM_F_ARC_SUBSYSTEM_VALID WM_F_HAS_AMT WM_F_HAS_MANAGE WM_F_WOL - Add wm_suspend() and wm_resume(). Give/get the control to/from the firmware. - Need more work for PCH. See wm_enable_phy_wakeup(). - Enable wm_get_hw_control() for 82574 and 82583. - Add Yet another workaround for ICH8. - Add wm_igp3_phy_powerdown_workaround_ich8lan() for power down problem on D3.
1.202	07-Mar-2010	msaitoh	Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working.
1.201	07-Mar-2010	msaitoh	Add the detach code.
1.200	25-Feb-2010	msaitoh	Merge if_wm.c rev. 1.162.4.13 and if_wmvar.h rev. 1.2.46.3 into maintrunk. Increase the timeout value to prevent timeout on 80003. The values are taken from Intel's e1000 driver. Tested by spz and me.
1.199	16-Feb-2010	msaitoh	Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems.
1.198	16-Feb-2010	msaitoh	indent, tabify and remove extra spaces.
1.197	04-Feb-2010	msaitoh	branches: 1.197.2; Don't increment when ICR_RXO is set because we count them by WMREG_MPC counter.
1.196	04-Feb-2010	msaitoh	- Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops.
1.195	22-Jan-2010	martin	Unify the name of the device property to hold a MAC address - there was no clear majority for either "mac-addr" vs. "mac-address", but a quick gallup poll among developers selected the latter.
1.194	21-Jan-2010	msaitoh	Remove an extra debug printf(). KNF. No functional change.
1.193	19-Jan-2010	pooka	Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.192	14-Jan-2010	msaitoh	Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.191	12-Jan-2010	msaitoh	- Add i82567LM-2 i82567LM-4 i82567V-3 LAN controller. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. Now we can linkup 1000BaseT on PCH. It seems that we have to do the same work for ICH9.
1.190	11-Jan-2010	msaitoh	Add partial support for PCH. We have to do more work for PCH. Especially I don't know what driver i82578 should be attached to. makephy? or atphy? MII_OUI() says that the PHY is from Attansic (== Atheros). Intel's e1000 driver says that it's close to makphy... I can't link up at 1000BaseT yet...
1.189	07-Jan-2010	msaitoh	Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9.
1.188	07-Jan-2010	msaitoh	bugfixes: - add missing break in wm_reset()... - fix the offset of WMREG_PBS... - fix length of some delay()s in wm_gmii_reset()
1.187	05-Jan-2010	msaitoh	Fix the checking of jumbo frame function though I don't know whether the function wokrs or not... Remove the extra macro definition for the offset 0x1a in EEPROM.
1.186	05-Jan-2010	msaitoh	Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9.
1.185	29-Dec-2009	msaitoh	- Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - Add support for i82583V. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574.
1.184	27-Dec-2009	msaitoh	Fix the bug that ICH9 can't found a PHY. This fix is not good, but it's the same as e1000 driver... Fixes PR#42237
1.183	16-Dec-2009	msaitoh	Sync with Intel's original em driver: - Add dspcode for igp3 and use it when the EEPROM isn't available. - Add some delays. - Stop the PHY transmitter before patching the DSP code and restart it after wrote. - Save and restore register 0x2f5b.
1.182	16-Dec-2009	msaitoh	Re-enable igphy's 82566 support. - Patch for the DSP code is only for 8254[17] and we have to apply the different patches between rev. 1 and rev. 2. - The workaround for analog fuse is only for 82547 rev. 1. - The workaround for smartspeed is only for 8254[17] see http://mail-index.netbsd.org/tech-net/2009/08/05/msg001546.html
1.181	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.180	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.179	06-Aug-2009	msaitoh	If the difference bettween last flag and new flag is only IFF_PROMISC or IFF_ALLMULTI, set multicast filter only to prevent link down. Tested by Mark Davies and me. Fixes PR#29126 for wm.
1.178	30-Jul-2009	msaitoh	Fix the bug that another lock mechanism is used to access Kumeran registers on i80003 and ICHs. Fix yet another i80003 ONLY workaround. The code to modifing TIPG register is only for i80003. Set the Re-Transmit on Late Collision(RTLC) flag for all devices. Fix a typo in a printf message.
1.177	14-Jul-2009	msaitoh	Some fixes for i80003 and ICH{8,9,10} from e1000 driver and document: Add setting for KABGTXD register for ICH{8,9,10}. ICH9 and ICH10 has no FCAL, FCAH and FCT like ICH8. Add special setting for FCTTV and TCTL_EXT register for i80003 The special setting for TIPG is only for i80003. Some of kumeran settings are only for i80003's bugs. Add some ICH10 fixes.
1.176	13-Jul-2009	msaitoh	Check PCI-X mode as e1000 driver.
1.175	29-May-2009	darran	Add vlan support and hardware offload capabilities to agr. These changes allow vlans to be layered above agr, with the attach and detach propogated to the member ports in the aggregation. Note the agr interface must be up before the vlan is attached. Adds SIOCINITIFADDR support to the wm driver for setting the AF_LINK address, necessary for agr to be able to set the mac addresses of each port to the agr address (i.e. so it can receive all intended traffic at the hardware level). Adds support for disabling the LACP protocol by setting LINK1 on the agr interface (e.g. ifconfig agr0 link1). In consultation with tls@.
1.174	07-Apr-2009	msaitoh	Reload sc_ctrl in wm_reset(). Add an ICH10 entry. Remove some obsolete comments.
1.173	07-Apr-2009	msaitoh	Fix about TBI mode. This fix doesn't influence MII mode. - Fix panic in mediachange. - Fix SWDPIN(1)'s polarity on some chips. - Fix flow control stuff (includes PR#32009). - Stop RXCFG storm. It ocours easily. - And more fix about autonego. Tested on PRO/1000F, PRO/1000XF and PRO/1000 MF.
1.172	31-Mar-2009	darran	Missed a small but important change to enable hardware VLAN support.
1.171	31-Mar-2009	darran	Enable the vlan code that thorpej@NetBSD.org added, and fix a bug where the vlan tag was read from the rx descriptor after it was reinitialized (and hence was always 0). Hardware vlan support looks good tested between two vmware clients. With support from tls@NetBSD.org.
1.170	29-Mar-2009	msaitoh	Add workaround for 82543GC. We need to force speed and duplex on the MAC equal to what the PHY speed and duplex configuration is. Fixes PR#36430.
1.169	20-Mar-2009	msaitoh	Examine the management mode and mark DRV_LOAD (for ICH{8,9,10},80003, 8257{1,2,3,4}). Add some error's printf(). Make the bank detect routine into the function.
1.168	20-Mar-2009	msaitoh	On i82563, the em driver says that the ready bit in the MDIC register may be incorrectly set. Insert delay(200) like the em driver. Fixes PR#41014.
1.167	10-Mar-2009	msaitoh	add an entry for 82567LF-3. fix the register access for ICH10DO.
1.166	28-Feb-2009	msaitoh	Fix the multicast hash bug on ICH9's wm. Now we can catch ff02::9 on ICH9's wm.
1.165	27-Feb-2009	sborrill	Add support for 82574L (as found on Desktop CT adaptor) based on 82573 support.
1.164	27-Jan-2009	markd	branches: 1.164.2; Add i82567LM-3
1.163	02-Dec-2008	sketch	Fix '\|\|' vs '&&' mix-up from when ich9 support was added. Fixes nasty EEPROM-trashing bug on ich8 and ich9 chipsets as described in http://mail-index.netbsd.org/current-users/2008/12/02/msg006435.html caused by setting the wrong type of EEPROM.
1.162	15-Oct-2008	bouyer	branches: 1.162.2; 1.162.4; Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.
1.161	09-Sep-2008	cegger	make an i386/ALL kernel build
1.160	08-Sep-2008	christos	cf_attach_decl_new
1.159	15-Aug-2008	simonb	When receiving a chain with multiple mbufs (eg a jumbo frame) make sure the last mbuf in the chain doesn't have a negative length when adjusting lengths when removing the FCS (which leads to a panic in sbcompress()). Fix contributed by Wasabi Systems.
1.158	10-Apr-2008	cegger	branches: 1.158.4; 1.158.6; 1.158.10; use aprint_*_dev and device_xname
1.157	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.156	11-Mar-2008	dyoung	Prepare for PMF self-suspension: in the if_stop() methods, clear IFF_UP and IFF_RUNNING before running the 'disable' step, instead of after. Soon I will handle the 'disable' step by calling into PMF, which may call if_stop(, 0). Ordinarily, that is harmless. This change lets the if_stop() routines exit early when they find on entry that IFF_RUNNING is not set.
1.155	23-Feb-2008	rafal	Request DMA-coherent memory for the rx/tx descriptors. Makes this work on sgimips (tested on my O2). Hint re: descriptor size << cacheline size from Tsutsui-san; thanks!
1.154	07-Feb-2008	dyoung	branches: 1.154.2; 1.154.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.153	29-Jan-2008	tls	Sigh. What was meant here was an ITR register value of 1500, for 2604 interrupts/sec -- not the other way around. Caught by yamt. When I can confirm that it won't lock the chip up on the models claimed to be problematic, I'll probably adjust the packet timers a bit further to see if I can get latency down under low load. But this should be better.
1.152	19-Jan-2008	dyoung	Make many ethernet drivers share the common code for MII media handling, ether_mediastatus() and ether_mediachange(). Check for a non-ENXIO error return from mii_mediachg(). (ENXIO indicates that a PHY is suspended.) This patch shrinks the source code size by 979 lines. There was a 5100-byte savings on the NetBSD/i386 kernel configuration, ALL. I have made a few miscellaneous changes, too: gem(4): use LIST_EMPTY(), LIST_FOREACH(). mtd(4): handle media ioctls, for a change! axe(4): do not track link status in sc->axe_link any longer nfe(4), aue(4), axe(4), udav(4), url(4): do not reset all PHYs on a change of media Except for the change to mtd(4), no functional changes are intended. XXX This patch affects more architectures than I can feasibly XXX compile and run. I have compiled macppc, sparc64, i386. I XXX have run the patches on i386 boxen with bnx(4) and sip(4). XXX Compiling and running on evbmips (MERAKI, ADM5120) is in XXX progress.
1.151	11-Jan-2008	ragge	Add entry to match the "Intel� PRO/1000 PT Quad Port Server Adapter".
1.150	14-Dec-2007	tls	Adjust interrupt moderation for newer chips to be saner according to the 82571 manual and Intel Application Note 450. Previously, we were setting RADV and TIDV/TADV values that didn't make any sense given the enormous ITR value we were setting (well outside the range recommended by Intel and quite possibly rejected silently by the chip as junk) and setting RADV without setting RDTR, which is explicitly documented as having no effect. A considerable performance improvement is achieved for TCP and UDP at gigabit speed. I need to revisit this to deal with the timer ticks being 4X as long when the chip's in 100mbit mode, and to set values for the older chips' interrupt timers that are more like what the appnote recommends. This should help for 82543 and newer, though.
1.149	09-Dec-2007	jmcneill	branches: 1.149.2; Merge jmcneill-pm branch.
1.148	30-Oct-2007	simonb	branches: 1.148.2; 1.148.4; 1.148.6; Some space-to-tab changes. XXX: All the recently added ich8 support functions use 4-space indent, not touching that...
1.147	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.146	30-Aug-2007	msaitoh	branches: 1.146.4; Add more work for the EEPROM reaload sequence. Fix typo.
1.145	30-Aug-2007	msaitoh	EECD_EE_AUTORD is only for 8257{123},80003,ICH{89}. Will fix "auto read from eeprom failed to complete" problem.
1.144	28-Aug-2007	msaitoh	add ICH9 need more work?
1.143	26-Aug-2007	dyoung	branches: 1.143.2; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.142	09-Jul-2007	ad	branches: 1.142.2; 1.142.6; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.141	29-May-2007	simonb	Fix a tyop.
1.140	30-Apr-2007	simonb	Restore $NetBSD$ tag.
1.139	29-Apr-2007	bouyer	Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2
1.138	27-Mar-2007	salo	fix typos in last, Eratta->Errata
1.137	27-Mar-2007	msaitoh	Workaround for 82541 Eratta 29 and 82547 Eratta 28. These devices have to reset the PHY before reset the MAC. Reported and tested by salo.
1.136	13-Mar-2007	msaitoh	branches: 1.136.2; KNF
1.135	04-Mar-2007	christos	branches: 1.135.2; 1.135.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.134	23-Feb-2007	msaitoh	fix media link issues with fiber-based card fixes PR#35797
1.133	23-Feb-2007	msaitoh	fix some negotiation problems on wm(4). will fix PR#30078, PR#30490, PR#30906, PR#33429, PR#35386.
1.132	21-Feb-2007	thorpej	Pick up some additional files that were missed before due to conflicts with newlock2 merge: Replace the Mach-derived boolean_t type with the C99 bool type. A future commit will replace use of TRUE and FALSE with true and false.
1.131	23-Nov-2006	yamt	branches: 1.131.2; 1.131.4; implement ipv6 TSO. partly from Matthias Scheler. tested by him.
1.130	16-Nov-2006	yamt	wm: add IPv6 rx tcp/udp checksum offloading.
1.129	16-Nov-2006	yamt	wm_tx_offload: in the case of "txtsopain", m_copyback ip header to the correct offset.
1.128	15-Nov-2006	yamt	remove __unused from function parameters.
1.127	21-Oct-2006	bouyer	Add support for the Intel 80003 Gigabit Ethernet controller (found e.g. in newer server chipsets) to wm(4), from the FreeBSD em(4) driver. While there, add a few other Intel Ethernet controller that should work as is. Properly update the RX error and TX collision counters. Add ikphy(4), a driver for the Intel i82563 Kumeran 10/100/1000 Ethernet PHYs
1.126	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.125	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.124	05-Jul-2006	wrstuden	branches: 1.124.4; 1.124.6; Fix len=-4 issues. When receiving a packet, if a receive buffer has length 0, treat it the same way we treat a failed mapping. Fix contributed by Wasabi Systems.
1.123	20-Jun-2006	jmcneill	Add powerhook for wm(4), from George Michaelson <ggm AT apnic DOT net>
1.122	17-Jun-2006	christos	re-factor the pci powestate api. reviewed by gimpy
1.121	16-Jun-2006	msaitoh	fix the MAC address of the second port of the 82571
1.120	10-Jun-2006	msaitoh	branches: 1.120.2; o i82573 doesn't support jumbo frame o set PBA for i8257[123] o set MULA when sc_type == i8257[123]
1.119	10-Jun-2006	uebayasi	KNF.
1.118	10-Jun-2006	msaitoh	fix compile error
1.117	10-Jun-2006	msaitoh	Enable i82573!!! TODO: add some code for FLASH fix phy stuff
1.116	05-Jun-2006	msaitoh	add support 82546GB_PCIE(PRO/100MT), 82541ER_LOM, 82547EI_MOBILE, 82571EB_COPPER, 82571EB_FIBER, 82572EI_COPPER, 82572EI_FIBER and 82572EI(PRO/1000PT) TODO: add i82573 support
1.115	05-May-2006	thorpej	branches: 1.115.2; Remove the devprop API and switch everthing over to the new proplib. Add a new device_properties() accessor for device_t that returns the device's property dictionary.
1.114	27-Feb-2006	thorpej	branches: 1.114.2; 1.114.4; 1.114.6; cfg1, cfg2, and swdpin are not generic properties the same way that mac-addr is. Prefix them with i82543-
1.113	22-Feb-2006	gavan	Convert MD hook to device property reads. Device properties, if present, take precedence over EEPROM values.
1.112	16-Feb-2006	gavan	Compute the EEPROM checksum to determine the presence of valid EEPROM data. If EEPROM data is invalid, allow a MD hook to supply EEPROM data instead.
1.111	07-Feb-2006	thorpej	In the receive interrupt handler, compute the packet length before resetting m to point to the first buffer in the packet. The bug would cause the length of jumbo frames to be incorrect.
1.110	24-Dec-2005	perry	branches: 1.110.2; 1.110.4; 1.110.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.109	11-Dec-2005	christos	merge ktrace-lwp.
1.108	29-Oct-2005	yamt	don't reset interface on rx overrun because it makes the situation worse, if the overrun is merely a symptom of heavy load. increament if_ierrors instead. PR/29831.
1.107	10-Aug-2005	yamt	branches: 1.107.2; add ipv6 tx checksum offloading support for wm driver. reviewed by Jason Thorpe.
1.106	07-Aug-2005	yamt	wm_rxintr: ignore h/w assisted checksum results if WRX_ST_IXSM is set.
1.105	04-Aug-2005	skrll	Fix a typo.
1.104	25-Jul-2005	ross	Fix compilation nits with WM_DEBUG.
1.103	02-May-2005	yamt	branches: 1.103.2; split IFCAP_CSUM_xxx to IFCAP_CSUM_xxx_Rx and IFCAP_CSUM_xxx_Tx.
1.102	26-Apr-2005	scw	In wm_stop(): - Clear the interrupt mask to ensure the device cannot assert its interrupt line. - Clear sc->sc_icr to ensure wm_intr() makes no attempt to service any currently pending or shared interrupt. This should address the problem reported in both PR/29903 and PR/22493. (wm(4) is susceptible to crashing in wm_intr() when the interface is downed)
1.101	19-Mar-2005	tron	Add support for Intel PRO/1000 GT. Change provided in private e-mail by unex at linija dot org.
1.100	11-Mar-2005	matt	branches: 1.100.2; Set ip_len to 0 in the wm driver when TSO is being used.
1.99	09-Mar-2005	matt	Add TCP Segment Offload support.
1.98	05-Mar-2005	thorpej	- Set IFCS on every data descriptor (hint from Linux driver). - Make sure IDE is set on context descriptors.
1.97	05-Mar-2005	thorpej	txpullup event counters are no longer needed.
1.96	27-Feb-2005	perry	nuke trailing whitespace
1.95	21-Feb-2005	thorpej	Define M_CSUM_DATA_IPv4_IPHL() and M_CSUM_DATA_IPv4_OFFSET() macros to extract data from csum_data, rather than just open-coding it.
1.94	20-Feb-2005	jdolecek	use VLAN_* macros for VLAN tag extraction/addition XXX the HW VLAN tagging code stays disabled, same as before
1.93	18-Feb-2005	thorpej	Re-apply changes that were mistakenly lost by the previous revision.
1.92	18-Feb-2005	briggs	Update the interrupt mitigation support. Actually initialize the interrupt throttling register. Actually initialize the tx and tx absolute interrupt delay timer regs. Update default values for the rx and tx delay timers. Inspired by Zdenek Salvet in PR kern/29373.
1.91	18-Feb-2005	heas	Adjust csum_data usage; mask off the high 16 bits for L4 header stuff offset and make use of the L3 header length from the high 16 bits such that an m_pullup is no londer necessary.
1.90	30-Jan-2005	thorpej	Eliminate use of M_HASFCS.
1.89	30-Nov-2004	briggs	branches: 1.89.4; 1.89.6; Minor tweak to add newline to an aprint_error().
1.88	24-Nov-2004	briggs	The 8254x apparently doesn't respond well to I/O accesses at 0. Do not attempt to map the I/O BAR if it is 0. This will have the side-effect of causing the wm_reset() routine to fall back to memory BAR accesses for those chips on which it would attempt I/O BAR accesses for the reset. Also update the comments about why we attempt the I/O BAR accesses for the reset in favor of the memory BAR accesses.
1.87	24-Nov-2004	thorpej	Don't bother caching the checksum context. Other drivers for this chip don't, and it seems as though doing it may cause some problems.
1.86	23-Nov-2004	thorpej	If we have to do an m_pullup() when setting up offload parameters, make sure that the pullup was done into trailing space at the end of the existing first-mbuf-in-chain. If not, log an error and drop the packet, because the mbuf chain no longer corresponds to the DMA map, and we are already committed to transmitting the packet.
1.85	23-Nov-2004	thorpej	Also toggle the second port's Ethernet address lsb on the WM_T_82546_3.
1.84	22-Nov-2004	thorpej	Use log() instead of printf().
1.83	30-Oct-2004	thorpej	When adding/deleting multicast addresses, only whack the address filter if the interface is marked RUNNING. Fixes kern/27678.
1.82	06-Oct-2004	thorpej	Define a WM_MAXTXDMA constant to represent the largest Tx DMA any given packet will require.
1.81	06-Oct-2004	thorpej	Sync the Tx DMA map after offload stuff has been dealt with, since the offload set-up digs into the packet.
1.80	06-Oct-2004	thorpej	Restructure the loop in wm_start() that passes over the DMA map segments slightly to make it easier to drop in errata work-arounds that require adding extra descriptors onto the end of the packet.
1.79	05-Oct-2004	thorpej	Largest Tx descriptor buffer size is actually 4K.
1.78	05-Oct-2004	thorpej	Work around another problem (that causes horrible performance) on the 82547: - Change the Tx:Rx on-chip RAM allocation ratio based on MTU: => 82547: > 8192 18K:22K, else 10K:30K => others: > 8192 24K:40K, else 16K:48K Values derived from Linux driver. - On the 82547, keep track of how much of the Tx FIFO has been used. When in half-duplex mode, don't let packets wrap around the FIFO ring buffer, because that causes the chip to croak. Detect this, stall the Tx queue, use a timer to wait for the packets to drain from the Tx FIFO, reset the internal FIFO pointers, and restart the Tx queue. Basic algorithm (and some magic numbers) derived from FreeBSD and Linux drivers.
1.77	14-Sep-2004	simonb	Remove an unused event counter.
1.76	21-Aug-2004	thorpej	Make the product table static.
1.75	05-Aug-2004	thorpej	- Dump the mbuf chain if we encounter a "too many DMA segments" error on transmit. - On 82544 and up, allocate 4096 Tx descriptors rather than just 256. 82543 and down must still use 256, due to errata. - Allow up to 256 DMA segments per packet. I have observed some truly pathological mbuf chains under certain (admittedly uncommon) workloads when jumbo frames are in use.
1.74	15-Jul-2004	tron	Reduce transmit queue len to 16 for the Intel i82547 to avoid FIFO overruns. This fixes the horrible send performance reported in PR kern/23782. Reviewed by Frank van der Linden and Jason R. Thorpe.
1.73	13-Jul-2004	tron	The Intel i82547EI gigabit ethernet controller isn't really a PCI device. So instead trying to find and print out the PCI bus properties mark it as a CSA device.
1.72	12-Jul-2004	tron	Make this build if WM_DEBUG is defined.
1.71	16-May-2004	thorpej	Add flow control support, based on work by HITOSHI Osada.
1.70	06-May-2004	ragge	Remove the "panic("wm_init: MCLBYTES > 2048 not supported.");" stuff, the driver works very well with larger MCLBYTES.
1.69	09-Apr-2004	thorpej	Add support for 64-bit PCI DMA addressing.
1.68	19-Feb-2004	thorpej	branches: 1.68.2; s/WTC_DTYP_D/WTX_DTYP_D/
1.67	19-Feb-2004	thorpej	Fix typo in debug message. From Jeffrey Hsu.
1.66	19-Feb-2004	thorpej	Add missing txseg20 counter name. Pointed out by Jeffrey Hsu.
1.65	14-Jan-2004	tsutsui	Fix hardware cksum problem on big endian machines. Tested on macppc by HATANO Hiromichi. Note I guess this bug could be fixed only adding one htole32(), but I'd rather clean up endianness handling: - Use htole32() only to accesses against DMA descriptors. - Don't use uint32_t union member with htole32()/le32toh() to access uint8_t/uint16_t descriptors. - Add le32toh() in some DPRINTF. (XXX: strictly speaking, bus_dmamap_sync() is needed for these DPRINTF)
1.64	04-Dec-2003	thorpej	Bump the Tx DMA segment count to 40. Apparently, it's common to have DMA segment counts in this ballpark when using jumbo frames (so says the "ti" driver).
1.63	22-Nov-2003	thorpej	Add a missing i82541 ID to the table.
1.62	07-Nov-2003	thorpej	Fix brain'o in PCIX speed detection. kern/23375 (John R. Shannon).
1.61	04-Nov-2003	thorpej	Enable i82541 and i82547 chips - it works well enough to get broader testing. Update the TODO list.
1.60	03-Nov-2003	ichiro	Fix uninitialized variable warnings
1.59	25-Oct-2003	christos	Fix uninitialized variable warnings
1.58	23-Oct-2003	ragge	Set snd queue size to max(WM_IFQUEUELEN, IFQ_MAXLEN), instead of hard to WM_IFQUEUELEN.
1.57	22-Oct-2003	thorpej	Add support for the SPI EEPROM found on the newer i82541 and i82547 chips. Add the i82541 and i82547 product IDs to the table, #if 0'd out, for now (there are still more changes to come for these new chips to work properly).
1.56	21-Oct-2003	thorpej	Add a TODO list item.
1.55	21-Oct-2003	thorpej	- Move the i82540EM (LOM) along with the other i82540 product IDs. - Match the i82545GM (1000BASE-T and 1000BASE-X) and i82546GB (1000BASE-T and 1000BASE-X) product IDs.
1.54	21-Oct-2003	thorpej	If we're on PCI-X, look up the PCI-X capability. Verify that the max memory read byte count in the command register is <= the max designed memory read byte count in the status register. If not, reset it to the status register's value. This is necessary because some BIOSes configure the wrong value, and the wrong value confused some versions of the PCI-X capable chips.
1.53	21-Oct-2003	thorpej	- Map I/O space on chips that support it. While it is not used for normal operation, it is required to implement a bug work-around on some chips. - When issuing a reset on the i82544, i82540, i82545, i82546, and i82541, use I/O space, if available, to work around a problem with the memory write cycle (problem only occurs when issuing a reset).
1.52	21-Oct-2003	thorpej	Record characteristics of the bus connection at attach time.
1.51	20-Oct-2003	thorpej	- Do better error checking on EEPROM access. - Let the EEPROM wire-protocol handler perform the loop.
1.50	20-Oct-2003	thorpej	Move all the code that handles the MicroWire EEPROM protocol into a separate function.
1.49	20-Oct-2003	thorpej	Fix a typo in the "failed to acquire eeprom" case.
1.48	20-Oct-2003	thorpej	Move the code that reads a series of bits from the EEPROM into a separate function.
1.47	20-Oct-2003	thorpej	Make all of the functions in this file static.
1.46	20-Oct-2003	thorpej	Factor out the code that sends bits to the EEPROM into a separate function.
1.45	20-Oct-2003	thorpej	Move the code that handle the EEPROM handshake into separate functions.
1.44	17-Oct-2003	thorpej	Store the number of EEPROM address bits in the softc.
1.43	17-Oct-2003	thorpej	Add internal representation for i82545 rev 3, i82546 rev3, i82541, i82541 rev 2, i82547, and i82547 rev 2.
1.42	17-Oct-2003	thorpej	Make the "align tweak" factor per-interface (since the MTU is per-interface).
1.41	10-Sep-2003	tls	Add support for jumbo (9K) frames. Add support for larger receive descriptor sizes if larger mbuf clusters are in use -- currently commented out because, for some reason, transmit doesn't work at all with large mbuf cluster sizes. Note that using actual 9K frames is a _lot_ less efficient than using frames that are a little smaller than 8K so that they fit in one or two pages (instead of two or three) and avoid the extra chained descriptor with only 1K of data in it, too. TCP will segment things just fine for an 8K total packet size, just as it does for 1500-byte packets. On my system, using "mtu 8000" with ifconfig, I get an almost 50% performance boost for TCP (930Mbit/sec instead of 660Mbit/sec) over 1500-byte frames. "mtu 9000" yields only 770Mbit/sec.
1.40	04-Sep-2003	thorpej	GMII works on the i82543 now.
1.39	29-Jul-2003	thorpej	Add support for the Intel PRO/1000 MT Quad Port server adapter; this is just 2 i82546EB chips behind an IBM 133 PCI-X bridge.
1.38	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.37	29-Apr-2003	thorpej	branches: 1.37.2; Use aprint*().
1.36	15-Apr-2003	tron	Pullup short mbuf in wm_tx_cksum() instead of dropping it which fixes PR kern/21190 by myself. Jason R. Thorpe code reviewed the changes.
1.35	15-Apr-2003	thorpej	wm_tx_cksum(): Set the initial offset correctly for packets using 802.1Q encap.
1.34	05-Apr-2003	kent	Support for 82540EP and 82540EP (LOM). Not tested.
1.33	04-Apr-2003	kent	Support for i82450EP.
1.32	04-Feb-2003	thorpej	Use bus_dmamap_load_mbuf() in wm_add_rxbuf().
1.31	21-Jan-2003	itojun	m_tag transition mistake (within #if 0)
1.30	17-Jan-2003	itojun	switch from kame-based m_aux mbuf auxiliary data, to openbsd m_tag implementation. it will simplify porting across *bsd (such as kame/altq), and make us more synchronized. from Joel Wilsson
1.29	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.28	21-Dec-2002	kristerw	wm_copy_small in not used anywhere. Remove.
1.27	23-Oct-2002	christos	change the macro params to longer names
1.26	21-Oct-2002	fair	Change the "dontcare bits" argument to ifmedia_init() to IFM_IMASK so that it is possible select PHY instances other than the first one (instance zero), if there is more than one PHY attached.
1.25	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.24	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.23	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.22	23-Aug-2002	itojun	#if 0 out call to rnd_add_uint32() in interrupt logic until thorpej measures performance differences.
1.21	23-Aug-2002	itojun	attach random number source.
1.20	17-Aug-2002	thorpej	* After testing for DD on a descriptor, and finding it clear, sync that descriptor with PREREAD to make sure that it is evicted from the data cache. From Allen Briggs. * With the above bug fixed, stop using BUS_DMA_COHERENT, resulting in a fairly decent performance improvement on systems where BUS_DMA_COHERENT causes descriptors to be accessed uncached (most painful in wm_start()).
1.19	15-Aug-2002	thorpej	Update a comment; TCP/IP checksum offloading is fixed!
1.18	15-Aug-2002	briggs	On transmit, zero the upper 32 bits of the address in the tx descriptor. The descriptor may have been used as as context descriptor in the past, in which case this field will be non-zero. h/w checksum offload works now.
1.17	08-Aug-2002	thorpej	Add support for the new i82540, i82545, and i82546 Gig-E chips. The i82546 chip has 2 Gig-E interfaces in a single chip. The main difference is these chips have a special handshaking protocol for accessing the EEPROM (due to the shared nature on the i82546).
1.16	07-Aug-2002	simonb	Fix a typo in a comment.
1.15	06-Aug-2002	simonb	Fix a typo in a comment.
1.14	14-Jul-2002	thorpej	Document exactly what the problem with TCP/UDP checksum offloading is.
1.13	14-Jul-2002	thorpej	And thus spake the manual... When initializing a TCP/IP context descriptor, even if a checksum offload feature is not going to be used, the IPCSS and TUCSS fields must be initialized.
1.12	09-Jul-2002	thorpej	* Shuffle the TODO list a bit. * Fix outbound IPv4 header checksums (missing add of an offset). * When enabling Rx TCP/UDP checksum offload, make sure that IPv4 header checksum Rx offload is enabled on the chip, as well.
1.11	09-Jul-2002	thorpej	Don't use internal names to identify chips. Add types for the i82540, i82545, and i82546 (but don't match them yet).
1.10	09-Jul-2002	thorpej	Make a few performance tweaks: * Bump the number of Rx descriptors from 128 to 256. * Don't use a sliding Tx interrupt window. Instead, just do reap-behind when we have <= 1/8 of our available descriptors in wm_start(). * Don't use Tx Queue Empty interrupts, and always set the Tx Interrupt Delay bit in the Tx descriptor. * In wm_intr(), always call wm_rxintr() and wm_txintr(), regardless of their respective ISR bits being set. We're here, might as well do some work. * Adjust the Tx and Rx interrupt delay timer values. New values from Intel's driver for FreeBSD via Allen Briggs. With these changes, NetBSD can sustain > 900Mb/s userland to userland without using TCP checksum offload using Intel PRO/1000 XT cards.
1.9	09-May-2002	thorpej	branches: 1.9.2; 1.9.4; 1.9.6; * Increase the number of Tx job queue entries from 32 to 64. * Garbage-collect some unused stuff. * Make the Tx window slide along the Tx job queue space, not the Tx descriptor space. We are more likely to run low on DMA maps than we are hardware descriptors.
1.8	09-May-2002	thorpej	* Instrument the number of times we force a Tx interrupt. * When forcing an interrupt, make sure that interrupt-delay-enable is cleared (necessary with last change to make sure it's set for all descriptors). * Crank up the Transmit Interrupt Delay Value to 1024 * 1.024 msec. We really want these to be deferred.
1.7	08-May-2002	thorpej	Don't set WTX_TCPIP_CMD_TCP or WTX_TCPIP_CMD_IP in the TCP/IP context descriptor. Those are apparently only valid if using the Cordova's "segmentation offload" feature.
1.6	08-May-2002	thorpej	* Make sure the interrupt-delay bit gets set on ALL descriptors, except for the one we want to force a Tx interrupt. * Tweak the Tx descriptor accounting a little.
1.5	08-May-2002	thorpej	Implement a cache for the outbound IP/TCP/UDP checksum offload context.
1.4	08-May-2002	thorpej	* Account for the descriptor consumed by the Tx checksum offload context setup. * Implement Matt Thomas's sliding Tx interrupt window algorithm, forcing an interrupt when the Tx desc list is 2/3 consumed. * Use the Report Packet Sent interrupt, rather than Report Status, since we use the Tx descriptor to count Tx errors.
1.3	02-May-2002	thorpej	Fix a comment.
1.2	02-May-2002	thorpej	* Deal with errata on the i82542 and i82543 chips: The size of the transmit and receive descriptor rings is limited to 256 descriptors. So, set the if_snd queue length to 256 to let the upper layers queue lots of packets, and let the driver handle up to 32 of them at a time. (We should probably make this change to most Ethernet drivers, since it actually saves some resources.) * Increase the number of Tx DMA segments from 8 to 16. * Clean up the way we count "how many times did I get a packet with N DMA segments". * Add a missing htole32() in wm_tx_cksum(). * Don't set both RS and RPS in the last Tx descriptor of a packet; just use RS. * Add some more information to the watchdog message.
1.1	28-Mar-2002	thorpej	branches: 1.1.2; Device driver for the Intel i82542, i82542, and i82544 Gigabit Ethernet interfaces.
1.1.2.11	17-Jan-2003	thorpej	Sync with HEAD.
1.1.2.10	29-Dec-2002	thorpej	Sync with HEAD.
1.1.2.9	11-Nov-2002	nathanw	Catch up to -current
1.1.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.7	27-Aug-2002	nathanw	Catch up to -current.
1.1.2.6	19-Aug-2002	thorpej	Sync with trunk.
1.1.2.5	13-Aug-2002	nathanw	Catch up to -current.
1.1.2.4	01-Aug-2002	nathanw	Catch up to -current.
1.1.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.2	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.1.2.1	28-Mar-2002	nathanw	file if_wm.c was added on branch nathanw_sa on 2002-04-01 07:46:27 +0000
1.9.6.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.9.6.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.9.6.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.6.1	09-May-2002	jdolecek	file if_wm.c was added on branch kqueue on 2002-06-23 17:47:46 +0000
1.9.4.10	01-May-2005	tron	Pull up revision 1.102 (requested by scw in ticket #5761): In wm_stop(): - Clear the interrupt mask to ensure the device cannot assert its interrupt line. - Clear sc->sc_icr to ensure wm_intr() makes no attempt to service any currently pending or shared interrupt. This should address the problem reported in both PR/29903 and PR/22493. (wm(4) is susceptible to crashing in wm_intr() when the interface is downed)
1.9.4.9	16-Jun-2003	grant	Pull up revision 1.36 (requested by tron in ticket #1256): Pullup short mbuf in wm_tx_cksum() instead of dropping it which fixes PR kern/21190 by myself. Jason R. Thorpe code reviewed the changes.
1.9.4.8	08-Nov-2002	tron	Pull up revision 1.17 (requested by thorpej in ticket #650): Add support for the new i82540, i82545, and i82546 Gig-E chips. The i82546 chip has 2 Gig-E interfaces in a single chip. The main difference is these chips have a special handshaking protocol for accessing the EEPROM (due to the shared nature on the i82546).
1.9.4.7	07-Nov-2002	tron	Pull up revision 1.11 (requested by thorpej in ticket #453): Don't use internal names to identify chips. Add types for the i82540, i82545, and i82546 (but don't match them yet).
1.9.4.6	17-Aug-2002	lukem	Pull up revision 1.19 (requested by thorpej in ticket #684): Update a comment; TCP/IP checksum offloading is fixed!
1.9.4.5	17-Aug-2002	lukem	Pull up revision 1.18 (requested by briggs in ticket #683): On transmit, zero the upper 32 bits of the address in the tx descriptor. The descriptor may have been used as as context descriptor in the past, in which case this field will be non-zero. h/w checksum offload works now.
1.9.4.4	15-Jul-2002	thorpej	pullup-1-6 ticket #501 (thorpej). Original log message: Document exactly what the problem with TCP/UDP checksum offloading is.
1.9.4.3	15-Jul-2002	thorpej	pullup-1-6 ticket #500 (thorpej). Original log message: And thus spake the manual... When initializing a TCP/IP context descriptor, even if a checksum offload feature is not going to be used, the IPCSS and TUCSS fields must be initialized.
1.9.4.2	10-Jul-2002	lukem	Pull up revision 1.12 (requested by thorpej in ticket #455): * Shuffle the TODO list a bit. * Fix outbound IPv4 header checksums (missing add of an offset). * When enabling Rx TCP/UDP checksum offload, make sure that IPv4 header checksum Rx offload is enabled on the chip, as well.
1.9.4.1	10-Jul-2002	lukem	Pull up revision 1.10 (requested by thorpej in ticket #451): Make a few performance tweaks: * Bump the number of Rx descriptors from 128 to 256. * Don't use a sliding Tx interrupt window. Instead, just do reap-behind when we have <= 1/8 of our available descriptors in wm_start(). * Don't use Tx Queue Empty interrupts, and always set the Tx Interrupt Delay bit in the Tx descriptor. * In wm_intr(), always call wm_rxintr() and wm_txintr(), regardless of their respective ISR bits being set. We're here, might as well do some work. * Adjust the Tx and Rx interrupt delay timer values. New values from Intel's driver for FreeBSD via Allen Briggs. With these changes, NetBSD can sustain > 900Mb/s userland to userland without using TCP checksum offload using Intel PRO/1000 XT cards.
1.9.2.2	29-Aug-2002	gehenna	catch up with -current.
1.9.2.1	15-Jul-2002	gehenna	catch up with -current.
1.37.2.14	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.37.2.13	01-Apr-2005	skrll	Sync with HEAD.
1.37.2.12	08-Mar-2005	skrll	Sync with HEAD.
1.37.2.11	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.37.2.10	04-Feb-2005	skrll	Sync with HEAD.
1.37.2.9	18-Dec-2004	skrll	Sync with HEAD.
1.37.2.8	29-Nov-2004	skrll	Sync with HEAD.
1.37.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.37.2.6	19-Oct-2004	skrll	Sync with HEAD
1.37.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.37.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.37.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.37.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.37.2.1	03-Aug-2004	skrll	Sync with HEAD
1.68.2.5	08-Jan-2005	jdc	Fix pull up merge error. Pointed out by Martin J. Laubach.
1.68.2.4	07-Jan-2005	jdc	Pull up revision 1.78 (requested by thorpej in ticket #897) Work around another problem (that causes horrible performance) on the 82547: - Change the Tx:Rx on-chip RAM allocation ratio based on MTU: => 82547: > 8192 18K:22K, else 10K:30K => others: > 8192 24K:40K, else 16K:48K Values derived from Linux driver. - On the 82547, keep track of how much of the Tx FIFO has been used. When in half-duplex mode, don't let packets wrap around the FIFO ring buffer, because that causes the chip to croak. Detect this, stall the Tx queue, use a timer to wait for the packets to drain from the Tx FIFO, reset the internal FIFO pointers, and restart the Tx queue. Basic algorithm (and some magic numbers) derived from FreeBSD and Linux drivers.
1.68.2.3	16-Jul-2004	he	branches: 1.68.2.3.2; Pull up revision 1.74 (requested by tron in ticket #670): Reduce transmit queue length to 16 for the Intel i82547 to avoid FIFO overruns. This fixes the horrible performance problems reported in PR#23782.
1.68.2.2	15-Jul-2004	he	Pull up revision 1.73 (requested by tron in ticket #665): The Intel i82547EI gigabit Ethernet controller isn't really a PCI device. So instead of trying to find and print the PCI bus properties mark it as a CSA device.
1.68.2.1	12-Jul-2004	he	Pull up revision 1.72 (requested by tron in ticket #640): Make this build if WM_DEBUG is defined.
1.68.2.3.2.9	21-Jul-2005	snj	Pull up revision 1.101 (requested by tron in ticket #5507): Add support for Intel PRO/1000 GT. Change provided in private e-mail by unex at linija dot org.
1.68.2.3.2.8	01-May-2005	tron	Pull up revision 1.102 (requested by scw in ticket #1481): In wm_stop(): - Clear the interrupt mask to ensure the device cannot assert its interrupt line. - Clear sc->sc_icr to ensure wm_intr() makes no attempt to service any currently pending or shared interrupt. This should address the problem reported in both PR/29903 and PR/22493. (wm(4) is susceptible to crashing in wm_intr() when the interface is downed)
1.68.2.3.2.7	23-Feb-2005	he	Pull up revision 1.85 (requested by thorpej in ticket #1001): Also toggle the second port's Ethernet address lsb on the WM_T_82546_3, so the two ports get different MAC addresses.
1.68.2.3.2.6	24-Jan-2005	he	Pull up revision 1.83 (requested by thorpej in ticket #939): When adding or deleting multicast addresses, only change the address filter if the interface is marked RUNNING. Fixes PR#27678.
1.68.2.3.2.5	13-Jan-2005	he	Revert previous commit because the result does not even compile. Further input on req ticket #899 requested.
1.68.2.3.2.4	12-Jan-2005	he	Pull up revision 1.80 (via patch, requested by thorpej in ticket #899): Restructure the loop in wm_start() that passes over the DMA map segments slightly to make it easier to drop in errata work-arounds that require adding extra descriptors onto the end of the packet.
1.68.2.3.2.3	12-Jan-2005	he	Pull up revision 1.79 (via patch, requested by thorpej in ticket #898): Largest Tx descriptor buffer size is actually 4K.
1.68.2.3.2.2	08-Jan-2005	jdc	Fix pull up merge error. Pointed out by Martin J. Laubach.
1.68.2.3.2.1	07-Jan-2005	jdc	Pull up revision 1.78 (requested by thorpej in ticket #897) Work around another problem (that causes horrible performance) on the 82547: - Change the Tx:Rx on-chip RAM allocation ratio based on MTU: => 82547: > 8192 18K:22K, else 10K:30K => others: > 8192 24K:40K, else 16K:48K Values derived from Linux driver. - On the 82547, keep track of how much of the Tx FIFO has been used. When in half-duplex mode, don't let packets wrap around the FIFO ring buffer, because that causes the chip to croak. Detect this, stall the Tx queue, use a timer to wait for the packets to drain from the Tx FIFO, reset the internal FIFO pointers, and restart the Tx queue. Basic algorithm (and some magic numbers) derived from FreeBSD and Linux drivers.
1.89.6.3	26-Mar-2005	yamt	sync with head.
1.89.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.89.6.1	12-Feb-2005	yamt	sync with head.
1.89.4.1	29-Apr-2005	kent	sync with -current
1.100.2.13	26-Mar-2009	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #2002): sys/dev/pci/if_wm.c: revision 1.166 Fix the multicast hash bug on ICH9's wm. Now we can catch ff02::9 on ICH9's wm. Fixes PR#37976.
1.100.2.12	13-Dec-2008	bouyer	Pull up following revision(s) (requested by sketch in ticket #1983): sys/dev/pci/if_wm.c: revision 1.163 Fix '\|\|' vs '&&' mix-up from when ich9 support was added. Fixes nasty EEPROM-trashing bug on ich8 and ich9 chipsets as described in http://mail-index.netbsd.org/current-users/2008/12/02/msg006435.html caused by setting the wrong type of EEPROM.
1.100.2.11	10-Nov-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #1976): sys/dev/pci/if_wm.c: revision 1.162 sys/dev/pci/pcidevs: revision 1.961, 1.962 via patch Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.
1.100.2.10	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1863): sys/dev/pci/if_wm.c 1.137-1.138,1.141,1.144-1.146 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 via patch share/man/man4/wm.4 1.18-1.19 Fix typo. Fix EEPROM reload sequence. Add ICH9 support. Workaround for 82541 Errata 29 and 82547 Errata 28: These devices have to reset the PHY before reset the MAC. Reported and tested by salo.
1.100.2.9	14-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1770): share/man/man4/wm.4: revision 1.17 sys/dev/pci/if_wmreg.h: revision 1.22 sys/dev/pci/if_wm.c: revision 1.139 via patch sys/dev/pci/if_wm.c: revision 1.140 via patch sys/dev/pci/pcidevs: patch Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2 Restore $NetBSD$ tag. Note ICH8 LAN support.
1.100.2.8	06-Jun-2007	liamjfoy	Pull up following revision(s) (requested by yamt in ticket #1750): sys/dev/pci/if_wm.c: revision 1.129 wm_tx_offload: in the case of "txtsopain", m_copyback ip header to the correct offset.
1.100.2.7	31-Mar-2007	bouyer	pullup the following revisions (requested by msaitoh in ticket 1681): sys/dev/pci/if_wm.c 1.104-1.105, 1.116-1.121, 1.127,1.133-1.134 via patch sys/dev/pci/if_wmreg.h 1.17-1.20 sys/dev/pci/pcidevs patch sys/dev/mii/igphy.c 1.11 sys/dev/mii/makphy.c 1.20, 1.23 sys/dev/mii/ikphy.c patch sys/dev/mii/ikphyreg.h patch sys/dev/mii/miidevs 1.68 sys/dev/mii/files.mii 1.39 sys/arch/i386/conf/GENERIC 1.788-1.789 via patch sys/arch/i386/conf/GENERIC.MPACPI patch sys/arch/i386/conf/GENERIC_LAPTOP 1.209 via patch sys/arch/i386/conf/INSTALL 1.291 via patch sys/arch/i386/conf/INSTALL_LAPTOP 1.104 via patch sys/arch/i386/conf/XEN2_DOM0 1.13 via patch share/man/man4/wm.4 1.14-1.16 Add support for many cards (include PCI-express based chips). Many bug fixes about auto negotiations (PR#30078, PR#30490, PR#30906, PR#33429 and PR#35386). Fix media link issues with fiber-based card (PR#35797).
1.100.2.6	11-Nov-2006	bouyer	Pull up following revision(s) (requested by hauke in ticket #1578): sys/dev/pci/if_wm.c: revision 1.108 don't reset interface on rx overrun because it makes the situation worse, if the overrun is merely a symptom of heavy load. increament if_ierrors instead. PR/29831.
1.100.2.5	07-Jul-2006	tron	Pull up following revision(s) (requested by lukem in ticket #1400): sys/dev/pci/if_wm.c: revision 1.124 Fix len=-4 issues. When receiving a packet, if a receive buffer has length 0, treat it the same way we treat a failed mapping. Fix contributed by Wasabi Systems.
1.100.2.4	08-Feb-2006	tron	Pull up following revision(s) (requested by thorpej in ticket #1161): sys/dev/pci/if_wm.c: revision 1.111 sys/dev/pci/if_sip.c: revision 1.104 In the receive interrupt handler, compute the packet length before resetting m to point to the first buffer in the packet. The bug would cause the length of jumbo frames to be incorrect.
1.100.2.3	28-Aug-2005	tron	Pull up following revision(s) (requested by yamt in ticket #700): sys/dev/pci/if_wm.c: revision 1.106 wm_rxintr: ignore h/w assisted checksum results if WRX_ST_IXSM is set.
1.100.2.2	28-Apr-2005	tron	Pull up revision 1.102 (requested by scw in ticket #207): In wm_stop(): - Clear the interrupt mask to ensure the device cannot assert its interrupt line. - Clear sc->sc_icr to ensure wm_intr() makes no attempt to service any currently pending or shared interrupt. This should address the problem reported in both PR/29903 and PR/22493. (wm(4) is susceptible to crashing in wm_intr() when the interface is downed)
1.100.2.1	07-Apr-2005	jmc	Pullup rev 1.101 (requested by tron in ticket #16) Add support for Intel PRO/1000 GT.
1.103.2.12	24-Mar-2008	yamt	sync with head.
1.103.2.11	17-Mar-2008	yamt	sync with head.
1.103.2.10	27-Feb-2008	yamt	sync with head.
1.103.2.9	11-Feb-2008	yamt	sync with head.
1.103.2.8	04-Feb-2008	yamt	sync with head.
1.103.2.7	21-Jan-2008	yamt	sync with head
1.103.2.6	15-Nov-2007	yamt	sync with head.
1.103.2.5	27-Oct-2007	yamt	sync with head.
1.103.2.4	03-Sep-2007	yamt	sync with head.
1.103.2.3	26-Feb-2007	yamt	sync with head.
1.103.2.2	30-Dec-2006	yamt	sync with head.
1.103.2.1	21-Jun-2006	yamt	sync with head.
1.107.2.1	02-Nov-2005	yamt	sync with head.
1.110.6.3	07-Jun-2006	kardel	Sync with head.
1.110.6.2	01-Jun-2006	kardel	Sync with head.
1.110.6.1	22-Apr-2006	simonb	Sync with head.
1.110.4.1	09-Sep-2006	rpaulo	sync with head
1.110.2.2	01-Mar-2006	yamt	sync with head.
1.110.2.1	18-Feb-2006	yamt	sync with head.
1.114.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.114.4.1	11-May-2006	elad	sync with head
1.114.2.4	15-Sep-2006	yamt	revert unrelated changes erroneously slipped in.
1.114.2.3	11-Aug-2006	yamt	sync with head
1.114.2.2	26-Jun-2006	yamt	sync with head.
1.114.2.1	24-May-2006	yamt	sync with head.
1.115.2.1	19-Jun-2006	chap	Sync with head.
1.120.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.124.6.2	10-Dec-2006	yamt	sync with head.
1.124.6.1	22-Oct-2006	yamt	sync with head
1.124.4.2	12-Jan-2007	ad	Sync with head.
1.124.4.1	18-Nov-2006	ad	Sync with head.
1.131.4.5	07-May-2007	yamt	sync with head.
1.131.4.4	15-Apr-2007	yamt	sync with head.
1.131.4.3	24-Mar-2007	yamt	sync with head.
1.131.4.2	12-Mar-2007	rmind	Sync with HEAD.
1.131.4.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.131.2.8	24-Mar-2009	bouyer	Pull up following revision(s) (requested by msaitoh in ticket #1291): sys/dev/pci/if_wm.c: revision 1.166 Fix the multicast hash bug on ICH9's wm. Now we can catch ff02::9 on ICH9's wm. Fixes PR#37976
1.131.2.7	13-Dec-2008	bouyer	Pull up following revision(s) (requested by sketch in ticket #1246): sys/dev/pci/if_wm.c: revision 1.163 Fix '\|\|' vs '&&' mix-up from when ich9 support was added. Fixes nasty EEPROM-trashing bug on ich8 and ich9 chipsets as described in http://mail-index.netbsd.org/current-users/2008/12/02/msg006435.html caused by setting the wrong type of EEPROM.
1.131.2.6	31-Oct-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #1215): sys/dev/pci/if_wm.c: revision 1.162 sys/dev/pci/pcidevs: revision 1.961,1.962 Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.' sys/dev/pci/if_wm.c sys/dev/pci/pcidevs
1.131.2.5	11-Sep-2007	xtraeme	branches: 1.131.2.5.4; Pull up following revision(s) (requested by msaitoh in ticket #865): sys/dev/pci/if_wm.c: revision 1.145 sys/dev/pci/if_wm.c: revision 1.146 share/man/man4/wm.4: revision 1.19 EECD_EE_AUTORD is only for 8257{123},80003,ICH{89}. Will fix "auto read from eeprom failed to complete" problem. Add more work for the EEPROM reaload sequence. Fix typo. Bump date for previous (ICH9 support).
1.131.2.4	29-Aug-2007	liamjfoy	Pull up following revision(s) (requested by masanobu in ticket #850): sys/dev/pci/if_wm.c 1.137 sys/dev/pci/if_wm.c 1.138 sys/dev/pci/if_wm.c 1.141 sys/dev/pci/if_wm.c 1.144 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphy.c 1.12 sys/dev/mii/miidevs 1.72 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/wm.4 1.18 Fix for 82541 and 82547's reset bug. Modify wm_reset() to make some device stable. Add ICH9 support to wm. Add I82566 support to igphy. Pulled done via patch.
1.131.2.3	03-May-2007	snj	branches: 1.131.2.3.2; Pull up following revision(s) (requested by bouyer in ticket #598): share/man/man4/wm.4: revision 1.17 sys/dev/pci/if_wmreg.h: revision 1.22 doc/CHANGES: revision 1.836 sys/dev/pci/if_wm.c: revision 1.139 sys/dev/pci/if_wm.c: revision 1.140 Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2 Restore $NetBSD$ tag. Note ICH8 LAN support. Note ICH8 LAN support.
1.131.2.2	26-Mar-2007	jdc	Pull up revision 1.134 (requested by msaitoh in ticket #515). fix media link issues with fiber-based card fixes PR#35797
1.131.2.1	26-Mar-2007	jdc	Pull up revision 1.133 (requested by msaitoh in ticket #514). fix some negotiation problems on wm(4). will fix PR#30078, PR#30490, PR#30906, PR#33429, PR#35386.
1.131.2.5.4.1	13-Dec-2008	bouyer	Pull up following revision(s) (requested by sketch in ticket #1246): sys/dev/pci/if_wm.c: revision 1.163 Fix '\|\|' vs '&&' mix-up from when ich9 support was added. Fixes nasty EEPROM-trashing bug on ich8 and ich9 chipsets as described in http://mail-index.netbsd.org/current-users/2008/12/02/msg006435.html caused by setting the wrong type of EEPROM.
1.131.2.3.2.2	23-Sep-2007	wrstuden	Sync with somewhat-recent netbsd-4.
1.131.2.3.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.135.4.1	11-Jul-2007	mjf	Sync with head.
1.135.2.7	23-Oct-2007	ad	Sync with head.
1.135.2.6	09-Oct-2007	ad	Sync with head.
1.135.2.5	01-Jul-2007	ad	Adapt to callout API change.
1.135.2.4	09-Jun-2007	ad	Sync with head.
1.135.2.3	27-May-2007	ad	Sync with head.
1.135.2.2	10-Apr-2007	ad	Sync with head.
1.135.2.1	13-Mar-2007	ad	Sync with head.
1.136.2.1	29-Mar-2007	reinoud	Pullup to -current
1.142.6.11	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.142.6.10	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.142.6.9	31-Oct-2007	joerg	Sync with HEAD.
1.142.6.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.142.6.7	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.142.6.6	04-Sep-2007	joerg	Register power management function only once.
1.142.6.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.142.6.4	21-Aug-2007	joerg	Explicitly stop interface before going to D3. Don't bother to init interface again if it is down. Assume wm_init is successful as all failure causes are critical errors anyway. Remove shutdown hooks.
1.142.6.3	10-Aug-2007	he	Add a missing "break;" in the power handler function before any more developers copy this bug over to other drivers. Pointed out by jmcneill@
1.142.6.2	04-Aug-2007	jmcneill	Make this compile again -- sys/php.h does not exist, and since sys/device.h pulls in sys/pnp.h anyway, don't bother.
1.142.6.1	04-Aug-2007	he	Convert the old powerhook to a pnp power management handler, based on the azalia power management handler and what was here before.
1.142.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.143.2.3	23-Mar-2008	matt	sync with HEAD
1.143.2.2	09-Jan-2008	matt	sync with HEAD
1.143.2.1	06-Nov-2007	matt	sync with HEAD
1.146.4.2	13-Nov-2007	bouyer	Sync with HEAD
1.146.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.148.6.1	11-Dec-2007	yamt	sync with head.
1.148.4.1	26-Dec-2007	ad	Sync with head.
1.148.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.149.2.3	20-Jan-2008	bouyer	Sync with HEAD
1.149.2.2	11-Jan-2008	bouyer	Sync with HEAD
1.149.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.154.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.154.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.154.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.154.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.154.2.1	24-Mar-2008	keiichi	sync with head.
1.158.10.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.158.10.1	19-Oct-2008	haad	Sync with HEAD.
1.158.6.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.158.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.158.4.7	11-Aug-2010	yamt	sync with head.
1.158.4.6	11-Mar-2010	yamt	sync with head
1.158.4.5	16-Sep-2009	yamt	sync with head
1.158.4.4	19-Aug-2009	yamt	sync with head.
1.158.4.3	18-Jul-2009	yamt	sync with head.
1.158.4.2	20-Jun-2009	yamt	sync with head
1.158.4.1	04-May-2009	yamt	sync with head.
1.162.4.21	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1962): sys/dev/pci/if_wm.c: revisions 1.259-1.266, 1.269, 1.271, 1.273-1.274, 1.277-1.278, 1.282, 1.284-1.285, 1.287, 1.290, 1.294, 1.297-1.298, 1.300-1.301, 1.304-1.307, 1.310, 1.312-1.314, 1.316 via patch sys/dev/pci/if_wmreg.c: revisions 1.54-1.57, 1.59-1.60, 1.62, 1.64-1.66 via patch sys/dev/pci/if_wmvar.c: revisions 1.17, 1.20-1.21 via patch sys/dev/pci/pcidevs: revisions 1.1172, 1.1195, 1.1201 via patch - Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. - Fix BMC related bugs. - Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth. - Add support for I354 DH89xxCC and some new I218 devices. - Fix definition of CTRL_GIO_M_DIS bit. - Insert completion barrier between register write and delay(). - Bump max TX DMA size to avoid pathological condition with TSO. From dyoung. - Fix semaphore related bugs. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - It's not required to print "failed to detect NVM bank" with aprint_error_dev(). Use DPRINTF(). Same as {Free,Open}BSD. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Remove extra debug message. - Remove unused variable. - Remove a duplicated error message. - Cleanup comments. - Fix debug message.
1.162.4.20	26-Apr-2015	martin	Apply patch, requested by msaitoh in ticket #1961: sys/dev/pci/if_wm.c: (patch) Fix a bug (in the backport for #1850) that passed the wrong pointer as the first argument of bpf_mtap() on 82575 and newer chip.
1.162.4.19	07-Sep-2013	bouyer	sys/dev/pci/if_wm.c 1.246-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 sys/dev/pci/pcidevs 1.1152-1.1153 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #1867]
1.162.4.18	19-Jun-2013	bouyer	Pullup the following revisions via patch, requested by msaitoh in ticket #1850: sys/dev/pci/if_wm.c 1.201, 1.203-1.204, 1.207-1.212, 1.215, 1.217-1.218, 1.220-1.223, 1.228, 1.232-245 sys/dev/pci/if_wmreg.h 1.40-1.45, 1.47-1.48 sys/dev/pci/if_wmvar.h 1.11-1.13 sys/dev/pci/pcidevs 1.1074, 1.1077, 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.6 sys/dev/mii/ihphy.c 1.1-1.2 sys/dev/mii/ihphyreg.h 1.1 sys/dev/mii/inbmphyreg.h 1.3 sys/dev/mii/files.mii 1.47 via patch sys/dev/mii/miidevs 1.97 and 1.100 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/arch/i386/conf/ALL 1.280 sys/arch/i386/conf/GENERIC 1.1001 sys/arch/i386/conf/INSTALL_FLOPPY 1.11 sys/arch/i386/conf/XEN2_DOM0 patch sys/arch/amd64/conf/GENERIC 1.293 sys/arch/amd64/conf/XEN3_DOM0 1.61 share/man/man4/wm.4 1.21-1.24 Apply almost all fixes and improvements from netbsd-6 except for the rev. 1.196's iqdrops' change. - Add the detach code. - Add code for WOL, ASF, IPMI and Intel AMT. WOL is disabled by default - Add Yet another workaround for ICH8. - 82576 is dual port, so check the FUNCID and increment the MAC address for the 2nd port. - Fix the names of 82577L[MC] LAN controllers (for mobile). - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add 82580 support. - 82571 quirk. Only 82571 shares port 0 of EEMNGCTL_CFGDONE. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. - Fix some register names. No functional change. - Omit U+00AE "REGISTERED SIGN" in a product name due to its non-ASCII nature. - Stop wm(4) from needlessly resetting when you add or delete a vlan(4). - Fix MAC address check on 8257[156] and 80003 case. Some cards have non 0xffff pointer but those don't use alternative MAC address in reality. So we check whether the broadcast bit is set or not like Intel's e1000 driver. Fixes PR kern/44072 reported by Jean-Yves Moulin. - Add PCH2(and 82579) support. Fixes PR#46487 - Add yet another 82567V support. - Add ICH10+HANKSVILL support. - Add support Intel I350 Ethernet. - Make vlan and all ip/ip6 checksum offload work for the I350. - Fix compile error with WM_DEBUG. - Fix a bug that PHY isn't set to low-power mode on PCH and PCH2. - Add WM_DEBUG_NVM. If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. - Skip 64bit BAR correctly. - Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350. - Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542. - Style fix. Fix typo in comment. Fix comments. Add comments.
1.162.4.17	17-Oct-2012	bouyer	Apply patch, requested by msaitoh in ticket 1800: sys/dev/pci/if_wm.c: revision 1.230-1.231 via patch Add workaround for QEMU and the variants that fail on EEPROM access. This problem was discovered a few years ago, but some variants and cloud services still have the bug. This problem is not NetBSD's bug but qemus' bug. For NetBSD users, existence of buggy virtual machines is sad thing, so we add a workaroud.
1.162.4.16	25-Jan-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #1699): sys/dev/pci/if_wm.c: revision 1.225 There is problem with 82576 chips (and probably 82575 too) with hardware vlan tagging: some packets are sent untagged on the wire. Follow OpenBSD and disable hardware vlan tagging for these chips (I couldn't find a hint in other open-source drivers at what could be wrong ...)
1.162.4.15	07-Mar-2011	riz	Pull up following revision(s) (requested by bouyer in ticket #1544): sys/dev/pci/if_wm.c: revision 1.219 wm_gmii_reset(): restore generic reset delays to what they were before rev 1.186. This makes the following hardware find its PHY again, and I can't see how these larger delays could be a problem for other hardware: wm0 at pci6 dev 7 function 0: Intel i82541GI 1000BASE-T Ethernet, rev. 5 wm0: interrupting at ioapic2 pin 0, event channel 5 wm0: 32-bit 66MHz PCI bus wm0: 65536 word (16 address bits) SPI EEPROM wm0: Ethernet address 00:13:72:54:ee:13 igphy0 at wm0 phy 1: Intel IGP01E1000 Gigabit PHY, rev. 0 igphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto
1.162.4.14	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1358): sys/dev/pci/if_wm.c 1.196-1.199,1.202,1.205 sys/dev/pci/if_wmvar.h 1.9 sys/dev/pci/if_wmreg.h 1.36-1.39 sys/dev/pci/pcireg.h 1.61-1.64 sys/dev/pci/pcidevs 1.1023 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen mii/igphy.c 1.21 mii/igphyvar.h 1.1 mii/inbmphyreg.h 1.2 - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems. - Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working. - (pcireg.h) Add PCIe config register definitions. - Note that the changes to count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops in rev. 1.196 of if_wm.c is not pulled up.
1.162.4.13	25-Feb-2010	sborrill	branches: 1.162.4.13.2; Fix regression caused by pullup #1277 requested by msaitoh.
1.162.4.12	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.162.4.11	23-Dec-2009	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: 1.176-1.179, 1.181-1.183 sys/dev/pci/if_wmreg.h: 1.28 sys/dev/pci/if_wmvar.h: 1.1-1.4 sys/dev/mii/igphy.c: 1.18-1.20 via patch Many bugfixes: - Some fixes for i80003 and ICH{8,9,10} from e1000 driver and document: - Add setting for KABGTXD register for ICH{8,9,10}. - ICH9 and ICH10 has no FCAL, FCAH and FCT like ICH8. - Add special setting for FCTTV and TCTL_EXT register for i80003 - The special setting for TIPG is only for i80003. - Some of kumeran settings are only for i80003's bugs. - Add some ICH10 fixes. - Fix the bug that another lock mechanism is used to access Kumeran registers on i80003 and ICHs. - Fix yet another i80003 ONLY workaround. The code to modifing TIPG register is only for i80003. - Set the Re-Transmit on Late Collision(RTLC) flag for all devices. - Fix a typo in a printf message. - If the difference bettween last flag and new flag is only IFF_PROMISC or IFF_ALLMULTI, set multicast filter only to prevent link down. Tested by Mark Davies and me. Fixes PR#29126 for wm. - Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line. - Fix igphy's 82566 support. - Patch for the DSP code is only for 8254[17] and we have to apply the different patches between rev. 1 and rev. 2. - The workaround for analog fuse is only for 82547 rev. 1. - The workaround for smartspeed is only for 8254[17] - Sync with Intel's original em driver: - Check PCI-X mode as e1000 driver. - Add dspcode for igp3 and use it when the EEPROM isn't available. - Add some delays. - Stop the PHY transmitter before patching the DSP code and restart it after writing. - Save and restore register 0x2f5b.
1.162.4.10	05-Jun-2009	snj	Pull up following revision(s) (requested by 792): sys/dev/pci/if_wm.c: revision 1.175 via patch sys/net/if_ethersubr.c: revision 1.172 via patch sys/net/agr/ieee8023ad_lacp.c: revision 1.9 via patch sys/net/agr/if_agr.c: revision 1.23 via patch sys/net/agr/if_agrether.c: revision 1.7 via patch sys/net/agr/if_agrvar_impl.h: revision 1.8 via patch Add vlan support and hardware offload capabilities to agr. These changes allow vlans to be layered above agr, with the attach and detach propogated to the member ports in the aggregation. Note the agr interface must be up before the vlan is attached. Adds SIOCSIFADDR support to the wm driver for setting the AF_LINK address, necessary for agr to be able to set the mac addresses of each port to the agr address (i.e. so it can receive all intended traffic at the hardware level). Adds support for disabling the LACP protocol by setting LINK1 on the agr interface (e.g. ifconfig agr0 link1). In consultation with tls@.
1.162.4.9	21-May-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #711): sys/dev/pci/pcidevs: revisions 1.975, 1.981, 1.982 via patch sys/dev/pci/if_wm.c: revisions 1.164, 1.167, 1.173, 1.174 via patch sys/dev/pci/if_wmreg.h: revisions 1.25, 1.27 via patch Add Intel 82567LM_3 ethernet -- Add i82567LM-3 -- add i82567LF-3 LAN Controller -- add an entry for 82567LF-3. fix the register access for ICH10DO. -- Fix about TBI mode. This fix doesn't influence MII mode. -- - Fix panic in mediachange. - Fix SWDPIN(1)'s polarity on some chips. - Fix flow control stuff (includes PR#32009). - Stop RXCFG storm. It ocours easily. - And more fix about autonego. -- add 82801J_D_BM_LF (ICH10) -- Reload sc_ctrl in wm_reset(). Add an ICH10 entry. Remove some obsolete comments.
1.162.4.8	11-May-2009	bouyer	Apply patch (requested by msaitoh in ticket #751): sys/dev/pci/if_wm.c: patch Remove extra printf()s.
1.162.4.7	03-May-2009	snj	Pull up following revision(s) (requested by msaitoh in ticket #630): sys/dev/pci/if_wm.c: revision 1.170 Add workaround for 82543GC. We need to force speed and duplex on the MAC equal to what the PHY speed and duplex configuration is. Fixes PR#36430.
1.162.4.6	03-May-2009	snj	Pull up following revision(s) (requested by tls in ticket #627): sys/dev/pci/if_wm.c: revision 1.172 Missed a small but important change to enable hardware VLAN support.
1.162.4.5	03-May-2009	snj	Pull up following revision(s) (requested by tls in ticket #627): sys/dev/pci/if_wm.c: revision 1.171 Enable the vlan code that thorpej@NetBSD.org added, and fix a bug where the vlan tag was read from the rx descriptor after it was reinitialized (and hence was always 0). Hardware vlan support looks good tested between two vmware clients. With support from tls@NetBSD.org.
1.162.4.4	03-May-2009	snj	Pull up following revision(s) (requested by spz in ticket #621): sys/dev/pci/if_wm.c: revisions 1.168, 1.169 via patch sys/dev/pci/if_wmreg.h: revision 1.26 via patch On i82563, the em driver says that the ready bit in the MDIC register may be incorrectly set. Insert delay(200) like the em driver. -- Examine the management mode and mark DRV_LOAD (for ICH{8,9,10},80003, 8257{1,2,3,4}). Add some error's printf(). Make the bank detect routine into the function.
1.162.4.3	20-Mar-2009	msaitoh	branches: 1.162.4.3.2; Pull up following revision(s) (requested by sborrill in ticket #587): sys/dev/pci/pcidevs: revision 1.978 sys/dev/pci/if_wm.c: revision 1.165 Add support for 82574L (as found on Desktop CT adaptor) based on 82573 support.
1.162.4.2	18-Mar-2009	snj	Pull up following revision(s) (requested by msaitoh in ticket #583): sys/dev/pci/if_wm.c: revision 1.166 Fix the multicast hash bug on ICH9's wm. Now we can catch ff02::9 on ICH9's wm.
1.162.4.1	07-Dec-2008	bouyer	Pull up following revision(s) (requested by sketch in ticket #160): sys/dev/pci/if_wm.c: revision 1.163 Fix '\|\|' vs '&&' mix-up from when ich9 support was added. Fixes nasty EEPROM-trashing bug on ich8 and ich9 chipsets as described in http://mail-index.netbsd.org/current-users/2008/12/02/msg006435.html caused by setting the wrong type of EEPROM.
1.162.4.13.2.1	07-Mar-2011	riz	Pull up following revision(s) (requested by bouyer in ticket #1544): sys/dev/pci/if_wm.c: revision 1.219 wm_gmii_reset(): restore generic reset delays to what they were before rev 1.186. This makes the following hardware find its PHY again, and I can't see how these larger delays could be a problem for other hardware: wm0 at pci6 dev 7 function 0: Intel i82541GI 1000BASE-T Ethernet, rev. 5 wm0: interrupting at ioapic2 pin 0, event channel 5 wm0: 32-bit 66MHz PCI bus wm0: 65536 word (16 address bits) SPI EEPROM wm0: Ethernet address 00:13:72:54:ee:13 igphy0 at wm0 phy 1: Intel IGP01E1000 Gigabit PHY, rev. 0 igphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, 1000baseT, 1000baseT-FDX, auto
1.162.4.3.2.1	19-Jul-2009	snj	branches: 1.162.4.3.2.1.2; Pull up following revision(s) (requested by msaitoh in ticket #862): sys/dev/pci/if_wm.c: revisions 1.168, 1.170, and 1.173 via patch sys/dev/pci/if_wmreg.h: revision 1.27 Fixes serious three bugs. 1) On i82563, the em driver says that the ready bit in the MDIC register may be incorrectly set. Insert delay(200) like the em driver. Fixes PR#41014 2) Add workaround for 82543GC. We need to force speed and duplex on the MAC equal to what the PHY speed and duplex configuration is. Fixes PR#36430. 3) Fix many problems and panic on TBI's cards (includes PR#32009).
1.162.4.3.2.1.2.2	15-Feb-2014	matt	sync pcireg.h with HEAD. update if_wm.c and ppb.c accordingly.
1.162.4.3.2.1.2.1	21-Apr-2010	matt	sync to netbsd-5
1.162.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.162.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.162.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.164.2.2	23-Jul-2009	jym	Sync with HEAD.
1.164.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.197.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.197.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.197.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.205.2.4	31-May-2011	rmind	sync with head
1.205.2.3	05-Mar-2011	rmind	sync with head
1.205.2.2	03-Jul-2010	rmind	sync with head
1.205.2.1	30-May-2010	rmind	sync with head
1.217.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.217.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.217.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.223.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.223.2.3	16-Jan-2013	yamt	sync with (a bit old) head
1.223.2.2	30-Oct-2012	yamt	sync with head
1.223.2.1	17-Apr-2012	yamt	sync with head
1.225.2.2	02-Jun-2012	mrg	sync to latest -current.
1.225.2.1	18-Feb-2012	mrg	merge to -current.
1.227.2.20	24-Sep-2016	bouyer	Apply patch, requested by martin in ticket #1407: sys/dev/pci/if_wm.c patch fix evbppc build, where the older gcc wrongly warns about uninitialized variable.
1.227.2.19	06-May-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1366): sys/dev/pci/if_wm.c: 1.281, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.387, 1.389 via patch sys/dev/pci/if_wmreg.h: 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: 1.22-1.23, 1.25-1.30 via patch sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 Sync wm(4) as of if_wm.c rev 1.389 except SERDES, MSI/MSI-X, multiqueue and NET_MPSAFE: - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvctl -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.227.2.18	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1296): sys/dev/pci/pcidevs: revision 1.1195 via patch sys/dev/pci/if_wm.c: revisions 1.290, 1.304 via patch sys/dev/pci/if_wmvar.h: revision 1.20 via patch - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - Support DH89xxCC device. - Add extra delay for 82580 and newer devices except DH89XXCC SGMII device. Same as FreeBSD.
1.227.2.17	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #1292): sys/dev/pci/if_wm.c 1.313-1.314 and 1.316 via patch Fix a bug that the first access to NVM is failed on 8254[17] which use SPI EEPROM. Observed on Dell PowerEdge [12]850. Thanks Tom Ivar Helbekkmo for debugging. Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789. Fix a bug that newer revision of I218-{LM,V} use wrong PHY access functions. The problem only occured on devices that the PCI device ID was 0x15a[0123]. [msaitoh, ticket #1292]
1.227.2.16	16-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c: revision 1.312 via patch sys/dev/pci/if_wmreg.h: revision 1.66 via patch - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment.
1.227.2.15	04-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #1244): sys/dev/pci/if_wm.c: revision 1.310 Fix a bug that a workaround for SWSM.SMBI bit doesn't work correctly. This problem was only occured with old boot ROM on 8257[12]. XXX pullup to netbsd-[67]
1.227.2.14	04-Dec-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: revisions 1.271, 1.273-1.274, 1.277-1.278, 1.280, 1.282, 1.284-1.285, 1.287, 1.293-1.294, 1.297-1.298, 1.300-1.301, 1.305-1.307 via patch sys/dev/pci/if_wmreg.h: revisions 1.57-1.62, 1.64-1.65 via patch sys/dev/pci/if_wmvar.h: revisions 1.19, 1.21 via patch - Acquire SW semaphore in wm_get_swsm_semaphore(). - Fix some bugs realted to semaphore. This change fixes a problem which was exposed in if_wm.c rev. 1.271. Tested by riastradh@. - Clear the SMBI bit in SWSM register before accessing NVM and PHY in wm_attach(). Same as FreeBSD. - Fix a bug that 82573 doesn't put the hardware semaphore. Same as FreeBSD r256200. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). Tested with 82543GC, 82544EI, 82545EM, 82546GB 82571EB and 82572EI fiber cards. - Don't use the RXCFG interrupt. It's not required and the interrupt is very heavy (a lot of interrupts). Same as {Free,Open}BSD. - Modify wm_tbi_mediachange() to be close to em_setup_fiber_serdes_link() of {Free,Open}BSD. At least, don't forget to set duplex setting. - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Don't check SWSM_SMBI bit if WM_F_LOCK_SWSM isn't set. Fix a problem when using vmware with e1000"e". With e1000e which is regarded as 82574L, wm_gmii_init() fails with "could not acquire SWSM SMBI" message without this change. This problem doesn't occur with real 82574L card. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Calculate NVM word size correctly. - Determine timeout value based on the NVM word size. - It's not required to print "failed to detect NVM bank" message. Only print while debugging. Same as {Free,Open}BSD. - Add some new I218 devices. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Move some NVM related macros from if_wm.c to if_wmreg.h. - Sort definitions in if_wmreg.h - move NVM related values to the bottom. - sort in register's address' order. - Simplify wm_read_mac_addr(). - Fix debug message. - Add missing prototypes. - Rename some functions for consistency and clarify. - Rename some macros for consistency. - Remove a duplicated error message. - Fix typo in comment. - Cleanup comments. - KNF.
1.227.2.13	09-Nov-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #1189): sys/dev/pci/if_wm.c: revision 1.270 sys/dev/mii/inphy.c: revision 1.53 sys/dev/mii/glxtphy.c: revision 1.24 sys/dev/mii/tlphy.c: revision 1.62 sys/dev/mii/iophy.c: revision 1.37 sys/dev/mii/brgphy.c: revision 1.70 sys/dev/mii/ihphy.c: revision 1.8 sys/dev/mii/bmtphy.c: revision 1.31 sys/dev/mii/urlphy.c: revision 1.30 sys/dev/mii/makphy.c: revision 1.40 sys/dev/mii/qsphy.c: revision 1.48 sys/dev/mii/igphy.c: revision 1.23 sys/dev/mii/nsphy.c: revision 1.58 sys/dev/mii/mvphy.c: revision 1.10 sys/dev/pci/if_txp.c: revision 1.41 sys/dev/mii/nsphy.c: revision 1.59 sys/dev/mii/rlphy.c: revision 1.28 sys/dev/mii/icsphy.c: revision 1.49 sys/dev/mii/rlphy.c: revision 1.29 sys/dev/mii/lxtphy.c: revision 1.49 sys/dev/mii/ciphyreg.h: revision 1.5 sys/dev/mii/nsphyter.c: revision 1.38 sys/dev/mii/sqphy.c: revision 1.50 sys/dev/mii/gentbi.c: revision 1.26 sys/dev/mii/gentbi.c: revision 1.27 sys/dev/mii/tqphy.c: revision 1.39 sys/dev/mii/ikphy.c: revision 1.10 sys/dev/mii/dmphy.c: revision 1.35 sys/dev/mii/amhphy.c: revision 1.20 sys/dev/mii/acphy.c: revision 1.24 sys/dev/mii/ciphy.c: revision 1.25 sys/dev/mii/brgphyreg.h: revision 1.8 sys/dev/mii/ukphy_subr.c: revision 1.12 sys/dev/ic/rtl80x9.c: revision 1.16 sys/arch/mips/adm5120/dev/if_admsw.c: revision 1.12 sys/dev/pci/if_kse.c: revision 1.28 sys/dev/mii/ukphy_subr.c: revision 1.13 sys/dev/mii/mii.h: revision 1.18 sys/dev/mii/gphyter.c: revision 1.29 No functional change: - Fix typo. - Remove trailing white spaces. - Capitalize comments. - Tabify. - KNF. IFM_FDX and IFM_HDX use different bit, so set IFM_HDX bit if it's not full duplex. For many drivers, it recognize half duplex if IFM_FDX isn't set, but not for others. Same as {Free\|Open}BSD.
1.227.2.12	09-Nov-2014	martin	Pullup the following revisions, requested by msaitoh in ticket #1188: sys/dev/pci/pcidevs 1.1172 sys/dev/pci/if_wm.c 1.263-1.266 via patch sys/dev/pci/if_wmreg.h 1.55-1.56 sys/dev/pci/if_wmvar.h 1.17 share/man/man4/wm.4 1.26-1.27 and 1.29 via patch - Add I354 support. - Insert completion barrier between register write and delay(). - Fix the definition of CTRL_GIO_M_DIS. This bit is not bit 3 but bit 2. - Cleanup
1.227.2.11	03-Jun-2014	msaitoh	Pull up following revision(s) (requested by tls in ticket #1070): sys/dev/pci/if_wm.c: revision 1.269 From dyoung@ -- bump max TX DMA size to avoid pathological condition with TSO.
1.227.2.10	29-Jul-2013	jdc	Pull up revisions: src/sys/dev/pci/if_wm.c revisions 1.259,1.260,1.261,1.262 src/sys/dev/pci/if_wmreg.h revision 1.54 (requested by msaitoh in ticket #918). Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. Sync the wm_enable_mng_pass_thru() function with FreeBSD. Don't check MANC_EN_MAC_ADDR_FILTER bit. Add 82574 and 82583 specific check. This modification may change the setting of WM_F_HAS_MANAGE flag on some machines. Sync the wm_release_manageablilty() fucntion with FreeBSD. Set MANC_ARP_EN. This change enables HW ARP function when entering suspend. When the chip is 82580(ER) or I350, set WM_F_ASF_FIRMWARE_PRES flag and check for the WM_F_ARC_SUBSYS_VALID flag. Same as FreeBSD. Move the location of wm_check_mng_mode() and wm_get_wakeup() in wm_attach(). Those functions access EEPROM, so they have to call after identifying EEPROM access type. This modification may change the behavior of BMC(IPMI). Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth.
1.227.2.9	14-Jul-2013	riz	Apply changes (requested by msaitoh in ticket #907): sys/dev/pci/if_wm.c 1.238, 1.244-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.50-1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 Various fixes to wm(4): Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Change style, add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #907]
1.227.2.8	18-Feb-2013	riz	branches: 1.227.2.8.2; Pull up following revision(s) (requested by msaitoh in ticket #824): sys/dev/pci/if_wm.c: revision 1.240 sys/dev/pci/if_wm.c: revision 1.241 sys/dev/pci/if_wm.c: revision 1.242 sys/dev/pci/if_wmreg.h: revision 1.49 - Add WM_DEBUG_NVM. - If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. Skip 64bit BAR correctly. I don't know if this bug causes a real problem. Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350.
1.227.2.7	18-Feb-2013	riz	sys/dev/pci/if_wm.c 1.235-1.236 via patch Use PRIx64 and explicitly cast to uint64_t to print a DMA address. This fixes a problem that if_wm.c can't compile with WM_DEBUG on non-64bit platforms. [msaitoh, ticket #823]
1.227.2.6	14-Feb-2013	jdc	Pull up revisions: src/sys/dev/pci/if_wm.c revision 1.243 src/sys/dev/pci/if_wmvar.h revision 1.14 (requested by msaitoh in ticket #820). Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542.
1.227.2.5	17-Dec-2012	riz	Pull up following revision(s) (requested by msaitoh in ticket #755): sys/dev/pci/if_wm.c: revision 1.239 - 82578DC is not PCH2 but PCH. 82579V is not PCH but PCH2. This bug was introduced in rev. 1.221. Reported by FUKAUMI Naoki. - Fix comment.
1.227.2.4	23-Nov-2012	riz	Pull up following revision(s) (requested by msaitoh in ticket #697): sys/dev/pci/if_wm.c: revision 1.237 Fix a bug that PHY isn't set to low-power mode on PCH and PCH2.
1.227.2.3	03-Sep-2012	riz	branches: 1.227.2.3.2; Pull up following revision(s) (requested by bouyer in ticket #534): sys/dev/pci/if_wmreg.h: revision 1.48 sys/dev/pci/if_wm.c: revision 1.232 sys/dev/pci/if_wm.c: revision 1.233 sys/dev/pci/if_wm.c: revision 1.234 Make vlan and all ip/ip6 checksum offload work for the I350. On newer devices, when using the legacy TX descriptors, vlan-related flags that were set on the last descriptor of a packet have to be set on the first one. For tso/checksum offloads, a new "advanced" descriptor format has to be used. Change wcd_txdescs to a union defining all types of descriptors (they are all 16-bytes wide). Define a new tx function wm_nq_start(), which handle newer devices. There is some code duplication with wm_start(), but adding support to the existing wm_start() would make it a if () {} else {} maze. This also allows to get rid of some workaround for older chips that are not needed here. Use wm_nq_start() instead of wm_start() for the I350 (this should probably be for all WM_F_NEWQUEUE devices, but I have no hardware but the I350 to test). Call ifp->if_start() instead of wm_start() where is matters. Tested on a I350, and a i80003 (which use the old format), both with and without vlans, with and without checksum offloads. Enable VLAN hardware tagging on all chips that have the new queue mechanism. Tested with 82575{EB,GB}, 82576, 82580, I350 and ICH9. Shut up gcc about some uninitialized variables.
1.227.2.2	09-Aug-2012	martin	branches: 1.227.2.2.2; Pull up following revision(s) (requested by msaitoh in ticket #471): sys/dev/pci/if_wm.c: revision 1.230 sys/dev/pci/if_wm.c: revision 1.231 Add workaround for QEMU and the variants that fail on EEPROM access. This problem was discovered a few years ago, but some variants and cloud services still have the bug. This problem is not NetBSD's bug but qemus's bug. For NetBSD users, existence of buggy virtual machines s sad thing, so we add a workaroud. Fix the check of the device type in last commit. Reported by Thomas Klausner.
1.227.2.1	28-Jun-2012	riz	Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/if_wm.c 1.228 sys/dev/pci/if_wmreg.h 1.47 sys/dev/pci/if_wmvar.h 1.13 sys/dev/pci/pcidevs 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/wm.4 1.22-1.25 Add support for Intel I350 Ethernet. Update document.
1.227.2.8.2.1	03-Jun-2014	msaitoh	Pull up following revision(s) (requested by tls in ticket #1070): sys/dev/pci/if_wm.c: revision 1.269 From dyoung@ -- bump max TX DMA size to avoid pathological condition with TSO.
1.227.2.3.2.1	03-Jun-2014	msaitoh	Pull up following revision(s) (requested by tls in ticket #1070): sys/dev/pci/if_wm.c: revision 1.269 From dyoung@ -- bump max TX DMA size to avoid pathological condition with TSO.
1.227.2.2.2.1	01-Nov-2012	matt	sync with netbsd-6-0-RELEASE.
1.234.2.5	03-Dec-2017	jdolecek	update from HEAD
1.234.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.234.2.3	23-Jun-2013	tls	resync from head
1.234.2.2	25-Feb-2013	tls	resync with head
1.234.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.261.2.2	18-May-2014	rmind	sync with head
1.261.2.1	28-Aug-2013	rmind	sync with head
1.268.2.2	10-Aug-2014	tls	Rebase.
1.268.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.289.2.18	28-Nov-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1656): sys/dev/pci/if_wm.c: revision 1.599 - Fix a bug that i82578 PHY can't detect correctly. - Print PHY OUI and model number if a PHY's can't be identified. XXX pullup-[78]
1.289.2.17	09-Nov-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1648): sys/dev/mii/inbmphyreg.h 1.11 sys/dev/pci/if_wm.c 1.586-1.588, 1.590-1.596 via patch sys/dev/pci/if_wmreg.h 1.108 - rename tu event counter to txunderrun. - Try m_defrag() to reduce the number of DMA segment if bus_dmamap_load_mbuf() returned EFBIG. When m_defrag() is called, txqNNdefrag event counter is incremented. If the 2nd try of bus_dmamap_load_mbuf() failed, txqNNtoomanyseg event counter is incremented. - Reduce the max number of DMA segments from 256 to 64 (it's the same value as other BSD's (EM_MAX_SCATTER) and more than before if_wm.c rev. 1.75's value (40)) because we do m_defrag() now. - 82574 and newer document says the status field has neither EC (Excessive Collision) bit nor LC (Late Collision) bit (reserved), so don't check the bit. - Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This workaround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. - Fix a PCH2 specific bug that wrong PHY register value can be read when boot. Same as FreeBSD and Linux. - After writing MDIC register, don't read quickly the same register but do delay(50). Same as other OSes. - Add missing wm_gate_hw_phy_config_ich8lan(false) in wm_phy_post_reset() on PCH2. wm_gate_hw_phy_config_ich8lan(true) is called in wm_reset(), so wm_gate_hw_phy_config_ich8lan(false) should be called after reset in wm_phy_post_reset(). - On PCH2, set the phy config counter to 50msec after (PHY) reset. - KNF & Use macro.
1.289.2.16	30-Oct-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1647): sys/dev/pci/if_wm.c: revision 1.589 Fix KASSERT to prevent panic on CNP (Intel 300 series + I219) with INTx interrupt (Xen dom0 or pre netbsd-8)
1.289.2.15	11-Aug-2018	martin	Pull up the following, requested by msaitoh in ticket #1628: share/man/man4/wm.4 1.40 via patch sys/dev/mii/ihphyreg.h 1.2 sys/dev/mii/inbmphyreg.h 1.10 sys/dev/pci/if_wm.c 1.504, 1.506, 1.510-1.535, 1.539-1.540, 1.546, 1.548, 1.551-1.552, 1.558, 1.565-1.573, 1.575, 1.579, 1.582, 1.584 via patch sys/dev/pci/if_wmreg.h 1.99-1.103, 1.106-1.107 via patch sys/dev/pci/if_wmvar.h 1.34-1.39 via patch sys/dev/pci/pcidevs 1.1327 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/pci/pcireg.h patch Sync wm(4) up to 2018/08/08 except MSI/MSI-X and NET_MPSAFE: - remove extra "+" - Fix a bug that non-GMII devices don't send a routing message when the link status is changed. - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes. - If a interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Print sc_flags with snprintb(). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange(). - Add missing "else" in wm_nvm_release(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - If TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function are required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Use new nvm.{acquire,release}() for semaphore. - Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register value from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally on I21[01]". PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Print NVM offset and word count when EERD polling failed. - On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux). - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Fix typo in comment. Reported by Shinichi Doyashiki in PR#52885. - Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki. - CID-1427779: Fix uninitialized variables. - Fix a bug that wm_pll_workaround_i210() is not called when a) Chip is I211 or b) Chip is I210 and it uses iNVM (not FLASH). - Do wm_reset_mdicnfg_82580() on 82580 only. - Fix FLASH access on PCH_SPT and newer. Their FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it must be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. - Make wm_nvm_valid_bank_detect_ich8lan() the same as other OSes. - If the extended configuration size in the EXTCNFSIZE register is 0, don't continue in wm_init_lcd_from_nvm(). - Add PCH_CNP support (I219 with Intel 300 series chipset). - Enable I219 support. - I354 uses an external PHY, so don't use wm_set_eee_i350(). - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. - Fix a bug that 82542 misunderstand fiber's signal detection. - Add debug printf()s. - Update comment. - Rename functions and variables. - Add diagnostic code. - Sort registers. - Lowercase hexadecimal values. - KNF.
1.289.2.14	19-Apr-2017	snj	Pull up following revision(s) (requested by kardel in ticket #1385): sys/dev/pci/if_wm.c: revision 1.497 PR kern/52039: use same safeguard as for the 82578
1.289.2.13	09-Mar-2017	snj	branches: 1.289.2.13.2; Pull up following revision(s) (requested by msaitoh in ticket #1372): sys/dev/pci/if_wm.c: revisions 1.462, 1.464-1.465 1.474 via patch sys/dev/pci/if_wmreg.h: revision 1.97 sys/dev/pci/if_wmvar.h: revision 1.33 - Set IPV6EXDIS bit in RFCTL register because of an Errata on 82575 and newer devices. - Linux and FreeBSD defines 0x10f5 as E1000_DEV_ID_ICH9_IGP_M_AMT. In reality, This is not IGP but BM. Add new case to identify PHY type device. Fixes PR#51924 reported byJarle Greipsland. - Use new wm_gmii_setup_phytype() function to setup sc_phytype and mii_{read\|write}reg. This change improves detection of PHY type. - Fix typo in comment.
1.289.2.12	18-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1327): sys/dev/pci/if_wm.c: revision 1.457 sys/dev/pci/if_wmreg.h: revision 1.94 - Change to use 2500Base-KX correctly on C2000(I354). It worked, but the output of ifconfig and if_baudrate was not good. Tested by nils@: - The STATUS_TBIMODE bit in the STATUS register is deleted since 82575, so check for 82575 and newer first and then check for old devices. - Check the 2P5_SKU and 2P5_SKU_OVER bit for KX. - Set IFM_2500_SX instead of IFM_1000_SX for 2.5G. - Check SERDES's speed directly from the PCS layer (PCS_LSTS register) for old devices. - Style fix.
1.289.2.11	12-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1302): sys/dev/mii/igphyreg.h: revisions 1.7-1.10 sys/dev/mii/ikphyreg.h: revisions 1.3 sys/dev/mii/inbmphyreg.h: revisions 1.4-1.9 sys/dev/mii/mii.h: revisions 1.19-1.20 sys/dev/pci/if_wm.c: revisions 1.390, 1.392-1.395, 1.397, 1.419-1.425, 1.427-1.428, 1.430-1.435, 1.437-1.453 via patch sys/dev/pci/if_wmreg.: revisions 1.89-1.93 via patch sys/dev/pci/if_wmvar.h: revisions 1.31-1.32 Update wm(4) up to if_wm.c rev. 1.453 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add I219 support. It's not stable so it's disabled by default. - wm_gate_hw_phy_config_ich8lan() is for younger than PCH2. - Drop the host wakeup bit after resetting PHY on PCH and newer devices. - Increase delay while toggling LANPHYPC - Move call of wm_reset() in wm_attach() after setting PHY and NVM related flags because those flags are used in wm_reset(). - Use mutex for NVM access on ICH8 and newer devices. Same as FreeBSD. - Rewrite PHY related lock stuff. Almost the same as FreeBSD. This change will fix a bug that PHY read/write fail on some cases. - Increase delay in wm_phy_resetisblocked(). Same as FreeBSD. - Use semaphore in wm_hv_phy_workaround_ich8lan() and wm_k1_gig_workaround_hv() - Use wm_gii_mdic_readreg/writereg() in wm_access_phy_wakeup_reg_bm() because these functions are called with taking lock. - 82567V_3 is BME1000_E_2(bm). Tested with Advantech AIMB-212 1st Ethernet port. - Use wm_gmii_82544_{read,write}reg() on non-82567 ICH8, 9 and 10. - Remove an 82578 workaround which was for PCH rev < 3. FreeBSD removed this workaround in r228386. - Add an 82578 workaround which is for PHY rev < 2. From FreeBSD and Linux. - Fix wm(4) input drop packet counter. WMREG_RNBC is incremented when there is no available buffers in host memory. However, ethernet controller can receive packets in such case if there is space in phy's FIFO. That is, ethernet controller drops packet only if there is no available buffers and there is no space in phy's FIFO. So, the number of dropped packets should be added WMREG_MPC only. - Use MII_ADDRMASK. - Define WMPHY_I217, WMPHY_VF and WMPHY_210. - Use BME1000_PHY_PAGE_SELECT in wm_gmii_bm_{read,write}reg(). This change has no effect because GG82563_PHY_PAGE_SELECT and BME1000_PHY_PAGE_SELECT have the same value. - Fix PHY access on 82567(ICH8 or ICH10), 82574 and 82583: - Use wm_gmii_bm_{read,write}reg() on 82574 and 82573. - Issue page select correctly on BM PHYs. - Fix workaround which did dummy read BM_WUC register. This code was changed to drop BM_WUC_HOST_WU_BIT of BM_PROT_GEN_CFG register in FreeBSD r228386. The code was added rev. 1.149, but the location was not the best. - wm_gmii_hv_{read/write}reg(): USE PHY address 1 for some special registers. - Add check code for an 82578 workaround. Not completed yet. - wm_release_hw_control(): Remove extra line. No any effect. - Add "10/100" into non-gigabit devices' name. - Call wm_enable_wakeup() in wm_detach() and wm_suspend(). Now wake on lan works on Thinkpad X61(ICH8). - Fix wm_access_phy_wakeup_reg_bm(). This change has no effect because this function is used for WUC register and our driver currenlty doesn't access to it. - Call wm_enable_phy_wakeup() on PCH2 and newer, too. Now these devices can do WOL. Tested with Thinkpad X220(PCH2). - Set CTRL_MEHE correctly (PCH_{LPT,SPT} only). - Add three workarounds for PCH_{LPT,SPT}. - Fix a bug that 8257[56], 82580, I35[04] and I21[01] didn't use wm_{get,release}_hw_control() correctly. - Sync wm_smbustopci() with Linux and FreeBSD. This change effects PCH and newer devices. - Move the location of wm_smbustopci() call. - Fix flag check in wm_get_wakeup() - 8254[17] and 8257[124] should not set WM_F_ARC_SUBSYS_VALID. - Add missing WM_T_82541_2 and WM_T_82547_2. - Fix WOL related setting of the WUC register for other than PCH* in wm_enable_wakeup(). Tested with 82567V(ICH8) and 82583V. - Use common MII_ADDRMASK. - igphy(4): No binary change: - s/IGPPHY/IGPHY/ - Fix the definition of PLHR_VALID_CHANNEL_* - Fix the definition of MSE_CHANNEL_* - Add MII_IGPHY_POWER_MGMT. - Add some KASSERT. - Add comment. Modify comment. - Add debug code.
1.289.2.10	09-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1300): sys/dev/pci/pcidevs: revision 1.1270 sys/dev/pci/if_wm.c: revision 1.426 i82567V-3 is not ICH9 but ICH8. -- 82567V-3 is not ICH9 but ICH8.
1.289.2.9	26-Feb-2016	snj	branches: 1.289.2.9.2; Pull up following revision(s) (requested by msaitoh in ticket #1102): sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 sys/dev/pci/if_wm.c: revisions 1.308, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.389 via patch sys/dev/pci/if_wmreg.h: revisions 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: revisions 1.22-1.23, 1.25-1.30 via patch Sync wm(4) as of if_wm.c rev 1.389 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add C2000 KX and 2.5G support. - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Fix a lot of bugs to make 82575 and newer SERDES based systems work. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvct -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.289.2.8	10-Jun-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #833): sys/dev/pci/if_wm.c: revisions 1.322, 1.323 sys/dev/pci/if_wmreg.h: revision 1.72 - Currently, WM_F_EEE bit is not set on all chips. It's intened to not to use all of EEE fuctions but wm_set_eee_i350() leaves IPCNFG_10BASE_TE bit and it causes link negotiation problem on some old switches. Disable 10BASE-Te function, too. - Call wm_set_eee_i350() on some chips, too. -- Fix a bug that flags related to semaphore were incorrectly checked in wm_kmrn_{readreg,writereg}. i80003, ICH* and PCH* had this problem.
1.289.2.7	19-May-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #784): sys/dev/pci/if_wm.c: revision 1.321 sys/dev/pci/if_wmreg.h: revision 1.71 sys/dev/pci/if_wmvar.h: 1.24 via patch Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines/NICs.
1.289.2.6	19-Apr-2015	riz	Pull up following revision(s) (requested by msaitoh in ticket #698): sys/dev/pci/if_wm.c: revision 1.316 Fix a bug that newer revision of I218-{LM,V} use wrong PHY access functions. The problem only occured on devices that the PCI device ID was 0x15a[0123].
1.289.2.5	16-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #694): sys/dev/pci/if_wm.c: revision 1.314 Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789.
1.289.2.4	18-Mar-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #614): sys/dev/pci/if_wm.c: revision 1.311-1.313 sys/dev/pci/if_wmreg.h: revision 1.66 Fix a bug that sc_mediatype is evaluated incorrectly. The real problem might be occured only on fiber and serdes cases. ##### - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment. ##### Fix a bug that the first access to NVM is failed on 8254[17] which use SPI EEPROM. Observed on Dell PowerEdge [12]850. Thanks Tom Ivar Helbekkmo for debugging.
1.289.2.3	04-Feb-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #490): sys/dev/pci/if_wm.c: revision 1.310 Fix a bug that a workaround for SWSM.SMBI bit doesn't work correctly. This problem was only occured with old boot ROM on 8257[12].
1.289.2.2	07-Nov-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #190): sys/dev/pci/pcidevs 1.1201-1.1202 via patch sys/dev/pci/if_wm.c 1.291-1.301, 1.304, 1.306-1.307 via patch sys/dev/pci/if_wmreg.h 1.61-1.65 via patch sys/dev/pci/if_wmvar.h 1.21 via patch share/man/man4/wm.4 1.31 via patch - Add Internal SERDES mode support newer than or equal to 82575. - Add new I218 devices. - Add DH89xxCC device. - Add some old devices. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Don't print "failed to detect NVM bank" message. - Delete 82580ER related code. - Fix typo in comment. - Fix debug message. - Cleanup
1.289.2.1	29-Aug-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #61): sys/dev/pci/if_wm.c: revision 1.290 sys/dev/pci/if_wmvar.h: revision 1.20 Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102.
1.289.2.13.2.1	06-Nov-2017	snj	Pull up following revision(s) (requested by kardel in ticket #1385): sys/dev/pci/if_wm.c: revision 1.497 PR kern/52039: use same safeguard as for the 82578
1.289.2.9.2.2	13-Mar-2017	skrll	Sync with netbsd-7-1-RELEASE
1.289.2.9.2.1	18-Jan-2017	skrll	Sync with netbsd-5
1.308.2.11	28-Aug-2017	skrll	Sync with HEAD
1.308.2.10	05-Feb-2017	skrll	Sync with HEAD
1.308.2.9	05-Dec-2016	skrll	Sync with HEAD
1.308.2.8	05-Oct-2016	skrll	Sync with HEAD
1.308.2.7	09-Jul-2016	skrll	Sync with HEAD
1.308.2.6	29-May-2016	skrll	Sync with HEAD
1.308.2.5	19-Mar-2016	skrll	Sync with HEAD
1.308.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.308.2.3	22-Sep-2015	skrll	Sync with HEAD
1.308.2.2	06-Jun-2015	skrll	Sync with HEAD
1.308.2.1	06-Apr-2015	skrll	Sync with HEAD
1.416.2.4	26-Apr-2017	pgoyette	Sync with HEAD
1.416.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.416.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.416.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.465.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.508.4.55	29-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1938: sys/dev/pci/if_wm.c 1.792,1.794-1.798 via patch sys/dev/pci/if_wmreg.h 1.131 - Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. - Drop frames if the RX descriptor ring has no room on multiqueue system. - Improve dmesg output. - Print RX packet buffer size. - Fix the upper 16bit of Image Unique ID(EtrackID). - Fix comment.
1.508.4.54	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #1932): sys/dev/pci/if_wm.c: revision 1.793 Add Intel I219-{LM,V}(20,21) support.
1.508.4.53	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1915: sys/dev/pci/pcidevs 1.1497 sys/dev/pci/if_wm.c 1.689,1.790-1.791 via patch sys/dev/pci/if_wmreg.h 1.120 via patch sys/dev/pci/if_wmvar.h 1.51 - Use 12K for packet buffer for jumbo frame on PCH2 and newer. - Add new workaround for Tiger Lake and newer to avoid packet loss. - Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.508.4.52	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1910): sys/dev/pci/if_wm.c: revision 1.785 sys/dev/pci/if_wm.c: revision 1.786 sys/dev/pci/if_wm.c: revision 1.787 sys/dev/pci/if_wm.c: revision 1.788 sys/dev/pci/if_wm.c: revision 1.789 wm(4): Use SCVPC and HRMPC for 82575 and newer. The SCVPC(SerDes/SGMII Code Violation Packet Count) register and the HRMPC (Header Redirection Missed Packet Count) register were added in if_wm.c rev. 1.776 but the location in the code were incorrect. Fix them. wm(4): Add some info to some event counters. - The difference between the RUC(Rx Undersize) and RFC(Rx Fragment) is whether the CRC is valid or not. - The difference between the ROC(Rx Oversize) and RJC(Rx Jabber) is whether the CRC is valid or not. - LENERRS(Length Errors) is for the length/type field <= 1500. wm(4): the IAC(Interrupt Assertion Count) register is for 82571 and newer. wm(4): Modify descriptions of flow control related event counters. wm(4): Replace /* nothing */ to __nothing for evcnt macros Prevent empty if bodies for !WM_EVENT_COUNTERS.
1.508.4.51	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1894): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783,1.784 via patch Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.508.4.50	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1847: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.508.4.49	21-Jun-2023	martin	Apply patch, requested by riastradh in ticket #1842: sys/dev/pci/if_wm.c (apply patch) Fix a bug introduced in ticket #1795.
1.508.4.48	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1795: sys/dev/pci/if_wm.c 1.764-1.767 via patch sys/dev/pci/if_wmreg.h 1.128 - Workaround for some hypervisor environments. The environments cannot stop e1000 interrupt immediately. - Rename nq_txdesc' member "nqrx_ctx" to "nqtx_ctx". No functional change. - Add comment. Modify comment. - KNF.
1.508.4.47	08-Sep-2022	martin	Undo accidental whitespace changes from applying the patch for ticket #1759 by applying a patch from msaitoh (NFC).
1.508.4.46	07-Sep-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1759: sys/dev/pci/if_wm.c 1.741-1.749,1.753-1.757, 1.762 via patch sys/dev/pci/if_wmreg.h 1.126-1.127 - Fix I219 workaround in wm_flush_desc_rings(). - Add more statistics counters. - To avoid releasing mutex temporally, use new wm_set_mdio_slow_mode_hv_locked(). - No functional changes: - Turn a locking botch (shouldn't drop lock on error) into a KASSERT in wm_deferred_start_locked(). - Remove unneeded header inclusion. - Use __BIT() a little. - Modify comment and debug messages. - Consistency use -1 instead of 1 for some error code. - KNF.
1.508.4.45	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1751: sys/dev/pci/if_wmreg.h 1.122-1.125 via patch sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720,1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.508.4.44	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1707: sys/dev/pci/if_wm.c 1.716-1.718 via patch sys/dev/pci/if_wmreg.h 1.121 sys/dev/mii/ihphy.c 1.20 - Fix a bug that device timeout still happens when the link is down on ICH/PCH. Fixes PR kern/56478. - Add some sysctl info for debugging. - ihphy(4): Don't power down the PHY when the interface goes down. - Use macro. Fix comment.
1.508.4.43	23-Oct-2021	martin	Pull up the following, requested by msaitoh in ticket #1701: share/man/man4/wm.4 1.42 sys/dev/pci/files.pci 1.434 sys/dev/pci/if_wm.c 1.686-1.687,1.692-1.693, 1.697,1.699-1.703, 1.706-1.715 via patch sys/dev/pci/if_wmvar.h 1.46 - Add missing drain for pcq in wm_stop_locked(). - Add support for I219V 15-19 and I219LM 16-19. - Fix Tx stall. - Use wm_flush_desc_rings() workaround more on I219. - Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. - Add some sysctl values for debugging TX/RX queues. - WM_EVENT_COUNTER is enabled by default on 64 bit architectures. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxq->rxq_ptr updating. - Stop legacy interrupts before calling softint. - Disable printf()s in wm_flush_desc_rings() because the code is verified. - Print I219's version number. - Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. - Fix return value of interrupt handler. - Only print an error about missing I/O BARs for chips that need it. - Do not return a void value from a void function. - Add new flag named WM_F_CRC_STRIP and use it. No functional change. - Whitespace fixes. Fix comments. No functional change.
1.508.4.42	16-Nov-2020	martin	Pull up following revision(s) (requested by rin in ticket #1623): sys/dev/pci/if_wm.c: revision 1.696 Fix little-endian dependence in wm_rxeof(), by which packets cannot be received by 82574 and successors on big-endian machines. Tested by aarch64eb with I210-T1 on ROCKPro64. Thanks msaitoh for discussion! XXX pullup to netbsd-9 and netbsd-8
1.508.4.41	04-Nov-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #1620): sys/dev/pci/if_wm.c: revision 1.694 sys/dev/pci/if_wm.c: revision 1.695 (via patch) sys/dev/pci/if_wmvar.h: revision 1.47 Add WMPHY_I350. Not used yet. Workaround for ihphy and atphy(ICH/PCH, 82580 and I350). These phys stop DMA while link is down which causes device timeout. Fix PR/kern 40981 Reviewed and tested by msaitoh@n.o, thanks. XXX pullup-[89]
1.508.4.40	16-Oct-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1617): sys/dev/pci/if_wm.c: revision 1.691 Fixes a problem that the attach function reported "wm_gmii_setup_phytype: Unknown PHY model. OUI=000000, model=0000" and "PHY type is still unknown." Don't call wm_gmii_setup_phytype() three times if the interface uses SGMII with internal MDIO. Tested with I354(Rangeley(SGMII(MDIO))) and I350(SERDES(SFP), SGMII(SFP)).
1.508.4.39	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1594: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678, 1.680-1.685 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 via patch sys/dev/pci/if_wmvar.c 1.45 via patch sys/dev/mii/igphy.c 1.35-1.36 via patch sys/dev/mii/igphyreg.h 1.12-1.13 sys/dev/mii/makphy.c 1.66 via patch sys/dev/mii/makphyreg.h 1.11 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler). Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Explicitly cast from uint16_t to uint32_t before shifting 16bit left when printing Image Unique ID to avoid undefined behavior. - Set if_baudrate for non-MII device. - Rename some macros and function. - KNF. Add comment.
1.508.4.38	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1492: sys/dev/pci/if_wm.c 1.650, 1.652-1.654 via patch sys/dev/pci/if_wmreg.h 1.116-1.117 - Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. - Simplify code by using "struct mii_data *mii" more. No functional change. - MSI-X doesn't use sc->sc_icr variable, so move the code into non-MSI-X part. No functional change intended. - Modify debug printfs a bit. - Rename macro. - Use __BIT() - Fix comment. Add comment. - KNF.
1.508.4.37	23-Jan-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #1488): sys/dev/pci/if_wm.c: revision 1.661 fix softint_disestablish() in wm_detach(). ok by msaitoh@n.o XXX pullup-8, pullup-9
1.508.4.36	06-Nov-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1428: sys/dev/pci/if_wm.c 1.647, 1.649 via patch - Use device_printf() instead of printf(). - Fix typo in comment.
1.508.4.35	04-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1317): sys/dev/pci/if_wm.c: revision 1.646 sys/dev/pci/if_wm.c: revision 1.612 Accept IFM_NONE. - Use unsigned to avoid undefined behavior in wm_i82543_mii_sendbits(). Found by kUBSan.
1.508.4.34	01-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1312): sys/dev/pci/if_wm.c: revision 1.644 sys/dev/pci/if_wm.c: revision 1.645 sys/dev/pci/if_wmreg.h: revision 1.115 Define RETA_ENTRY_MASK_Q(used in RSS redirection table) correctly. Found by UBSan. Use aprint_debug_dev() instead of printf() in wm_ulp_disable(). - rnd_add_*() call must be serialized. Unserialized accesses broke the sc->rnd_source structure and caused off by one in kern_rndq.c's rnd_add_data_ts(). This problem was found by kUBSan. Make krndsource_t's data per queue-pair and call rnd_add_uint32() in wm_txrxintr_msix(). One of the reason why rnd_add_uint32() calls were moved from wm_{tx,rx}eof() is that rnd(9) manual page says "To allow the system to gather the timing information accurately, this call should be placed within the actual hardware interrupt service routine." - Pass a better value to rnd_add_uint32() in wm_intr_legacy(). XXX pullup-8.
1.508.4.33	17-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1298: sys/dev/pci/if_wm.c 1.633, 1.637-1.641 sys/dev/pci/if_wmreg.h 1.114 - Use unsigned to avoid undefined behavior. - Print ICH/PCH's NVM version. - Fix typo in comment. u_int_t -> uint_t. Whitespace fix.
1.508.4.32	14-May-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1267): sys/arch/arm/imx/if_enet.c: revision 1.19 sys/dev/pci/if_wm.c: revision 1.634 SIOCS is 'S'et function and the ioctl argument is ifreq. SIOCG is 'G'et function and the ioctl armument is ifmediareq. Before this change, SIOCG modify sc->sc_flowflags unexpectedly. Don't hook SIOCGIFMEDIA becauise this driver uses MII(4) and ether_ioctl has the hook(if_ethersubr.c rev. 1.160). real bug because the modification is overriden in ifmedia_ioctl(). This driver might require some additional fixes for SIOCSIFMTU and other ioctl()s. XXX pullup-[78].
1.508.4.31	07-Mar-2019	martin	Pull up the following, requested by msaitoh in ticket #1206: sys/dev/pci/if_wmreg.h 1.113 sys/dev/pci/if_wmvar.h 1.43-1.44 sys/dev/pci/if_wm.c 1.626-1.627, 1.629-1.636 via patch - Add support for I210 SGMII Flash-less device. - Add I219 variations for Cannon Lake. - Add missing rnd_detach_source(). - Use do { ... } while (0) for DPRINTF(x, y). - Swap enumeration of 82578 and 82577 PHY in if_wmvar.h. No functional change. - Add KASSERT. - Update TODO list. Add comment. Fix typo in comment.
1.508.4.30	11-Feb-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1189): sys/dev/pci/if_wm.c: revision 1.622 sys/dev/pci/if_wm.c: revision 1.624 sys/dev/pci/if_wm.c: revision 1.625 We now have IFM_2500_KX and IFM_1000_KX, so use them. - I21[01]'s RAL size is not 15 but 16(== 82575's value). Same as other OSes. - Fix a bug that WOL didn't work on some chips since if_wm.c rev. 1.610. Set WUC_APME bit older than PCH. Will fixes PR kern/53945 reported by kardel@. Tested with my own 82574 card.
1.508.4.29	31-Jan-2019	martin	Pull up the following, requested by msaitoh in ticket #1179: sys/dev/pci/if_wm.c 1.603-1.605,1.607-1.611, 1.613,1.615,1.618-1.620 via patch sys/dev/pci/if_wmreg.h 1.110-1.111 sys/dev/pci/if_wmvar.h 1.40-1.42 sys/dev/mii/inbmphyreg.h 1.13-1.15 - Add some code for suspend/resume: - Rename wm_smbustopci() to wm_init_phy_workarounds_pchlan(). It will also called when resume. - Call wm_phy_resetisblocked() after PHY reset in wm_init_phy_workarounds_pchlan() to wait for the PHY to quiesce to an accessible state. - Add new wm_resume_workarounds_pchlan() function and use it in wm_resume(). This workaround is only for PCH2 and newer. - Don't call wm_disable_aspm() neither in wm_attach() nor in wm_resume() but in wm_reset(). - Do some initialization in wm_resume() when IFF_UP is NOT set. - Don't continue when it failed to acquire semaphore in wm_ulp_disable(). - Print CLSEM workaround bit correctly. - Fix availability detection of WoL on some chips. - Print the WUS (WakeUp Status) register bits when resume. - Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220. - Remove an extra register read in wm_kmrn_lock_loss_workaround_ich8lan(). - Don't leave the MDICNFG register modified when the Power Management capability offset can't get. - Reduce indent level of wm_linkintr_gmii(). No functional change. - 80003's SERDES is not the same as 82575's but the same as legacy devices. Use the old methods on 80003. - Use __nothing for null DPRINTF(). - Rename functions. Add comment.
1.508.4.28	18-Dec-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #1138): sys/dev/pci/if_wm.c: revision 1.606 Fix txqueue assignment. Pointed out by yamaguchi@n.o, thanks. E.g. When ncpu is six and nqueue is four, the sequence error occurs. XXX pullup-8
1.508.4.27	04-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1117): sys/dev/pci/if_wmreg.h: revision 1.109 sys/dev/pci/if_wm.c: revision 1.597 sys/dev/pci/if_wm.c: revision 1.598 sys/dev/mii/inbmphyreg.h: revision 1.12 sys/dev/pci/if_wm.c: revision 1.600 sys/dev/pci/if_wm.c: revision 1.601 sys/dev/pci/if_wm.c: revision 1.602 - Add new wm_gmii_{hv,i82544}_{read,write}reg_locked() and use them in wm_gmii_{hv,i82544}_{read,write}reg(). _locked() functions are not mii(4) API functions, so it's not required to keep the mii API. Change the PHY register type from int to uint16_t. It also change the usage of return value. It returns zero on success and non-zero on error. - Check the return value of _locked() function and treat it. - Use *writereg_locked() function to reduce race condition in wm_init_lcd_from_nvm(). - Add comment. - Control TX/RX descriptor snooping control bits on ICH8 and newer. Only on ICH8, No-snoop bits are opposite polarity. On my Thinkpad X61, the default value of this bits are all zero, so this commit changes the snoop function enable on the machine. I tested with some other PCH machines and those bits are all zero (enable snoop by default), so this commit won't affect to some machines. - Disable relax ordering on 82546GB(Device ID 0x1099 and 0x10b5) or >= ICH8. Same as other OSes. - Add wm_oem_bits_config_ich8lan() to control LPLU and GbE setting base on the NVM's info. - Modify wm_enable_wakeup() to reduce difference against FreeBSD and Linux. This modification affects to ICH8 and newer devices. I217 Rapid Start Technology support have not written yet (it's TODO). - Add wm_k1_workaround_lv() from FreeBSD. It's PCH2 specific: Workaround to set the K1 beacon duration for 82579 parts in 10Mbps. Disable K1 for 1000 and 100 speeds. - Make wm_link_stall_workaround_hv() and move an 82578 specific code into it. Don't apply the workaround if BMCR_LOOP bit is set. Same as FreeBSD. - Add comment. Modify comment. Fix comment. No functional change.
1.508.4.26	26-Nov-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #1108): sys/dev/pci/if_wm.c: revision 1.599 - Fix a bug that i82578 PHY can't detect correctly. - Print PHY OUI and model number if a PHY's can't be identified.
1.508.4.25	08-Nov-2018	martin	Pull up the following, requested by msaitoh in tickt #1085: sys/dev/mii/inbmphyreg.h 1.11 sys/dev/pci/if_wmreg.h 1.108 sys/dev/pci/if_wm.c 1.590-1.596 via patch - 82574 and newer document says the status field has neither EC (Excessive Collision) bit nor LC (Late Collision) bit (reserved), so don't check the bit. - Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This workaround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. - Fix a PCH2 specific bug that wrong PHY register value can be read when boot. Same as FreeBSD and Linux. - After writing MDIC register, don't read quickly the same register but do delay(50). Same as other OSes. - Add missing wm_gate_hw_phy_config_ich8lan(false) in wm_phy_post_reset() on PCH2. wm_gate_hw_phy_config_ich8lan(true) is called in wm_reset(), so wm_gate_hw_phy_config_ich8lan(false) should be called after reset in wm_phy_post_reset(). - On PCH2, set the phy config counter to 50msec after (PHY) reset. - KNF & Use macro.
1.508.4.24	30-Oct-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1075): sys/dev/pci/if_wm.c: revision 1.589 Fix KASSERT to prevent panic on CNP (Intel 300 series + I219) with INTx interrupt (Xen dom0 or pre netbsd-8)
1.508.4.23	23-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1027): sys/dev/pci/if_wm.c: revision 1.586 sys/dev/pci/if_wm.c: revision 1.587 sys/dev/pci/if_wm.c: revision 1.588 - Split txdrop evcnt into pcqdrop and descdrop. - Simplify evcnt name by removing duplicated "tx" or "rx". - s/tu/underrun/ - Try m_defrag() to reduce the number of DMA segment if bus_dmamap_load_mbuf() returned EFBIG. When m_defrag() is called, txqNNdefrag event counter is incremented. If the 2nd try of bus_dmamap_load_mbuf() failed, txqNNtoomanyseg event counter is incremented. - Reduce the max number of DMA segments from 256 to 64 (it's the same value as other BSD's (EM_MAX_SCATTER) and more than before if_wm.c rev. 1.75's value (40)) because we do m_defrag() now. Fix comment.
1.508.4.22	11-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #969): sys/dev/pci/if_wm.c: revision 1.584 sys/dev/pci/if_wmvar.h: revision 1.39 Add MDIC_WA and CLSEMWA to snprintb() string for WM_F_* flags. Fix two bugs for non-SERDES fiber. Tested on 82544EI(0x1009), 82545GM(0x1027), 82546GB(0x107a), 82571GB(0x105f), 82572EI(0x107e). Tested on SERDES 82580(0x150e (Winyao SERDES SFP)) and 82580(0x150e) as regression test: - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. When the signal is detected & STATUS_LU == 0 & RXCW == 0, it's required to set CTRL_SLU and CTRL_FD bit. wm_check_for_link() is the function to do it. If a link status change interrupt occurred, call wm_check_for_link() before checking current link status. - Fix a bug that 82542 misunderstand fiber's signal detection. CTRL_SWDPIN(1) is used for fiber's signal and it's value is reversed only on 82543 and 82544. Simplify the code and move it into a new function wm_tbi_havesignal() and use it.
1.508.4.21	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #928): sys/dev/pci/if_wm.c: 1.573-1.576, 1.579 and 1.582 via patch wm(4): - Count timeout correctly. This change reduce timeout value for 80003 as expected. Reported by mouse@. - Print "device timeout (lost interrupt)" when all descriptors in a queue are free. - Rename txq_watchdog to txq_sending to make the meaning clear. - KNF. No functional change.
1.508.4.20	07-Jun-2018	martin	Pull up following revision(s) (requested by ozaki-r in ticket #843): sys/dev/pci/ixgbe/ixv.c: revision 1.101 sys/net/if_bridge.c: revision 1.156 sys/net/if_pppoe.c: revision 1.138 sys/dev/pci/if_wm.c: revision 1.580 sys/dev/pci/ixgbe/ixgbe.c: revision 1.156 sys/net/if_gif.c: revision 1.142 Ensure to call if_register after interface initializations finish
1.508.4.19	17-May-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #827): sys/dev/pci/if_wm.c: revision 1.577 rxipsum and rxtusum are not interrupt counter, so use EVCNT_TYPE_MISC instead of EVCNT_TYPE_INTR.
1.508.4.18	09-May-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #816): sys/dev/pci/if_wm.c: revision 1.578 Fix a bug that TX might stall because WM_TXQ_NO_SPACE is not cleared in if_init() (though I've never seen this problem). Clear txq->txq_flags in wm_init_tx_queue(). OK'd by knakahara.
1.508.4.17	16-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #764): sys/dev/pci/if_wm.c: revision 1.567 sys/dev/pci/if_wm.c: revision 1.568 sys/dev/pci/if_wm.c: revision 1.569 sys/dev/pci/if_wmvar.h: revision 1.38 sys/dev/pci/if_wm.c: revision 1.570 sys/dev/pci/if_wm.c: revision 1.571 sys/dev/pci/if_wm.c: revision 1.572 share/man/man4/wm.4: revision 1.40 sys/dev/pci/if_wmreg.h: revision 1.106 sys/dev/pci/if_wmreg.h: revision 1.107 SW PHY Config Enable bit for ICH8 B0 stepping is not bit 1 but bit 0. No binary change: - Sort registers. - Lowercase hexadecimal value. On PCH_SPT (and newer), FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it should be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. Our PCH_SPT part of wm_nvm_valid_bank_detect_ich8lan() was based on FreeBSD r287467. After that, they reverted it and committed the different code in r287762. r287762's bank detect code didn't work for us because our wm dirver had a problem in flash access. The problem was fixed in if_wm.c rev. 1.567, so we can use the new way now. If the extended configration size in the EXTCNFSIZE register is 0, don't continue. Add PCH_CNP support (I219 with Intel 300 series chipset). It's required more test, so it's disabled by default. Enable I219. I354 uses an external PHY, so don't use wm_set_eee_i350().
1.508.4.16	06-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #604): sys/dev/pci/if_wm.c: revision 1.565-1.566 - Fix a bug that wm_pll_workaround_i210() is not called on some cases. As I wrote in if_wm.c rev 1.329's commit message, There are three cases: a) Chip is I211 b) Chip is I210 and it uses INVM (not FLASH) c) Chip is I210 (it uses FLASH) and NVM image version < 3.25 Before this commit, the workaround function was not called on a) and b). - Rename function. Do wm_reset_mdicnfg_82580() on 82580 only.
1.508.4.15	26-Feb-2018	snj	Pull up following revision(s) (requested by knakahara in ticket #567): distrib/sets/lists/comp/mi: 1.2182-1.2183 sys/dev/pci/if_wm.c: 1.564 sys/dev/pci/ixgbe/ixgbe.c: 1.122 sys/dev/pci/ixgbe/ixgbe_rss.h: 1.3 sys/dev/pci/ixgbe/ixv.c: 1.78 sys/net/Makefile: 1.35-1.36 sys/net/files.net: 1.15 sys/net/rss_config.c: 1.1 sys/net/rss_config.h: 1.1 Introduce very simple Receive Side Scaling (RSS) utility. ok by msaitoh@n.o. -- Apply RSS utility to wm(4). ok by msaitoh@n.o. -- Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o. -- Fix build failure, sorry. -- Currently, it is not necessary to install rss_config.h. Pointed out by msaitoh@n.o.
1.508.4.14	26-Feb-2018	snj	Pull up following revision(s) (requested by knakahara in ticket #564): sys/dev/pci/if_wm.c: 1.563 Fix a bug that RX and TX may stall on heavy load on wm(4) like ixgbe.c:r1.121. wm_rxeof() and wm_txeof() have loop limit and the function returns true if a packet still exists.
1.508.4.13	05-Feb-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #529): sys/dev/pci/if_wm.c: revision 1.560 sys/dev/pci/if_wm.c: revision 1.561 sys/dev/pci/if_wm.c: revision 1.562 share/man/man4/wm.4: revision 1.37 share/man/man4/wm.4: revision 1.38 sys/dev/pci/if_wm.c: revision 1.551 sys/dev/pci/if_wm.c: revision 1.553 sys/dev/pci/if_wm.c: revision 1.554 sys/dev/pci/if_wm.c: revision 1.555 sys/dev/pci/if_wm.c: revision 1.556 sys/dev/pci/if_wm.c: revision 1.557 sys/dev/pci/if_wm.c: revision 1.558 sys/dev/pci/if_wm.c: revision 1.559 PR/52885 - Shinichi Doyashiki -- typo in comment Fix legacy Tx descriptors printing when WM_DEBUG is enabled. improve comments Fix wm_watchdog_txq() lock region. Not only wm_txeof() but also wm_watchdog_txq() itself requires txq_lock as it reads Tx descriptor management variables such as "txq_free". There is almost no influence on performance. Fix duplicated "rxintr" evcnt counting. Pointed out by ozaki-r@n.o, thanks. wm_txeof() can limit the loop count the same as wm_rxeof() now. add WM_TX_PROCESS_LIMIT_DEFAULT and WM_TX_INTR_PROCESS_LIMIT_DEFAULT man. More markup. CID-1427779: Fix uninitialized variables Fix 82574 MSI-X mode cannot receive packets after 82574 receives high rate traffic. In short, 82574 MSI-X mode does not cause RXQ MSI-X vector when 82574's phys FIFO overflows. I don't know why but 82574 causes not RXQ MSI-X vector but OTHER MSI-X vector at the situation. see: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v4.15-rc9&id=4aea7a5c5e940c1723add439f4088844cd26196d advised by msaitoh@n.o, thanks. Fix if_wm.c:r1.557 merge miss, sorry. Fix unmatched return type. The return value of wm_txeof() is not useded yet. Make wm(4) watchdog MP-safe. There is almost no influence on performance. wm(4) does not use ifp->if_watchdog now, that is, it does not touch ifp->if_timer. It also uses own callout(wm_tick) as watchdog now. The watchdog uses per-queue counter to check timeout. So, global lock is not required.
1.508.4.12	13-Jan-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #491): sys/dev/pci/if_wm.c: 1.550, 1.552 Don't use MSI-X if we can use only one queue to save interrupt resource. Written by knakahara and tested by me. -- Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki.
1.508.4.11	02-Jan-2018	snj	Pull up following revision(s) (requested by ozaki-r in ticket #456): sys/arch/arm/sunxi/sunxi_emac.c: 1.9 sys/dev/ic/dwc_gmac.c: 1.43-1.44 sys/dev/pci/if_iwm.c: 1.75 sys/dev/pci/if_wm.c: 1.543 sys/dev/pci/ixgbe/ixgbe.c: 1.112 sys/dev/pci/ixgbe/ixv.c: 1.74 sys/kern/sys_socket.c: 1.75 sys/net/agr/if_agr.c: 1.43 sys/net/bpf.c: 1.219 sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416 sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257 sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146 sys/net/if_etherip.c: 1.40 sys/net/if_ethersubr.c: 1.243, 1.246 sys/net/if_faith.c: 1.57 sys/net/if_gif.c: 1.132 sys/net/if_l2tp.c: 1.15, 1.17 sys/net/if_loop.c: 1.98-1.101 sys/net/if_media.c: 1.35 sys/net/if_pppoe.c: 1.131-1.132 sys/net/if_spppsubr.c: 1.176-1.177 sys/net/if_tun.c: 1.142 sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121 sys/net/npf/npf_ifaddr.c: 1.3 sys/net/npf/npf_os.c: 1.8-1.9 sys/net/rtsock.c: 1.230 sys/netcan/if_canloop.c: 1.3-1.5 sys/netinet/if_arp.c: 1.255 sys/netinet/igmp.c: 1.65 sys/netinet/in.c: 1.210-1.211 sys/netinet/in_pcb.c: 1.180 sys/netinet/ip_carp.c: 1.92, 1.94 sys/netinet/ip_flow.c: 1.81 sys/netinet/ip_input.c: 1.362 sys/netinet/ip_mroute.c: 1.147 sys/netinet/ip_output.c: 1.283, 1.285, 1.287 sys/netinet6/frag6.c: 1.61 sys/netinet6/in6.c: 1.251, 1.255 sys/netinet6/in6_pcb.c: 1.162 sys/netinet6/ip6_flow.c: 1.35 sys/netinet6/ip6_input.c: 1.183 sys/netinet6/ip6_output.c: 1.196 sys/netinet6/mld6.c: 1.90 sys/netinet6/nd6.c: 1.239-1.240 sys/netinet6/nd6_nbr.c: 1.139 sys/netinet6/nd6_rtr.c: 1.136 sys/netipsec/ipsec_output.c: 1.65 sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10 kmem_intr_free kmem_intr_[z]alloced memory the underlying pools are the same but api-wise those should match Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@ Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..." scattered all over the source code and makes it easy to identify remaining KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE. No functional change Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE If IFEF_MPSAFE is set, hold the lock and otherwise don't hold. This change requires additions of KERNEL_LOCK to subsequence functions from if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe components. Proposed on tech-kern@ and tech-net@ Ensure to hold if_ioctl_lock when calling if_flags_set Fix locking against myself on ifpromisc vlan_unconfig_locked could be called with holding if_ioctl_lock. Ensure to not turn on IFF_RUNNING of an interface until its initialization completes And ensure to turn off it before destruction as per IFF_RUNNING's description "resource allocated". (The description is a bit doubtful though, I believe the change is still proper.) Ensure to hold if_ioctl_lock on if_up and if_down One exception for if_down is if_detach; in the case the lock isn't needed because it's guaranteed that no other one can access ifp at that point. Make if_link_queue MP-safe if IFEF_MPSAFE if_link_queue is a queue to store events of link state changes, which is used to pass events from (typically) an interrupt handler to if_link_state_change softint. The queue was protected by KERNEL_LOCK so far, but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it by a spin mutex. Additionally with this change KERNEL_LOCK of if_link_state_change softint is omitted if NET_MPSAFE is enabled. Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of if_timer (see the comment). Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH At that point no other one modifies the list so IFADDR_READER_FOREACH is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though, if we try to detect contract violations of pserialize, using it violates the contract. So avoid using it makes life easy. Ensure to call if_addr_init with holding if_ioctl_lock Get rid of outdated comments Fix build of kernels without ether By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that created a unnecessary dependency from if.c to if_ethersubr.c. PR kern/52790 Rename IFNET_LOCK to IFNET_GLOBAL_LOCK IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then). Wrap if_ioctl_lock with IFNET_* macros (NFC) Also if_ioctl_lock perhaps needs to be renamed to something because it's now not just for ioctl... Reorder some destruction routines in if_detach - Destroy if_ioctl_lock at the end of the if_detach because it's used in various destruction routines - Move psref_target_destroy after pr_purgeif because we want to use psref in pr_purgeif (otherwise destruction procedures can be tricky) Ensure to call if_mcast_op with holding IFNET_LOCK Note that CARP doesn't deal with IFNET_LOCK yet. Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held Describe which lock is used to protect each member variable of struct ifnet Requested by skrll@ Write a guideline for converting an interface to IFEF_MPSAFE Requested by skrll@ Note that IFNET_LOCK must not be held in softint Don't set IFEF_MPSAFE unless NET_MPSAFE at this point Because recent investigations show that interfaces with IFEF_MPSAFE need to follow additional restrictions to work with the flag safely. We should enable it on an interface by default only if the interface surely satisfies the restrictions, which are described in if.h. Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because the network stack is still serialized by the big kernel locks by default.
1.508.4.10	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #432): sys/dev/pci/if_wm.c: revision 1.548 Fix a bug that 8257[56] can't receive packet reported by Bert Kiers in PR#52717. For 82575 and 82576, the RX descriptors must be initialized after the setting of RCTL.EN in wm_set_filter(). This bug was added in if_wm.c rev. 1.515.
1.508.4.9	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.508.4.8	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #424): sys/dev/pci/if_wm.c: 1.545-1.546 Don't allocate MSI-X interrupt on 82583. 82583 chip has a MSI-X capability in the PCI configuration space but it doesn't support it. At least the document doesn't say anything about MSI-X. Fixes PR#52767 reported by Shinichi Doyashiki. -- - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Cleanup comment.
1.508.4.7	24-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #389): sys/net/if_ether.h: revision 1.69 sys/net/if_vlan.c: revision 1.108 sys/dev/pci/if_bge.c: revision 1.313 sys/net/if_vlanvar.h: revision 1.11 sys/net/if_vlanvar.h: revision 1.12 sys/net/if_ether.h: revision 1.70 sys/net/if_vlan.c: revision 1.110 sys/dev/pci/if_wm.c: revision 1.544 sys/dev/pci/if_wmreg.h: revision 1.105 Fix a bug that a vlan packet which has priority or CFI bit in the tag causes panic. Revert part of if_bge.c 1.312. It's not required to mask other than VLAN ID bits in VLAN tag. Revert if_wmreg.h 1.104 and if_wm.c 1.542. It's not required to mask other than VLAN ID bits in VLAN tag. No functional change: - u_int16_t -> uint16_t - u_short -> uint16_t - tag_hash_func -> vlan_tag_hash - 0 -> NULL because vlr_parent is a pointer.
1.508.4.6	22-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #379): sys/dev/pci/if_wm.c: revision 1.540 On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux).
1.508.4.5	06-Nov-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #348): sys/dev/pci/if_wm.c: revision 1.542 sys/dev/pci/if_wmreg.h: revision 1.104 fix wm(4) vlan panic. Reported and tested by Tom Ivar Helbekkmo, thanks. wm(4) used PRI bits and CFI bit as vlantag by mistake. It is found out by if_ether.h:r1.67.
1.508.4.4	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.508.4.3	15-Oct-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #306): sys/dev/pci/if_wm.c: revision 1.539 - Correct NVM word size check in wm_nvm_version() to prevent "EERD polling failed" error on attach. if_wm.c rev. 1.534 was not perfect. Reported by ozaki-r@. - Print NVM offset and word count when EERD polling failed.
1.508.4.2	01-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #173): sys/dev/pci/if_wmreg.h: 1.99-1.103 sys/dev/pci/if_wmvar.h: 1.34-1.37 sys/dev/pci/if_wm.c: 1.510-1.537 sys/dev/mii/ihphyreg.h: 1.2 sys/dev/mii/inbmphyreg.h: 1.10 sys/dev/mii/igphy.c: 1.27 Sync wm(4) up to if_wm.c rev. 1.537: - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). - If a legacy interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Flush writing WMREG_CTRL in wm_gmii_mediachange(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Add I219 specific workaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for workaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - IF TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function is required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Configure the LCD with the extended configuration region in NVM if it's required. Tested with Thinkpad X220. - 8257[12]: Don't directly access SPI but use EERD register. - 82575-I354: If the size of SPI ROM >= 32K words, use direct SPI access instead of EERD register access. - Add wm_nvm_eec_clock_raise() and wm_nvm_eec_clock_lower() and use them for Microwire/SPI bus control. Same as Linux and FreeBSD. - Reduce timeout value for 80003 in wm_get_swfw_semaphore(). Same as Linux and FreeBSD. - Change API of kmrn_{read,write}reg() and check the return value. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally." on I21[01]. PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check NVM word size before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Print internal driver flags when attaching. - Add debug printf()s. - Rename variables. - Add comment, update comment and remove wrong comment.
1.508.4.1	04-Jul-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #81): sys/dev/pci/ixgbe/ixgbe.c: revision 1.89 sys/dev/pci/if_wm.c: revision 1.509 sys/dev/pci/ixgbe/ixgbe.c: revision 1.90 Use IFM_1000_KX and IFM_2500_KX. Use IFM_1000_KX more.
1.566.2.13	26-Jan-2019	pgoyette	Sync with HEAD
1.566.2.12	18-Jan-2019	pgoyette	Synch with HEAD
1.566.2.11	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.566.2.10	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.566.2.9	20-Oct-2018	pgoyette	Sync with head
1.566.2.8	30-Sep-2018	pgoyette	Ssync with HEAD
1.566.2.7	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.566.2.6	28-Jul-2018	pgoyette	Sync with HEAD
1.566.2.5	25-Jun-2018	pgoyette	Sync with HEAD
1.566.2.4	21-May-2018	pgoyette	Sync with HEAD
1.566.2.3	02-May-2018	pgoyette	Synch with HEAD
1.566.2.2	22-Apr-2018	pgoyette	Sync with HEAD
1.566.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.583.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.583.2.1	10-Jun-2019	christos	Sync with HEAD
1.645.2.21	23-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1802: sys/dev/pci/if_wm.c 1.794-1.798 via patch sys/dev/pci/if_wmreg.h 1.131 - Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. - Drop frames if the RX descriptor ring has no room on multiqueue system. - Improve dmesg output: - Print RX packet buffer size. - Fix the upper 16bit of Image Unique ID(EtrackID).
1.645.2.20	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #1791): sys/dev/pci/if_wm.c: revision 1.793 Add Intel I219-{LM,V}(20,21) support.
1.645.2.19	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1754): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.645.2.18	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1746): sys/dev/pci/if_wm.c: revision 1.785 sys/dev/pci/if_wm.c: revision 1.786 sys/dev/pci/if_wm.c: revision 1.787 sys/dev/pci/if_wm.c: revision 1.788 sys/dev/pci/if_wm.c: revision 1.789 wm(4): Use SCVPC and HRMPC for 82575 and newer. The SCVPC(SerDes/SGMII Code Violation Packet Count) register and the HRMPC (Header Redirection Missed Packet Count) register were added in if_wm.c rev. 1.776 but the location in the code were incorrect. Fix them. wm(4): Add some info to some event counters. - The difference between the RUC(Rx Undersize) and RFC(Rx Fragment) is whether the CRC is valid or not. - The difference between the ROC(Rx Oversize) and RJC(Rx Jabber) is whether the CRC is valid or not. - LENERRS(Length Errors) is for the length/type field <= 1500. wm(4): the IAC(Interrupt Assertion Count) register is for 82571 and newer. wm(4): Modify descriptions of flow control related event counters. wm(4): Replace /* nothing */ to __nothing for evcnt macros Prevent empty if bodies for !WM_EVENT_COUNTERS.
1.645.2.17	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #344): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783,1.784 via patch Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.645.2.16	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1656: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.645.2.15	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1578: sys/dev/pci/if_wm.c 1.764-1.767 via patch sys/dev/pci/if_wmreg.h 1.128 - Workaround for some hypervisor environments. The environments cannot stop e1000 interrupt immediately. - Rename nq_txdesc' member "nqrx_ctx" to "nqtx_ctx". No functional change. - Add comment. Modify comment. - KNF.
1.645.2.14	07-Sep-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1515: sys/dev/pci/if_wm.c 1.741-1.749, 1.753-1.757, 1.762 via patch sys/dev/pci/if_wmreg.h 1.126-1.127 - Fix I219 workaround in wm_flush_desc_rings(). - Add more statistics counters. - To avoid releasing mutex temporally, use new wm_set_mdio_slow_mode_hv_locked(). - No functional changes: - Turn a locking botch (shouldn't drop lock on error) into a KASSERT in wm_deferred_start_locked(). - Remove unneeded header inclusion. - Use __BIT() a little. - Modify comment and debug messages. - Consistency use -1 instead of 1 for some error code. - KNF.
1.645.2.13	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1477: sys/dev/pci/if_wmreg.h 1.122-1.125 sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720, 1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.645.2.12	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1373: sys/dev/pci/if_wm.c 1.716-1.718 via patch sys/dev/pci/if_wmreg.h 1.121 sys/dev/mii/ihphy.c 1.20 - Fix a bug that device timeout still happens when the link is down on ICH/PCH. Fixes PR kern/56478. - Add some sysctl info for debugging. - ihphy(4): Don't power down the PHY when the interface goes down. - Use macro. Fix comment.
1.645.2.11	21-Oct-2021	martin	Pull up the following, requested by msaitoh in ticket #1363: share/man/man4/wm.4 1.42 sys/dev/pci/files.pci 1.434 sys/dev/pci/if_wm.c 1.690,1.692-1.693,1.697-1.704, 1.706-1.715 via patch - Add missing drain for pcq in wm_stop_locked(). - Add support for I219V 15-19 and I219LM 16-19. - Fix Tx stall. - Use wm_flush_desc_rings() workaround more on I219. - Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. - Add some sysctl values for debugging TX/RX queues. - Use atomic_{load,store}_relaxed for evcnt 64 bit counter. - WM_EVENT_COUNTER is enabled by default on 64 bit architectures. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxq->rxq_ptr updating. - Stop legacy interrupts before calling softint. - Disable printf()s in wm_flush_desc_rings() because the code is verified. - Print I219's version number. - Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. - Fix return value of interrupt handler. - Only print an error about missing I/O BARs for chips that need it. - Do not return a void value from a void function. - Fix comment. - Whitespace fixes. No functional change.
1.645.2.10	16-Nov-2020	martin	Pull up following revision(s) (requested by rin in ticket #1133): sys/dev/pci/if_wm.c: revision 1.696 Fix little-endian dependence in wm_rxeof(), by which packets cannot be received by 82574 and successors on big-endian machines. Tested by aarch64eb with I210-T1 on ROCKPro64. Thanks msaitoh for discussion! XXX pullup to netbsd-9 and netbsd-8
1.645.2.9	04-Nov-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #1126): sys/dev/pci/if_wm.c: revision 1.694 sys/dev/pci/if_wm.c: revision 1.695 (via patch) sys/dev/pci/if_wmvar.h: revision 1.47 Add WMPHY_I350. Not used yet. Workaround for ihphy and atphy(ICH/PCH, 82580 and I350). These phys stop DMA while link is down which causes device timeout. Fix PR/kern 40981 Reviewed and tested by msaitoh@n.o, thanks. XXX pullup-[89]
1.645.2.8	16-Oct-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1114): sys/dev/pci/if_wm.c: revision 1.691 Fixes a problem that the attach function reported "wm_gmii_setup_phytype: Unknown PHY model. OUI=000000, model=0000" and "PHY type is still unknown." Don't call wm_gmii_setup_phytype() three times if the interface uses SGMII with internal MDIO. Tested with I354(Rangeley(SGMII(MDIO))) and I350(SERDES(SFP), SGMII(SFP)).
1.645.2.7	23-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1093): sys/dev/pci/if_wmreg.h: revision 1.120 sys/dev/pci/if_wmvar.h: revision 1.46 sys/dev/pci/if_wm.c: revision 1.686 sys/dev/pci/if_wm.c: revision 1.687 sys/dev/mii/inbmphyreg.h: revision 1.20 sys/dev/pci/if_wm.c: revision 1.688 sys/dev/pci/if_wm.c: revision 1.689 s/ressource/resource/. Found by knakahara. Add new flag named WM_F_CRC_STRIP and use it. No functional change. This change also sets the RCTL_SECRC bit on I211 but it doesn't change the behavior because I211 always strips CRC like I35[04] and I210. Add a workaround for jumbo frame on PCH2 and newer. Tested by chs@. - Add wm_lv_jumbo_workaround_ich8lan() and use it. From FreeBSD. XXX For KUMCTRLSTA_OFFSET_HD_CTRL register modification, it's doubtful. FreeBSD and Linux do the same thing that they set the same value on both jumbo frame's enable case and the disable case. It seems the default value is 0x0b0c and it's not changed on the enable case, so it might be a bug on the enable case or the modification is not required. - Rename I219_UNKNOWN1 to I82579_UNKNOWN1. Use 12K for packet buffer for jumbo frame on PCH2 and newer. XXX Note that Linux Use 14K.
1.645.2.6	05-Aug-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1040): sys/dev/mii/igphy.c: revision 1.35 sys/dev/mii/igphy.c: revision 1.36 sys/dev/mii/igphyreg.h: revision 1.12 sys/dev/mii/igphyreg.h: revision 1.13 sys/dev/mii/makphyreg.h: revision 1.11 sys/dev/pci/if_wm.c: revision 1.682 sys/dev/pci/if_wm.c: revision 1.683 sys/dev/pci/if_wm.c: revision 1.684 sys/dev/pci/if_wm.c: revision 1.685 sys/dev/mii/makphy.c: revision 1.66 s/MII_IGPHY_/IGPHY_/. No functional change. Rename PSSR_* to MAKPHY_PSSR_* and IGPHY_PSSR_* to avoid conflict. No functional change. Setup PCS and SGMII for SFP correctly. It still doesn't support SFP insertion/removal. Copper: wm2: SGMII(SFP) wm2: 0x1043c440<SPI,IOH_VALID,PCIE,SGMII,NEWQUEUE,ASF_FIRM,EEE,SFP> makphy0 at wm2 phy 6: Marvell 88E1111 Gigabit PHY, rev. 1 Fiber: wm3: SERDES(SFP) wm3: 0x10034440<SPI,IOH_VALID,PCIE,NEWQUEUE,ASF_FIRM,SFP> wm3: 1000baseSX, 1000baseSX-FDX, auto Explicitly cast from uint16_t to uint32_t before shifting 16bit left when printing Image Unique ID. Found by kUBSan. Set if_baudrate for non-MII device. Before this commit, it was 0.
1.645.2.5	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #994: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678,1.680-1.681 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 sys/dev/pci/if_wmvar.c 1.45 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler. Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Rename some macros and function. - KNF. Add comment.
1.645.2.4	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #650: sys/dev/pci/if_wm.c 1.650, 1.652-1.654 sys/dev/pci/if_wmreg.h 1.116-1.117 - Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. - Simplify code by using "struct mii_data *mii" more. No functional change. - MSI-X doesn't use sc->sc_icr variable, so move the code into non-MSI-X part. No functional change intended. - Modify debug printfs a bit. - Rename macro. - Use __BIT() - Fix comment. Add comment. - KNF.
1.645.2.3	23-Jan-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #643): sys/dev/pci/if_wm.c: revision 1.661 fix softint_disestablish() in wm_detach(). ok by msaitoh@n.o XXX pullup-8, pullup-9
1.645.2.2	11-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #541): sys/dev/pci/if_wm.c: revision 1.651 Fix a bug that wm_sgmii_readreg_locked() may return error even if error isn't occurred.
1.645.2.1	06-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #404): sys/dev/pci/if_wm.c: revision 1.646 sys/dev/pci/if_wm.c: revision 1.647 sys/dev/pci/if_wm.c: revision 1.649 Use unsigned to avoid undefined behavior in wm_i82543_mii_sendbits(). Found by kUBSan. printf -> device_printf Fix typo in comment.
1.658.2.2	29-Feb-2020	ad	Sync with head.
1.658.2.1	25-Jan-2020	ad	Sync with head.
1.670.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.694.2.2	03-Apr-2021	thorpej	Sync with HEAD.
1.694.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.702.4.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.702.4.1	13-May-2021	thorpej	Sync with HEAD.
1.703.2.1	31-May-2021	cjep	sync with head
1.767.2.8	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #607): sys/dev/pci/if_wm.c: revision 1.795 sys/dev/pci/if_wm.c: revision 1.796 sys/dev/pci/if_wm.c: revision 1.797 sys/dev/pci/if_wmreg.h: revision 1.131 wm(4): Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. The iqdrops counter should include not only MPC(Missed Packet Count) but also RQDPC(Receive Queue Drop Packet Count). Same as ixgbe(4) and igc(4). Note that the RQDPC is not currently counted because SRRCTL_DROP_EN is not set. wm(4): Drop frames if the RX descriptor ring has no room on multiqueue system. Drop frames if the RX descriptor ring has no room. This is enabled only on multiqueue system to avoid bad influence to other queues. The drop count can be seen by the RQDPC counter (wmN rxqXXdrop evcnt). wm(4): Fix compile error without WM_EVENT_COUNTERS.
1.767.2.7	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #606): sys/dev/pci/if_wm.c: revision 1.794 sys/dev/pci/if_wm.c: revision 1.798 wm(4): Print RX packet buffer size. wm(4): Fix upper 16bit of Image Unique ID(EtrackID). Don't override uid1 variable while reading option ROM version to print Image Unique ID correctly.
1.767.2.6	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #564): sys/dev/pci/if_wm.c: revision 1.793 Add Intel I219-{LM,V}(20,21) support.
1.767.2.5	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #421): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.767.2.4	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #396): sys/dev/pci/if_wm.c: revision 1.785 sys/dev/pci/if_wm.c: revision 1.786 sys/dev/pci/if_wm.c: revision 1.787 sys/dev/pci/if_wm.c: revision 1.788 sys/dev/pci/if_wm.c: revision 1.789 wm(4): Use SCVPC and HRMPC for 82575 and newer. The SCVPC(SerDes/SGMII Code Violation Packet Count) register and the HRMPC (Header Redirection Missed Packet Count) register were added in if_wm.c rev. 1.776 but the location in the code were incorrect. Fix them. wm(4): Add some info to some event counters. - The difference between the RUC(Rx Undersize) and RFC(Rx Fragment) is whether the CRC is valid or not. - The difference between the ROC(Rx Oversize) and RJC(Rx Jabber) is whether the CRC is valid or not. - LENERRS(Length Errors) is for the length/type field <= 1500. wm(4): the IAC(Interrupt Assertion Count) register is for 82571 and newer. wm(4): Modify descriptions of flow control related event counters. wm(4): Replace /* nothing */ to __nothing for evcnt macros Prevent empty if bodies for !WM_EVENT_COUNTERS.
1.767.2.3	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #344): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783 sys/dev/pci/if_wm.c: revision 1.784 Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.767.2.2	27-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #216): sys/dev/pci/if_wm.c: revision 1.782 Add missing read to count Circuit Breaker Rx Dropped Packet correctly.
1.767.2.1	22-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #213): sys/dev/pci/if_wm.c: revision 1.770 sys/dev/pci/if_wm.c: revision 1.771 sys/dev/pci/if_wm.c: revision 1.772 sys/dev/pci/if_wm.c: revision 1.773 sys/dev/pci/if_wm.c: revision 1.774 sys/dev/pci/if_wm.c: revision 1.775 sys/dev/pci/if_wm.c: revision 1.776 sys/dev/pci/if_wmreg.h: revision 1.129 sys/dev/pci/if_wm.c: revision 1.777 sys/dev/pci/if_wm.c: revision 1.778 sys/dev/pci/if_wmvar.h: revision 1.49 sys/dev/pci/if_wm.c: revision 1.779 sys/dev/pci/if_wm.c: revision 1.768 sys/dev/pci/if_wm.c: revision 1.769 sys/dev/pci/if_wmreg.h: revision 1.130 sys/dev/pci/if_wm.c: revision 1.780 sys/dev/pci/if_wm.c: revision 1.781 Count some 64bit counters correctly. - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Found by knakahara. Add note for the TORL register. The TOR register includes error, flow control and broadcast rejected. Sort lines. No functional change. Rearrange the order of the registers so that they are roughly in ascending order. Sort lines. No functional change. Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. IC{TX,RX}*C registers are for older than 82575. Fix a bug that the transmit underrun counter is incorrectly counted. The transmit underrun bit in the transmit status filed is only for 82544 (and older?), so don't use the counter for newer chips. The bit is reserved for newer chips, but the bit sometimes set on 82575 at least. Don't add "Count" for event counter's description. Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero. Add some new event counters. Add the following counters for 82575 and newer except 80003, ICHs and PCHs: - Only 82576 document describes about the circuit breaker, so the following two might be only for 82575: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - 82575's document doesn't describe the following two, but we can see the same value as GO{T,R}C have: - Host Good Octets RX - Host Good Octets TX - 82575's document doesn't describe the LENERRS (Length Errors) counter. I don't know if it has. - Perhaps Non-SerDes/SGMII devices don't have SCVPC (SerDes/SGMII Code Violation Packet) register. We don't care if it's SerDes/SGMII or not for now. - HRMPC (Header Redirection Missed Packet) appears only once in 8257[56]'s datasheet. FreeBSD's igb counts it, so we do, too. - Count the following two for I350 and newer. I don't know if PCHs have: - EEE TX LPI - EEE RX LPI Move statistics updating code from wm_tick() to new wm_update_stats(). - To reuse. - No functional change. Add SOICZIFDATA (ifconfig -z) support for evcnt(9). First update the statistics data, then clear the event counters, and finally copy and clear if_data via ether_ioctl(). Fix prc511's comment and description. Use WM_IS_ICHPCH(). No functional change. Fix typo. s/ictxact/ictxatc/. No functional change.
1.799.2.1	02-Aug-2025	perseant	Sync with HEAD
1.131	29-Jan-2024	msaitoh	wm(4): Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. The iqdrops counter should include not only MPC(Missed Packet Count) but also RQDPC(Receive Queue Drop Packet Count). Same as ixgbe(4) and igc(4). Note that the RQDPC is not currently counted because SRRCTL_DROP_EN is not set.
1.130	11-May-2023	msaitoh	Add some new event counters. Add the following counters for 82575 and newer except 80003, ICHs and PCHs: - Only 82576 document describes about the circuit breaker, so the following two might be only for 82575: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - 82575's document doesn't describe the following two, but we can see the same value as GO{T,R}C have: - Host Good Octets RX - Host Good Octets TX - 82575's document doesn't describe the LENERRS (Length Errors) counter. I don't know if it has. - Perhaps Non-SerDes/SGMII devices don't have SCVPC (SerDes/SGMII Code Violation Packet) register. We don't care if it's SerDes/SGMII or not for now. - HRMPC (Header Redirection Missed Packet) appears only once in 8257[56]'s datasheet. FreeBSD's igb counts it, so we do, too. - Count the following two for I350 and newer. I don't know if PCHs have: - EEE TX LPI - EEE RX LPI
1.129	11-May-2023	msaitoh	Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero.
1.128	19-Oct-2022	msaitoh	branches: 1.128.2; Fix typo. No functional change. OK'd by knakahara.
1.127	22-Jul-2022	msaitoh	Add more statistics countes. - Add many statics counters that the chip has. - Attach event counters only if available.
1.126	20-Jul-2022	msaitoh	Modify comment. Use __BIT() a little. No functional change.
1.125	22-Dec-2021	skrll	Fixup NQTXC_CMD_{TCP,UDP} after previous
1.124	22-Dec-2021	skrll	In a close run thing use #define<tab> everywhere for consistency
1.123	22-Dec-2021	skrll	Use __BIT() some more. Same code before and after.
1.122	20-Dec-2021	skrll	Trailing whitespace
1.121	05-Nov-2021	msaitoh	Fix comment. No functional change.
1.120	16-Sep-2020	msaitoh	Add a workaround for jumbo frame on PCH2 and newer. Tested by chs@. - Add wm_lv_jumbo_workaround_ich8lan() and use it. From FreeBSD. XXX For KUMCTRLSTA_OFFSET_HD_CTRL register modification, it's doubtful. FreeBSD and Linux do the same thing that they set the same value on both jumbo frame's enable case and the disable case. It seems the default value is 0x0b0c and it's not changed on the enable case, so it might be a bug on the enable case or the modification is not required. - Rename I219_UNKNOWN1 to I82579_UNKNOWN1.
1.119	08-Jan-2020	msaitoh	- Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Define SFF_SFP_ETH_FLAGS_100LX with 0x10.
1.118	11-Dec-2019	msaitoh	Add SFP support part 1. The code for SerDes and SGMII setting is not include in this commit. SFP module removal interrupt detects but not used yet: - Detect SFP's 1000BASE-SX, 1000BASE-LX, 100BASE-FX and 1000BASE-T correctly. - Detect the Media Auto Sense feature. Not supported yet. - Add comment.
1.117	11-Dec-2019	msaitoh	- Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. XXX We should check for other NVM reads if they have the same problem. One of the solution to avoid this type of bug is to check the register layout in wm_nvm_read(). - Use __BIT() - Add comment
1.116	11-Dec-2019	msaitoh	No functional change: - Add defintion of CONNSW register and PCS_NPTX (not used yet). - Add some bit definitions of PCS_LCTL. - Renae macro. - Fix comment. Add comment. - KNF.
1.115	23-Jul-2019	msaitoh	branches: 1.115.2; Define RETA_ENTRY_MASK_Q(used in RSS redirection table) correctly. Found by UBSan.
1.114	07-Jun-2019	msaitoh	Use unsigned. Found by UBSan. OK'd by kamil.
1.113	28-Feb-2019	khorben	Typo (s/vaule/value/)
1.112	31-Jan-2019	msaitoh	Add some workarounds which use EMI register or EEE related: - PCH only: Add workaround for link disconnects on a busy hub in half duplex. - PCH and PCH2 only: Set MSE higher to enable link to stay up when noise is high. - PCH2 only: Drop link after 5 times MSE threshold was reached. - PCH2 only: Set EEE LPI Update Timer to 200usec. - For PCH2 and newer: When connected at 10Mbps half-duplex, some parts are excessively aggressive resulting in many collisions. To avoid this, increase the IPG and reduce Rx latency in the PHY. - For I21[789] and if EEE is enabled: Disable LPLU if both link partners support 100BaseT EEE and 100Full is advertised on both ends of the link, and enable Auto Enable LPI since there will be no driver to enable LPI while in Sx.
1.111	20-Dec-2018	msaitoh	- Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220.
1.110	14-Dec-2018	msaitoh	- Fix availability detection of WoL on some chips. This change has no effect other than dmesg because WM_F_WOL is currently not used to change the behavior: - For ICH/PCH, check the capability not from NVM but from the WUC register. Check the value before clearing the register. - 82580 and newer have per-port NVM block, so read the area correctly. Note that 82580, I350 and I354 may have PCI function 2 and 3. - Some devices can't detect WoL capability neither from NVM nor from WUC. Use PCI device ID and the function number. - Print the WUS (WakeUp Status) register bits when resume.
1.109	19-Nov-2018	msaitoh	- Control TX/RX descriptor snooping control bits on ICH8 and newer. Only on ICH8, No-snoop bits are opposite polarity. On my Thinkpas X61, the default value of this bits are all zero, so this commit changes the snoop function enable on the machine. I tested with some other PCH machines and those bits are all zero (enable snoop by default), so this commit won't affect to some machines. - Disable relax ordering on 82546GB(Device ID 0x1099 and 0x10b5) or >= ICH8. Same as other OSes.
1.108	02-Nov-2018	msaitoh	Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This woraround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. I observed this problem on my Shuttle DS57U's I218 port six times in two months.
1.107	12-Apr-2018	msaitoh	branches: 1.107.2; No binary change: - Sort registers. - Lowercase hexadecimal value.
1.106	10-Apr-2018	msaitoh	SW PHY Config Enable bit for ICH8 B0 stepping is not bit 1 but bit 0.
1.105	22-Nov-2017	msaitoh	branches: 1.105.2; Revert if_wmreg.h 1.104 and if_wm.c 1.542. It's not required to mask other than VLAN ID bits in VLAN tag.
1.104	23-Oct-2017	knakahara	fix wm(4) vlan panic. Reported and tested by Tom Ivar Helbekkmo, thanks. wm(4) used PRI bits and CFI bit as vlantag by mistake. It is found out by if_ether.h:r1.67. XXX need pullup-8
1.103	26-Jul-2017	msaitoh	- Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register vaule from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg.
1.102	13-Jul-2017	msaitoh	- Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - Add debug printf.
1.101	12-Jul-2017	msaitoh	Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register.
1.100	26-Jun-2017	msaitoh	- Make new wm_phy_post_reset() and use this function at all location after reseting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer.
1.99	26-Jun-2017	msaitoh	- Fix wm_set_ral(): - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. - Remove wrong comment.
1.98	28-Feb-2017	knakahara	branches: 1.98.6; fix EITR setting. + 82574 - add EITR setting, that is MSI-X mode interrupt interval + 82575 - fix EITR value. 82575's EITR usage is the same as legacy (not NEWQUEUE) controllers - apply workaround which 82575's EITR does not have CNT_INGR bit + other NEWQUEUE controllers - fix interrupt interval field. NEWQUEUE (include 82575) controllers' interrupt interval field is 2:14 - use CNT_INGR bit which avoid to overwrite counter tested 82574, 82575 and I354. ok by msaitoh@n.o.
1.97	02-Feb-2017	msaitoh	Fix typo in comment.
1.96	01-Feb-2017	knakahara	tabify
1.95	16-Jan-2017	knakahara	support {extended, advanced} receive descriptor for wm(4). support below two receive descriptors. - extended descriptor (used by 82574 only) - advanced descriptor (used by 82575 and newer) SPH(split header buffer) is not supported yet. ok by msaitoh@n.o.
1.94	13-Dec-2016	msaitoh	branches: 1.94.2; - Change to use 2500Base-KX correctly on C2000(I354). It worked, but the output of ifconfig and if_baudrate was not good. Tested by nils@: - The STATUS_TBIMODE bit in the STATUS register is deleted since 82575, so check for 82575 and newer first and then check for old devices. - Check the 2P5_SKU and 2P5_SKU_OVER bit for KX. - Set IFM_2500_SX instead of IFM_1000_SX for 2.5G. - Check SERDES's speed directly from the PCS layer (PCS_LSTS register) for old devices. - Style fix.
1.93	16-Nov-2016	msaitoh	Sync wm_smbustopci() with Linux and FreeBSD. This change effects PCH and newer devices.
1.92	16-Nov-2016	msaitoh	Add three workarounds for PCH_{LPT,SPT}.
1.91	14-Nov-2016	msaitoh	Sort registers. No functional change.
1.90	11-Nov-2016	msaitoh	The MEHE bit in the CTRL register is not bit 17 but 19.
1.89	06-May-2016	msaitoh	branches: 1.89.2; Basic support for I219. It doesn't work on I219, so it's disabled.
1.88	25-Dec-2015	msaitoh	- Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Fix typo in comment and modify comment by tnn@. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change.
1.87	27-Oct-2015	msaitoh	Move the definition of CTRL_EXT_PHYPDEN and use __BIT(). No binary change.
1.86	23-Oct-2015	msaitoh	PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken.
1.85	13-Oct-2015	knakahara	fix RXCSUM_CRCOFL bit of WMREG_RXCSUM register.
1.84	13-Oct-2015	knakahara	support RX multiqueue. ok by msaitoh@n.o
1.83	13-Oct-2015	knakahara	change const value register macros to macro expressions to support multiqueue.
1.82	08-Oct-2015	msaitoh	- ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition have not used yet. - Add ACK data Disable bit's definition.
1.81	30-Sep-2015	knakahara	fix IVARs offset for 82580 and newer.
1.80	30-Sep-2015	msaitoh	- Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Print the NVM image version on 82583, too. Pointed out by joerg@.
1.79	07-Sep-2015	msaitoh	- Check iNVM's image version and print it. - Update TODO.
1.78	13-Jun-2015	msaitoh	Add MSI/MSI-X support written by Kengo Nakahara. Some old devices' support is written by me. It's disabled by default. If you'd like to use, define WM_MSI_MSIX. Tested with: 8254[3405617] (INTx even if it has MSI CAP because of a errata) 8257[12], 82583 ICH8, ICH10, PCH2, PCH_LPT(I21[78]) (MSI) 8257[456], 82580, I35[04], I21[01] (MSI-X) Not tested: 82542, 82573, 80003, ICH9, PCH,
1.77	06-Jun-2015	msaitoh	- Add workaround for I210 Errata 25 and I211 Errata 10. - Add wm_gmii_gs40g_{read\|write}reg() and use it to access non-standatrd page. - Add wm_pll_workaround_i210() and call it when chip is i211 chip is i210 and it use INVM chip is i210 and NVM image version < 3.25 - Add comment - Rename macros.
1.76	06-Jun-2015	msaitoh	Print NVM image version.
1.75	06-Jun-2015	msaitoh	Revert previos. Sorry, I committed in another working directory...
1.74	06-Jun-2015	msaitoh	Print NVM image version.
1.73	02-Jun-2015	msaitoh	Fix a lot of bugs to make 82575 and newer's SERDES based systems work. - Add SERDES specific functions. - Fix IO pin configuration. - Reset autonego timer when link becomes up. TODO: - Fix a bug that SFP ROM can't read. - Perhaps some work is required for 8257[12] serdes systems. - Remove duplicated code in TBI's link related functions.
1.72	22-May-2015	msaitoh	- Currently, WM_F_EEE bit is not set on all chips. it's intended to not to use all of EEE fuctions but wm_set_eee_i350() leaves IPCNFG_10BASE_TE bit and it causes link negotiation problem on some old switches. Disable 10BASE-Te function, too. - Call wm_set_eee_i350() on some chips, too.
1.71	16-May-2015	msaitoh	Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines and NICs. Tested with Shuttle DS57U(iNVM based) and other non iNVM based I210 machines.
1.70	15-May-2015	msaitoh	Fix typo in comment.
1.69	04-May-2015	msaitoh	For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way.
1.68	04-May-2015	msaitoh	Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the Same as other OSes
1.67	04-May-2015	msaitoh	Remove WMREG_TQSA_LO and WMREG_TQSA_HIGH. Those registers are not described in documents and other OS's drivers don't access it. (I have no the first chip(82542)'s document. Those registers might be described in the document).
1.66	15-Feb-2015	msaitoh	- Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment.
1.65	24-Oct-2014	msaitoh	branches: 1.65.2; Simplify. No functional change.
1.64	24-Oct-2014	msaitoh	Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location.
1.63	03-Sep-2014	msaitoh	Read SFF SFP ROM and configure driver from the value.
1.62	01-Sep-2014	msaitoh	Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Calculate NVM word size correctly. - Determine timeout value based on the NVM word size.
1.61	29-Aug-2014	msaitoh	No binary change: - Move some NVM related macros from if_wm.c to if_wmreg.h. - Rename some macros for consistency.
1.60	31-Jul-2014	msaitoh	branches: 1.60.2; Fix fiber link problem (PR#44776 and PR#30880). Tested with 82543GC, 82544EI, 82545EM, 82546GB 82571EB and 82572EI fiber cards. - Don't use the RXCFG interrupt. It's not required and the interrupt is very heavy (a lot of interrupts). Same as {Free,Open}BSD. - Modify wm_tbi_mediachange() to be close to em_setup_fiber_serdes_link() of {Free,Open}BSD. At least, don't forget to set duplex setting. - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. Tested with my own 82545EM card.
1.59	25-Jul-2014	msaitoh	- Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD.
1.58	25-Jul-2014	msaitoh	No functional change: - move NVM related values to the bottom. - sort in register's address.
1.57	11-Jul-2014	msaitoh	Fix some bugs realted to semaphore. This change fixes a problem which was exposed in if_wm.c rev. 1.271. Tested by riastradh@. - Clear the SMBI bit in SWSM register before accessing NVM and PHY in wm_attach(). Same as FreeBSD. - Fix a bug that 82573 doesn't put the hardware semaphore. Same as FreeBSD r256200.
1.56	11-Apr-2014	msaitoh	Fix the definition of CTRL_GIO_M_DIS. This bit is not bit 3 but bit 2. Reported by pooka@.
1.55	29-Dec-2013	msaitoh	branches: 1.55.2; Add support For I354(C2000 interna Ethernet controller): - Add I354 support. - Fix SGMII+MDIO case. SGMII+I2C is not supported yet. - Not tested well. - Sprinkle "XXX" to check later.
1.54	25-Jun-2013	msaitoh	branches: 1.54.2; Sync the wm_enable_mng_pass_thru() function with FreeBSD. Don't check MANC_EN_MAC_ADDR_FILTER bit. Add 82574 and 82583 specific check. This modification may change the setting of WM_F_HAS_MANAGE flag on some machines. Sync the wm_release_manageablilty() fucntion with FreeBSD. Set MANC_ARP_EN. This change enables HW ARP function when entering suspend. When the chip is 82580(ER) or I350, set WM_F_ASF_FIRMWARE_PRES flag and check for the WM_F_ARC_SUBSYS_VALID flag. Same as FreeBSD.
1.53	19-Jun-2013	msaitoh	Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC.
1.52	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.51	18-Apr-2013	msaitoh	Add support I21[01]. Only tested with my own I210-T1 Ethernet Server Adapter.
1.50	13-Feb-2013	msaitoh	No functional change. - Fix comment. - Use macro.
1.49	07-Feb-2013	msaitoh	Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350.
1.48	29-Aug-2012	bouyer	branches: 1.48.2; Make vlan and all ip/ip6 checksum offload work for the I350. On newer devices, when using the legacy TX descriptors, vlan-related flags that were set on the last descriptor of a packet have to be set on the first one. For tso/checksum offloads, a new "advanced" descriptor format has to be used. Change wcd_txdescs to a union defining all types of descriptors (they are all 16-bytes wide). Define a new tx function wm_nq_start(), which handle newer devices. There is some code duplication with wm_start(), but adding support to the existing wm_start() would make it a if () {} else {} maze. This also allows to get rid of some workaround for older chips that are not needed here. Use wm_nq_start() instead of wm_start() for the I350 (this should probably be for all WM_F_NEWQUEUE devices, but I have no hardware but the I350 to test). Call ifp->if_start() instead of wm_start() where is matters. Tested on a I350, and a i80003 (which use the old format), both with and without vlans, with and without checksum offloads.
1.47	25-May-2012	msaitoh	Add support Intel I350 Ethernet.
1.46	20-Dec-2011	dyoung	branches: 1.46.2; Mark all of the Rx descriptor fields 'volatile' so that the compiler will not re-order accesses. Some versions of GCC (such as one in NetBSD 5.x) definitely do re-order reads from these fields if they're not marked volatile.
1.45	20-May-2011	msaitoh	branches: 1.45.4; 1.45.8; Add PCH2 support.
1.44	14-Jul-2010	msaitoh	branches: 1.44.2; - s/TBDA/TDBA/. It stands for Transmit Descriptor Base Address. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. TODO: - ip4csum doesn't work on 82575 and newer devices (reported by Paul Goyette), so we have to fix it.
1.43	25-Jun-2010	msaitoh	fix comment.
1.42	25-Jun-2010	msaitoh	Add some code to support 82580[ER]. Tested on my own I340-T4. - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add some code to support 82580. TODO: - ukphy -> somephy
1.41	25-Jun-2010	msaitoh	The GIO master enable bit in STATUS register is not bit 16 but bit 19. It will fix a problem in the reset sequence on PCI-e chips.
1.40	07-Mar-2010	msaitoh	branches: 1.40.2; - Add code for WOL, ASF, IPMI and Intel AMT. - wm_enable_wakeup() is disabled by default. If you want to use WOL with the Magic Packet, define WM_WOL. - Add the following flags: WM_F_ASF_FIRMWARE_PRESENT WM_F_ARC_SUBSYSTEM_VALID WM_F_HAS_AMT WM_F_HAS_MANAGE WM_F_WOL - Add wm_suspend() and wm_resume(). Give/get the control to/from the firmware. - Need more work for PCH. See wm_enable_phy_wakeup(). - Enable wm_get_hw_control() for 82574 and 82583. - Add Yet another workaround for ICH8. - Add wm_igp3_phy_powerdown_workaround_ich8lan() for power down problem on D3.
1.39	07-Mar-2010	msaitoh	Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working.
1.38	16-Feb-2010	msaitoh	Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems.
1.37	16-Feb-2010	msaitoh	indent, tabify and remove extra spaces.
1.36	04-Feb-2010	msaitoh	branches: 1.36.2; - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops.
1.35	14-Jan-2010	msaitoh	Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.34	12-Jan-2010	msaitoh	- Add i82567LM-2 i82567LM-4 i82567V-3 LAN controller. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. Now we can linkup 1000BaseT on PCH. It seems that we have to do the same work for ICH9.
1.33	11-Jan-2010	msaitoh	Add partial support for PCH. We have to do more work for PCH. Especially I don't know what driver i82578 should be attached to. makephy? or atphy? MII_OUI() says that the PHY is from Attansic (== Atheros). Intel's e1000 driver says that it's close to makphy... I can't link up at 1000BaseT yet...
1.32	07-Jan-2010	msaitoh	Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9.
1.31	07-Jan-2010	msaitoh	bugfixes: - add missing break in wm_reset()... - fix the offset of WMREG_PBS... - fix length of some delay()s in wm_gmii_reset()
1.30	05-Jan-2010	msaitoh	Fix the checking of jumbo frame function though I don't know whether the function wokrs or not... Remove the extra macro definition for the offset 0x1a in EEPROM.
1.29	29-Dec-2009	msaitoh	- Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - Add support for i82583V. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574.
1.28	14-Jul-2009	msaitoh	Some fixes for i80003 and ICH{8,9,10} from e1000 driver and document: Add setting for KABGTXD register for ICH{8,9,10}. ICH9 and ICH10 has no FCAL, FCAH and FCT like ICH8. Add special setting for FCTTV and TCTL_EXT register for i80003 The special setting for TIPG is only for i80003. Some of kumeran settings are only for i80003's bugs. Add some ICH10 fixes.
1.27	07-Apr-2009	msaitoh	Fix about TBI mode. This fix doesn't influence MII mode. - Fix SWDPIN(1)'s polarity on some chips. - Fix flow control stuff (includes PR#32009). - Stop RXCFG storm. It ocours easily. - And more fix about autonego. Tested on PRO/1000F, PRO/1000XF and PRO/1000 MF.
1.26	20-Mar-2009	msaitoh	Examine the management mode and mark DRV_LOAD (for ICH{8,9,10},80003, 8257{1,2,3,4}). Add some error's printf(). Make the bank detect routine into the function.
1.25	10-Mar-2009	msaitoh	add an entry for 82567LF-3. fix the register access for ICH10DO.
1.24	25-Dec-2007	perry	branches: 1.24.10; 1.24.18; 1.24.20; 1.24.24; 1.24.26; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.23	28-Aug-2007	msaitoh	branches: 1.23.6; 1.23.8; 1.23.12; add ICH9 need more work?
1.22	29-Apr-2007	bouyer	branches: 1.22.2; 1.22.6; 1.22.8; Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2
1.21	16-Nov-2006	yamt	branches: 1.21.2; 1.21.4; 1.21.8; 1.21.10; wm: add IPv6 rx tcp/udp checksum offloading.
1.20	21-Oct-2006	bouyer	Add support for the Intel 80003 Gigabit Ethernet controller (found e.g. in newer server chipsets) to wm(4), from the FreeBSD em(4) driver. While there, add a few other Intel Ethernet controller that should work as is. Properly update the RX error and TX collision counters. Add ikphy(4), a driver for the Intel i82563 Kumeran 10/100/1000 Ethernet PHYs
1.19	10-Jun-2006	msaitoh	branches: 1.19.6; 1.19.8; o i82573 doesn't support jumbo frame o set PBA for i8257[123] o set MULA when sc_type == i8257[123]
1.18	10-Jun-2006	uebayasi	KNF.
1.17	10-Jun-2006	msaitoh	fix TX_COLLISION_DISTANCE_FDX and TX_COLLISION_DISTANCE_HDX (swapped). add some register definitions (EERD, EEWR).
1.16	11-Dec-2005	christos	branches: 1.16.4; 1.16.8; 1.16.14; merge ktrace-lwp.
1.15	07-Aug-2005	yamt	fix a typo(?) in an unused macro. (WRX_ST_ISXM -> WRX_ST_IXSM)
1.14	18-Feb-2005	briggs	branches: 1.14.4; 1.14.6; Update the interrupt mitigation support. Actually initialize the interrupt throttling register. Actually initialize the tx and tx absolute interrupt delay timer regs. Update default values for the rx and tx delay timers. Inspired by Zdenek Salvet in PR kern/29373.
1.13	05-Oct-2004	thorpej	branches: 1.13.4; 1.13.6; Largest Tx descriptor buffer size is actually 4K.
1.12	05-Oct-2004	thorpej	- Add some useful constants related to the Packet Buffer Allocation register. - Add the Receive and Transmit Data FIFO registers.
1.11	16-May-2004	thorpej	Add flow control support, based on work by HITOSHI Osada.
1.10	19-Feb-2004	thorpej	branches: 1.10.2; 1.10.4; s/WTC_DTYP_D/WTX_DTYP_D/
1.9	14-Jan-2004	tsutsui	Fix hardware cksum problem on big endian machines. Tested on macppc by HATANO Hiromichi. Note I guess this bug could be fixed only adding one htole32(), but I'd rather clean up endianness handling: - Use htole32() only to accesses against DMA descriptors. - Don't use uint32_t union member with htole32()/le32toh() to access uint8_t/uint16_t descriptors. - Add le32toh() in some DPRINTF. (XXX: strictly speaking, bus_dmamap_sync() is needed for these DPRINTF)
1.8	07-Nov-2003	thorpej	Fix brain'o in PCIX speed detection. kern/23375 (John R. Shannon).
1.7	21-Oct-2003	thorpej	Fix a typo.
1.6	21-Oct-2003	thorpej	Add some additional register definitions needed for newer chips: - Control register shadow - SPI EEPROM stuff
1.5	30-May-2003	matt	branches: 1.5.2; Use proper shift marcos for EXT. This makes the PHY on i83543 get recognized. From Doug Donsbach <dldonsbach at tds dot net>
1.4	18-Oct-2002	yamt	fix typos; "wiseman" is 82542, not 82452.
1.3	08-Aug-2002	thorpej	Add support for the new i82540, i82545, and i82546 Gig-E chips. The i82546 chip has 2 Gig-E interfaces in a single chip. The main difference is these chips have a special handshaking protocol for accessing the EEPROM (due to the shared nature on the i82546).
1.2	14-Jul-2002	thorpej	Correct the description of two bits in the TCP context descriptor.
1.1	28-Mar-2002	thorpej	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Device driver for the Intel i82542, i82542, and i82544 Gigabit Ethernet interfaces.
1.1.8.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.1.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.8.1	28-Mar-2002	jdolecek	file if_wmreg.h was added on branch kqueue on 2002-06-23 17:47:47 +0000
1.1.6.4	20-Jun-2003	msaitoh	Pull up revision 1.5 (requested by matt in ticket #1299): FIx to recognize the PHY on i83543 correctly.
1.1.6.3	21-Nov-2002	tron	Pull up revision 1.4 (requested by yamt in ticket #925): fix typos; "wiseman" is 82542, not 82452.
1.1.6.2	08-Nov-2002	tron	Pull up revision 1.3 (requested by thorpej in ticket #650): Add support for the new i82540, i82545, and i82546 Gig-E chips. The i82546 chip has 2 Gig-E interfaces in a single chip. The main difference is these chips have a special handshaking protocol for accessing the EEPROM (due to the shared nature on the i82546).
1.1.6.1	15-Jul-2002	thorpej	pullup-1-6 ticket #499 (thorpej). Original log message: Correct the description of two bits in the TCP context descriptor.
1.1.4.2	29-Aug-2002	gehenna	catch up with -current.
1.1.4.1	15-Jul-2002	gehenna	catch up with -current.
1.1.2.5	11-Nov-2002	nathanw	Catch up to -current
1.1.2.4	13-Aug-2002	nathanw	Catch up to -current.
1.1.2.3	01-Aug-2002	nathanw	Catch up to -current.
1.1.2.2	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.1.2.1	28-Mar-2002	nathanw	file if_wmreg.h was added on branch nathanw_sa on 2002-04-01 07:46:28 +0000
1.5.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.2.4	19-Oct-2004	skrll	Sync with HEAD
1.5.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.2.1	03-Aug-2004	skrll	Sync with HEAD
1.10.4.3	12-Jan-2005	he	Pull up revision 1.13 (requested by thorpej in ticket #898): Largest Tx descriptor buffer size is actually 4K.
1.10.4.2	08-Jan-2005	jdc	Fix pull up merge error. Pointed out by Martin J. Laubach.
1.10.4.1	07-Jan-2005	jdc	Pull up revision 1.12 (requested by thorpej in ticket #897) - Add some useful constants related to the Packet Buffer Allocation register. - Add the Receive and Transmit Data FIFO registers.
1.10.2.2	08-Jan-2005	jdc	Fix pull up merge error. Pointed out by Martin J. Laubach.
1.10.2.1	07-Jan-2005	jdc	Pull up revision 1.12 (requested by thorpej in ticket #897) - Add some useful constants related to the Packet Buffer Allocation register. - Add the Receive and Transmit Data FIFO registers.
1.13.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.4.1	29-Apr-2005	kent	sync with -current
1.14.6.4	21-Jan-2008	yamt	sync with head
1.14.6.3	03-Sep-2007	yamt	sync with head.
1.14.6.2	30-Dec-2006	yamt	sync with head.
1.14.6.1	21-Jun-2006	yamt	sync with head.
1.14.4.4	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1863): sys/dev/pci/if_wm.c 1.137-1.138,1.141,1.144-1.146 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 via patch share/man/man4/wm.4 1.18-1.19 Fix typo. Fix EEPROM reload sequence. Add ICH9 support. Workaround for 82541 Errata 29 and 82547 Errata 28: These devices have to reset the PHY before reset the MAC. Reported and tested by salo.
1.14.4.3	14-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1770): share/man/man4/wm.4: revision 1.17 sys/dev/pci/if_wmreg.h: revision 1.22 sys/dev/pci/if_wm.c: revision 1.139 via patch sys/dev/pci/if_wm.c: revision 1.140 via patch sys/dev/pci/pcidevs: patch Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2 Restore $NetBSD$ tag. Note ICH8 LAN support.
1.14.4.2	31-Mar-2007	bouyer	pullup the following revisions (requested by msaitoh in ticket 1681): sys/dev/pci/if_wm.c 1.104-1.105, 1.116-1.121, 1.127,1.133-1.134 via patch sys/dev/pci/if_wmreg.h 1.17-1.20 sys/dev/pci/pcidevs patch sys/dev/mii/igphy.c 1.11 sys/dev/mii/makphy.c 1.20, 1.23 sys/dev/mii/ikphy.c patch sys/dev/mii/ikphyreg.h patch sys/dev/mii/miidevs 1.68 sys/dev/mii/files.mii 1.39 sys/arch/i386/conf/GENERIC 1.788-1.789 via patch sys/arch/i386/conf/GENERIC.MPACPI patch sys/arch/i386/conf/GENERIC_LAPTOP 1.209 via patch sys/arch/i386/conf/INSTALL 1.291 via patch sys/arch/i386/conf/INSTALL_LAPTOP 1.104 via patch sys/arch/i386/conf/XEN2_DOM0 1.13 via patch share/man/man4/wm.4 1.14-1.16 Add support for many cards (include PCI-express based chips). Many bug fixes about auto negotiations (PR#30078, PR#30490, PR#30906, PR#33429 and PR#35386). Fix media link issues with fiber-based card (PR#35797).
1.14.4.1	28-Aug-2005	tron	Pull up following revision(s) (requested by yamt in ticket #699): sys/dev/pci/if_wmreg.h: revision 1.15 fix a typo(?) in an unused macro. (WRX_ST_ISXM -> WRX_ST_IXSM)
1.16.14.1	19-Jun-2006	chap	Sync with head.
1.16.8.1	26-Jun-2006	yamt	sync with head.
1.16.4.1	09-Sep-2006	rpaulo	sync with head
1.19.8.2	10-Dec-2006	yamt	sync with head.
1.19.8.1	22-Oct-2006	yamt	sync with head
1.19.6.1	18-Nov-2006	ad	Sync with head.
1.21.10.1	11-Jul-2007	mjf	Sync with head.
1.21.8.2	09-Oct-2007	ad	Sync with head.
1.21.8.1	27-May-2007	ad	Sync with head.
1.21.4.1	07-May-2007	yamt	sync with head.
1.21.2.2	29-Aug-2007	liamjfoy	Pull up following revision(s) (requested by masanobu in ticket #850): sys/dev/pci/if_wm.c 1.137 sys/dev/pci/if_wm.c 1.138 sys/dev/pci/if_wm.c 1.141 sys/dev/pci/if_wm.c 1.144 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphy.c 1.12 sys/dev/mii/miidevs 1.72 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/wm.4 1.18 Fix for 82541 and 82547's reset bug. Modify wm_reset() to make some device stable. Add ICH9 support to wm. Add I82566 support to igphy. Pulled done via patch.
1.21.2.1	03-May-2007	snj	branches: 1.21.2.1.2; Pull up following revision(s) (requested by bouyer in ticket #598): share/man/man4/wm.4: revision 1.17 sys/dev/pci/if_wmreg.h: revision 1.22 doc/CHANGES: revision 1.836 sys/dev/pci/if_wm.c: revision 1.139 sys/dev/pci/if_wm.c: revision 1.140 Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2 Restore $NetBSD$ tag. Note ICH8 LAN support. Note ICH8 LAN support.
1.21.2.1.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.22.8.2	09-Jan-2008	matt	sync with HEAD
1.22.8.1	06-Nov-2007	matt	sync with HEAD
1.22.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.22.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.23.12.1	02-Jan-2008	bouyer	Sync with HEAD
1.23.8.1	26-Dec-2007	ad	Sync with head.
1.23.6.1	18-Feb-2008	mjf	Sync with HEAD.
1.24.26.1	19-Jul-2009	snj	branches: 1.24.26.1.2; Pull up following revision(s) (requested by msaitoh in ticket #862): sys/dev/pci/if_wm.c: revisions 1.168, 1.170, and 1.173 via patch sys/dev/pci/if_wmreg.h: revision 1.27 Fixes serious three bugs. 1) On i82563, the em driver says that the ready bit in the MDIC register may be incorrectly set. Insert delay(200) like the em driver. Fixes PR#41014 2) Add workaround for 82543GC. We need to force speed and duplex on the MAC equal to what the PHY speed and duplex configuration is. Fixes PR#36430. 3) Fix many problems and panic on TBI's cards (includes PR#32009).
1.24.26.1.2.1	21-Apr-2010	matt	sync to netbsd-5
1.24.24.2	23-Jul-2009	jym	Sync with HEAD.
1.24.24.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.24.20.9	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1962): sys/dev/pci/if_wm.c: revisions 1.259-1.266, 1.269, 1.271, 1.273-1.274, 1.277-1.278, 1.282, 1.284-1.285, 1.287, 1.290, 1.294, 1.297-1.298, 1.300-1.301, 1.304-1.307, 1.310, 1.312-1.314, 1.316 via patch sys/dev/pci/if_wmreg.c: revisions 1.54-1.57, 1.59-1.60, 1.62, 1.64-1.66 via patch sys/dev/pci/if_wmvar.c: revisions 1.17, 1.20-1.21 via patch sys/dev/pci/pcidevs: revisions 1.1172, 1.1195, 1.1201 via patch - Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. - Fix BMC related bugs. - Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth. - Add support for I354 DH89xxCC and some new I218 devices. - Fix definition of CTRL_GIO_M_DIS bit. - Insert completion barrier between register write and delay(). - Bump max TX DMA size to avoid pathological condition with TSO. From dyoung. - Fix semaphore related bugs. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - It's not required to print "failed to detect NVM bank" with aprint_error_dev(). Use DPRINTF(). Same as {Free,Open}BSD. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Remove extra debug message. - Remove unused variable. - Remove a duplicated error message. - Cleanup comments. - Fix debug message.
1.24.20.8	07-Sep-2013	bouyer	sys/dev/pci/if_wm.c 1.246-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 sys/dev/pci/pcidevs 1.1152-1.1153 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #1867]
1.24.20.7	19-Jun-2013	bouyer	Pullup the following revisions via patch, requested by msaitoh in ticket #1850: sys/dev/pci/if_wm.c 1.201, 1.203-1.204, 1.207-1.212, 1.215, 1.217-1.218, 1.220-1.223, 1.228, 1.232-245 sys/dev/pci/if_wmreg.h 1.40-1.45, 1.47-1.48 sys/dev/pci/if_wmvar.h 1.11-1.13 sys/dev/pci/pcidevs 1.1074, 1.1077, 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.6 sys/dev/mii/ihphy.c 1.1-1.2 sys/dev/mii/ihphyreg.h 1.1 sys/dev/mii/inbmphyreg.h 1.3 sys/dev/mii/files.mii 1.47 via patch sys/dev/mii/miidevs 1.97 and 1.100 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/arch/i386/conf/ALL 1.280 sys/arch/i386/conf/GENERIC 1.1001 sys/arch/i386/conf/INSTALL_FLOPPY 1.11 sys/arch/i386/conf/XEN2_DOM0 patch sys/arch/amd64/conf/GENERIC 1.293 sys/arch/amd64/conf/XEN3_DOM0 1.61 share/man/man4/wm.4 1.21-1.24 Apply almost all fixes and improvements from netbsd-6 except for the rev. 1.196's iqdrops' change. - Add the detach code. - Add code for WOL, ASF, IPMI and Intel AMT. WOL is disabled by default - Add Yet another workaround for ICH8. - 82576 is dual port, so check the FUNCID and increment the MAC address for the 2nd port. - Fix the names of 82577L[MC] LAN controllers (for mobile). - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add 82580 support. - 82571 quirk. Only 82571 shares port 0 of EEMNGCTL_CFGDONE. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. - Fix some register names. No functional change. - Omit U+00AE "REGISTERED SIGN" in a product name due to its non-ASCII nature. - Stop wm(4) from needlessly resetting when you add or delete a vlan(4). - Fix MAC address check on 8257[156] and 80003 case. Some cards have non 0xffff pointer but those don't use alternative MAC address in reality. So we check whether the broadcast bit is set or not like Intel's e1000 driver. Fixes PR kern/44072 reported by Jean-Yves Moulin. - Add PCH2(and 82579) support. Fixes PR#46487 - Add yet another 82567V support. - Add ICH10+HANKSVILL support. - Add support Intel I350 Ethernet. - Make vlan and all ip/ip6 checksum offload work for the I350. - Fix compile error with WM_DEBUG. - Fix a bug that PHY isn't set to low-power mode on PCH and PCH2. - Add WM_DEBUG_NVM. If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. - Skip 64bit BAR correctly. - Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350. - Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542. - Style fix. Fix typo in comment. Fix comments. Add comments.
1.24.20.6	05-Jan-2012	sborrill	Pull up the following revisions(s) (requested by dyoung in ticket #1703): sys/dev/pci/if_wmreg.h: revision 1.46 Mark all of the wm(4) Rx descriptor fields 'volatile' so that the compiler will not re-order accesses. Some versions of GCC (such as one in NetBSD 5.x) definitely do re-order reads from these fields if they're not marked volatile.
1.24.20.5	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1358): sys/dev/pci/if_wm.c 1.196-1.199,1.202,1.205 sys/dev/pci/if_wmvar.h 1.9 sys/dev/pci/if_wmreg.h 1.36-1.39 sys/dev/pci/pcireg.h 1.61-1.64 sys/dev/pci/pcidevs 1.1023 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen mii/igphy.c 1.21 mii/igphyvar.h 1.1 mii/inbmphyreg.h 1.2 - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems. - Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working. - (pcireg.h) Add PCIe config register definitions. - Note that the changes to count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops in rev. 1.196 of if_wm.c is not pulled up.
1.24.20.4	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.24.20.3	23-Dec-2009	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: 1.176-1.179, 1.181-1.183 sys/dev/pci/if_wmreg.h: 1.28 sys/dev/pci/if_wmvar.h: 1.1-1.4 sys/dev/mii/igphy.c: 1.18-1.20 via patch Many bugfixes: - Some fixes for i80003 and ICH{8,9,10} from e1000 driver and document: - Add setting for KABGTXD register for ICH{8,9,10}. - ICH9 and ICH10 has no FCAL, FCAH and FCT like ICH8. - Add special setting for FCTTV and TCTL_EXT register for i80003 - The special setting for TIPG is only for i80003. - Some of kumeran settings are only for i80003's bugs. - Add some ICH10 fixes. - Fix the bug that another lock mechanism is used to access Kumeran registers on i80003 and ICHs. - Fix yet another i80003 ONLY workaround. The code to modifing TIPG register is only for i80003. - Set the Re-Transmit on Late Collision(RTLC) flag for all devices. - Fix a typo in a printf message. - If the difference bettween last flag and new flag is only IFF_PROMISC or IFF_ALLMULTI, set multicast filter only to prevent link down. Tested by Mark Davies and me. Fixes PR#29126 for wm. - Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line. - Fix igphy's 82566 support. - Patch for the DSP code is only for 8254[17] and we have to apply the different patches between rev. 1 and rev. 2. - The workaround for analog fuse is only for 82547 rev. 1. - The workaround for smartspeed is only for 8254[17] - Sync with Intel's original em driver: - Check PCI-X mode as e1000 driver. - Add dspcode for igp3 and use it when the EEPROM isn't available. - Add some delays. - Stop the PHY transmitter before patching the DSP code and restart it after writing. - Save and restore register 0x2f5b.
1.24.20.2	21-May-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #711): sys/dev/pci/pcidevs: revisions 1.975, 1.981, 1.982 via patch sys/dev/pci/if_wm.c: revisions 1.164, 1.167, 1.173, 1.174 via patch sys/dev/pci/if_wmreg.h: revisions 1.25, 1.27 via patch Add Intel 82567LM_3 ethernet -- Add i82567LM-3 -- add i82567LF-3 LAN Controller -- add an entry for 82567LF-3. fix the register access for ICH10DO. -- Fix about TBI mode. This fix doesn't influence MII mode. -- - Fix panic in mediachange. - Fix SWDPIN(1)'s polarity on some chips. - Fix flow control stuff (includes PR#32009). - Stop RXCFG storm. It ocours easily. - And more fix about autonego. -- add 82801J_D_BM_LF (ICH10) -- Reload sc_ctrl in wm_reset(). Add an ICH10 entry. Remove some obsolete comments.
1.24.20.1	03-May-2009	snj	Pull up following revision(s) (requested by spz in ticket #621): sys/dev/pci/if_wm.c: revisions 1.168, 1.169 via patch sys/dev/pci/if_wmreg.h: revision 1.26 via patch On i82563, the em driver says that the ready bit in the MDIC register may be incorrectly set. Insert delay(200) like the em driver. -- Examine the management mode and mark DRV_LOAD (for ICH{8,9,10},80003, 8257{1,2,3,4}). Add some error's printf(). Make the bank detect routine into the function.
1.24.18.1	28-Apr-2009	skrll	Sync with HEAD.
1.24.10.4	11-Aug-2010	yamt	sync with head.
1.24.10.3	11-Mar-2010	yamt	sync with head
1.24.10.2	18-Jul-2009	yamt	sync with head.
1.24.10.1	04-May-2009	yamt	sync with head.
1.36.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.36.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.40.2.3	31-May-2011	rmind	sync with head
1.40.2.2	05-Mar-2011	rmind	sync with head
1.40.2.1	03-Jul-2010	rmind	sync with head
1.44.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.45.8.2	02-Jun-2012	mrg	sync to latest -current.
1.45.8.1	18-Feb-2012	mrg	merge to -current.
1.45.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.45.4.2	30-Oct-2012	yamt	sync with head
1.45.4.1	17-Apr-2012	yamt	sync with head
1.46.2.9	06-May-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1366): sys/dev/pci/if_wm.c: 1.281, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.387, 1.389 via patch sys/dev/pci/if_wmreg.h: 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: 1.22-1.23, 1.25-1.30 via patch sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 Sync wm(4) as of if_wm.c rev 1.389 except SERDES, MSI/MSI-X, multiqueue and NET_MPSAFE: - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvctl -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.46.2.8	16-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c: revision 1.312 via patch sys/dev/pci/if_wmreg.h: revision 1.66 via patch - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment.
1.46.2.7	04-Dec-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: revisions 1.271, 1.273-1.274, 1.277-1.278, 1.280, 1.282, 1.284-1.285, 1.287, 1.293-1.294, 1.297-1.298, 1.300-1.301, 1.305-1.307 via patch sys/dev/pci/if_wmreg.h: revisions 1.57-1.62, 1.64-1.65 via patch sys/dev/pci/if_wmvar.h: revisions 1.19, 1.21 via patch - Acquire SW semaphore in wm_get_swsm_semaphore(). - Fix some bugs realted to semaphore. This change fixes a problem which was exposed in if_wm.c rev. 1.271. Tested by riastradh@. - Clear the SMBI bit in SWSM register before accessing NVM and PHY in wm_attach(). Same as FreeBSD. - Fix a bug that 82573 doesn't put the hardware semaphore. Same as FreeBSD r256200. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). Tested with 82543GC, 82544EI, 82545EM, 82546GB 82571EB and 82572EI fiber cards. - Don't use the RXCFG interrupt. It's not required and the interrupt is very heavy (a lot of interrupts). Same as {Free,Open}BSD. - Modify wm_tbi_mediachange() to be close to em_setup_fiber_serdes_link() of {Free,Open}BSD. At least, don't forget to set duplex setting. - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Don't check SWSM_SMBI bit if WM_F_LOCK_SWSM isn't set. Fix a problem when using vmware with e1000"e". With e1000e which is regarded as 82574L, wm_gmii_init() fails with "could not acquire SWSM SMBI" message without this change. This problem doesn't occur with real 82574L card. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Calculate NVM word size correctly. - Determine timeout value based on the NVM word size. - It's not required to print "failed to detect NVM bank" message. Only print while debugging. Same as {Free,Open}BSD. - Add some new I218 devices. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Move some NVM related macros from if_wm.c to if_wmreg.h. - Sort definitions in if_wmreg.h - move NVM related values to the bottom. - sort in register's address' order. - Simplify wm_read_mac_addr(). - Fix debug message. - Add missing prototypes. - Rename some functions for consistency and clarify. - Rename some macros for consistency. - Remove a duplicated error message. - Fix typo in comment. - Cleanup comments. - KNF.
1.46.2.6	09-Nov-2014	martin	Pullup the following revisions, requested by msaitoh in ticket #1188: sys/dev/pci/pcidevs 1.1172 sys/dev/pci/if_wm.c 1.263-1.266 via patch sys/dev/pci/if_wmreg.h 1.55-1.56 sys/dev/pci/if_wmvar.h 1.17 share/man/man4/wm.4 1.26-1.27 and 1.29 via patch - Add I354 support. - Insert completion barrier between register write and delay(). - Fix the definition of CTRL_GIO_M_DIS. This bit is not bit 3 but bit 2. - Cleanup
1.46.2.5	29-Jul-2013	jdc	Pull up revisions: src/sys/dev/pci/if_wm.c revisions 1.259,1.260,1.261,1.262 src/sys/dev/pci/if_wmreg.h revision 1.54 (requested by msaitoh in ticket #918). Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. Sync the wm_enable_mng_pass_thru() function with FreeBSD. Don't check MANC_EN_MAC_ADDR_FILTER bit. Add 82574 and 82583 specific check. This modification may change the setting of WM_F_HAS_MANAGE flag on some machines. Sync the wm_release_manageablilty() fucntion with FreeBSD. Set MANC_ARP_EN. This change enables HW ARP function when entering suspend. When the chip is 82580(ER) or I350, set WM_F_ASF_FIRMWARE_PRES flag and check for the WM_F_ARC_SUBSYS_VALID flag. Same as FreeBSD. Move the location of wm_check_mng_mode() and wm_get_wakeup() in wm_attach(). Those functions access EEPROM, so they have to call after identifying EEPROM access type. This modification may change the behavior of BMC(IPMI). Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth.
1.46.2.4	14-Jul-2013	riz	Apply changes (requested by msaitoh in ticket #907): sys/dev/pci/if_wm.c 1.238, 1.244-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.50-1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 Various fixes to wm(4): Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Change style, add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #907]
1.46.2.3	18-Feb-2013	riz	Pull up following revision(s) (requested by msaitoh in ticket #824): sys/dev/pci/if_wm.c: revision 1.240 sys/dev/pci/if_wm.c: revision 1.241 sys/dev/pci/if_wm.c: revision 1.242 sys/dev/pci/if_wmreg.h: revision 1.49 - Add WM_DEBUG_NVM. - If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. Skip 64bit BAR correctly. I don't know if this bug causes a real problem. Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350.
1.46.2.2	03-Sep-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #534): sys/dev/pci/if_wmreg.h: revision 1.48 sys/dev/pci/if_wm.c: revision 1.232 sys/dev/pci/if_wm.c: revision 1.233 sys/dev/pci/if_wm.c: revision 1.234 Make vlan and all ip/ip6 checksum offload work for the I350. On newer devices, when using the legacy TX descriptors, vlan-related flags that were set on the last descriptor of a packet have to be set on the first one. For tso/checksum offloads, a new "advanced" descriptor format has to be used. Change wcd_txdescs to a union defining all types of descriptors (they are all 16-bytes wide). Define a new tx function wm_nq_start(), which handle newer devices. There is some code duplication with wm_start(), but adding support to the existing wm_start() would make it a if () {} else {} maze. This also allows to get rid of some workaround for older chips that are not needed here. Use wm_nq_start() instead of wm_start() for the I350 (this should probably be for all WM_F_NEWQUEUE devices, but I have no hardware but the I350 to test). Call ifp->if_start() instead of wm_start() where is matters. Tested on a I350, and a i80003 (which use the old format), both with and without vlans, with and without checksum offloads. Enable VLAN hardware tagging on all chips that have the new queue mechanism. Tested with 82575{EB,GB}, 82576, 82580, I350 and ICH9. Shut up gcc about some uninitialized variables.
1.46.2.1	28-Jun-2012	riz	branches: 1.46.2.1.2; Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/if_wm.c 1.228 sys/dev/pci/if_wmreg.h 1.47 sys/dev/pci/if_wmvar.h 1.13 sys/dev/pci/pcidevs 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/wm.4 1.22-1.25 Add support for Intel I350 Ethernet. Update document.
1.46.2.1.2.1	01-Nov-2012	matt	sync with netbsd-6-0-RELEASE.
1.48.2.4	03-Dec-2017	jdolecek	update from HEAD
1.48.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.48.2.2	23-Jun-2013	tls	resync from head
1.48.2.1	25-Feb-2013	tls	resync with head
1.54.2.1	18-May-2014	rmind	sync with head
1.55.2.1	10-Aug-2014	tls	Rebase.
1.60.2.10	09-Nov-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1648): sys/dev/mii/inbmphyreg.h 1.11 sys/dev/pci/if_wm.c 1.586-1.588, 1.590-1.596 via patch sys/dev/pci/if_wmreg.h 1.108 - rename tu event counter to txunderrun. - Try m_defrag() to reduce the number of DMA segment if bus_dmamap_load_mbuf() returned EFBIG. When m_defrag() is called, txqNNdefrag event counter is incremented. If the 2nd try of bus_dmamap_load_mbuf() failed, txqNNtoomanyseg event counter is incremented. - Reduce the max number of DMA segments from 256 to 64 (it's the same value as other BSD's (EM_MAX_SCATTER) and more than before if_wm.c rev. 1.75's value (40)) because we do m_defrag() now. - 82574 and newer document says the status field has neither EC (Excessive Collision) bit nor LC (Late Collision) bit (reserved), so don't check the bit. - Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This workaround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. - Fix a PCH2 specific bug that wrong PHY register value can be read when boot. Same as FreeBSD and Linux. - After writing MDIC register, don't read quickly the same register but do delay(50). Same as other OSes. - Add missing wm_gate_hw_phy_config_ich8lan(false) in wm_phy_post_reset() on PCH2. wm_gate_hw_phy_config_ich8lan(true) is called in wm_reset(), so wm_gate_hw_phy_config_ich8lan(false) should be called after reset in wm_phy_post_reset(). - On PCH2, set the phy config counter to 50msec after (PHY) reset. - KNF & Use macro.
1.60.2.9	11-Aug-2018	martin	Pull up the following, requested by msaitoh in ticket #1628: share/man/man4/wm.4 1.40 via patch sys/dev/mii/ihphyreg.h 1.2 sys/dev/mii/inbmphyreg.h 1.10 sys/dev/pci/if_wm.c 1.504, 1.506, 1.510-1.535, 1.539-1.540, 1.546, 1.548, 1.551-1.552, 1.558, 1.565-1.573, 1.575, 1.579, 1.582, 1.584 via patch sys/dev/pci/if_wmreg.h 1.99-1.103, 1.106-1.107 via patch sys/dev/pci/if_wmvar.h 1.34-1.39 via patch sys/dev/pci/pcidevs 1.1327 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/pci/pcireg.h patch Sync wm(4) up to 2018/08/08 except MSI/MSI-X and NET_MPSAFE: - remove extra "+" - Fix a bug that non-GMII devices don't send a routing message when the link status is changed. - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes. - If a interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Print sc_flags with snprintb(). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange(). - Add missing "else" in wm_nvm_release(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - If TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function are required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Use new nvm.{acquire,release}() for semaphore. - Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register value from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally on I21[01]". PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Print NVM offset and word count when EERD polling failed. - On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux). - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Fix typo in comment. Reported by Shinichi Doyashiki in PR#52885. - Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki. - CID-1427779: Fix uninitialized variables. - Fix a bug that wm_pll_workaround_i210() is not called when a) Chip is I211 or b) Chip is I210 and it uses iNVM (not FLASH). - Do wm_reset_mdicnfg_82580() on 82580 only. - Fix FLASH access on PCH_SPT and newer. Their FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it must be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. - Make wm_nvm_valid_bank_detect_ich8lan() the same as other OSes. - If the extended configuration size in the EXTCNFSIZE register is 0, don't continue in wm_init_lcd_from_nvm(). - Add PCH_CNP support (I219 with Intel 300 series chipset). - Enable I219 support. - I354 uses an external PHY, so don't use wm_set_eee_i350(). - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. - Fix a bug that 82542 misunderstand fiber's signal detection. - Add debug printf()s. - Update comment. - Rename functions and variables. - Add diagnostic code. - Sort registers. - Lowercase hexadecimal values. - KNF.
1.60.2.8	09-Mar-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #1372): sys/dev/pci/if_wm.c: revisions 1.462, 1.464-1.465 1.474 via patch sys/dev/pci/if_wmreg.h: revision 1.97 sys/dev/pci/if_wmvar.h: revision 1.33 - Set IPV6EXDIS bit in RFCTL register because of an Errata on 82575 and newer devices. - Linux and FreeBSD defines 0x10f5 as E1000_DEV_ID_ICH9_IGP_M_AMT. In reality, This is not IGP but BM. Add new case to identify PHY type device. Fixes PR#51924 reported byJarle Greipsland. - Use new wm_gmii_setup_phytype() function to setup sc_phytype and mii_{read\|write}reg. This change improves detection of PHY type. - Fix typo in comment.
1.60.2.7	18-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1327): sys/dev/pci/if_wm.c: revision 1.457 sys/dev/pci/if_wmreg.h: revision 1.94 - Change to use 2500Base-KX correctly on C2000(I354). It worked, but the output of ifconfig and if_baudrate was not good. Tested by nils@: - The STATUS_TBIMODE bit in the STATUS register is deleted since 82575, so check for 82575 and newer first and then check for old devices. - Check the 2P5_SKU and 2P5_SKU_OVER bit for KX. - Set IFM_2500_SX instead of IFM_1000_SX for 2.5G. - Check SERDES's speed directly from the PCS layer (PCS_LSTS register) for old devices. - Style fix.
1.60.2.6	12-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1302): sys/dev/mii/igphyreg.h: revisions 1.7-1.10 sys/dev/mii/ikphyreg.h: revisions 1.3 sys/dev/mii/inbmphyreg.h: revisions 1.4-1.9 sys/dev/mii/mii.h: revisions 1.19-1.20 sys/dev/pci/if_wm.c: revisions 1.390, 1.392-1.395, 1.397, 1.419-1.425, 1.427-1.428, 1.430-1.435, 1.437-1.453 via patch sys/dev/pci/if_wmreg.: revisions 1.89-1.93 via patch sys/dev/pci/if_wmvar.h: revisions 1.31-1.32 Update wm(4) up to if_wm.c rev. 1.453 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add I219 support. It's not stable so it's disabled by default. - wm_gate_hw_phy_config_ich8lan() is for younger than PCH2. - Drop the host wakeup bit after resetting PHY on PCH and newer devices. - Increase delay while toggling LANPHYPC - Move call of wm_reset() in wm_attach() after setting PHY and NVM related flags because those flags are used in wm_reset(). - Use mutex for NVM access on ICH8 and newer devices. Same as FreeBSD. - Rewrite PHY related lock stuff. Almost the same as FreeBSD. This change will fix a bug that PHY read/write fail on some cases. - Increase delay in wm_phy_resetisblocked(). Same as FreeBSD. - Use semaphore in wm_hv_phy_workaround_ich8lan() and wm_k1_gig_workaround_hv() - Use wm_gii_mdic_readreg/writereg() in wm_access_phy_wakeup_reg_bm() because these functions are called with taking lock. - 82567V_3 is BME1000_E_2(bm). Tested with Advantech AIMB-212 1st Ethernet port. - Use wm_gmii_82544_{read,write}reg() on non-82567 ICH8, 9 and 10. - Remove an 82578 workaround which was for PCH rev < 3. FreeBSD removed this workaround in r228386. - Add an 82578 workaround which is for PHY rev < 2. From FreeBSD and Linux. - Fix wm(4) input drop packet counter. WMREG_RNBC is incremented when there is no available buffers in host memory. However, ethernet controller can receive packets in such case if there is space in phy's FIFO. That is, ethernet controller drops packet only if there is no available buffers and there is no space in phy's FIFO. So, the number of dropped packets should be added WMREG_MPC only. - Use MII_ADDRMASK. - Define WMPHY_I217, WMPHY_VF and WMPHY_210. - Use BME1000_PHY_PAGE_SELECT in wm_gmii_bm_{read,write}reg(). This change has no effect because GG82563_PHY_PAGE_SELECT and BME1000_PHY_PAGE_SELECT have the same value. - Fix PHY access on 82567(ICH8 or ICH10), 82574 and 82583: - Use wm_gmii_bm_{read,write}reg() on 82574 and 82573. - Issue page select correctly on BM PHYs. - Fix workaround which did dummy read BM_WUC register. This code was changed to drop BM_WUC_HOST_WU_BIT of BM_PROT_GEN_CFG register in FreeBSD r228386. The code was added rev. 1.149, but the location was not the best. - wm_gmii_hv_{read/write}reg(): USE PHY address 1 for some special registers. - Add check code for an 82578 workaround. Not completed yet. - wm_release_hw_control(): Remove extra line. No any effect. - Add "10/100" into non-gigabit devices' name. - Call wm_enable_wakeup() in wm_detach() and wm_suspend(). Now wake on lan works on Thinkpad X61(ICH8). - Fix wm_access_phy_wakeup_reg_bm(). This change has no effect because this function is used for WUC register and our driver currenlty doesn't access to it. - Call wm_enable_phy_wakeup() on PCH2 and newer, too. Now these devices can do WOL. Tested with Thinkpad X220(PCH2). - Set CTRL_MEHE correctly (PCH_{LPT,SPT} only). - Add three workarounds for PCH_{LPT,SPT}. - Fix a bug that 8257[56], 82580, I35[04] and I21[01] didn't use wm_{get,release}_hw_control() correctly. - Sync wm_smbustopci() with Linux and FreeBSD. This change effects PCH and newer devices. - Move the location of wm_smbustopci() call. - Fix flag check in wm_get_wakeup() - 8254[17] and 8257[124] should not set WM_F_ARC_SUBSYS_VALID. - Add missing WM_T_82541_2 and WM_T_82547_2. - Fix WOL related setting of the WUC register for other than PCH* in wm_enable_wakeup(). Tested with 82567V(ICH8) and 82583V. - Use common MII_ADDRMASK. - igphy(4): No binary change: - s/IGPPHY/IGPHY/ - Fix the definition of PLHR_VALID_CHANNEL_* - Fix the definition of MSE_CHANNEL_* - Add MII_IGPHY_POWER_MGMT. - Add some KASSERT. - Add comment. Modify comment. - Add debug code.
1.60.2.5	26-Feb-2016	snj	branches: 1.60.2.5.2; Pull up following revision(s) (requested by msaitoh in ticket #1102): sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 sys/dev/pci/if_wm.c: revisions 1.308, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.389 via patch sys/dev/pci/if_wmreg.h: revisions 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: revisions 1.22-1.23, 1.25-1.30 via patch Sync wm(4) as of if_wm.c rev 1.389 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add C2000 KX and 2.5G support. - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Fix a lot of bugs to make 82575 and newer SERDES based systems work. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvct -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.60.2.4	10-Jun-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #833): sys/dev/pci/if_wm.c: revisions 1.322, 1.323 sys/dev/pci/if_wmreg.h: revision 1.72 - Currently, WM_F_EEE bit is not set on all chips. It's intened to not to use all of EEE fuctions but wm_set_eee_i350() leaves IPCNFG_10BASE_TE bit and it causes link negotiation problem on some old switches. Disable 10BASE-Te function, too. - Call wm_set_eee_i350() on some chips, too. -- Fix a bug that flags related to semaphore were incorrectly checked in wm_kmrn_{readreg,writereg}. i80003, ICH* and PCH* had this problem.
1.60.2.3	19-May-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #784): sys/dev/pci/if_wm.c: revision 1.321 sys/dev/pci/if_wmreg.h: revision 1.71 sys/dev/pci/if_wmvar.h: 1.24 via patch Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines/NICs.
1.60.2.2	18-Mar-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #614): sys/dev/pci/if_wm.c: revision 1.311-1.313 sys/dev/pci/if_wmreg.h: revision 1.66 Fix a bug that sc_mediatype is evaluated incorrectly. The real problem might be occured only on fiber and serdes cases. ##### - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Add comment. - Fix typo in comment. ##### Fix a bug that the first access to NVM is failed on 8254[17] which use SPI EEPROM. Observed on Dell PowerEdge [12]850. Thanks Tom Ivar Helbekkmo for debugging.
1.60.2.1	07-Nov-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #190): sys/dev/pci/pcidevs 1.1201-1.1202 via patch sys/dev/pci/if_wm.c 1.291-1.301, 1.304, 1.306-1.307 via patch sys/dev/pci/if_wmreg.h 1.61-1.65 via patch sys/dev/pci/if_wmvar.h 1.21 via patch share/man/man4/wm.4 1.31 via patch - Add Internal SERDES mode support newer than or equal to 82575. - Add new I218 devices. - Add DH89xxCC device. - Add some old devices. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Don't print "failed to detect NVM bank" message. - Delete 82580ER related code. - Fix typo in comment. - Fix debug message. - Cleanup
1.60.2.5.2.2	13-Mar-2017	skrll	Sync with netbsd-7-1-RELEASE
1.60.2.5.2.1	18-Jan-2017	skrll	Sync with netbsd-5
1.65.2.8	28-Aug-2017	skrll	Sync with HEAD
1.65.2.7	05-Feb-2017	skrll	Sync with HEAD
1.65.2.6	05-Dec-2016	skrll	Sync with HEAD
1.65.2.5	29-May-2016	skrll	Sync with HEAD
1.65.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.65.2.3	22-Sep-2015	skrll	Sync with HEAD
1.65.2.2	06-Jun-2015	skrll	Sync with HEAD
1.65.2.1	06-Apr-2015	skrll	Sync with HEAD
1.89.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.89.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.94.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.98.6.19	29-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1938: sys/dev/pci/if_wm.c 1.792,1.794-1.798 via patch sys/dev/pci/if_wmreg.h 1.131 - Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. - Drop frames if the RX descriptor ring has no room on multiqueue system. - Improve dmesg output. - Print RX packet buffer size. - Fix the upper 16bit of Image Unique ID(EtrackID). - Fix comment.
1.98.6.18	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1915: sys/dev/pci/pcidevs 1.1497 sys/dev/pci/if_wm.c 1.689,1.790-1.791 via patch sys/dev/pci/if_wmreg.h 1.120 via patch sys/dev/pci/if_wmvar.h 1.51 - Use 12K for packet buffer for jumbo frame on PCH2 and newer. - Add new workaround for Tiger Lake and newer to avoid packet loss. - Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.98.6.17	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1847: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.98.6.16	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1795: sys/dev/pci/if_wm.c 1.764-1.767 via patch sys/dev/pci/if_wmreg.h 1.128 - Workaround for some hypervisor environments. The environments cannot stop e1000 interrupt immediately. - Rename nq_txdesc' member "nqrx_ctx" to "nqtx_ctx". No functional change. - Add comment. Modify comment. - KNF.
1.98.6.15	07-Sep-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1759: sys/dev/pci/if_wm.c 1.741-1.749,1.753-1.757, 1.762 via patch sys/dev/pci/if_wmreg.h 1.126-1.127 - Fix I219 workaround in wm_flush_desc_rings(). - Add more statistics counters. - To avoid releasing mutex temporally, use new wm_set_mdio_slow_mode_hv_locked(). - No functional changes: - Turn a locking botch (shouldn't drop lock on error) into a KASSERT in wm_deferred_start_locked(). - Remove unneeded header inclusion. - Use __BIT() a little. - Modify comment and debug messages. - Consistency use -1 instead of 1 for some error code. - KNF.
1.98.6.14	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1751: sys/dev/pci/if_wmreg.h 1.122-1.125 via patch sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720,1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.98.6.13	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1707: sys/dev/pci/if_wm.c 1.716-1.718 via patch sys/dev/pci/if_wmreg.h 1.121 sys/dev/mii/ihphy.c 1.20 - Fix a bug that device timeout still happens when the link is down on ICH/PCH. Fixes PR kern/56478. - Add some sysctl info for debugging. - ihphy(4): Don't power down the PHY when the interface goes down. - Use macro. Fix comment.
1.98.6.12	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1594: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678, 1.680-1.685 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 via patch sys/dev/pci/if_wmvar.c 1.45 via patch sys/dev/mii/igphy.c 1.35-1.36 via patch sys/dev/mii/igphyreg.h 1.12-1.13 sys/dev/mii/makphy.c 1.66 via patch sys/dev/mii/makphyreg.h 1.11 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler). Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Explicitly cast from uint16_t to uint32_t before shifting 16bit left when printing Image Unique ID to avoid undefined behavior. - Set if_baudrate for non-MII device. - Rename some macros and function. - KNF. Add comment.
1.98.6.11	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1492: sys/dev/pci/if_wm.c 1.650, 1.652-1.654 via patch sys/dev/pci/if_wmreg.h 1.116-1.117 - Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. - Simplify code by using "struct mii_data *mii" more. No functional change. - MSI-X doesn't use sc->sc_icr variable, so move the code into non-MSI-X part. No functional change intended. - Modify debug printfs a bit. - Rename macro. - Use __BIT() - Fix comment. Add comment. - KNF.
1.98.6.10	01-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1312): sys/dev/pci/if_wm.c: revision 1.644 sys/dev/pci/if_wm.c: revision 1.645 sys/dev/pci/if_wmreg.h: revision 1.115 Define RETA_ENTRY_MASK_Q(used in RSS redirection table) correctly. Found by UBSan. Use aprint_debug_dev() instead of printf() in wm_ulp_disable(). - rnd_add_*() call must be serialized. Unserialized accesses broke the sc->rnd_source structure and caused off by one in kern_rndq.c's rnd_add_data_ts(). This problem was found by kUBSan. Make krndsource_t's data per queue-pair and call rnd_add_uint32() in wm_txrxintr_msix(). One of the reason why rnd_add_uint32() calls were moved from wm_{tx,rx}eof() is that rnd(9) manual page says "To allow the system to gather the timing information accurately, this call should be placed within the actual hardware interrupt service routine." - Pass a better value to rnd_add_uint32() in wm_intr_legacy(). XXX pullup-8.
1.98.6.9	17-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1298: sys/dev/pci/if_wm.c 1.633, 1.637-1.641 sys/dev/pci/if_wmreg.h 1.114 - Use unsigned to avoid undefined behavior. - Print ICH/PCH's NVM version. - Fix typo in comment. u_int_t -> uint_t. Whitespace fix.
1.98.6.8	07-Mar-2019	martin	Pull up the following, requested by msaitoh in ticket #1206: sys/dev/pci/if_wmreg.h 1.113 sys/dev/pci/if_wmvar.h 1.43-1.44 sys/dev/pci/if_wm.c 1.626-1.627, 1.629-1.636 via patch - Add support for I210 SGMII Flash-less device. - Add I219 variations for Cannon Lake. - Add missing rnd_detach_source(). - Use do { ... } while (0) for DPRINTF(x, y). - Swap enumeration of 82578 and 82577 PHY in if_wmvar.h. No functional change. - Add KASSERT. - Update TODO list. Add comment. Fix typo in comment.
1.98.6.7	31-Jan-2019	martin	Pull up the following, requested by msaitoh in ticket #1179: sys/dev/pci/if_wm.c 1.603-1.605,1.607-1.611, 1.613,1.615,1.618-1.620 via patch sys/dev/pci/if_wmreg.h 1.110-1.111 sys/dev/pci/if_wmvar.h 1.40-1.42 sys/dev/mii/inbmphyreg.h 1.13-1.15 - Add some code for suspend/resume: - Rename wm_smbustopci() to wm_init_phy_workarounds_pchlan(). It will also called when resume. - Call wm_phy_resetisblocked() after PHY reset in wm_init_phy_workarounds_pchlan() to wait for the PHY to quiesce to an accessible state. - Add new wm_resume_workarounds_pchlan() function and use it in wm_resume(). This workaround is only for PCH2 and newer. - Don't call wm_disable_aspm() neither in wm_attach() nor in wm_resume() but in wm_reset(). - Do some initialization in wm_resume() when IFF_UP is NOT set. - Don't continue when it failed to acquire semaphore in wm_ulp_disable(). - Print CLSEM workaround bit correctly. - Fix availability detection of WoL on some chips. - Print the WUS (WakeUp Status) register bits when resume. - Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220. - Remove an extra register read in wm_kmrn_lock_loss_workaround_ich8lan(). - Don't leave the MDICNFG register modified when the Power Management capability offset can't get. - Reduce indent level of wm_linkintr_gmii(). No functional change. - 80003's SERDES is not the same as 82575's but the same as legacy devices. Use the old methods on 80003. - Use __nothing for null DPRINTF(). - Rename functions. Add comment.
1.98.6.6	04-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1117): sys/dev/pci/if_wmreg.h: revision 1.109 sys/dev/pci/if_wm.c: revision 1.597 sys/dev/pci/if_wm.c: revision 1.598 sys/dev/mii/inbmphyreg.h: revision 1.12 sys/dev/pci/if_wm.c: revision 1.600 sys/dev/pci/if_wm.c: revision 1.601 sys/dev/pci/if_wm.c: revision 1.602 - Add new wm_gmii_{hv,i82544}_{read,write}reg_locked() and use them in wm_gmii_{hv,i82544}_{read,write}reg(). _locked() functions are not mii(4) API functions, so it's not required to keep the mii API. Change the PHY register type from int to uint16_t. It also change the usage of return value. It returns zero on success and non-zero on error. - Check the return value of _locked() function and treat it. - Use *writereg_locked() function to reduce race condition in wm_init_lcd_from_nvm(). - Add comment. - Control TX/RX descriptor snooping control bits on ICH8 and newer. Only on ICH8, No-snoop bits are opposite polarity. On my Thinkpad X61, the default value of this bits are all zero, so this commit changes the snoop function enable on the machine. I tested with some other PCH machines and those bits are all zero (enable snoop by default), so this commit won't affect to some machines. - Disable relax ordering on 82546GB(Device ID 0x1099 and 0x10b5) or >= ICH8. Same as other OSes. - Add wm_oem_bits_config_ich8lan() to control LPLU and GbE setting base on the NVM's info. - Modify wm_enable_wakeup() to reduce difference against FreeBSD and Linux. This modification affects to ICH8 and newer devices. I217 Rapid Start Technology support have not written yet (it's TODO). - Add wm_k1_workaround_lv() from FreeBSD. It's PCH2 specific: Workaround to set the K1 beacon duration for 82579 parts in 10Mbps. Disable K1 for 1000 and 100 speeds. - Make wm_link_stall_workaround_hv() and move an 82578 specific code into it. Don't apply the workaround if BMCR_LOOP bit is set. Same as FreeBSD. - Add comment. Modify comment. Fix comment. No functional change.
1.98.6.5	08-Nov-2018	martin	Pull up the following, requested by msaitoh in tickt #1085: sys/dev/mii/inbmphyreg.h 1.11 sys/dev/pci/if_wmreg.h 1.108 sys/dev/pci/if_wm.c 1.590-1.596 via patch - 82574 and newer document says the status field has neither EC (Excessive Collision) bit nor LC (Late Collision) bit (reserved), so don't check the bit. - Add workaround for DMA hang problem which result in TX device timeout on PCH_LPT with I218. Same as FreeBSD and Linux. This workaround is only for device ID 0x155a, 0x15a2, 0x1559 and 0x15a3. - Fix a PCH2 specific bug that wrong PHY register value can be read when boot. Same as FreeBSD and Linux. - After writing MDIC register, don't read quickly the same register but do delay(50). Same as other OSes. - Add missing wm_gate_hw_phy_config_ich8lan(false) in wm_phy_post_reset() on PCH2. wm_gate_hw_phy_config_ich8lan(true) is called in wm_reset(), so wm_gate_hw_phy_config_ich8lan(false) should be called after reset in wm_phy_post_reset(). - On PCH2, set the phy config counter to 50msec after (PHY) reset. - KNF & Use macro.
1.98.6.4	16-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #764): sys/dev/pci/if_wm.c: revision 1.567 sys/dev/pci/if_wm.c: revision 1.568 sys/dev/pci/if_wm.c: revision 1.569 sys/dev/pci/if_wmvar.h: revision 1.38 sys/dev/pci/if_wm.c: revision 1.570 sys/dev/pci/if_wm.c: revision 1.571 sys/dev/pci/if_wm.c: revision 1.572 share/man/man4/wm.4: revision 1.40 sys/dev/pci/if_wmreg.h: revision 1.106 sys/dev/pci/if_wmreg.h: revision 1.107 SW PHY Config Enable bit for ICH8 B0 stepping is not bit 1 but bit 0. No binary change: - Sort registers. - Lowercase hexadecimal value. On PCH_SPT (and newer), FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it should be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. Our PCH_SPT part of wm_nvm_valid_bank_detect_ich8lan() was based on FreeBSD r287467. After that, they reverted it and committed the different code in r287762. r287762's bank detect code didn't work for us because our wm dirver had a problem in flash access. The problem was fixed in if_wm.c rev. 1.567, so we can use the new way now. If the extended configration size in the EXTCNFSIZE register is 0, don't continue. Add PCH_CNP support (I219 with Intel 300 series chipset). It's required more test, so it's disabled by default. Enable I219. I354 uses an external PHY, so don't use wm_set_eee_i350().
1.98.6.3	24-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #389): sys/net/if_ether.h: revision 1.69 sys/net/if_vlan.c: revision 1.108 sys/dev/pci/if_bge.c: revision 1.313 sys/net/if_vlanvar.h: revision 1.11 sys/net/if_vlanvar.h: revision 1.12 sys/net/if_ether.h: revision 1.70 sys/net/if_vlan.c: revision 1.110 sys/dev/pci/if_wm.c: revision 1.544 sys/dev/pci/if_wmreg.h: revision 1.105 Fix a bug that a vlan packet which has priority or CFI bit in the tag causes panic. Revert part of if_bge.c 1.312. It's not required to mask other than VLAN ID bits in VLAN tag. Revert if_wmreg.h 1.104 and if_wm.c 1.542. It's not required to mask other than VLAN ID bits in VLAN tag. No functional change: - u_int16_t -> uint16_t - u_short -> uint16_t - tag_hash_func -> vlan_tag_hash - 0 -> NULL because vlr_parent is a pointer.
1.98.6.2	06-Nov-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #348): sys/dev/pci/if_wm.c: revision 1.542 sys/dev/pci/if_wmreg.h: revision 1.104 fix wm(4) vlan panic. Reported and tested by Tom Ivar Helbekkmo, thanks. wm(4) used PRI bits and CFI bit as vlantag by mistake. It is found out by if_ether.h:r1.67.
1.98.6.1	01-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #173): sys/dev/pci/if_wmreg.h: 1.99-1.103 sys/dev/pci/if_wmvar.h: 1.34-1.37 sys/dev/pci/if_wm.c: 1.510-1.537 sys/dev/mii/ihphyreg.h: 1.2 sys/dev/mii/inbmphyreg.h: 1.10 sys/dev/mii/igphy.c: 1.27 Sync wm(4) up to if_wm.c rev. 1.537: - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). - If a legacy interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Flush writing WMREG_CTRL in wm_gmii_mediachange(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Add I219 specific workaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for workaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - IF TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function is required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Configure the LCD with the extended configuration region in NVM if it's required. Tested with Thinkpad X220. - 8257[12]: Don't directly access SPI but use EERD register. - 82575-I354: If the size of SPI ROM >= 32K words, use direct SPI access instead of EERD register access. - Add wm_nvm_eec_clock_raise() and wm_nvm_eec_clock_lower() and use them for Microwire/SPI bus control. Same as Linux and FreeBSD. - Reduce timeout value for 80003 in wm_get_swfw_semaphore(). Same as Linux and FreeBSD. - Change API of kmrn_{read,write}reg() and check the return value. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally." on I21[01]. PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check NVM word size before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Print internal driver flags when attaching. - Add debug printf()s. - Rename variables. - Add comment, update comment and remove wrong comment.
1.105.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.105.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.105.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.107.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.107.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.107.2.1	10-Jun-2019	christos	Sync with HEAD
1.115.2.9	23-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1802: sys/dev/pci/if_wm.c 1.794-1.798 via patch sys/dev/pci/if_wmreg.h 1.131 - Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. - Drop frames if the RX descriptor ring has no room on multiqueue system. - Improve dmesg output: - Print RX packet buffer size. - Fix the upper 16bit of Image Unique ID(EtrackID).
1.115.2.8	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1656: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.115.2.7	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1578: sys/dev/pci/if_wm.c 1.764-1.767 via patch sys/dev/pci/if_wmreg.h 1.128 - Workaround for some hypervisor environments. The environments cannot stop e1000 interrupt immediately. - Rename nq_txdesc' member "nqrx_ctx" to "nqtx_ctx". No functional change. - Add comment. Modify comment. - KNF.
1.115.2.6	07-Sep-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1515: sys/dev/pci/if_wm.c 1.741-1.749, 1.753-1.757, 1.762 via patch sys/dev/pci/if_wmreg.h 1.126-1.127 - Fix I219 workaround in wm_flush_desc_rings(). - Add more statistics counters. - To avoid releasing mutex temporally, use new wm_set_mdio_slow_mode_hv_locked(). - No functional changes: - Turn a locking botch (shouldn't drop lock on error) into a KASSERT in wm_deferred_start_locked(). - Remove unneeded header inclusion. - Use __BIT() a little. - Modify comment and debug messages. - Consistency use -1 instead of 1 for some error code. - KNF.
1.115.2.5	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1477: sys/dev/pci/if_wmreg.h 1.122-1.125 sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720, 1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.115.2.4	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1373: sys/dev/pci/if_wm.c 1.716-1.718 via patch sys/dev/pci/if_wmreg.h 1.121 sys/dev/mii/ihphy.c 1.20 - Fix a bug that device timeout still happens when the link is down on ICH/PCH. Fixes PR kern/56478. - Add some sysctl info for debugging. - ihphy(4): Don't power down the PHY when the interface goes down. - Use macro. Fix comment.
1.115.2.3	23-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1093): sys/dev/pci/if_wmreg.h: revision 1.120 sys/dev/pci/if_wmvar.h: revision 1.46 sys/dev/pci/if_wm.c: revision 1.686 sys/dev/pci/if_wm.c: revision 1.687 sys/dev/mii/inbmphyreg.h: revision 1.20 sys/dev/pci/if_wm.c: revision 1.688 sys/dev/pci/if_wm.c: revision 1.689 s/ressource/resource/. Found by knakahara. Add new flag named WM_F_CRC_STRIP and use it. No functional change. This change also sets the RCTL_SECRC bit on I211 but it doesn't change the behavior because I211 always strips CRC like I35[04] and I210. Add a workaround for jumbo frame on PCH2 and newer. Tested by chs@. - Add wm_lv_jumbo_workaround_ich8lan() and use it. From FreeBSD. XXX For KUMCTRLSTA_OFFSET_HD_CTRL register modification, it's doubtful. FreeBSD and Linux do the same thing that they set the same value on both jumbo frame's enable case and the disable case. It seems the default value is 0x0b0c and it's not changed on the enable case, so it might be a bug on the enable case or the modification is not required. - Rename I219_UNKNOWN1 to I82579_UNKNOWN1. Use 12K for packet buffer for jumbo frame on PCH2 and newer. XXX Note that Linux Use 14K.
1.115.2.2	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #994: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678,1.680-1.681 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 sys/dev/pci/if_wmvar.c 1.45 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler. Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Rename some macros and function. - KNF. Add comment.
1.115.2.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #650: sys/dev/pci/if_wm.c 1.650, 1.652-1.654 sys/dev/pci/if_wmreg.h 1.116-1.117 - Set CTRL_ILOS(Invert loss of signal) bit correctly on 82580 port 1, 2, 3 and newer chips. This change fixes a bug that some fiber, serdes or SFP devices don't detect the link status correctly. - Simplify code by using "struct mii_data *mii" more. No functional change. - MSI-X doesn't use sc->sc_icr variable, so move the code into non-MSI-X part. No functional change intended. - Modify debug printfs a bit. - Rename macro. - Use __BIT() - Fix comment. Add comment. - KNF.
1.128.2.2	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #607): sys/dev/pci/if_wm.c: revision 1.795 sys/dev/pci/if_wm.c: revision 1.796 sys/dev/pci/if_wm.c: revision 1.797 sys/dev/pci/if_wmreg.h: revision 1.131 wm(4): Add RQDPC(Receive Queue Drop Packet Count) to iqdrops. The iqdrops counter should include not only MPC(Missed Packet Count) but also RQDPC(Receive Queue Drop Packet Count). Same as ixgbe(4) and igc(4). Note that the RQDPC is not currently counted because SRRCTL_DROP_EN is not set. wm(4): Drop frames if the RX descriptor ring has no room on multiqueue system. Drop frames if the RX descriptor ring has no room. This is enabled only on multiqueue system to avoid bad influence to other queues. The drop count can be seen by the RQDPC counter (wmN rxqXXdrop evcnt). wm(4): Fix compile error without WM_EVENT_COUNTERS.
1.128.2.1	22-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #213): sys/dev/pci/if_wm.c: revision 1.770 sys/dev/pci/if_wm.c: revision 1.771 sys/dev/pci/if_wm.c: revision 1.772 sys/dev/pci/if_wm.c: revision 1.773 sys/dev/pci/if_wm.c: revision 1.774 sys/dev/pci/if_wm.c: revision 1.775 sys/dev/pci/if_wm.c: revision 1.776 sys/dev/pci/if_wmreg.h: revision 1.129 sys/dev/pci/if_wm.c: revision 1.777 sys/dev/pci/if_wm.c: revision 1.778 sys/dev/pci/if_wmvar.h: revision 1.49 sys/dev/pci/if_wm.c: revision 1.779 sys/dev/pci/if_wm.c: revision 1.768 sys/dev/pci/if_wm.c: revision 1.769 sys/dev/pci/if_wmreg.h: revision 1.130 sys/dev/pci/if_wm.c: revision 1.780 sys/dev/pci/if_wm.c: revision 1.781 Count some 64bit counters correctly. - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Found by knakahara. Add note for the TORL register. The TOR register includes error, flow control and broadcast rejected. Sort lines. No functional change. Rearrange the order of the registers so that they are roughly in ascending order. Sort lines. No functional change. Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. IC{TX,RX}*C registers are for older than 82575. Fix a bug that the transmit underrun counter is incorrectly counted. The transmit underrun bit in the transmit status filed is only for 82544 (and older?), so don't use the counter for newer chips. The bit is reserved for newer chips, but the bit sometimes set on 82575 at least. Don't add "Count" for event counter's description. Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero. Add some new event counters. Add the following counters for 82575 and newer except 80003, ICHs and PCHs: - Only 82576 document describes about the circuit breaker, so the following two might be only for 82575: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - 82575's document doesn't describe the following two, but we can see the same value as GO{T,R}C have: - Host Good Octets RX - Host Good Octets TX - 82575's document doesn't describe the LENERRS (Length Errors) counter. I don't know if it has. - Perhaps Non-SerDes/SGMII devices don't have SCVPC (SerDes/SGMII Code Violation Packet) register. We don't care if it's SerDes/SGMII or not for now. - HRMPC (Header Redirection Missed Packet) appears only once in 8257[56]'s datasheet. FreeBSD's igb counts it, so we do, too. - Count the following two for I350 and newer. I don't know if PCHs have: - EEE TX LPI - EEE RX LPI Move statistics updating code from wm_tick() to new wm_update_stats(). - To reuse. - No functional change. Add SOICZIFDATA (ifconfig -z) support for evcnt(9). First update the statistics data, then clear the event counters, and finally copy and clear if_data via ether_ioctl(). Fix prc511's comment and description. Use WM_IS_ICHPCH(). No functional change. Fix typo. s/ictxact/ictxatc/. No functional change.
1.51	11-Oct-2023	msaitoh	wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss.
1.50	25-Aug-2023	msaitoh	Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@.
1.49	11-May-2023	msaitoh	Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero.
1.48	20-Dec-2021	skrll	branches: 1.48.4; Trailing whitespace
1.47	30-Oct-2020	msaitoh	Add WMPHY_I350. Not used yet.
1.46	15-Sep-2020	msaitoh	Add new flag named WM_F_CRC_STRIP and use it. No functional change. This change also sets the RCTL_SECRC bit on I211 but it doesn't change the behavior because I211 always strips CRC like I35[04] and I210.
1.45	11-Dec-2019	msaitoh	Add SFP support part 1. The code for SerDes and SGMII setting is not include in this commit. SFP module removal interrupt detects but not used yet: - Detect SFP's 1000BASE-SX, 1000BASE-LX, 100BASE-FX and 1000BASE-T correctly. - Detect the Media Auto Sense feature. Not supported yet. - Add comment.
1.44	19-Feb-2019	msaitoh	branches: 1.44.4; Modify PHY tree in comment.
1.43	07-Feb-2019	msaitoh	Swap enumeration of 82578 and 82577. Same as FreeBSD. Chip Model number of 82578 is +1 from 82577 but 82577 is functionally newer than 82578 (and 82577's MII_MODEL(0x05) is greater than 82578's (0x04)). This change doesn't affect any behavior to if_wm.c (NFCI). Add comment.
1.42	25-Dec-2018	msaitoh	Add comment.
1.41	20-Dec-2018	msaitoh	- Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220.
1.40	14-Dec-2018	msaitoh	Print CLSEM workaround bit correctly.
1.39	08-Aug-2018	msaitoh	Add MDIC_WA and CLSEMWA to snprintb() string for WM_F_* flags.
1.38	12-Apr-2018	msaitoh	branches: 1.38.2; Add PCH_CNP support (I219 with Intel 300 series chipset). It's required more test, so it's disabled by default.
1.37	27-Jul-2017	msaitoh	branches: 1.37.2; Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally." on I21[01]. PXE boot is one of the case. XXX pullup-[678].
1.36	26-Jul-2017	msaitoh	- Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register vaule from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg.
1.35	25-Jul-2017	msaitoh	Use new nvm.{acquire,release}() for semaphore. Almost the same except the following: - 8257[12]: Don't directly access SPI but use EERD register. - 82575-I354: If the size of SPI ROM >= 32K words, use direct SPI access instead of EERD register access. - Add wm_nvm_eec_clock_raise() and wm_nvm_eec_clock_lower() and use them for Microwire/SPI bus control. Same as Linux and FreeBSD. - Redude timeout value for 80003 in wm_get_swfw_semaphore(). Same as Linux and FreeBSD.
1.34	26-Jun-2017	msaitoh	Print sc_flags with snprintb().
1.33	01-Feb-2017	msaitoh	branches: 1.33.6; Use new wm_gmii_setup_phytype() function to setup sc_phytype and mii_{read\|write}reg. This function is called twie. To identify PHY type, correct read/write function should be selected. To select correct read/write function, PCI ID or MAC type are required without accessing PHY registers. On the first call of this function, PHY ID is not known yet. Check PCI ID or MAC type. The list of the PCI ID may not be perfect, so the result might be incorrect. In the second call, PHY OUI and model are used to identify PHY type. It might not be perfpect because of the lack of compared entry, but it would be better than the first call. If the detected new result and previous assumption is different, diagnous message will be printed.
1.32	28-Oct-2016	msaitoh	branches: 1.32.2; - Define WMPHY_I217, WMPHY_VF and WMPHY_210. - Use BME1000_PHY_PAGE_SELECT in wm_gmii_bm_{read,write}reg(). This change has no effect because GG82563_PHY_PAGE_SELECT and BME1000_PHY_PAGE_SELECT have the same value.
1.31	06-May-2016	msaitoh	branches: 1.31.2; Basic support for I219. It doesn't work on I219, so it's disabled.
1.30	25-Dec-2015	msaitoh	- Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Fix typo in comment and modify comment by tnn@. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change.
1.29	06-Jun-2015	msaitoh	- Add workaround for I210 Errata 25 and I211 Errata 10. - Add wm_gmii_gs40g_{read\|write}reg() and use it to access non-standatrd page. - Add wm_pll_workaround_i210() and call it when chip is i211 chip is i210 and it use INVM chip is i210 and NVM image version < 3.25 - Add comment - Rename macros.
1.28	06-Jun-2015	msaitoh	Print NVM image version.
1.27	06-Jun-2015	msaitoh	Revert previos. Sorry, I committed in another working directory...
1.26	06-Jun-2015	msaitoh	Print NVM image version.
1.25	02-Jun-2015	msaitoh	Fix a lot of bugs to make 82575 and newer's SERDES based systems work. - Add SERDES specific functions. - Fix IO pin configuration. - Reset autonego timer when link becomes up. TODO: - Fix a bug that SFP ROM can't read. - Perhaps some work is required for 8257[12] serdes systems. - Remove duplicated code in TBI's link related functions.
1.24	16-May-2015	msaitoh	Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines and NICs. Tested with Shuttle DS57U(iNVM based) and other non iNVM based I210 machines.
1.23	27-Nov-2014	msaitoh	branches: 1.23.2; Fix typo in comment.
1.22	25-Nov-2014	msaitoh	Add note about the variations of Intel gigabit Ethernet controller.
1.21	06-Oct-2014	msaitoh	Delete 82580ER related code. It was from FreeBSD and was removed in r203049.
1.20	24-Aug-2014	msaitoh	Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102.
1.19	14-Jul-2014	msaitoh	branches: 1.19.2; Cleanup comments. No functional change.
1.18	11-Jul-2014	msaitoh	Rename some WM_F_* flags to clarify. Renumbering flags. No functional change. WM_F_EEPROM_HANDSHAKE -> WM_F_LOCK_EECD WM_F_EEPROM_SEMAPHORE -> WM_F_LOCK_SWSM WM_F_SWFW_SYNC -> WM_F_LOCK_SWFW WM_F_SWFWHW_SYNC -> WM_F_LOCK_EXTCNF
1.17	29-Dec-2013	msaitoh	branches: 1.17.2; Add support For I354(C2000 interna Ethernet controller): - Add I354 support. - Fix SGMII+MDIO case. SGMII+I2C is not supported yet. - Not tested well. - Sprinkle "XXX" to check later.
1.16	02-Jun-2013	msaitoh	branches: 1.16.2; Add I217 (and I218) support. - The name of I217 is similar to I210, but the function is rather similar to PCH2. - Not tested well. Tested with my own Intel DQ87PG which has I217LM onboard. - It seems that PCH2 and PCH_LPT specific function for RAL should be written. - Quick hack for the NVM checksum mismatch. if_wm.c currently has no wm_write_eeprom(), so it cannot update NVM's "updated bit". To avoid this problem, check only the last 12bits of the checksum. My own DQ87PG's updated bit is not set, and I could avoid the problem using with this hack.
1.15	18-Apr-2013	msaitoh	Add support I21[01]. Only tested with my own I210-T1 Ethernet Server Adapter.
1.14	12-Feb-2013	msaitoh	Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542.
1.13	25-May-2012	msaitoh	branches: 1.13.2; Add support Intel I350 Ethernet.
1.12	20-May-2011	msaitoh	branches: 1.12.4; 1.12.8; 1.12.10; Add PCH2 support.
1.11	07-Mar-2010	msaitoh	branches: 1.11.2; 1.11.4; 1.11.6; - Add code for WOL, ASF, IPMI and Intel AMT. - wm_enable_wakeup() is disabled by default. If you want to use WOL with the Magic Packet, define WM_WOL. - Add the following flags: WM_F_ASF_FIRMWARE_PRESENT WM_F_ARC_SUBSYSTEM_VALID WM_F_HAS_AMT WM_F_HAS_MANAGE WM_F_WOL - Add wm_suspend() and wm_resume(). Give/get the control to/from the firmware. - Need more work for PCH. See wm_enable_phy_wakeup(). - Enable wm_get_hw_control() for 82574 and 82583. - Add Yet another workaround for ICH8. - Add wm_igp3_phy_powerdown_workaround_ich8lan() for power down problem on D3.
1.10	25-Feb-2010	msaitoh	Merge if_wm.c rev. 1.162.4.13 and if_wmvar.h rev. 1.2.46.3 into maintrunk. Increase the timeout value to prevent timeout on 80003. The values are taken from Intel's e1000 driver. Tested by spz and me.
1.9	16-Feb-2010	msaitoh	Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems.
1.8	14-Jan-2010	msaitoh	branches: 1.8.2; Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.7	11-Jan-2010	msaitoh	Add partial support for PCH. We have to do more work for PCH. Especially I don't know what driver i82578 should be attached to. makephy? or atphy? MII_OUI() says that the PHY is from Attansic (== Atheros). Intel's e1000 driver says that it's close to makphy... I can't link up at 1000BaseT yet...
1.6	07-Jan-2010	msaitoh	Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9.
1.5	29-Dec-2009	msaitoh	- Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - Add support for i82583V. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574.
1.4	16-Dec-2009	msaitoh	Sync with Intel's original em driver: - Add dspcode for igp3 and use it when the EEPROM isn't available. - Add some delays. - Stop the PHY transmitter before patching the DSP code and restart it after wrote. - Save and restore register 0x2f5b.
1.3	16-Dec-2009	msaitoh	Re-enable igphy's 82566 support. - Patch for the DSP code is only for 8254[17] and we have to apply the different patches between rev. 1 and rev. 2. - The workaround for analog fuse is only for 82547 rev. 1. - The workaround for smartspeed is only for 8254[17] see http://mail-index.netbsd.org/tech-net/2009/08/05/msg001546.html
1.2	22-Feb-2006	gavan	branches: 1.2.40; 1.2.46; Convert MD hook to device property reads. Device properties, if present, take precedence over EEPROM values.
1.1	16-Feb-2006	gavan	branches: 1.1.2; Compute the EEPROM checksum to determine the presence of valid EEPROM data. If EEPROM data is invalid, allow a MD hook to supply EEPROM data instead.
1.1.2.3	01-Mar-2006	yamt	sync with head.
1.1.2.2	18-Feb-2006	yamt	sync with head.
1.1.2.1	16-Feb-2006	yamt	file if_wmvar.h was added on branch yamt-uio_vmspace on 2006-02-18 15:39:08 +0000
1.2.46.7	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1962): sys/dev/pci/if_wm.c: revisions 1.259-1.266, 1.269, 1.271, 1.273-1.274, 1.277-1.278, 1.282, 1.284-1.285, 1.287, 1.290, 1.294, 1.297-1.298, 1.300-1.301, 1.304-1.307, 1.310, 1.312-1.314, 1.316 via patch sys/dev/pci/if_wmreg.c: revisions 1.54-1.57, 1.59-1.60, 1.62, 1.64-1.66 via patch sys/dev/pci/if_wmvar.c: revisions 1.17, 1.20-1.21 via patch sys/dev/pci/pcidevs: revisions 1.1172, 1.1195, 1.1201 via patch - Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. - Fix BMC related bugs. - Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth. - Add support for I354 DH89xxCC and some new I218 devices. - Fix definition of CTRL_GIO_M_DIS bit. - Insert completion barrier between register write and delay(). - Bump max TX DMA size to avoid pathological condition with TSO. From dyoung. - Fix semaphore related bugs. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - It's not required to print "failed to detect NVM bank" with aprint_error_dev(). Use DPRINTF(). Same as {Free,Open}BSD. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Remove extra debug message. - Remove unused variable. - Remove a duplicated error message. - Cleanup comments. - Fix debug message.
1.2.46.6	07-Sep-2013	bouyer	sys/dev/pci/if_wm.c 1.246-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 sys/dev/pci/pcidevs 1.1152-1.1153 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #1867]
1.2.46.5	19-Jun-2013	bouyer	Pullup the following revisions via patch, requested by msaitoh in ticket #1850: sys/dev/pci/if_wm.c 1.201, 1.203-1.204, 1.207-1.212, 1.215, 1.217-1.218, 1.220-1.223, 1.228, 1.232-245 sys/dev/pci/if_wmreg.h 1.40-1.45, 1.47-1.48 sys/dev/pci/if_wmvar.h 1.11-1.13 sys/dev/pci/pcidevs 1.1074, 1.1077, 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.6 sys/dev/mii/ihphy.c 1.1-1.2 sys/dev/mii/ihphyreg.h 1.1 sys/dev/mii/inbmphyreg.h 1.3 sys/dev/mii/files.mii 1.47 via patch sys/dev/mii/miidevs 1.97 and 1.100 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/arch/i386/conf/ALL 1.280 sys/arch/i386/conf/GENERIC 1.1001 sys/arch/i386/conf/INSTALL_FLOPPY 1.11 sys/arch/i386/conf/XEN2_DOM0 patch sys/arch/amd64/conf/GENERIC 1.293 sys/arch/amd64/conf/XEN3_DOM0 1.61 share/man/man4/wm.4 1.21-1.24 Apply almost all fixes and improvements from netbsd-6 except for the rev. 1.196's iqdrops' change. - Add the detach code. - Add code for WOL, ASF, IPMI and Intel AMT. WOL is disabled by default - Add Yet another workaround for ICH8. - 82576 is dual port, so check the FUNCID and increment the MAC address for the 2nd port. - Fix the names of 82577L[MC] LAN controllers (for mobile). - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add 82580 support. - 82571 quirk. Only 82571 shares port 0 of EEMNGCTL_CFGDONE. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. - Fix some register names. No functional change. - Omit U+00AE "REGISTERED SIGN" in a product name due to its non-ASCII nature. - Stop wm(4) from needlessly resetting when you add or delete a vlan(4). - Fix MAC address check on 8257[156] and 80003 case. Some cards have non 0xffff pointer but those don't use alternative MAC address in reality. So we check whether the broadcast bit is set or not like Intel's e1000 driver. Fixes PR kern/44072 reported by Jean-Yves Moulin. - Add PCH2(and 82579) support. Fixes PR#46487 - Add yet another 82567V support. - Add ICH10+HANKSVILL support. - Add support Intel I350 Ethernet. - Make vlan and all ip/ip6 checksum offload work for the I350. - Fix compile error with WM_DEBUG. - Fix a bug that PHY isn't set to low-power mode on PCH and PCH2. - Add WM_DEBUG_NVM. If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. - Skip 64bit BAR correctly. - Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350. - Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542. - Style fix. Fix typo in comment. Fix comments. Add comments.
1.2.46.4	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1358): sys/dev/pci/if_wm.c 1.196-1.199,1.202,1.205 sys/dev/pci/if_wmvar.h 1.9 sys/dev/pci/if_wmreg.h 1.36-1.39 sys/dev/pci/pcireg.h 1.61-1.64 sys/dev/pci/pcidevs 1.1023 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen mii/igphy.c 1.21 mii/igphyvar.h 1.1 mii/inbmphyreg.h 1.2 - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems. - Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working. - (pcireg.h) Add PCIe config register definitions. - Note that the changes to count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops in rev. 1.196 of if_wm.c is not pulled up.
1.2.46.3	25-Feb-2010	sborrill	Fix regression caused by pullup #1277 requested by msaitoh.
1.2.46.2	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.2.46.1	23-Dec-2009	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: 1.176-1.179, 1.181-1.183 sys/dev/pci/if_wmreg.h: 1.28 sys/dev/pci/if_wmvar.h: 1.1-1.4 sys/dev/mii/igphy.c: 1.18-1.20 via patch Many bugfixes: - Some fixes for i80003 and ICH{8,9,10} from e1000 driver and document: - Add setting for KABGTXD register for ICH{8,9,10}. - ICH9 and ICH10 has no FCAL, FCAH and FCT like ICH8. - Add special setting for FCTTV and TCTL_EXT register for i80003 - The special setting for TIPG is only for i80003. - Some of kumeran settings are only for i80003's bugs. - Add some ICH10 fixes. - Fix the bug that another lock mechanism is used to access Kumeran registers on i80003 and ICHs. - Fix yet another i80003 ONLY workaround. The code to modifing TIPG register is only for i80003. - Set the Re-Transmit on Late Collision(RTLC) flag for all devices. - Fix a typo in a printf message. - If the difference bettween last flag and new flag is only IFF_PROMISC or IFF_ALLMULTI, set multicast filter only to prevent link down. Tested by Mark Davies and me. Fixes PR#29126 for wm. - Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line. - Fix igphy's 82566 support. - Patch for the DSP code is only for 8254[17] and we have to apply the different patches between rev. 1 and rev. 2. - The workaround for analog fuse is only for 82547 rev. 1. - The workaround for smartspeed is only for 8254[17] - Sync with Intel's original em driver: - Check PCI-X mode as e1000 driver. - Add dspcode for igp3 and use it when the EEPROM isn't available. - Add some delays. - Stop the PHY transmitter before patching the DSP code and restart it after writing. - Save and restore register 0x2f5b.
1.2.40.1	11-Mar-2010	yamt	sync with head
1.8.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.11.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.11.4.2	21-Apr-2010	matt	sync to netbsd-5
1.11.4.1	07-Mar-2010	matt	file if_wmvar.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:42 +0000
1.11.2.1	31-May-2011	rmind	sync with head
1.12.10.7	06-May-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1366): sys/dev/pci/if_wm.c: 1.281, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.387, 1.389 via patch sys/dev/pci/if_wmreg.h: 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: 1.22-1.23, 1.25-1.30 via patch sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 Sync wm(4) as of if_wm.c rev 1.389 except SERDES, MSI/MSI-X, multiqueue and NET_MPSAFE: - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvctl -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.12.10.6	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1296): sys/dev/pci/pcidevs: revision 1.1195 via patch sys/dev/pci/if_wm.c: revisions 1.290, 1.304 via patch sys/dev/pci/if_wmvar.h: revision 1.20 via patch - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - Support DH89xxCC device. - Add extra delay for 82580 and newer devices except DH89XXCC SGMII device. Same as FreeBSD.
1.12.10.5	04-Dec-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #1203): sys/dev/pci/if_wm.c: revisions 1.271, 1.273-1.274, 1.277-1.278, 1.280, 1.282, 1.284-1.285, 1.287, 1.293-1.294, 1.297-1.298, 1.300-1.301, 1.305-1.307 via patch sys/dev/pci/if_wmreg.h: revisions 1.57-1.62, 1.64-1.65 via patch sys/dev/pci/if_wmvar.h: revisions 1.19, 1.21 via patch - Acquire SW semaphore in wm_get_swsm_semaphore(). - Fix some bugs realted to semaphore. This change fixes a problem which was exposed in if_wm.c rev. 1.271. Tested by riastradh@. - Clear the SMBI bit in SWSM register before accessing NVM and PHY in wm_attach(). Same as FreeBSD. - Fix a bug that 82573 doesn't put the hardware semaphore. Same as FreeBSD r256200. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). Tested with 82543GC, 82544EI, 82545EM, 82546GB 82571EB and 82572EI fiber cards. - Don't use the RXCFG interrupt. It's not required and the interrupt is very heavy (a lot of interrupts). Same as {Free,Open}BSD. - Modify wm_tbi_mediachange() to be close to em_setup_fiber_serdes_link() of {Free,Open}BSD. At least, don't forget to set duplex setting. - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Don't check SWSM_SMBI bit if WM_F_LOCK_SWSM isn't set. Fix a problem when using vmware with e1000"e". With e1000e which is regarded as 82574L, wm_gmii_init() fails with "could not acquire SWSM SMBI" message without this change. This problem doesn't occur with real 82574L card. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Calculate NVM word size correctly. - Determine timeout value based on the NVM word size. - It's not required to print "failed to detect NVM bank" message. Only print while debugging. Same as {Free,Open}BSD. - Add some new I218 devices. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Move some NVM related macros from if_wm.c to if_wmreg.h. - Sort definitions in if_wmreg.h - move NVM related values to the bottom. - sort in register's address' order. - Simplify wm_read_mac_addr(). - Fix debug message. - Add missing prototypes. - Rename some functions for consistency and clarify. - Rename some macros for consistency. - Remove a duplicated error message. - Fix typo in comment. - Cleanup comments. - KNF.
1.12.10.4	09-Nov-2014	martin	Pullup the following revisions, requested by msaitoh in ticket #1188: sys/dev/pci/pcidevs 1.1172 sys/dev/pci/if_wm.c 1.263-1.266 via patch sys/dev/pci/if_wmreg.h 1.55-1.56 sys/dev/pci/if_wmvar.h 1.17 share/man/man4/wm.4 1.26-1.27 and 1.29 via patch - Add I354 support. - Insert completion barrier between register write and delay(). - Fix the definition of CTRL_GIO_M_DIS. This bit is not bit 3 but bit 2. - Cleanup
1.12.10.3	14-Jul-2013	riz	Apply changes (requested by msaitoh in ticket #907): sys/dev/pci/if_wm.c 1.238, 1.244-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.50-1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 Various fixes to wm(4): Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Change style, add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #907]
1.12.10.2	14-Feb-2013	jdc	Pull up revisions: src/sys/dev/pci/if_wm.c revision 1.243 src/sys/dev/pci/if_wmvar.h revision 1.14 (requested by msaitoh in ticket #820). Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542.
1.12.10.1	28-Jun-2012	riz	Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/if_wm.c 1.228 sys/dev/pci/if_wmreg.h 1.47 sys/dev/pci/if_wmvar.h 1.13 sys/dev/pci/pcidevs 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/wm.4 1.22-1.25 Add support for Intel I350 Ethernet. Update document.
1.12.8.1	02-Jun-2012	mrg	sync to latest -current.
1.12.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.4.1	30-Oct-2012	yamt	sync with head
1.13.2.4	03-Dec-2017	jdolecek	update from HEAD
1.13.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.13.2.2	23-Jun-2013	tls	resync from head
1.13.2.1	25-Feb-2013	tls	resync with head
1.16.2.1	18-May-2014	rmind	sync with head
1.17.2.1	10-Aug-2014	tls	Rebase.
1.19.2.7	11-Aug-2018	martin	Pull up the following, requested by msaitoh in ticket #1628: share/man/man4/wm.4 1.40 via patch sys/dev/mii/ihphyreg.h 1.2 sys/dev/mii/inbmphyreg.h 1.10 sys/dev/pci/if_wm.c 1.504, 1.506, 1.510-1.535, 1.539-1.540, 1.546, 1.548, 1.551-1.552, 1.558, 1.565-1.573, 1.575, 1.579, 1.582, 1.584 via patch sys/dev/pci/if_wmreg.h 1.99-1.103, 1.106-1.107 via patch sys/dev/pci/if_wmvar.h 1.34-1.39 via patch sys/dev/pci/pcidevs 1.1327 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/pci/pcireg.h patch Sync wm(4) up to 2018/08/08 except MSI/MSI-X and NET_MPSAFE: - remove extra "+" - Fix a bug that non-GMII devices don't send a routing message when the link status is changed. - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes. - If a interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Print sc_flags with snprintb(). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange(). - Add missing "else" in wm_nvm_release(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - If TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function are required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Use new nvm.{acquire,release}() for semaphore. - Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register value from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally on I21[01]". PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Print NVM offset and word count when EERD polling failed. - On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux). - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Fix typo in comment. Reported by Shinichi Doyashiki in PR#52885. - Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki. - CID-1427779: Fix uninitialized variables. - Fix a bug that wm_pll_workaround_i210() is not called when a) Chip is I211 or b) Chip is I210 and it uses iNVM (not FLASH). - Do wm_reset_mdicnfg_82580() on 82580 only. - Fix FLASH access on PCH_SPT and newer. Their FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it must be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. - Make wm_nvm_valid_bank_detect_ich8lan() the same as other OSes. - If the extended configuration size in the EXTCNFSIZE register is 0, don't continue in wm_init_lcd_from_nvm(). - Add PCH_CNP support (I219 with Intel 300 series chipset). - Enable I219 support. - I354 uses an external PHY, so don't use wm_set_eee_i350(). - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. - Fix a bug that 82542 misunderstand fiber's signal detection. - Add debug printf()s. - Update comment. - Rename functions and variables. - Add diagnostic code. - Sort registers. - Lowercase hexadecimal values. - KNF.
1.19.2.6	09-Mar-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #1372): sys/dev/pci/if_wm.c: revisions 1.462, 1.464-1.465 1.474 via patch sys/dev/pci/if_wmreg.h: revision 1.97 sys/dev/pci/if_wmvar.h: revision 1.33 - Set IPV6EXDIS bit in RFCTL register because of an Errata on 82575 and newer devices. - Linux and FreeBSD defines 0x10f5 as E1000_DEV_ID_ICH9_IGP_M_AMT. In reality, This is not IGP but BM. Add new case to identify PHY type device. Fixes PR#51924 reported byJarle Greipsland. - Use new wm_gmii_setup_phytype() function to setup sc_phytype and mii_{read\|write}reg. This change improves detection of PHY type. - Fix typo in comment.
1.19.2.5	12-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1302): sys/dev/mii/igphyreg.h: revisions 1.7-1.10 sys/dev/mii/ikphyreg.h: revisions 1.3 sys/dev/mii/inbmphyreg.h: revisions 1.4-1.9 sys/dev/mii/mii.h: revisions 1.19-1.20 sys/dev/pci/if_wm.c: revisions 1.390, 1.392-1.395, 1.397, 1.419-1.425, 1.427-1.428, 1.430-1.435, 1.437-1.453 via patch sys/dev/pci/if_wmreg.: revisions 1.89-1.93 via patch sys/dev/pci/if_wmvar.h: revisions 1.31-1.32 Update wm(4) up to if_wm.c rev. 1.453 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add I219 support. It's not stable so it's disabled by default. - wm_gate_hw_phy_config_ich8lan() is for younger than PCH2. - Drop the host wakeup bit after resetting PHY on PCH and newer devices. - Increase delay while toggling LANPHYPC - Move call of wm_reset() in wm_attach() after setting PHY and NVM related flags because those flags are used in wm_reset(). - Use mutex for NVM access on ICH8 and newer devices. Same as FreeBSD. - Rewrite PHY related lock stuff. Almost the same as FreeBSD. This change will fix a bug that PHY read/write fail on some cases. - Increase delay in wm_phy_resetisblocked(). Same as FreeBSD. - Use semaphore in wm_hv_phy_workaround_ich8lan() and wm_k1_gig_workaround_hv() - Use wm_gii_mdic_readreg/writereg() in wm_access_phy_wakeup_reg_bm() because these functions are called with taking lock. - 82567V_3 is BME1000_E_2(bm). Tested with Advantech AIMB-212 1st Ethernet port. - Use wm_gmii_82544_{read,write}reg() on non-82567 ICH8, 9 and 10. - Remove an 82578 workaround which was for PCH rev < 3. FreeBSD removed this workaround in r228386. - Add an 82578 workaround which is for PHY rev < 2. From FreeBSD and Linux. - Fix wm(4) input drop packet counter. WMREG_RNBC is incremented when there is no available buffers in host memory. However, ethernet controller can receive packets in such case if there is space in phy's FIFO. That is, ethernet controller drops packet only if there is no available buffers and there is no space in phy's FIFO. So, the number of dropped packets should be added WMREG_MPC only. - Use MII_ADDRMASK. - Define WMPHY_I217, WMPHY_VF and WMPHY_210. - Use BME1000_PHY_PAGE_SELECT in wm_gmii_bm_{read,write}reg(). This change has no effect because GG82563_PHY_PAGE_SELECT and BME1000_PHY_PAGE_SELECT have the same value. - Fix PHY access on 82567(ICH8 or ICH10), 82574 and 82583: - Use wm_gmii_bm_{read,write}reg() on 82574 and 82573. - Issue page select correctly on BM PHYs. - Fix workaround which did dummy read BM_WUC register. This code was changed to drop BM_WUC_HOST_WU_BIT of BM_PROT_GEN_CFG register in FreeBSD r228386. The code was added rev. 1.149, but the location was not the best. - wm_gmii_hv_{read/write}reg(): USE PHY address 1 for some special registers. - Add check code for an 82578 workaround. Not completed yet. - wm_release_hw_control(): Remove extra line. No any effect. - Add "10/100" into non-gigabit devices' name. - Call wm_enable_wakeup() in wm_detach() and wm_suspend(). Now wake on lan works on Thinkpad X61(ICH8). - Fix wm_access_phy_wakeup_reg_bm(). This change has no effect because this function is used for WUC register and our driver currenlty doesn't access to it. - Call wm_enable_phy_wakeup() on PCH2 and newer, too. Now these devices can do WOL. Tested with Thinkpad X220(PCH2). - Set CTRL_MEHE correctly (PCH_{LPT,SPT} only). - Add three workarounds for PCH_{LPT,SPT}. - Fix a bug that 8257[56], 82580, I35[04] and I21[01] didn't use wm_{get,release}_hw_control() correctly. - Sync wm_smbustopci() with Linux and FreeBSD. This change effects PCH and newer devices. - Move the location of wm_smbustopci() call. - Fix flag check in wm_get_wakeup() - 8254[17] and 8257[124] should not set WM_F_ARC_SUBSYS_VALID. - Add missing WM_T_82541_2 and WM_T_82547_2. - Fix WOL related setting of the WUC register for other than PCH* in wm_enable_wakeup(). Tested with 82567V(ICH8) and 82583V. - Use common MII_ADDRMASK. - igphy(4): No binary change: - s/IGPPHY/IGPHY/ - Fix the definition of PLHR_VALID_CHANNEL_* - Fix the definition of MSE_CHANNEL_* - Add MII_IGPHY_POWER_MGMT. - Add some KASSERT. - Add comment. Modify comment. - Add debug code.
1.19.2.4	26-Feb-2016	snj	branches: 1.19.2.4.2; Pull up following revision(s) (requested by msaitoh in ticket #1102): sys/dev/mii/igphy.c: 1.25 sys/dev/mii/ukphy.c: 1.48 sys/dev/pci/if_wm.c: revisions 1.308, 1.318, 1.320, 1.324-1.332, 1.334, 1.336, 1.343-1.344, 1.347-1.348, 1.350, 1.376-1.382, 1.386-1.389 via patch sys/dev/pci/if_wmreg.h: revisions 1.68-1.70, 1.73-1.77, 1.79-1.80, 1.82, 1.86-1.88 via patch sys/dev/pci/if_wmvar.h: revisions 1.22-1.23, 1.25-1.30 via patch Sync wm(4) as of if_wm.c rev 1.389 except MSI/MSI-X, multiqueue and NET_MPSAFE: - Add C2000 KX and 2.5G support. - Set ICH9 and ICH10's PBA size to 14K if the RX buffer size is more than 4096. Almost the same as other OSes. - For 82576 and newer devices, the PBA register is deleted. Don't write PBA for those chips. Also change the calculation of RX packet buffer size in new way. - Fix a lot of bugs to make 82575 and newer SERDES based systems work. - Print NVM image version and option ROM version. - Add workaround for I210 Errata 25 and I211 Errata 10 (PLL bug). This workaround is required if the NVM image version < 3.25. - Fix a bug that wm_detach() didn't unmap the FHASH's area. Now "drvct -d wm0" -> "drvctl -r pci0" works on ICH* and PCH. - Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets" and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image version 2.1.4 and newer have this workaround. - Check PHY type correctly. This change is required to use igphy(4) device correctly. - Disable LPLU (Low Power Link Up) on D0 state on 82574, 82583 and ICH too. - Call wm_get_hw_control() correctly. This change fixes a bug that some AMT based systems doesn't linkup at 1000BaseT. The problem was observed on HP Compaq dc7700. A lot of fixes have been done for wm(4) and igphy(4), so now PR#44893 should be fixed. - Call wm_get_wakeup(sc) before checking WM_F_HAS_AMT. It's required to check the existence of AMT correctly. - Fix a problem that wm_gate_hw_phy_config_ich8lan() isn't called in wm_reset() on PCH2. - Clear WMREG_WUC in wm_reset() if the chip >= 82544. This might fix the behavior on suspend/resume. - Fix logic of wm_check_reset_block() on ICH* and PCH*. This change might fix a problem that PHY's read/write functions can't get semaphore. - On ICH8, call wm_gig_downshift_workaround_ich8lan() when link changed down. - Drop PHY_CTRL_GBE_DIS explicitly in wm_lplu_d0_disable() in case BIOS sets this bit. - Fix two bugs in wm_kmrn_lock_loss_workaround_ich8lan(). Now the function checks the status correctly but it causes linkdown up to 10 times, so it's disabled for the time being. - PR/50527: David Binderman: Fix impossible code. Odd offsets need special treatment. - Fix RAL table's size of PCH2 and PCH_LPT. - PCH_LPT (and newer device) is required to check FWSM_WLOCK_MAC bit to determine the range of the RAL. - Use sc->sc_itr instead of hard-coded number. - Rename wm_tbi_check_link() to wm_tbi_tick() because this function acts as mii_tick(). - ACK Accelerate Disable in the RFCTL register is not bit 13 but 12. No binary change because this definition has not used yet. - Add ACK data Disable bit's definition (not used yet). - PHY_CTRL_GBE_DIS is not bit 4 but bit 6. This change has no any effect by default because WM_WOL is not defined yet and m_kmrn_lock_loss_workaround_ich8lan() is broken. - Fix wm_check_mng_mode_ich8lan(). This function is used only when WM_WOL is defined and it's disabled by default. - Rename wm_check_reset_block() to wm_phy_resetisblocked() and make it returns bool. No functional change. - Reorder function definitions and macro definitions. No functional change. - Fix comment. Add comment. Update comment. - KNF.
1.19.2.3	19-May-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #784): sys/dev/pci/if_wm.c: revision 1.321 sys/dev/pci/if_wmreg.h: revision 1.71 sys/dev/pci/if_wmvar.h: 1.24 via patch Add support iNVM (integrated Non-Volatile Memory) for I21[01]. This change fixes a bug that a MAC address is wrongly set on iNVM machines/NICs.
1.19.2.2	07-Nov-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #190): sys/dev/pci/pcidevs 1.1201-1.1202 via patch sys/dev/pci/if_wm.c 1.291-1.301, 1.304, 1.306-1.307 via patch sys/dev/pci/if_wmreg.h 1.61-1.65 via patch sys/dev/pci/if_wmvar.h 1.21 via patch share/man/man4/wm.4 1.31 via patch - Add Internal SERDES mode support newer than or equal to 82575. - Add new I218 devices. - Add DH89xxCC device. - Add some old devices. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Don't print "failed to detect NVM bank" message. - Delete 82580ER related code. - Fix typo in comment. - Fix debug message. - Cleanup
1.19.2.1	29-Aug-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #61): sys/dev/pci/if_wm.c: revision 1.290 sys/dev/pci/if_wmvar.h: revision 1.20 Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102.
1.19.2.4.2.2	13-Mar-2017	skrll	Sync with netbsd-7-1-RELEASE
1.19.2.4.2.1	18-Jan-2017	skrll	Sync with netbsd-5
1.23.2.6	28-Aug-2017	skrll	Sync with HEAD
1.23.2.5	05-Feb-2017	skrll	Sync with HEAD
1.23.2.4	05-Dec-2016	skrll	Sync with HEAD
1.23.2.3	29-May-2016	skrll	Sync with HEAD
1.23.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.23.2.1	06-Jun-2015	skrll	Sync with HEAD
1.31.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.31.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.32.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.33.6.12	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1915: sys/dev/pci/pcidevs 1.1497 sys/dev/pci/if_wm.c 1.689,1.790-1.791 via patch sys/dev/pci/if_wmreg.h 1.120 via patch sys/dev/pci/if_wmvar.h 1.51 - Use 12K for packet buffer for jumbo frame on PCH2 and newer. - Add new workaround for Tiger Lake and newer to avoid packet loss. - Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.33.6.11	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1894): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783,1.784 via patch Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.33.6.10	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1847: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.33.6.9	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1751: sys/dev/pci/if_wmreg.h 1.122-1.125 via patch sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720,1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.33.6.8	23-Oct-2021	martin	Pull up the following, requested by msaitoh in ticket #1701: share/man/man4/wm.4 1.42 sys/dev/pci/files.pci 1.434 sys/dev/pci/if_wm.c 1.686-1.687,1.692-1.693, 1.697,1.699-1.703, 1.706-1.715 via patch sys/dev/pci/if_wmvar.h 1.46 - Add missing drain for pcq in wm_stop_locked(). - Add support for I219V 15-19 and I219LM 16-19. - Fix Tx stall. - Use wm_flush_desc_rings() workaround more on I219. - Change DMA physical address in wm_flush_desc_rings() to match other OSes. - Check return value correctly in wm_lv_jumbo_workaround_ich8lan(). - Add new sysctl hw.wmN.debug_flags. This sysctl can be used if WM_DEBUG is set. - Add some sysctl values for debugging TX/RX queues. - WM_EVENT_COUNTER is enabled by default on 64 bit architectures. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxq->rxq_ptr updating. - Stop legacy interrupts before calling softint. - Disable printf()s in wm_flush_desc_rings() because the code is verified. - Print I219's version number. - Uniform INTx/MSI handler's Tx/Rx behavior to MSI-X's one. - Fix return value of interrupt handler. - Only print an error about missing I/O BARs for chips that need it. - Do not return a void value from a void function. - Add new flag named WM_F_CRC_STRIP and use it. No functional change. - Whitespace fixes. Fix comments. No functional change.
1.33.6.7	04-Nov-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #1620): sys/dev/pci/if_wm.c: revision 1.694 sys/dev/pci/if_wm.c: revision 1.695 (via patch) sys/dev/pci/if_wmvar.h: revision 1.47 Add WMPHY_I350. Not used yet. Workaround for ihphy and atphy(ICH/PCH, 82580 and I350). These phys stop DMA while link is down which causes device timeout. Fix PR/kern 40981 Reviewed and tested by msaitoh@n.o, thanks. XXX pullup-[89]
1.33.6.6	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1594: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678, 1.680-1.685 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 via patch sys/dev/pci/if_wmvar.c 1.45 via patch sys/dev/mii/igphy.c 1.35-1.36 via patch sys/dev/mii/igphyreg.h 1.12-1.13 sys/dev/mii/makphy.c 1.66 via patch sys/dev/mii/makphyreg.h 1.11 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler). Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Explicitly cast from uint16_t to uint32_t before shifting 16bit left when printing Image Unique ID to avoid undefined behavior. - Set if_baudrate for non-MII device. - Rename some macros and function. - KNF. Add comment.
1.33.6.5	07-Mar-2019	martin	Pull up the following, requested by msaitoh in ticket #1206: sys/dev/pci/if_wmreg.h 1.113 sys/dev/pci/if_wmvar.h 1.43-1.44 sys/dev/pci/if_wm.c 1.626-1.627, 1.629-1.636 via patch - Add support for I210 SGMII Flash-less device. - Add I219 variations for Cannon Lake. - Add missing rnd_detach_source(). - Use do { ... } while (0) for DPRINTF(x, y). - Swap enumeration of 82578 and 82577 PHY in if_wmvar.h. No functional change. - Add KASSERT. - Update TODO list. Add comment. Fix typo in comment.
1.33.6.4	31-Jan-2019	martin	Pull up the following, requested by msaitoh in ticket #1179: sys/dev/pci/if_wm.c 1.603-1.605,1.607-1.611, 1.613,1.615,1.618-1.620 via patch sys/dev/pci/if_wmreg.h 1.110-1.111 sys/dev/pci/if_wmvar.h 1.40-1.42 sys/dev/mii/inbmphyreg.h 1.13-1.15 - Add some code for suspend/resume: - Rename wm_smbustopci() to wm_init_phy_workarounds_pchlan(). It will also called when resume. - Call wm_phy_resetisblocked() after PHY reset in wm_init_phy_workarounds_pchlan() to wait for the PHY to quiesce to an accessible state. - Add new wm_resume_workarounds_pchlan() function and use it in wm_resume(). This workaround is only for PCH2 and newer. - Don't call wm_disable_aspm() neither in wm_attach() nor in wm_resume() but in wm_reset(). - Do some initialization in wm_resume() when IFF_UP is NOT set. - Don't continue when it failed to acquire semaphore in wm_ulp_disable(). - Print CLSEM workaround bit correctly. - Fix availability detection of WoL on some chips. - Print the WUS (WakeUp Status) register bits when resume. - Don't setup WoL on non-WoL capable port. - Setup PHY wakeup feature on PCH and newer. Tested on Thinkpad X220. - Remove an extra register read in wm_kmrn_lock_loss_workaround_ich8lan(). - Don't leave the MDICNFG register modified when the Power Management capability offset can't get. - Reduce indent level of wm_linkintr_gmii(). No functional change. - 80003's SERDES is not the same as 82575's but the same as legacy devices. Use the old methods on 80003. - Use __nothing for null DPRINTF(). - Rename functions. Add comment.
1.33.6.3	11-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #969): sys/dev/pci/if_wm.c: revision 1.584 sys/dev/pci/if_wmvar.h: revision 1.39 Add MDIC_WA and CLSEMWA to snprintb() string for WM_F_* flags. Fix two bugs for non-SERDES fiber. Tested on 82544EI(0x1009), 82545GM(0x1027), 82546GB(0x107a), 82571GB(0x105f), 82572EI(0x107e). Tested on SERDES 82580(0x150e (Winyao SERDES SFP)) and 82580(0x150e) as regression test: - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. When the signal is detected & STATUS_LU == 0 & RXCW == 0, it's required to set CTRL_SLU and CTRL_FD bit. wm_check_for_link() is the function to do it. If a link status change interrupt occurred, call wm_check_for_link() before checking current link status. - Fix a bug that 82542 misunderstand fiber's signal detection. CTRL_SWDPIN(1) is used for fiber's signal and it's value is reversed only on 82543 and 82544. Simplify the code and move it into a new function wm_tbi_havesignal() and use it.
1.33.6.2	16-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #764): sys/dev/pci/if_wm.c: revision 1.567 sys/dev/pci/if_wm.c: revision 1.568 sys/dev/pci/if_wm.c: revision 1.569 sys/dev/pci/if_wmvar.h: revision 1.38 sys/dev/pci/if_wm.c: revision 1.570 sys/dev/pci/if_wm.c: revision 1.571 sys/dev/pci/if_wm.c: revision 1.572 share/man/man4/wm.4: revision 1.40 sys/dev/pci/if_wmreg.h: revision 1.106 sys/dev/pci/if_wmreg.h: revision 1.107 SW PHY Config Enable bit for ICH8 B0 stepping is not bit 1 but bit 0. No binary change: - Sort registers. - Lowercase hexadecimal value. On PCH_SPT (and newer), FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it should be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. Our PCH_SPT part of wm_nvm_valid_bank_detect_ich8lan() was based on FreeBSD r287467. After that, they reverted it and committed the different code in r287762. r287762's bank detect code didn't work for us because our wm dirver had a problem in flash access. The problem was fixed in if_wm.c rev. 1.567, so we can use the new way now. If the extended configration size in the EXTCNFSIZE register is 0, don't continue. Add PCH_CNP support (I219 with Intel 300 series chipset). It's required more test, so it's disabled by default. Enable I219. I354 uses an external PHY, so don't use wm_set_eee_i350().
1.33.6.1	01-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #173): sys/dev/pci/if_wmreg.h: 1.99-1.103 sys/dev/pci/if_wmvar.h: 1.34-1.37 sys/dev/pci/if_wm.c: 1.510-1.537 sys/dev/mii/ihphyreg.h: 1.2 sys/dev/mii/inbmphyreg.h: 1.10 sys/dev/mii/igphy.c: 1.27 Sync wm(4) up to if_wm.c rev. 1.537: - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). - If a legacy interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Flush writing WMREG_CTRL in wm_gmii_mediachange(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Add I219 specific workaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for workaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - IF TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function is required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Configure the LCD with the extended configuration region in NVM if it's required. Tested with Thinkpad X220. - 8257[12]: Don't directly access SPI but use EERD register. - 82575-I354: If the size of SPI ROM >= 32K words, use direct SPI access instead of EERD register access. - Add wm_nvm_eec_clock_raise() and wm_nvm_eec_clock_lower() and use them for Microwire/SPI bus control. Same as Linux and FreeBSD. - Reduce timeout value for 80003 in wm_get_swfw_semaphore(). Same as Linux and FreeBSD. - Change API of kmrn_{read,write}reg() and check the return value. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally." on I21[01]. PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check NVM word size before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Print internal driver flags when attaching. - Add debug printf()s. - Rename variables. - Add comment, update comment and remove wrong comment.
1.37.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.37.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.37.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.38.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.38.2.1	10-Jun-2019	christos	Sync with HEAD
1.44.4.7	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1754): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.44.4.6	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #344): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783,1.784 via patch Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.44.4.5	27-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in #1656: sys/dev/pci/if_wm.c 1.768-1.782 via patch sys/dev/pci/if_wmreg.h 1.129-1.130 sys/dev/pci/if_wmvar.h 1.49 wm(4): - Rework for event counters: - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Rearrange the order of the registers so that they are roughly in ascending order. - Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. - IC{TX,RX}*C registers are for older than 82575. - Fix a bug that the transmit underrun counter is incorrectly counted. - Don't add "Count" for event counter's description. - Some statistics registers were replaced with new counters on newer chips. Treat 0x403c(CEXTERR->HTDPMC), 0x40fc(TSCTFC->CBRMPC), 0x4124(ICRXOC->HTCBDPC) and from 0x4104 to 0x4124. - Add some new counters: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - Host Good Octets RX - Host Good Octets TX - Length Errors - SerDes/SGMII Code Violation Packet - Header Redirection Missed Packet - EEE TX LPI - EEE RX LPI - Fix prc511's comment and description. - Add SOICZIFDATA (ifconfig -z) support for evcnt(9). - Use WM_IS_ICHPCH(). No functional change. - Fix typo. s/ictxact/ictxatc/. No functional change. - Add comment.
1.44.4.4	11-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1477: sys/dev/pci/if_wmreg.h 1.122-1.125 sys/dev/pci/if_wmvar.h 1.48 sys/dev/pci/if_wm.c 1.719-1.720, 1.722-1.725, 1.727-1.740 via patch - wm_tick: Add missing splx(s) when not WM_MPSAFE. - Print DMA range info if the system is booting in the verbose mode. - Micro optimization: - Call m_freem(m) only if m0 == NULL. - Call wm_xxeof() only when limit > 0. - Don't set the more flag when there is no packet to process. - No functional changes: - Call txeof first, then rxeof for the consistency. - Remove duplicated break. - Remove stray semicolons from struct declaration. - Fix value return from void function. - Use macros. - Modify comment. - KNF.
1.44.4.3	04-Nov-2020	martin	Pull up following revision(s) (requested by knakahara in ticket #1126): sys/dev/pci/if_wm.c: revision 1.694 sys/dev/pci/if_wm.c: revision 1.695 (via patch) sys/dev/pci/if_wmvar.h: revision 1.47 Add WMPHY_I350. Not used yet. Workaround for ihphy and atphy(ICH/PCH, 82580 and I350). These phys stop DMA while link is down which causes device timeout. Fix PR/kern 40981 Reviewed and tested by msaitoh@n.o, thanks. XXX pullup-[89]
1.44.4.2	23-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1093): sys/dev/pci/if_wmreg.h: revision 1.120 sys/dev/pci/if_wmvar.h: revision 1.46 sys/dev/pci/if_wm.c: revision 1.686 sys/dev/pci/if_wm.c: revision 1.687 sys/dev/mii/inbmphyreg.h: revision 1.20 sys/dev/pci/if_wm.c: revision 1.688 sys/dev/pci/if_wm.c: revision 1.689 s/ressource/resource/. Found by knakahara. Add new flag named WM_F_CRC_STRIP and use it. No functional change. This change also sets the RCTL_SECRC bit on I211 but it doesn't change the behavior because I211 always strips CRC like I35[04] and I210. Add a workaround for jumbo frame on PCH2 and newer. Tested by chs@. - Add wm_lv_jumbo_workaround_ich8lan() and use it. From FreeBSD. XXX For KUMCTRLSTA_OFFSET_HD_CTRL register modification, it's doubtful. FreeBSD and Linux do the same thing that they set the same value on both jumbo frame's enable case and the disable case. It seems the default value is 0x0b0c and it's not changed on the enable case, so it might be a bug on the enable case or the modification is not required. - Rename I219_UNKNOWN1 to I82579_UNKNOWN1. Use 12K for packet buffer for jumbo frame on PCH2 and newer. XXX Note that Linux Use 14K.
1.44.4.1	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #994: sys/dev/pci/if_wm.c 1.655-1.658, 1.660, 1.662, 1.664-1.668, 1.671-1.674, 1.678,1.680-1.681 via patch sys/dev/pci/if_wmreg.c 1.118-1.119 sys/dev/pci/if_wmvar.c 1.45 - Add SFP support. Module insertion/removal is not supported yet. Currently, SFP detection is only done in the driver's attach phase. - Detect the Media Auto Sense feature. Not supported yet. - Fix SFF_SFP_ETH_FLAGS_100FX. It's not 0x10 but 0x20. - Add extra delay in wm_serdes_power_up_link_82575(). - Add Intel I219 LM10-LM15 and V10-V14. - wm(4) can use workqueue as deferred Rx/Tx handler. Set hw.wm.txrx_workqueue=1 to use workqueue instead of softint. The default value of hw.wm.txrx_workqueue is 0 which use softint as before. - Unset RSS UDP flags like ixg(4) and other OSes. To handle IP fragmented UDP, first packet and second packet should be processed in the same Rx queue. - It's useless to not to set PCI_PMCSR_PME_STS bit when writing because the bit is W1C. Instead, always write PCI_PMCSR_PME_STS bit to clear in case it's already set. - Actually writing always the checksum offload context descriptor makes the HW do extra processing, avoid doing that if possible. - Fix a bug that the WMREG_EEARBC_I210 register is incorrectly set if the system uses iNVM. - "wmX: 0" on 82542 is difficult to understand, so don't print it. - Rename some macros and function. - KNF. Add comment.
1.48.4.3	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #421): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.48.4.2	04-Sep-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #344): sys/dev/pci/if_wmvar.h: revision 1.50 sys/dev/pci/if_wm.c: revision 1.783 sys/dev/pci/if_wm.c: revision 1.784 Delay sending LINK_STATE_UP to prevent dropping packets on I35[04] and I21[01]. Some (not all) systems use I35[04] or I21[01] don't send packet soon after linkup. The MAC send a packet to the PHY and any error is not observed. This behavior causes a problem that gratuitous ARP and/or IPv6 DAD packet are silently dropped. To avoid this problem, don't call mii_pollstat() here which will send LINK_STATE_UP notification to the upper layer. Instead, mii_pollstat() will be called in wm_gmii_mediastatus() or mii_tick() will be called in wm_tick(). Note that the similar workaround is in Linux's igb driver though it's only for I21[01]. OK'd by hikaru@ and knakahara@. Fix #ifdef WM_DEBUG code in wm_gmii_i82544_{read,write}reg_locked.
1.48.4.1	22-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #213): sys/dev/pci/if_wm.c: revision 1.770 sys/dev/pci/if_wm.c: revision 1.771 sys/dev/pci/if_wm.c: revision 1.772 sys/dev/pci/if_wm.c: revision 1.773 sys/dev/pci/if_wm.c: revision 1.774 sys/dev/pci/if_wm.c: revision 1.775 sys/dev/pci/if_wm.c: revision 1.776 sys/dev/pci/if_wmreg.h: revision 1.129 sys/dev/pci/if_wm.c: revision 1.777 sys/dev/pci/if_wm.c: revision 1.778 sys/dev/pci/if_wmvar.h: revision 1.49 sys/dev/pci/if_wm.c: revision 1.779 sys/dev/pci/if_wm.c: revision 1.768 sys/dev/pci/if_wm.c: revision 1.769 sys/dev/pci/if_wmreg.h: revision 1.130 sys/dev/pci/if_wm.c: revision 1.780 sys/dev/pci/if_wm.c: revision 1.781 Count some 64bit counters correctly. - Fix calculation of GORC, GOTC, TOR and TOT counters correctly. - Found by knakahara. Add note for the TORL register. The TOR register includes error, flow control and broadcast rejected. Sort lines. No functional change. Rearrange the order of the registers so that they are roughly in ascending order. Sort lines. No functional change. Reorder evcnt_attach_dynamic(), WM_EVCNT_ADD() and evcnt_detach() to match. IC{TX,RX}*C registers are for older than 82575. Fix a bug that the transmit underrun counter is incorrectly counted. The transmit underrun bit in the transmit status filed is only for 82544 (and older?), so don't use the counter for newer chips. The bit is reserved for newer chips, but the bit sometimes set on 82575 at least. Don't add "Count" for event counter's description. Some statistics registers were replaced with new counters. - 0x403c was CEXTERR(Carrier Extension Error). 82575 and newer except 80003, ICHs and PCHs have HTDPMC(Host Tx Discarded Packets by MAC). I don't really know for 82575. The 82575 datasheet say nothing about it. - The following two are changed for circuit breaker. Only 82576's datasheet describes abut it, so the registers might be only for 82576. Use those registers for 82575 and newer except 80003, ICHs and PCHs just in case. - 0x40fc was TSCTFC(TCP Segmentation Context Tx Fail). It was replaced by the CBRMPC(Circuit Breaker Rx Manageability Packet) register. - 0x4124 was ICRXOC(Interrupt Cause Receiver Overrun). It was replaced by the HTCBDPC(Host Tx Circuit Breaker Dropped Packet) register. - From 0x4104 to 0x4124: - For 0x4124, 82575's datasheet says it's not changed(ICRXOC). I don't know if it's correct. We use new HTCBDPC register for 82575. - 82576 and newer changed the meaning. - I don't know for 80003, ICHs and PCHs. Don't count those registers. At least, those registers in PCH2 and PCH_CNP are all zero. Add some new event counters. Add the following counters for 82575 and newer except 80003, ICHs and PCHs: - Only 82576 document describes about the circuit breaker, so the following two might be only for 82575: - Circuit Breaker TX Manageability Packet - Circuit Breaker RX Dropped Packet - 82575's document doesn't describe the following two, but we can see the same value as GO{T,R}C have: - Host Good Octets RX - Host Good Octets TX - 82575's document doesn't describe the LENERRS (Length Errors) counter. I don't know if it has. - Perhaps Non-SerDes/SGMII devices don't have SCVPC (SerDes/SGMII Code Violation Packet) register. We don't care if it's SerDes/SGMII or not for now. - HRMPC (Header Redirection Missed Packet) appears only once in 8257[56]'s datasheet. FreeBSD's igb counts it, so we do, too. - Count the following two for I350 and newer. I don't know if PCHs have: - EEE TX LPI - EEE RX LPI Move statistics updating code from wm_tick() to new wm_update_stats(). - To reuse. - No functional change. Add SOICZIFDATA (ifconfig -z) support for evcnt(9). First update the statistics data, then clear the event counters, and finally copy and clear if_data via ether_ioctl(). Fix prc511's comment and description. Use WM_IS_ICHPCH(). No functional change. Fix typo. s/ictxact/ictxatc/. No functional change.
1.92	05-Dec-2021	msaitoh	s/persistant/persistent/ in comment.
1.91	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.90	05-Feb-2021	christos	branches: 1.90.4; PR/55975: Riccardo Mottola: Don't try to lock a mutex from an interrupt context.
1.89	20-Mar-2020	sevan	branches: 1.89.4; Apply the same change as for if_iwi.c r1.114 here. This driver sleeps during wpi_media_change(), and thus requires an adaptive mutex for the media lock.
1.88	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.87	10-Nov-2019	chs	branches: 1.87.2; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.86	24-Jun-2019	jakllsch	Put back correct recieve ring allocation size which was lost just over five years ago mae culpa. Found by msaitoh@ Should fix PR kern/54320 once pulled up.
1.85	22-Dec-2018	maxv	Replace: M_COPY_PKTHDR -> m_copy_pkthdr. No functional change, since the former is a macro to the latter.
1.84	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.83	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.82	20-Aug-2018	riastradh	Suspend the getrfkill thread while we're suspended. Otherwise it tries to futz with device registers, which doesn't work, and who knows, maybe is bad.
1.81	20-Aug-2018	riastradh	What we have created, we must destroy.
1.80	26-Jun-2018	msaitoh	branches: 1.80.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.79	23-Oct-2017	msaitoh	branches: 1.79.2; If if_initialize() failed in the attach function, free resources and return.
1.78	23-May-2017	ozaki-r	branches: 1.78.2; Apply deferred if_start to more drivers And annotate some XXX_start as it runs in softint to clarify that it doesn't need deferred if_start.
1.77	02-Feb-2017	nonaka	wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.76	02-Feb-2017	jakllsch	wpi(4): use MSI if available.
1.75	08-Dec-2016	ozaki-r	branches: 1.75.2; Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.74	10-Jun-2016	ozaki-r	branches: 1.74.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.73	26-May-2016	ozaki-r	Introduce M_CLEARCTX and use it instead of open-coding rcvif No functional change.
1.72	26-May-2016	ozaki-r	Use M_GETCTX No functional change.
1.71	09-Jan-2015	bouyer	Take sc_rsw_mtx before calling wpi_getrfkill() from wpi_init(), Problem reported and patch tested by chris at chriswareham.net
1.70	06-Jan-2015	bouyer	As proposed on tech-net@, introduce a new switch type, PSWITCH_TYPE_RADIO, to be used to report to userland hardware radio switch changes. powerd(8) will call a "radio_button" script to handle the event. This script can e.g. start or stop wpa_supplicant. Update wpi(4) to report PSWITCH_TYPE_RADIO events to sysmon.
1.69	19-Dec-2014	bouyer	Consistently take the interface down when the radio swicth is off.
1.68	08-Aug-2014	jmcneill	branches: 1.68.2; 1.68.4; wpi_fix_channel hack no longer needed with recent scanning changes
1.67	07-Aug-2014	jmcneill	Scan one channel at a time instead of trying to do them all at once and trying to outsmart the ieee80211 state machine.
1.66	07-Aug-2014	jmcneill	scanning fixes
1.65	07-Aug-2014	jmcneill	simplify wpi_rx_intr rbuf handling and add some more bus_dmamap_sync; fixes stalls when downloading large files for me
1.64	05-Aug-2014	jmcneill	bus_dmamap_sync for rx descriptors too, derived from openbsd driver
1.63	05-Aug-2014	jmcneill	Sprinkle bus_dmamap_sync to make this work on my Thinkpad T61 (amd64) with 4GB RAM.
1.62	05-Jul-2014	jakllsch	Use M_ZERO. From OpenBSD if_wpi.c 1.53.
1.61	05-Jul-2014	jakllsch	Reduce wpi(4) diff to OpenBSD circa September 2007; mostly by following KNF. The only functional changes in this commit are some %d to %u debugging printf format specifier adjustments so as to match the OpenBSD code.
1.60	02-Jul-2014	jakllsch	No need to duplicate the members of the wpi_cmd_data structure within the wpi_scan_hdr structure when we can just put the wpi_cmd_data structure within the wpi_scan_hdr structure. This also brings our if_wpireg.h mostly in line with OpenBSD's 1.18.
1.59	16-Jun-2014	jakllsch	wpi(4): mostly cosmetic changes to source, also a little bit of diff reduction relative to OpenBSD as of late 2007.
1.58	29-Mar-2014	christos	branches: 1.58.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.57	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.56	30-Mar-2013	christos	branches: 1.56.4; replace function with macro
1.55	30-Mar-2013	christos	remove trailing whitespace
1.54	25-Nov-2012	riastradh	Rework firmware reference counting and error messages in wpi(4). . Clarify the shared firmware abstraction in wpi_cached_firmware and its new sibling wpi_release_firmware. . Fix typo in wpa_cache_firmware error branch leading to free NULL. . Fix leak in wpi_load_firmware error branch. . Sprinkle some kasserts to executably document invariants. . A little KNF here and there. Based on a patch from dh in PR kern/44144.
1.53	04-Aug-2012	riastradh	branches: 1.53.2; Fix error branch in wpi(4) firmware loading. Fixes panic if firmware is not available.
1.52	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.51	12-May-2012	khorben	No longer wrongly advertise ad-hoc (IBSS) mode as being supported. Fixes kern/46101 No objection from current-users@
1.50	30-Jan-2012	drochner	branches: 1.50.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.49	02-Apr-2011	mbalmer	branches: 1.49.4; 1.49.8; Fix misplaced parenthesis. From henning.petersen@t-online.de, thanks.
1.48	15-Nov-2010	uebayasi	branches: 1.48.2; tsleep needs sys/proc.h.
1.47	05-Apr-2010	joerg	Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.46	24-Feb-2010	dyoung	branches: 1.46.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.45	19-Jan-2010	pooka	branches: 1.45.2; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.44	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.43	05-Sep-2009	tsutsui	Invert logic around nested pmf(9) registrations for readability.
1.42	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.41	12-Nov-2008	joerg	branches: 1.41.4; Cache wpi firmware in memory, free it only on invalid content or when the last user is detached. This stops wpi from accessing the disk on resume.
1.40	07-Nov-2008	dyoung	* Summary * When a link-layer address changes (e.g., ifconfig ex0 link 02:de:ad:be:ef:02 active), send a gratuitous ARP and/or a Neighbor Advertisement to update the network-/link-layer address bindings on our LAN peers. Refuse a change of ethernet address to the address 00:00:00:00:00:00 or to any multicast/broadcast address. (Thanks matt@.) Reorder ifnet ioctl operations so that driver ioctls may inherit the functions of their "class"---ether_ioctl(), fddi_ioctl(), et cetera---and the class ioctls may inherit from the generic ioctl, ifioctl_common(), but both driver- and class-ioctls may override the generic behavior. Make network drivers share more code. Distinguish a "factory" link-layer address from others for the purposes of both protecting that address from deletion and computing EUI64. Return consistent, appropriate error codes from network drivers. Improve readability. KNF. * Details * In if_attach(), always initialize the interface ioctl routine, ifnet->if_ioctl, if the driver has not already initialized it. Delete if_ioctl == NULL tests everywhere else, because it cannot happen. In the ioctl routines of network interfaces, inherit common ioctl behaviors by calling either ifioctl_common() or whichever ioctl routine is appropriate for the class of interface---e.g., ether_ioctl() for ethernets. Stop (ab)using SIOCSIFADDR and start to use SIOCINITIFADDR. In the user->kernel interface, SIOCSIFADDR's argument was an ifreq, but on the protocol->ifnet interface, SIOCSIFADDR's argument was an ifaddr. That was confusing, and it would work against me as I make it possible for a network interface to overload most ioctls. On the protocol->ifnet interface, replace SIOCSIFADDR with SIOCINITIFADDR. In ifioctl(), return EPERM if userland tries to invoke SIOCINITIFADDR. In ifioctl(), give the interface the first shot at handling most interface ioctls, and give the protocol the second shot, instead of the other way around. Finally, let compatibility code (COMPAT_OSOCK) take a shot. Pull device initialization out of switch statements under SIOCINITIFADDR. For example, pull ..._init() out of any switch statement that looks like this: switch (...->sa_family) { case ...: ..._init(); ... break; ... default: ..._init(); ... break; } Rewrite many if-else clauses that handle all permutations of IFF_UP and IFF_RUNNING to use a switch statement, switch (x & (IFF_UP\|IFF_RUNNING)) { case 0: ... break; case IFF_RUNNING: ... break; case IFF_UP: ... break; case IFF_UP\|IFF_RUNNING: ... break; } unifdef lots of code containing #ifdef FreeBSD, #ifdef NetBSD, and #ifdef SIOCSIFMTU, especially in fwip(4) and in ndis(4). In ipw(4), remove an if_set_sadl() call that is out of place. In nfe(4), reuse the jumbo MTU logic in ether_ioctl(). Let ethernets register a callback for setting h/w state such as promiscuous mode and the multicast filter in accord with a change in the if_flags: ether_set_ifflags_cb() registers a callback that returns ENETRESET if the caller should reset the ethernet by calling if_init(), 0 on success, != 0 on failure. Pull common code from ex(4), gem(4), nfe(4), sip(4), tlp(4), vge(4) into ether_ioctl(), and register if_flags callbacks for those drivers. Return ENOTTY instead of EINVAL for inappropriate ioctls. In zyd(4), use ENXIO instead of ENOTTY to indicate that the device is not any longer attached. Add to if_set_sadl() a boolean 'factory' argument that indicates whether a link-layer address was assigned by the factory or some other source. In a comment, recommend using the factory address for generating an EUI64, and update in6_get_hw_ifid() to prefer a factory address to any other link-layer address. Add a routing message, RTM_LLINFO_UPD, that tells protocols to update the binding of network-layer addresses to link-layer addresses. Implement this message in IPv4 and IPv6 by sending a gratuitous ARP or a neighbor advertisement, respectively. Generate RTM_LLINFO_UPD messages on a change of an interface's link-layer address. In ether_ioctl(), do not let SIOCALIFADDR set a link-layer address that is broadcast/multicast or equal to 00:00:00:00:00:00. Make ether_ioctl() call ifioctl_common() to handle ioctls that it does not understand. In gif(4), initialize if_softc and use it, instead of assuming that the gif_softc and ifp overlap. Let ifioctl_common() handle SIOCGIFADDR. Sprinkle rtcache_invariants(), which checks on DIAGNOSTIC kernels that certain invariants on a struct route are satisfied. In agr(4), rewrite agr_ioctl_filter() to be a bit more explicit about the ioctls that we do not allow on an agr(4) member interface. bzero -> memset. Delete unnecessary casts to void . Use sockaddr_in_init() and sockaddr_in6_init(). Compare pointers with NULL instead of "testing truth". Replace some instances of (type )0 with NULL. Change some K&R prototypes to ANSI C, and join lines.
1.39	02-Jul-2008	cube	branches: 1.39.2; 1.39.4; 1.39.6; Protect private mbuf external data storage allocater with a mutex, as it was done in nfe(4), sk(4) and msk(4). And maybe some others. Confirmed to fix the usual issue of data corruption by Jared McNeill.
1.38	28-Apr-2008	drochner	branches: 1.38.2; 1.38.4; fix error handling after m_pullup()
1.37	11-Mar-2008	dyoung	branches: 1.37.2; 1.37.4; pci_disable_retry() is now a no-op, so don't call it any more.
1.36	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.35	19-Jan-2008	simonb	branches: 1.35.2; 1.35.6; Use M_80211_NODE as malloc type for a 80211 node as rest of net80211 code does, instead of M_DEVBUF. Fixes panics if KMEMSTATS are enabled. Also use M_ZERO instead of memset on malloc()'s results.
1.34	09-Jan-2008	degroote	Add some sysctl to retrieve the radio state (and the debug level). While here, fixe the return value in case where radio is off (suggested by joerg@).
1.33	09-Dec-2007	jmcneill	branches: 1.33.2; Merge jmcneill-pm branch.
1.32	01-Dec-2007	jmcneill	branches: 1.32.2; 1.32.4; Prefix the regulatory domain and address output with the device name.
1.31	28-Nov-2007	degroote	In wpi_init, check for the status of radio switch and print a useful message in this case.
1.30	23-Nov-2007	plunky	when using CFATTACH_DECL_NEW(), we must remember to record the device_t if we wish to use it. sc->sc_dev = self;
1.29	23-Nov-2007	joerg	If memory for the firmware image couldn't be allocated, print only one error message.
1.28	16-Nov-2007	degroote	Use device_t instead of struct device* and use associated functions Use callout_setfunc and callout_scheduler instead of callout_reset No functionnal change expected
1.27	07-Nov-2007	ad	Merge from vmlocking: - pool_cache changes. - Debugger/procfs locking fixes. - Other minor changes.
1.26	21-Oct-2007	degroote	branches: 1.26.2; 802.11 header may have different sizes. Deals correctly with this.
1.25	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.24	02-Sep-2007	degroote	branches: 1.24.4; The service queue isn't used anymore so don't allocate it. It saves a bit of memory and reduces diff with other BSD.
1.23	02-Sep-2007	degroote	Fix scanning code for wpi based on the iwi code. Keep track of the status of the scan. On a transition IEEE80211_S_SCAN -> IEEE80211_S_SCAN, don't stop the previous scan, finish the current scan. When we receive some frames in the IEEE80211_S_SCAN state, set current_channel based on the value advertised in beacons or probre reponse. Moreover, it fixes WPA issues for me.
1.22	01-Sep-2007	dyoung	Instead of IF_POLL()'ing and IF_DEQUEUE()'ing, just IF_DEQUEUE().
1.21	01-Sep-2007	dyoung	Change a bazillion occurrences of code resembling this, error = (cmd == SIOCADDMULTI) ? ether_addmulti(ifr, &sc->sc_ec) : ether_delmulti(ifr, &sc->sc_ec); if (error == ENETRESET) { to this, if ((error = ether_ioctl(ifp, cmd, data)) == ENETRESET) { which does the same thing. (A bazillion is a very large number. This seems to make the i386 ALL kernel smaller by 3kB to 4kB.) Use ifreq_getaddr() twice in es(4). Whitespace nits.
1.20	26-Aug-2007	dyoung	branches: 1.20.2; Constify: LLADDR -> CLLADDR. I'm aiming here to make it easier to identify sockaddr_dl abuse that remains in the kernel, especially the potential for overwriting memory past the end of a sockaddr_dl with, e.g., memcpy(LLADDR(), ...).
1.19	15-Aug-2007	degroote	Fix possible "packet corruption" which appears on high load
1.18	04-Aug-2007	degroote	branches: 1.18.2; Call wpi_stop at the beginning of wpi_init so we are sure that wpi is really stopped (in particular interrupt are disabled).
1.17	18-Jul-2007	degroote	branches: 1.17.4; change firmware loading code to adopt the new 2.14.4 firmware layout (from OpenBSD). While here, remove some dead code I added when I ported the code from OpenBSD. IMPORTANT : You must download the 2.14.4 firmware or update your sysutils/wpi-firmware2 to the last revision (2.14.4) or the driver will stop working.
1.16	11-Jul-2007	degroote	branches: 1.16.2; Fix more build issues on amd64 (exposed by WPI_DEBUG)
1.15	11-Jul-2007	xtraeme	Use %zu to print a size_t, fixes the build on amd64.
1.14	09-Jul-2007	ad	Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.13	09-Jul-2007	degroote	Fix some conflicts in my previous commit. Sorry for the bad commit Thanks to dogcow@ for the report.
1.12	09-Jul-2007	degroote	Sync the driver with the OpenBSD one 1/ Update the driver to use the new firmware images from Intel (2.14.3.) 2/ Read the list of supported channels from the EEPROM instead of having it hard-coded in the driver. 3/ Limit output power to what is specified in EEPROM. 4/ Decrease output power for highest OFDM rates to reduce distortion. 5/ Automatically adjust output power to temperature changes for increased throughput and range. 6/ Attach the adapter's onboard thermal sensor to the sensor framework. 7/ Replace 'magic' fields in structures with their correct definitions. 8/ Rewrite the firmware load in order to reduce the diff with OpenBSD one NOTE2: you must install sysutils/wpi-firmware2 in order to use the new driver NOTE2: if you are using a channel not allowed by the regulatory domain of your adapter, you will no longer be able to associate. Thanks a lot for his hard work to damien@OpenBSD.org
1.11	04-Jul-2007	pooka	wsize is size_t = unsigned and should be printed with %zu instead of %zd. cosmetics for PR 36591 by khorben.
1.10	18-Jun-2007	degroote	Add a workaround in the case where we have low number of rbuf. It seems to fix problem of frozen network with wpi. ok by @joerg.
1.9	04-Mar-2007	christos	branches: 1.9.2; 1.9.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.8	25-Jan-2007	njoly	branches: 1.8.2; Fix compilation on platforms where sizeof int/size_t differs. Successfully tested on amd64. ok by tron@
1.7	13-Jan-2007	degroote	Sync the wpi driver with the openbsd one. Thanks to Jean-Baptiste Campesato for his work. Fix PR/34463 ok dyoung@ joerg@
1.6	16-Nov-2006	christos	branches: 1.6.2; 1.6.4; 1.6.6; __unused removal on arguments; approved by core.
1.5	31-Oct-2006	joerg	Move AMRR code out of wpi(4) and ural(4) into net80211 itself. From OpenBSD.
1.4	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.3	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.2	13-Aug-2006	oster	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; Apply a change from damien in OpenBSD: fix a use-after-free (read) of a mbuf in wpi_tx_intr().
1.1	13-Aug-2006	simonb	Add a port of the OpenBSD Intel 3945ABG wpi(4) wireless driver, by Jean-Baptiste Campesato. From PR kern/33778.
1.2.8.2	10-Dec-2006	yamt	sync with head.
1.2.8.1	22-Oct-2006	yamt	sync with head
1.2.6.2	09-Sep-2006	rpaulo	sync with head
1.2.6.1	13-Aug-2006	rpaulo	file if_wpi.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:18 +0000
1.2.4.2	01-Feb-2007	ad	Sync with head.
1.2.4.1	18-Nov-2006	ad	Sync with head.
1.2.2.2	03-Sep-2006	yamt	sync with head.
1.2.2.1	13-Aug-2006	yamt	file if_wpi.c was added on branch yamt-pdpolicy on 2006-09-03 15:24:22 +0000
1.6.6.2	30-Sep-2007	wrstuden	Catch up on netbsd-4 as of a few days ago.
1.6.6.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.6.4.9	17-Mar-2008	yamt	sync with head.
1.6.4.8	21-Jan-2008	yamt	sync with head
1.6.4.7	07-Dec-2007	yamt	sync with head
1.6.4.6	15-Nov-2007	yamt	sync with head.
1.6.4.5	27-Oct-2007	yamt	sync with head.
1.6.4.4	03-Sep-2007	yamt	sync with head.
1.6.4.3	26-Feb-2007	yamt	sync with head.
1.6.4.2	30-Dec-2006	yamt	sync with head.
1.6.4.1	16-Nov-2006	yamt	file if_wpi.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:46 +0000
1.6.2.3	27-Sep-2007	xtraeme	Pull up following revision(s) (requested by degroote in ticket #860): sys/dev/pci/if_wpivar.h: revision 1.7 (via patch) sys/dev/pci/if_wpi.c: revision 1.18 (via patch) sys/dev/pci/if_wpi.c: revision 1.19 (via patch) sys/dev/pci/if_wpi.c: revision 1.23 (via patch) Call wpi_stop at the beginning of wpi_init so we are sure that wpi is really stopped (in particular interrupt are disabled). Fix possible "packet corruption" which appears on high load Fix scanning code for wpi based on the iwi code. Keep track of the status of the scan. On a transition IEEE80211_S_SCAN -> IEEE80211_S_SCAN, don't stop the previous scan, finish the current scan. When we receive some frames in the IEEE80211_S_SCAN state, set current_channel based on the value advertised in beacons or probre reponse. Moreover, it fixes WPA issues for me.
1.6.2.2	27-Jul-2007	liamjfoy	Pullup requested by degroote in ticket 795: Rewrite of the rbuf management Sync with the last OpenBSD driver Use the last firmware available from Intel (2.4.14). Use legal limitations and Tx regulation embedded in the eeprom. Rewrite the way we handle the firmware. Pullup the following revisions of this files sys/dev/pci/if_wpi.c rev 1.7 1.8 1.10 1.12 1.13 1.15 1.16 1.17 via patch sys/dev/pci/if_wpireg.h rev 1.3 1.4 1.5 via patch sys/dev/pci/if_wpivar.h rev 1.3 1.5 1.6 via patch
1.6.2.1	06-Jul-2007	liamjfoy	Pull up following revision(s) (requested by pooka in ticket #756) sys/dev/pci/if_wpi.c: revision 1.11 wsize is size_t = unsigned and should be printed with %zu instead of %zd. cosmetics for PR 36591 by khorben.
1.8.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.9.4.1	11-Jul-2007	mjf	Sync with head.
1.9.2.7	23-Oct-2007	ad	Sync with head.
1.9.2.6	09-Oct-2007	ad	Sync with head.
1.9.2.5	01-Sep-2007	ad	Update for pool_cache API changes.
1.9.2.4	20-Aug-2007	ad	Sync with HEAD.
1.9.2.3	15-Jul-2007	ad	Sync with head.
1.9.2.2	15-Jul-2007	ad	Sync with head.
1.9.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.16.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.16.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.17.4.15	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.17.4.14	03-Dec-2007	joerg	Sync with HEAD.
1.17.4.13	01-Dec-2007	jmcneill	Sync with HEAD.
1.17.4.12	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.17.4.11	21-Nov-2007	joerg	Sync with HEAD.
1.17.4.10	11-Nov-2007	joerg	Sync with HEAD.
1.17.4.9	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.17.4.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.17.4.7	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.17.4.6	04-Sep-2007	degroote	Convert wpi_power to use pci_net_generic_power. With this, I can use the card after resume.
1.17.4.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.17.4.4	02-Sep-2007	jmcneill	Sync with HEAD.
1.17.4.3	16-Aug-2007	jmcneill	Sync with HEAD.
1.17.4.2	09-Aug-2007	jmcneill	Sync with HEAD.
1.17.4.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.18.2.2	04-Aug-2007	degroote	Call wpi_stop at the beginning of wpi_init so we are sure that wpi is really stopped (in particular interrupt are disabled).
1.18.2.1	04-Aug-2007	degroote	file if_wpi.c was added on branch matt-mips64 on 2007-08-04 18:24:25 +0000
1.20.2.4	23-Mar-2008	matt	sync with HEAD
1.20.2.3	09-Jan-2008	matt	sync with HEAD
1.20.2.2	08-Nov-2007	matt	sync with -HEAD
1.20.2.1	06-Nov-2007	matt	sync with HEAD
1.24.4.3	18-Nov-2007	bouyer	Sync with HEAD
1.24.4.2	13-Nov-2007	bouyer	Sync with HEAD
1.24.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.26.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.26.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.26.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.26.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.32.4.1	11-Dec-2007	yamt	sync with head.
1.32.2.1	26-Dec-2007	ad	Sync with head.
1.33.2.2	19-Jan-2008	bouyer	Sync with HEAD
1.33.2.1	10-Jan-2008	bouyer	Sync with HEAD
1.35.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.35.6.3	02-Jul-2008	mjf	Sync with HEAD.
1.35.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.35.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.35.2.1	24-Mar-2008	keiichi	sync with head.
1.37.4.6	11-Aug-2010	yamt	sync with head.
1.37.4.5	11-Mar-2010	yamt	sync with head
1.37.4.4	16-Sep-2009	yamt	sync with head
1.37.4.3	16-May-2009	yamt	sync with head
1.37.4.2	04-May-2009	yamt	sync with head.
1.37.4.1	16-May-2008	yamt	sync with head.
1.37.2.1	18-May-2008	yamt	sync with head.
1.38.4.1	03-Jul-2008	simonb	Sync with head.
1.38.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.39.6.1	16-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #62): sys/dev/pci/if_wpivar.h: revision 1.13 sys/dev/pci/if_wpi.c: revision 1.41 Cache wpi firmware in memory, free it only on invalid content or when the last user is detached. This stops wpi from accessing the disk on resume.
1.39.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.39.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.41.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.45.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.46.2.3	21-Apr-2011	rmind	sync with head
1.46.2.2	05-Mar-2011	rmind	sync with head
1.46.2.1	30-May-2010	rmind	sync with head
1.48.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.49.8.2	02-Jun-2012	mrg	sync to latest -current.
1.49.8.1	18-Feb-2012	mrg	merge to -current.
1.49.4.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.49.4.4	16-Jan-2013	yamt	sync with (a bit old) head
1.49.4.3	30-Oct-2012	yamt	sync with head
1.49.4.2	23-May-2012	yamt	sync with head.
1.49.4.1	17-Apr-2012	yamt	sync with head
1.50.2.4	08-Feb-2013	riz	Pull up following revision(s) (requested by riastradh in ticket #784): sys/dev/pci/if_wpi.c: revision 1.54 Rework firmware reference counting and error messages in wpi(4). . Clarify the shared firmware abstraction in wpi_cached_firmware and its new sibling wpi_release_firmware. . Fix typo in wpa_cache_firmware error branch leading to free NULL. . Fix leak in wpi_load_firmware error branch. . Sprinkle some kasserts to executably document invariants. . A little KNF here and there. Based on a patch from dh in PR kern/44144.
1.50.2.3	12-Aug-2012	martin	Pull up following revision(s) (requested by riastradh in ticket #478): sys/dev/pci/if_wpi.c: revision 1.53 Fix error branch in wpi(4) firmware loading. Fixes panic if firmware is not available.
1.50.2.2	22-May-2012	riz	Pull up following revision(s) (requested by khorben in ticket #266): sys/dev/pci/if_wpi.c: patch Fix harmless typo which was previously pulled up.
1.50.2.1	19-May-2012	riz	Pull up following revision(s) (requested by khorben in ticket #266): sys/dev/pci/if_wpi.c: revision 1.51 No longer wrongly advertise ad-hoc (IBSS) mode as being supported. Fixes kern/46101 No objection from current-users@
1.53.2.4	03-Dec-2017	jdolecek	update from HEAD
1.53.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.53.2.2	23-Jun-2013	tls	resync from head
1.53.2.1	25-Feb-2013	tls	resync with head
1.56.4.1	18-May-2014	rmind	sync with head
1.58.2.1	10-Aug-2014	tls	Rebase.
1.68.4.5	28-Aug-2017	skrll	Sync with HEAD
1.68.4.4	05-Feb-2017	skrll	Sync with HEAD
1.68.4.3	09-Jul-2016	skrll	Sync with HEAD
1.68.4.2	29-May-2016	skrll	Sync with HEAD
1.68.4.1	06-Apr-2015	skrll	Sync with HEAD
1.68.2.2	12-Jan-2015	snj	Pull up following revision(s) (requested by bouyer in ticket #415): share/man/man9/sysmon_pswitch.9: revision 1.6 sys/dev/pci/if_wpi.c: revision 1.70, 1.71 sys/dev/pci/if_wpivar.h: revision 1.19 sys/dev/sysmon/sysmon_power.c: revision 1.49 sys/sys/power.h: revision 1.20 As proposed on tech-net@, introduce a new switch type, PSWITCH_TYPE_RADIO, to be used to report to userland hardware radio switch changes. powerd(8) will call a "radio_button" script to handle the event. This script can e.g. start or stop wpa_supplicant. Update wpi(4) to report PSWITCH_TYPE_RADIO events to sysmon. -- Take sc_rsw_mtx before calling wpi_getrfkill() from wpi_init(), Problem reported and patch tested by chris at chriswareham.net
1.68.2.1	21-Dec-2014	snj	Pull up following revision(s) (requested by bouyer in ticket #338): sys/dev/pci/if_wpi.c: revision 1.69 Consistently take the interface down when the radio swicth is off.
1.74.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.74.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.75.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.78.2.2	27-Jul-2019	martin	Pull up following revision(s) (requested by jakllsch in ticket #1308): sys/dev/pci/if_wpi.c: revision 1.86 Put back correct recieve ring allocation size which was lost just over five years ago mae culpa. Found by msaitoh@ Should fix PR kern/54320 once pulled up.
1.78.2.1	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.79.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.79.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.79.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.80.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.80.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.80.2.1	10-Jun-2019	christos	Sync with HEAD
1.87.2.1	29-Feb-2020	ad	Sync with head.
1.89.4.1	03-Apr-2021	thorpej	Sync with HEAD.
1.90.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.13	02-Jul-2014	jakllsch	fix boot code maximum size. From OpenBSD if_wpireg.h 1.19.
1.12	02-Jul-2014	jakllsch	No need to duplicate the members of the wpi_cmd_data structure within the wpi_scan_hdr structure when we can just put the wpi_cmd_data structure within the wpi_scan_hdr structure. This also brings our if_wpireg.h mostly in line with OpenBSD's 1.18.
1.11	30-Jun-2014	jakllsch	Apply OpenBSD src/sys/dev/pci/if_wpireg.h 1.17.
1.10	30-Jun-2014	jakllsch	Sync with OpenBSD src/sys/dev/pci/if_wpireg.h 1.16.
1.9	25-Dec-2007	perry	branches: 1.9.54; 1.9.70; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.8	28-Nov-2007	degroote	branches: 1.8.2; 1.8.6; In wpi_init, check for the status of radio switch and print a useful message in this case.
1.7	21-Oct-2007	degroote	branches: 1.7.2; 802.11 header may have different sizes. Deals correctly with this.
1.6	02-Sep-2007	degroote	branches: 1.6.4; The service queue isn't used anymore so don't allocate it. It saves a bit of memory and reduces diff with other BSD.
1.5	18-Jul-2007	degroote	branches: 1.5.4; 1.5.6; 1.5.8; change firmware loading code to adopt the new 2.14.4 firmware layout (from OpenBSD). While here, remove some dead code I added when I ported the code from OpenBSD. IMPORTANT : You must download the 2.14.4 firmware or update your sysutils/wpi-firmware2 to the last revision (2.14.4) or the driver will stop working.
1.4	09-Jul-2007	degroote	branches: 1.4.2; Sync the driver with the OpenBSD one 1/ Update the driver to use the new firmware images from Intel (2.14.3.) 2/ Read the list of supported channels from the EEPROM instead of having it hard-coded in the driver. 3/ Limit output power to what is specified in EEPROM. 4/ Decrease output power for highest OFDM rates to reduce distortion. 5/ Automatically adjust output power to temperature changes for increased throughput and range. 6/ Attach the adapter's onboard thermal sensor to the sensor framework. 7/ Replace 'magic' fields in structures with their correct definitions. 8/ Rewrite the firmware load in order to reduce the diff with OpenBSD one NOTE2: you must install sysutils/wpi-firmware2 in order to use the new driver NOTE2: if you are using a channel not allowed by the regulatory domain of your adapter, you will no longer be able to associate. Thanks a lot for his hard work to damien@OpenBSD.org
1.3	13-Jan-2007	degroote	branches: 1.3.6; 1.3.8; Sync the wpi driver with the openbsd one. Thanks to Jean-Baptiste Campesato for his work. Fix PR/34463 ok dyoung@ joerg@
1.2	31-Oct-2006	joerg	branches: 1.2.2; 1.2.4; 1.2.6; Move AMRR code out of wpi(4) and ural(4) into net80211 itself. From OpenBSD.
1.1	13-Aug-2006	simonb	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Add a port of the OpenBSD Intel 3945ABG wpi(4) wireless driver, by Jean-Baptiste Campesato. From PR kern/33778.
1.1.8.1	10-Dec-2006	yamt	sync with head.
1.1.6.2	09-Sep-2006	rpaulo	sync with head
1.1.6.1	13-Aug-2006	rpaulo	file if_wpireg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:18 +0000
1.1.4.2	01-Feb-2007	ad	Sync with head.
1.1.4.1	18-Nov-2006	ad	Sync with head.
1.1.2.2	03-Sep-2006	yamt	sync with head.
1.1.2.1	13-Aug-2006	yamt	file if_wpireg.h was added on branch yamt-pdpolicy on 2006-09-03 15:24:22 +0000
1.2.6.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.2.4.7	21-Jan-2008	yamt	sync with head
1.2.4.6	07-Dec-2007	yamt	sync with head
1.2.4.5	27-Oct-2007	yamt	sync with head.
1.2.4.4	03-Sep-2007	yamt	sync with head.
1.2.4.3	26-Feb-2007	yamt	sync with head.
1.2.4.2	30-Dec-2006	yamt	sync with head.
1.2.4.1	31-Oct-2006	yamt	file if_wpireg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:46 +0000
1.2.2.1	27-Jul-2007	liamjfoy	Pullup requested by degroote in ticket 795: Rewrite of the rbuf management Sync with the last OpenBSD driver Use the last firmware available from Intel (2.4.14). Use legal limitations and Tx regulation embedded in the eeprom. Rewrite the way we handle the firmware. Pullup the following revisions of this files sys/dev/pci/if_wpi.c rev 1.7 1.8 1.10 1.12 1.13 1.15 1.16 1.17 via patch sys/dev/pci/if_wpireg.h rev 1.3 1.4 1.5 via patch sys/dev/pci/if_wpivar.h rev 1.3 1.5 1.6 via patch
1.3.8.1	11-Jul-2007	mjf	Sync with head.
1.3.6.3	09-Oct-2007	ad	Sync with head.
1.3.6.2	20-Aug-2007	ad	Sync with HEAD.
1.3.6.1	15-Jul-2007	ad	Sync with head.
1.4.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.4.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.5.8.2	18-Jul-2007	degroote	change firmware loading code to adopt the new 2.14.4 firmware layout (from OpenBSD). While here, remove some dead code I added when I ported the code from OpenBSD. IMPORTANT : You must download the 2.14.4 firmware or update your sysutils/wpi-firmware2 to the last revision (2.14.4) or the driver will stop working.
1.5.8.1	18-Jul-2007	degroote	file if_wpireg.h was added on branch matt-mips64 on 2007-07-18 18:49:18 +0000
1.5.6.2	09-Jan-2008	matt	sync with HEAD
1.5.6.1	06-Nov-2007	matt	sync with HEAD
1.5.4.3	03-Dec-2007	joerg	Sync with HEAD.
1.5.4.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.4.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.6.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.7.2.2	18-Feb-2008	mjf	Sync with HEAD.
1.7.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.8.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.8.2.1	26-Dec-2007	ad	Sync with head.
1.9.70.1	10-Aug-2014	tls	Rebase.
1.9.54.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.23	20-Mar-2020	sevan	Apply the same change as for if_iwi.c r1.114 here. This driver sleeps during wpi_media_change(), and thus requires an adaptive mutex for the media lock.
1.22	20-Aug-2018	riastradh	Suspend the getrfkill thread while we're suspended. Otherwise it tries to futz with device registers, which doesn't work, and who knows, maybe is bad.
1.21	02-Feb-2017	nonaka	branches: 1.21.12; 1.21.14; wlan interfaces make interrupt routine running on softint context. see http://mail-index.netbsd.org/tech-kern/2016/12/06/msg021281.html tested device: * ath at pci: AR5212, AR5424 * athn at pci: AR9287 * ipw at pci: 2100BG * iwi at pci: 2915ABG * iwm at pci: 3165, 7260, 8260 * iwn at pci: 4945, 6235 * ral at pci: RT2560 * rtwn at pci: RTL8192CE
1.20	02-Feb-2017	jakllsch	wpi(4): use MSI if available.
1.19	06-Jan-2015	bouyer	branches: 1.19.2; 1.19.4; As proposed on tech-net@, introduce a new switch type, PSWITCH_TYPE_RADIO, to be used to report to userland hardware radio switch changes. powerd(8) will call a "radio_button" script to handle the event. This script can e.g. start or stop wpa_supplicant. Update wpi(4) to report PSWITCH_TYPE_RADIO events to sysmon.
1.18	09-Aug-2014	jmcneill	branches: 1.18.2; 1.18.4; increase WPI_RBUF_COUNT
1.17	07-Aug-2014	jmcneill	simplify wpi_rx_intr rbuf handling and add some more bus_dmamap_sync; fixes stalls when downloading large files for me
1.16	05-Aug-2014	jmcneill	bus_dmamap_sync for rx descriptors too, derived from openbsd driver
1.15	19-Jan-2010	pooka	branches: 1.15.22; 1.15.36; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.14	17-Jan-2010	pooka	Forward declare struct bpf_if and use that as the type for bpf_if instead of "void *". Buys us oo times the type-safety for 0 times the price. (no functional change)
1.13	12-Nov-2008	joerg	Cache wpi firmware in memory, free it only on invalid content or when the last user is detached. This stops wpi from accessing the disk on resume.
1.12	02-Jul-2008	cube	branches: 1.12.2; 1.12.4; 1.12.6; Protect private mbuf external data storage allocater with a mutex, as it was done in nfe(4), sk(4) and msk(4). And maybe some others. Confirmed to fix the usual issue of data corruption by Jared McNeill.
1.11	09-Jan-2008	degroote	branches: 1.11.6; 1.11.10; 1.11.12; 1.11.14; Add some sysctl to retrieve the radio state (and the debug level). While here, fixe the return value in case where radio is off (suggested by joerg@).
1.10	09-Dec-2007	jmcneill	branches: 1.10.2; Merge jmcneill-pm branch.
1.9	16-Nov-2007	degroote	branches: 1.9.2; 1.9.4; Use device_t instead of struct device* and use associated functions Use callout_setfunc and callout_scheduler instead of callout_reset No functionnal change expected
1.8	02-Sep-2007	degroote	branches: 1.8.4; 1.8.6; The service queue isn't used anymore so don't allocate it. It saves a bit of memory and reduces diff with other BSD.
1.7	02-Sep-2007	degroote	Fix scanning code for wpi based on the iwi code. Keep track of the status of the scan. On a transition IEEE80211_S_SCAN -> IEEE80211_S_SCAN, don't stop the previous scan, finish the current scan. When we receive some frames in the IEEE80211_S_SCAN state, set current_channel based on the value advertised in beacons or probre reponse. Moreover, it fixes WPA issues for me.
1.6	09-Jul-2007	degroote	branches: 1.6.2; 1.6.6; 1.6.8; Sync the driver with the OpenBSD one 1/ Update the driver to use the new firmware images from Intel (2.14.3.) 2/ Read the list of supported channels from the EEPROM instead of having it hard-coded in the driver. 3/ Limit output power to what is specified in EEPROM. 4/ Decrease output power for highest OFDM rates to reduce distortion. 5/ Automatically adjust output power to temperature changes for increased throughput and range. 6/ Attach the adapter's onboard thermal sensor to the sensor framework. 7/ Replace 'magic' fields in structures with their correct definitions. 8/ Rewrite the firmware load in order to reduce the diff with OpenBSD one NOTE2: you must install sysutils/wpi-firmware2 in order to use the new driver NOTE2: if you are using a channel not allowed by the regulatory domain of your adapter, you will no longer be able to associate. Thanks a lot for his hard work to damien@OpenBSD.org
1.5	18-Jun-2007	degroote	Add a workaround in the case where we have low number of rbuf. It seems to fix problem of frozen network with wpi. ok by @joerg.
1.4	04-Mar-2007	christos	branches: 1.4.2; 1.4.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	13-Jan-2007	degroote	branches: 1.3.2; Sync the wpi driver with the openbsd one. Thanks to Jean-Baptiste Campesato for his work. Fix PR/34463 ok dyoung@ joerg@
1.2	31-Oct-2006	joerg	branches: 1.2.2; 1.2.4; 1.2.6; Move AMRR code out of wpi(4) and ural(4) into net80211 itself. From OpenBSD.
1.1	13-Aug-2006	simonb	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Add a port of the OpenBSD Intel 3945ABG wpi(4) wireless driver, by Jean-Baptiste Campesato. From PR kern/33778.
1.1.8.1	10-Dec-2006	yamt	sync with head.
1.1.6.2	09-Sep-2006	rpaulo	sync with head
1.1.6.1	13-Aug-2006	rpaulo	file if_wpivar.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:18 +0000
1.1.4.2	01-Feb-2007	ad	Sync with head.
1.1.4.1	18-Nov-2006	ad	Sync with head.
1.1.2.2	03-Sep-2006	yamt	sync with head.
1.1.2.1	13-Aug-2006	yamt	file if_wpivar.h was added on branch yamt-pdpolicy on 2006-09-03 15:24:22 +0000
1.2.6.2	30-Sep-2007	wrstuden	Catch up on netbsd-4 as of a few days ago.
1.2.6.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.2.4.6	21-Jan-2008	yamt	sync with head
1.2.4.5	07-Dec-2007	yamt	sync with head
1.2.4.4	03-Sep-2007	yamt	sync with head.
1.2.4.3	26-Feb-2007	yamt	sync with head.
1.2.4.2	30-Dec-2006	yamt	sync with head.
1.2.4.1	31-Oct-2006	yamt	file if_wpivar.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:46 +0000
1.2.2.2	27-Sep-2007	xtraeme	Pull up following revision(s) (requested by degroote in ticket #860): sys/dev/pci/if_wpivar.h: revision 1.7 (via patch) sys/dev/pci/if_wpi.c: revision 1.18 (via patch) sys/dev/pci/if_wpi.c: revision 1.19 (via patch) sys/dev/pci/if_wpi.c: revision 1.23 (via patch) Call wpi_stop at the beginning of wpi_init so we are sure that wpi is really stopped (in particular interrupt are disabled). Fix possible "packet corruption" which appears on high load Fix scanning code for wpi based on the iwi code. Keep track of the status of the scan. On a transition IEEE80211_S_SCAN -> IEEE80211_S_SCAN, don't stop the previous scan, finish the current scan. When we receive some frames in the IEEE80211_S_SCAN state, set current_channel based on the value advertised in beacons or probre reponse. Moreover, it fixes WPA issues for me.
1.2.2.1	27-Jul-2007	liamjfoy	Pullup requested by degroote in ticket 795: Rewrite of the rbuf management Sync with the last OpenBSD driver Use the last firmware available from Intel (2.4.14). Use legal limitations and Tx regulation embedded in the eeprom. Rewrite the way we handle the firmware. Pullup the following revisions of this files sys/dev/pci/if_wpi.c rev 1.7 1.8 1.10 1.12 1.13 1.15 1.16 1.17 via patch sys/dev/pci/if_wpireg.h rev 1.3 1.4 1.5 via patch sys/dev/pci/if_wpivar.h rev 1.3 1.5 1.6 via patch
1.3.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.4.4.1	11-Jul-2007	mjf	Sync with head.
1.4.2.2	09-Oct-2007	ad	Sync with head.
1.4.2.1	15-Jul-2007	ad	Sync with head.
1.6.8.3	23-Mar-2008	matt	sync with HEAD
1.6.8.2	09-Jan-2008	matt	sync with HEAD
1.6.8.1	06-Nov-2007	matt	sync with HEAD
1.6.6.7	21-Nov-2007	joerg	Sync with HEAD.
1.6.6.6	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.6.6.5	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.6.6.4	04-Sep-2007	degroote	Convert wpi_power to use pci_net_generic_power. With this, I can use the card after resume.
1.6.6.3	03-Sep-2007	jmcneill	Sync with HEAD.
1.6.6.2	02-Sep-2007	jmcneill	Sync with HEAD.
1.6.6.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.6.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.8.6.3	18-Feb-2008	mjf	Sync with HEAD.
1.8.6.2	27-Dec-2007	mjf	Sync with HEAD.
1.8.6.1	19-Nov-2007	mjf	Sync with HEAD.
1.8.4.1	18-Nov-2007	bouyer	Sync with HEAD
1.9.4.1	11-Dec-2007	yamt	sync with head.
1.9.2.1	26-Dec-2007	ad	Sync with head.
1.10.2.1	10-Jan-2008	bouyer	Sync with HEAD
1.11.14.1	03-Jul-2008	simonb	Sync with head.
1.11.12.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.11.10.2	11-Mar-2010	yamt	sync with head
1.11.10.1	04-May-2009	yamt	sync with head.
1.11.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.11.6.1	02-Jul-2008	mjf	Sync with HEAD.
1.12.6.1	16-Nov-2008	snj	Pull up following revision(s) (requested by joerg in ticket #62): sys/dev/pci/if_wpivar.h: revision 1.13 sys/dev/pci/if_wpi.c: revision 1.41 Cache wpi firmware in memory, free it only on invalid content or when the last user is detached. This stops wpi from accessing the disk on resume.
1.12.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.12.2.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.15.36.1	10-Aug-2014	tls	Rebase.
1.15.22.2	03-Dec-2017	jdolecek	update from HEAD
1.15.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.18.4.2	05-Feb-2017	skrll	Sync with HEAD
1.18.4.1	06-Apr-2015	skrll	Sync with HEAD
1.18.2.1	12-Jan-2015	snj	Pull up following revision(s) (requested by bouyer in ticket #415): share/man/man9/sysmon_pswitch.9: revision 1.6 sys/dev/pci/if_wpi.c: revision 1.70, 1.71 sys/dev/pci/if_wpivar.h: revision 1.19 sys/dev/sysmon/sysmon_power.c: revision 1.49 sys/sys/power.h: revision 1.20 As proposed on tech-net@, introduce a new switch type, PSWITCH_TYPE_RADIO, to be used to report to userland hardware radio switch changes. powerd(8) will call a "radio_button" script to handle the event. This script can e.g. start or stop wpa_supplicant. Update wpi(4) to report PSWITCH_TYPE_RADIO events to sysmon. -- Take sc_rsw_mtx before calling wpi_getrfkill() from wpi_init(), Problem reported and patch tested by chris at chriswareham.net
1.19.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.19.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.21.14.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.21.14.1	10-Jun-2019	christos	Sync with HEAD
1.21.12.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.35	24-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.34	01-Mar-2020	thorpej	Use the 64-bit PCI DMA tag if available.
1.33	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.32	29-May-2019	msaitoh	branches: 1.32.4; Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.31	28-May-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() for all ethernet drivers to protect ec_multi*.
1.30	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.29	26-Apr-2019	msaitoh	No functional change: - u_int_{8,16,32}_t -> uint_{8,16,32}_t - KNF. - Tabify. - Remove extra space.
1.28	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.27	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.26	26-Jun-2018	msaitoh	branches: 1.26.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.25	15-Dec-2016	ozaki-r	branches: 1.25.14; Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.24	08-Dec-2016	ozaki-r	Apply deferred if_start framework if_schedule_deferred_start checks if the if_snd queue contains packets, so drivers don't need to check it by themselves.
1.23	14-Jul-2016	msaitoh	branches: 1.23.2; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.22	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.21	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.20	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.19	29-Mar-2014	christos	branches: 1.19.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	30-Mar-2013	christos	branches: 1.18.4; remove trailing whitespace
1.17	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.16	02-Feb-2012	tls	branches: 1.16.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.15	05-Apr-2010	joerg	branches: 1.15.8; 1.15.12; Push the bpf_ops usage back into bpf.h. Push the common ifp->if_bpf check into the inline functions as well the fourth argument for bpf_attach.
1.14	19-Jan-2010	pooka	branches: 1.14.2; 1.14.4; Redefine bpf linkage through an always present op vector, i.e. #if NBPFILTER is no longer required in the client. This change doesn't yet add support for loading bpf as a module, since drivers can register before bpf is attached. However, callers of bpf can now be modularized. Dynamically loadable bpf could probably be done fairly easily with coordination from the stub driver and the real driver by registering attachments in the stub before the real driver is loaded and doing a handoff. ... and I'm not going to ponder the depths of unload here. Tested with i386/MONOLITHIC, modified MONOLITHIC without bpf and rump.
1.13	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.12	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.11	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.10	16-Dec-2008	christos	branches: 1.10.2; replace bitmask_snprintf(9) with snprintb(3)
1.9	10-Apr-2008	cegger	branches: 1.9.4; 1.9.12; use aprint_*_dev and device_xname
1.8	07-Feb-2008	dyoung	branches: 1.8.6; Start patching up the kernel so that a network driver always has the opportunity to handle an ioctl before generic ifioctl handling occurs. This will ease extending the kernel and sharing of code between drivers. First steps: Make the signature of ifioctl_common() match struct ifinet->if_ioctl. Convert SIOCSIFCAP and SIOCSIFMTU to the new ifioctl() regime, throughout the kernel.
1.7	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.6	19-Oct-2007	ad	branches: 1.6.2; 1.6.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	04-Mar-2007	christos	branches: 1.5.2; 1.5.14; 1.5.16; 1.5.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.4	16-Nov-2006	christos	branches: 1.4.4; __unused removal on arguments; approved by core.
1.3	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.2	11-Dec-2005	christos	branches: 1.2.18; 1.2.22; 1.2.24; merge ktrace-lwp.
1.1	09-Sep-2005	ragge	branches: 1.1.2; 1.1.8; Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.1.8.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.8.1	09-Sep-2005	skrll	file if_xge.c was added on branch ktrace-lwp on 2005-11-10 14:06:02 +0000
1.1.2.2	13-Sep-2005	tron	Pull up following revision(s) (requested by ragge in ticket #768): sys/dev/pci/if_xge.c: revision 1.1 via patch sys/dev/pci/if_xgereg.h: revision 1.1 sys/dev/pci/files.pci: revision 1.238 Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.1.2.1	09-Sep-2005	tron	file if_xge.c was added on branch netbsd-3 on 2005-09-13 20:59:37 +0000
1.2.24.2	10-Dec-2006	yamt	sync with head.
1.2.24.1	22-Oct-2006	yamt	sync with head
1.2.22.1	18-Nov-2006	ad	Sync with head.
1.2.18.6	11-Feb-2008	yamt	sync with head.
1.2.18.5	21-Jan-2008	yamt	sync with head
1.2.18.4	27-Oct-2007	yamt	sync with head.
1.2.18.3	03-Sep-2007	yamt	sync with head.
1.2.18.2	21-Jun-2006	yamt	sync with head.
1.2.18.1	11-Dec-2005	yamt	file if_xge.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:05 +0000
1.4.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.5.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.16.3	23-Mar-2008	matt	sync with HEAD
1.5.16.2	09-Jan-2008	matt	sync with HEAD
1.5.16.1	06-Nov-2007	matt	sync with HEAD
1.5.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.2.1	23-Oct-2007	ad	Sync with head.
1.6.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.6.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.8.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.8.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.9.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.9.4.4	11-Aug-2010	yamt	sync with head.
1.9.4.3	11-Mar-2010	yamt	sync with head
1.9.4.2	16-May-2009	yamt	sync with head
1.9.4.1	04-May-2009	yamt	sync with head.
1.10.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.14.4.1	30-May-2010	rmind	sync with head
1.14.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.15.12.1	18-Feb-2012	mrg	merge to -current.
1.15.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.15.8.2	30-Oct-2012	yamt	sync with head
1.15.8.1	17-Apr-2012	yamt	sync with head
1.16.6.4	03-Dec-2017	jdolecek	update from HEAD
1.16.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.16.6.2	23-Jun-2013	tls	resync from head
1.16.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.18.4.1	18-May-2014	rmind	sync with head
1.19.6.5	05-Feb-2017	skrll	Sync with HEAD
1.19.6.4	05-Oct-2016	skrll	Sync with HEAD
1.19.6.3	09-Jul-2016	skrll	Sync with HEAD
1.19.6.2	19-Mar-2016	skrll	Sync with HEAD
1.19.6.1	06-Jun-2015	skrll	Sync with HEAD
1.23.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.25.14.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.25.14.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.25.14.1	28-Jul-2018	pgoyette	Sync with HEAD
1.26.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.26.2.1	10-Jun-2019	christos	Sync with HEAD
1.32.4.1	29-Feb-2020	ad	Sync with head.
1.3	29-Aug-2019	msaitoh	Add missing NUL to prevent buffer overrun.
1.2	11-Dec-2005	christos	branches: 1.2.18; 1.2.158; 1.2.166; 1.2.172; merge ktrace-lwp.
1.1	09-Sep-2005	ragge	branches: 1.1.2; 1.1.8; Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.1.8.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.8.1	09-Sep-2005	skrll	file if_xgereg.h was added on branch ktrace-lwp on 2005-11-10 14:06:02 +0000
1.1.2.2	13-Sep-2005	tron	Pull up following revision(s) (requested by ragge in ticket #768): sys/dev/pci/if_xge.c: revision 1.1 via patch sys/dev/pci/if_xgereg.h: revision 1.1 sys/dev/pci/files.pci: revision 1.238 Device driver for the Neterion (S2io) Xframe-I 10Gbit ethernet card. Still missing: VLAN + IPv6 checksum support.
1.1.2.1	09-Sep-2005	tron	file if_xgereg.h was added on branch netbsd-3 on 2005-09-13 20:59:37 +0000
1.2.172.1	01-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #145): sys/arch/sparc/include/ctlreg.h: revision 1.30 sys/dev/pci/if_xgereg.h: revision 1.3 Add missing NUL to prevent buffer overrun.
1.2.166.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.2.158.1	26-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1387): sys/arch/sparc/include/ctlreg.h: revision 1.30 sys/dev/pci/if_xgereg.h: revision 1.3 Add missing NUL to prevent buffer overrun.
1.2.18.2	21-Jun-2006	yamt	sync with head.
1.2.18.1	11-Dec-2005	yamt	file if_xgereg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:05 +0000
1.33	22-Sep-2018	maxv	Remove ifpci(4). It is part of ISDN, which we are retiring.
1.32	29-Mar-2014	christos	branches: 1.32.28; 1.32.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.31	27-Oct-2012	chs	branches: 1.31.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.30	06-Dec-2009	dyoung	branches: 1.30.12; 1.30.22; Simplify device-activation hooks.
1.29	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.28	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.27	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.26	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.25	10-Apr-2008	cegger	branches: 1.25.4; 1.25.18; use aprint_*_dev and device_xname
1.24	19-Oct-2007	ad	branches: 1.24.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.23	09-Jul-2007	ad	branches: 1.23.6; 1.23.8; 1.23.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.22	04-Mar-2007	christos	branches: 1.22.2; 1.22.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.21	16-Nov-2006	christos	branches: 1.21.4; __unused removal on arguments; approved by core.
1.20	16-Oct-2006	pooka	* __unused games + other relevant fixes * give me ANSI C or give me death (at least where it could be easily spotted)
1.19	10-Sep-2006	simonb	branches: 1.19.2; Add NetBSD RCS tag to top of file.
1.18	11-Dec-2005	christos	branches: 1.18.8; 1.18.20; merge ktrace-lwp.
1.17	09-Jul-2005	christos	make the isdn drivers compile again with -Wconst, -Wshadow
1.16	27-Feb-2005	perry	branches: 1.16.4; nuke trailing whitespace
1.15	28-Oct-2003	pooka	branches: 1.15.8; 1.15.10; fix uninitialized uses caught by gcc
1.14	03-Oct-2003	pooka	Ready the kernel side of i4b for primary rate interface support by removing assumptions that there are only two B channels and by adding support for a varying number of channels. Due to this, rename previously used isdn identified "bri" to "isdnif", which better describes the current situation.
1.13	02-Oct-2002	thorpej	branches: 1.13.6; Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.9	21-May-2002	martin	branches: 1.9.4; Uniformly pass a "struct isdn_l3_driver *" through layer 2, so we do not have to search for that driver later while we already knew it at layer 1.
1.8	10-Apr-2002	martin	branches: 1.8.2; Remove some redundant tries to disable interupts. We already avoided enabling them. Fix a stupid typo when enabling/disabling interrupts: we clear pending interrupts when enabling, not disbling isac interrupts.
1.7	08-Apr-2002	martin	Combine sc_enabled and sc_dying into one field sc_intr_valid, they are used for the same purpose (ignoring invalid interrupts). For cards that are not able to stop all interrupts (or we don't know a way to do that in software, at least) run the clearirq callback even when ignoring an interrupt because we are not enabled. Otherwise the card would stop interrupting. Reserve a driver specific callout handle and an int value in the generic isic_softc to allow card drivers to implement fancy blinkenlights.
1.6	06-Apr-2002	martin	Ooops, forgot to adopt this to the previous change of isic_isac_intr.
1.5	30-Mar-2002	martin	branches: 1.5.2; Be a nice citizen, enable interrupts only when we are ready to talk to the ISDN (i.e. when userland has /dev/isdn open).
1.4	30-Mar-2002	martin	Avoid duplicate expensive lookups by passing a pointer to the call descriptor/a pointer to the layer 3 state directly to driver functions, instead of their ID/index.
1.3	29-Mar-2002	martin	Split BRI attaching into two phases, so lower layer drivers can get their BRI identifier and L3 driver state early on, then finish initializing and announce the controller to userland when it's ready.
1.2	27-Mar-2002	martin	After Jason Thorpe explained to me how it is supposed to work, implement (de)activate for pcmcia cards. Implement detach/(de)activate for PCI cards. Clean up internal state (free call-descriptors) if a controller is detached while it has open connections.
1.1	25-Mar-2002	martin	Now that we have all pieces in place (and enough granularity to specify B-channel and D-channel drivers separately) split the Fritz!PCI card driver out of the isic driver. The new device is called "ifpci" and uses the same D-channel driver as the isic devices, but has it's own B-channel driver.
1.5.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.5.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.5.2.3	17-Apr-2002	nathanw	Catch up to -current.
1.5.2.2	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.5.2.1	30-Mar-2002	nathanw	file ifpci.c was added on branch nathanw_sa on 2002-04-01 07:46:29 +0000
1.8.2.1	30-May-2002	gehenna	Catch up with -current.
1.9.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.9.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.4.1	21-May-2002	jdolecek	file ifpci.c was added on branch kqueue on 2002-06-23 17:47:47 +0000
1.13.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.13.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.6.1	03-Aug-2004	skrll	Sync with HEAD
1.15.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.15.8.1	29-Apr-2005	kent	sync with -current
1.16.4.4	27-Oct-2007	yamt	sync with head.
1.16.4.3	03-Sep-2007	yamt	sync with head.
1.16.4.2	30-Dec-2006	yamt	sync with head.
1.16.4.1	21-Jun-2006	yamt	sync with head.
1.18.20.1	18-Nov-2006	ad	Sync with head.
1.18.8.1	14-Sep-2006	yamt	sync with head.
1.19.2.2	10-Dec-2006	yamt	sync with head.
1.19.2.1	22-Oct-2006	yamt	sync with head
1.21.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.22.4.1	11-Jul-2007	mjf	Sync with head.
1.22.2.2	23-Oct-2007	ad	Sync with head.
1.22.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.23.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.23.8.1	06-Nov-2007	matt	sync with HEAD
1.23.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.24.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.25.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.25.4.2	11-Mar-2010	yamt	sync with head
1.25.4.1	16-May-2009	yamt	sync with head
1.30.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.30.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.30.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.30.12.1	30-Oct-2012	yamt	sync with head
1.31.2.1	18-May-2014	rmind	sync with head
1.32.30.1	10-Jun-2019	christos	Sync with HEAD
1.32.28.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.23	22-Sep-2018	maxv	Remove the "ifritz" driver (no man page). It is part of ISDN, which we are retiring.
1.22	29-Mar-2014	christos	branches: 1.22.28; 1.22.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.21	23-Mar-2014	christos	remove unused
1.20	27-Oct-2012	chs	branches: 1.20.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.19	06-Dec-2009	dyoung	branches: 1.19.12; 1.19.22; Simplify device-activation hooks.
1.18	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.17	27-Oct-2009	martin	Fix endianess issues when accessing the B-channel fifos.
1.16	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.15	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.14	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.13	10-Apr-2008	cegger	branches: 1.13.4; 1.13.14; 1.13.18; 1.13.22; use aprint_*_dev and device_xname
1.12	19-Oct-2007	ad	branches: 1.12.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	09-Jul-2007	ad	branches: 1.11.6; 1.11.8; 1.11.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.10	04-Mar-2007	christos	branches: 1.10.2; 1.10.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.9	16-Nov-2006	christos	branches: 1.9.4; __unused removal on arguments; approved by core.
1.8	16-Oct-2006	pooka	* __unused games + other relevant fixes * give me ANSI C or give me death (at least where it could be easily spotted)
1.7	11-Dec-2005	christos	branches: 1.7.20; 1.7.22; merge ktrace-lwp.
1.6	01-Jun-2005	drochner	branches: 1.6.2; quell cast-qual and shadow warnings
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	06-Jul-2004	martin	branches: 1.4.4; 1.4.6; Give a magic value a name - inspired by a post on the FreeBSD isdn mailing list.
1.3	28-Oct-2003	pooka	fix uninitialized uses caught by gcc
1.2	03-Oct-2003	pooka	Ready the kernel side of i4b for primary rate interface support by removing assumptions that there are only two B channels and by adding support for a varying number of channels. Due to this, rename previously used isdn identified "bri" to "isdnif", which better describes the current situation.
1.1	25-Oct-2002	leo	branches: 1.1.2; 1.1.8; Fritz!PCI v2.0 ISDN driver from FreeBSD.
1.1.8.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.8.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.8.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.8.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.8.1	03-Aug-2004	skrll	Sync with HEAD
1.1.2.2	11-Nov-2002	nathanw	Catch up to -current
1.1.2.1	25-Oct-2002	nathanw	file ifpci2.c was added on branch nathanw_sa on 2002-11-11 22:11:15 +0000
1.4.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.4.1	29-Apr-2005	kent	sync with -current
1.6.2.3	27-Oct-2007	yamt	sync with head.
1.6.2.2	03-Sep-2007	yamt	sync with head.
1.6.2.1	30-Dec-2006	yamt	sync with head.
1.7.22.2	10-Dec-2006	yamt	sync with head.
1.7.22.1	22-Oct-2006	yamt	sync with head
1.7.20.1	18-Nov-2006	ad	Sync with head.
1.9.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.10.4.1	11-Jul-2007	mjf	Sync with head.
1.10.2.2	23-Oct-2007	ad	Sync with head.
1.10.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.11.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.8.1	06-Nov-2007	matt	sync with HEAD
1.11.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.13.22.1	21-Apr-2010	matt	sync to netbsd-5
1.13.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.13.14.1	31-Oct-2009	sborrill	Pull up the following revisions(s) (requested by martin in ticket #1119): sys/dev/pci/ifpci2.c: revision 1.17 Fix endianess issues with ifritz(4) when accessing the B-channel fifos.
1.13.4.2	11-Mar-2010	yamt	sync with head
1.13.4.1	16-May-2009	yamt	sync with head
1.19.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.19.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.19.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.19.12.1	30-Oct-2012	yamt	sync with head
1.20.2.1	18-May-2014	rmind	sync with head
1.22.30.1	10-Jun-2019	christos	Sync with HEAD
1.22.28.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.6	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.5	24-Apr-2021	thorpej	branches: 1.5.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.4	22-Dec-2019	thorpej	branches: 1.4.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.3	14-Feb-2016	chs	branches: 1.3.18; zero the i2c_attach_args structure before filling it in. fixes occasional crashes in iic_attach().
1.2	13-Jul-2014	mlelstv	branches: 1.2.4; 1.2.6; The vga driver now detaches wscons.
1.1	21-Jan-2014	mlelstv	branches: 1.1.2; 1.1.4; 1.1.6; wscons driver for Intel Graphics Media Accelerator. Initial commit that already works for a couple of Notebooks based on G35, G45, Sandy Bridge and Ivy Bridge chips. Despite the word 'Accelerator' there is nothing acclerated yet.
1.1.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.1	21-Jan-2014	yamt	file igma.c was added on branch yamt-pagecache on 2014-05-22 11:40:25 +0000
1.1.4.2	18-May-2014	rmind	sync with head
1.1.4.1	21-Jan-2014	rmind	file igma.c was added on branch rmind-smpnet on 2014-05-18 17:45:40 +0000
1.1.2.1	10-Aug-2014	tls	Rebase.
1.2.6.1	19-Mar-2016	skrll	Sync with HEAD
1.2.4.3	03-Dec-2017	jdolecek	update from HEAD
1.2.4.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.4.1	13-Jul-2014	tls	file igma.c was added on branch tls-maxphys on 2014-08-20 00:03:43 +0000
1.3.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.4.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.5.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	21-Jan-2014	mlelstv	branches: 1.1.4; 1.1.6; 1.1.10; wscons driver for Intel Graphics Media Accelerator. Initial commit that already works for a couple of Notebooks based on G35, G45, Sandy Bridge and Ivy Bridge chips. Despite the word 'Accelerator' there is nothing acclerated yet.
1.1.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.10.1	21-Jan-2014	tls	file igmareg.h was added on branch tls-maxphys on 2014-08-20 00:03:43 +0000
1.1.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.1	21-Jan-2014	yamt	file igmareg.h was added on branch yamt-pagecache on 2014-05-22 11:40:25 +0000
1.1.4.2	18-May-2014	rmind	sync with head
1.1.4.1	21-Jan-2014	rmind	file igmareg.h was added on branch rmind-smpnet on 2014-05-18 17:45:40 +0000
1.1	21-Jan-2014	mlelstv	branches: 1.1.4; 1.1.6; 1.1.10; wscons driver for Intel Graphics Media Accelerator. Initial commit that already works for a couple of Notebooks based on G35, G45, Sandy Bridge and Ivy Bridge chips. Despite the word 'Accelerator' there is nothing acclerated yet.
1.1.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.10.1	21-Jan-2014	tls	file igmavar.h was added on branch tls-maxphys on 2014-08-20 00:03:43 +0000
1.1.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.1	21-Jan-2014	yamt	file igmavar.h was added on branch yamt-pagecache on 2014-05-22 11:40:25 +0000
1.1.4.2	18-May-2014	rmind	sync with head
1.1.4.1	21-Jan-2014	rmind	file igmavar.h was added on branch rmind-smpnet on 2014-05-18 17:45:40 +0000
1.25	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.24	14-Jul-2016	msaitoh	branches: 1.24.18; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.23	30-Jan-2012	drochner	branches: 1.23.6; 1.23.24; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.22	27-Aug-2011	martin	branches: 1.22.2; 1.22.6; I suppose mrg did mean to include this change in yesterday commit: use CFATTACH_DECL_NEW for real.
1.21	26-Jul-2011	mrg	convert to device_t, cfdata_t and CFATTACH_DECL_NEW.
1.20	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.19	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.18	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.17	19-Oct-2007	ad	branches: 1.17.20; 1.17.34; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	13-Jan-2007	cube	branches: 1.16.6; 1.16.18; 1.16.20; 1.16.24; Not all archs have dma tags that can be compared with "==", so use memcmp instead.
1.15	20-May-2006	uwe	branches: 1.15.6; Redo previous fix properly. Original code had a typo caused by very similar function and variable names. We need to protect tag comparison b/c we don't want to perform it against an uninitialized tag.
1.14	20-May-2006	mrg	in code that looks like: igsfb_pci_is_console(pci_chipset_tag_t pc, pcitag_t tag) { return igsfb_pci_is_console && (tag == igsfb_pci_constag); } delete "igsfb_pci_is_console &&" as the address of igsfb_pci_is_console is always true. thanks gcc4.
1.13	05-Apr-2006	uwe	branches: 1.13.2; KNF: no parentheses around return value.
1.12	05-Apr-2006	uwe	ANSI'ify.
1.11	24-Feb-2006	uwe	branches: 1.11.2; 1.11.4; 1.11.6; Whitespace cleanup (hi, macallan).
1.10	23-Feb-2006	macallan	Add support for virtual consoles to igsfb. Works fine on a Krups, needs testing on other ports.
1.9	11-Dec-2005	christos	branches: 1.9.2; 1.9.4; 1.9.6; merge ktrace-lwp.
1.8	23-Apr-2004	itojun	branches: 1.8.12; pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.7	10-May-2003	uwe	branches: 1.7.2; Redo attachment to support proper cnattach. wscons on netwinder almost works now.
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	24-Sep-2002	uwe	Add some preliminary support for setting up the CyberPro in Netwinder. Setup sequence obtained from Krups OFW with some CyberPro-specific magic from Linux driver. The driver still has a lot of hardcoded stuff, but it is useful enough to bring up wscons on netwinder. XXX: Proper console attachment needs to be written (the driver was originally developed on sparc, where our approach to attaching console is totally different). Caveat emptor!
1.3	21-Jul-2002	uwe	Add definitions of IGS coprocessor registers (not used yet). Always init chip via I/O space. Map only I/O registers we actually need. Automatically detect if we need to do bswapping Add (untested) code to use automagic bswap on CyberPro. Report video memory amount. Report if/how bswapping is done.
1.2	04-Apr-2002	uwe	branches: 1.2.2; 1.2.4; 1.2.8; Nuke ugly ms-IIep specific hack now that mspcic knows about phys->pci address translation.
1.1	30-Mar-2002	uwe	First cut at framebuffer driver for Integraphics Systems IGA1682. It's not finished yet, but it can do enough to get wscons working for Krups JavaStation.
1.2.8.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.8.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.2.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.8.1	04-Apr-2002	jdolecek	file igsfb_pci.c was added on branch kqueue on 2002-06-23 17:47:47 +0000
1.2.4.1	29-Aug-2002	gehenna	catch up with -current.
1.2.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.2.2.3	01-Aug-2002	nathanw	Catch up to -current.
1.2.2.2	17-Apr-2002	nathanw	Catch up to -current.
1.2.2.1	04-Apr-2002	nathanw	file igsfb_pci.c was added on branch nathanw_sa on 2002-04-17 00:06:02 +0000
1.7.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.1	03-Aug-2004	skrll	Sync with HEAD
1.8.12.3	27-Oct-2007	yamt	sync with head.
1.8.12.2	26-Feb-2007	yamt	sync with head.
1.8.12.1	21-Jun-2006	yamt	sync with head.
1.9.6.2	01-Jun-2006	kardel	Sync with head.
1.9.6.1	22-Apr-2006	simonb	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.9.2.1	01-Mar-2006	yamt	sync with head.
1.11.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.11.4.1	19-Apr-2006	elad	sync with head.
1.11.2.2	24-May-2006	yamt	sync with head.
1.11.2.1	11-Apr-2006	yamt	sync with head
1.13.2.1	19-Jun-2006	chap	Sync with head.
1.15.6.1	01-Feb-2007	ad	Sync with head.
1.16.24.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.20.1	06-Nov-2007	matt	sync with HEAD
1.16.18.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.6.1	23-Oct-2007	ad	Sync with head.
1.17.34.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.17.20.1	16-May-2009	yamt	sync with head
1.22.6.1	18-Feb-2012	mrg	merge to -current.
1.22.2.1	17-Apr-2012	yamt	sync with head
1.23.24.1	05-Oct-2016	skrll	Sync with HEAD
1.23.6.1	03-Dec-2017	jdolecek	update from HEAD
1.24.18.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1	10-May-2003	uwe	Redo attachment to support proper cnattach. wscons on netwinder almost works now.
1.20	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.19	29-Mar-2014	christos	branches: 1.19.28; 1.19.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	30-Jan-2012	drochner	branches: 1.18.6; 1.18.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.17	13-Nov-2010	uebayasi	branches: 1.17.8; 1.17.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.16	14-May-2008	tsutsui	branches: 1.16.20; Normalize my licenses.
1.15	12-Apr-2008	tsutsui	branches: 1.15.2; 1.15.4; 1.15.6; Split device_t/softc.
1.14	10-Apr-2008	cegger	use aprint_*_dev and device_xname
1.13	16-Nov-2006	christos	branches: 1.13.48; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	16-Jun-2005	tsutsui	branches: 1.10.2; Rename iha_pci_probe() -> iha_pci_match() because we don't have to probe PCI devices.
1.9	25-Sep-2004	tsutsui	Use ANSI function decls.
1.8	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.7	14-Nov-2002	tsutsui	branches: 1.7.6; Sync with copyright verbiage update of OpenBSD's iha(4).
1.6	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	13-Nov-2001	lukem	add RCSID
1.2	27-Sep-2001	tsutsui	Include uvm/uvm_extern.h for PAGE_SIZE.
1.1	03-Jun-2001	tsutsui	branches: 1.1.2; 1.1.4; 1.1.6; Add a driver for the Initio INIC-940/950 PCI SCSI controllers. Based on OpenBSD's iha driver, and modified some structures by me. Tested on arc, i386 and macppc.
1.1.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.1.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.6	11-Dec-2002	thorpej	Sync with HEAD.
1.1.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.1.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.1	03-Jun-2001	nathanw	file iha_pci.c was added on branch nathanw_sa on 2001-06-21 20:04:54 +0000
1.7.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.6.4	19-Oct-2004	skrll	Sync with HEAD
1.7.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.6.1	03-Aug-2004	skrll	Sync with HEAD
1.10.2.1	30-Dec-2006	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.13.48.1	02-Jun-2008	mjf	Sync with HEAD.
1.15.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.20.1	05-Mar-2011	rmind	sync with head
1.17.12.1	18-Feb-2012	mrg	merge to -current.
1.17.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.17.8.1	17-Apr-2012	yamt	sync with head
1.18.10.1	18-May-2014	rmind	sync with head
1.18.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.19.30.1	10-Jun-2019	christos	Sync with HEAD
1.19.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.30	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.29	14-Jul-2016	msaitoh	branches: 1.29.16; 1.29.18; Use aprint*() instead of printf() in the attach function. KNF.
1.28	29-Mar-2014	christos	branches: 1.28.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.27	27-Oct-2012	chs	branches: 1.27.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.26	12-May-2009	cegger	branches: 1.26.12; 1.26.22; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.25	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.24	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.23	28-Apr-2008	martin	branches: 1.23.14; Remove clause 3 and 4 from TNF licenses
1.22	10-Apr-2008	cegger	branches: 1.22.2; 1.22.4; use aprint_*_dev and device_xname
1.21	19-Oct-2007	ad	branches: 1.21.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.20	09-Jul-2007	ad	branches: 1.20.6; 1.20.8; 1.20.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.19	16-Jun-2007	ad	opt_i2o.h is gone.
1.18	16-Nov-2006	christos	branches: 1.18.8; 1.18.10; __unused removal on arguments; approved by core.
1.17	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.16	11-Dec-2005	christos	branches: 1.16.20; 1.16.22; merge ktrace-lwp.
1.15	27-Feb-2005	perry	branches: 1.15.4; nuke trailing whitespace
1.14	09-Dec-2003	ad	branches: 1.14.8; 1.14.10; Match Promise controllers.
1.13	15-Nov-2002	itojun	branches: 1.13.6; typo
1.12	15-Nov-2002	ad	Cosmetic changes.
1.11	08-Nov-2002	msaitoh	Add support for Adaptec 200[05]S reviewed by Andrew Doran
1.10	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.9	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.8	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.7	13-Nov-2001	lukem	branches: 1.7.10; add RCSID
1.6	22-Sep-2001	ad	branches: 1.6.4; Save some info about the board for dpti.
1.5	01-Apr-2001	ad	branches: 1.5.2; 1.5.4; - Allocate any private space thats requested. - Just use mapreg_map().
1.4	20-Mar-2001	ad	+#include <dev/i2o/iopio.h>
1.3	28-Dec-2000	sommerfeld	branches: 1.3.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.2	03-Dec-2000	ad	- Force the memory mapping uncacheable. - We don't need a record of the I/O space mapping.
1.1	08-Nov-2000	ad	branches: 1.1.2; Add support for I2O adapters and SCSI/FC-AL/BSA class devices. This needs more testing with different configurations, and work in a number of areas (which I'm not able to do for a little while), but is at least functional and stable on i386 with DPT adapters.
1.1.2.6	01-Apr-2001	ad	Sync w/-current.
1.1.2.5	27-Mar-2001	bouyer	Sync with HEAD.
1.1.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.1.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.1.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.1	08-Nov-2000	bouyer	file iop_pci.c was added on branch thorpej_scsipi on 2000-11-22 16:04:08 +0000
1.3.2.6	27-Nov-2002	christos	make this compile again [sync to current]
1.3.2.5	11-Nov-2002	nathanw	Catch up to -current
1.3.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.2	26-Sep-2001	nathanw	Catch up to -current. Again.
1.3.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.5.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.5.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.5.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.4.3	09-Dec-2001	he	Pull up revision 1.6 (requested by ad): Add driver for DPT/Adaptec I2O RAID management interface.
1.6.4.2	25-Oct-2001	he	Pull up revisions 1.1-1.5 (new, via patch, requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.6.4.1	22-Sep-2001	he	file iop_pci.c was added on branch netbsd-1-5 on 2001-10-25 18:02:16 +0000
1.7.10.1	12-Dec-2002	he	Pull up revisions 1.11-1.13 (requested by ad in ticket #1030): Add support for Adaptec 2005S. Handle DPT ioctls() from native binaries. Fix SCSI probe problem.
1.13.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.6.1	03-Aug-2004	skrll	Sync with HEAD
1.14.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.8.1	29-Apr-2005	kent	sync with -current
1.15.4.3	27-Oct-2007	yamt	sync with head.
1.15.4.2	03-Sep-2007	yamt	sync with head.
1.15.4.1	30-Dec-2006	yamt	sync with head.
1.16.22.2	10-Dec-2006	yamt	sync with head.
1.16.22.1	22-Oct-2006	yamt	sync with head
1.16.20.1	18-Nov-2006	ad	Sync with head.
1.18.10.1	11-Jul-2007	mjf	Sync with head.
1.18.8.3	23-Oct-2007	ad	Sync with head.
1.18.8.2	15-Jul-2007	ad	Sync with head.
1.18.8.1	27-May-2007	ad	Checkpoint work in progress. Seems to break block I/O but I am not sure why yet: - Avoid a NULL pointer dereference in bus_dmamap_sync. - MP safety changes, incomplete.
1.20.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.20.8.1	06-Nov-2007	matt	sync with HEAD
1.20.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.21.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.22.4.2	16-May-2009	yamt	sync with head
1.22.4.1	16-May-2008	yamt	sync with head.
1.22.2.1	18-May-2008	yamt	sync with head.
1.23.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.26.22.3	03-Dec-2017	jdolecek	update from HEAD
1.26.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.26.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.26.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.26.12.1	30-Oct-2012	yamt	sync with head
1.27.2.1	18-May-2014	rmind	sync with head
1.28.6.1	05-Oct-2016	skrll	Sync with HEAD
1.29.18.1	10-Jun-2019	christos	Sync with HEAD
1.29.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7	08-Jan-2024	chs	ips: fix a couple more device_t/softc split bugs
1.6	12-Feb-2022	riastradh	branches: 1.6.4; ips(4): Missed a couple spots in device_t/softc split.
1.5	12-Feb-2022	riastradh	ips(4): device_t/softc split
1.4	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.3	24-Apr-2021	thorpej	branches: 1.3.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.2	10-Nov-2019	chs	branches: 1.2.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.1	03-Dec-2017	jdolecek	branches: 1.1.4; port ips(4) driver from OpenBSD; needs a lot more work, right now just compilable
1.1.4.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.2.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.3.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.6.4.1	14-Jan-2024	martin	Pull up following revision(s) (requested by chs in ticket #541): sys/dev/pci/ips.c: revision 1.7 ips: fix a couple more device_t/softc split bugs
1.42	22-Sep-2018	maxv	Remove isic(4). It is part of ISDN, which we are now retiring.
1.41	14-Jul-2016	msaitoh	branches: 1.41.16; 1.41.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.40	29-Mar-2014	christos	branches: 1.40.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.39	27-Oct-2012	chs	branches: 1.39.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.38	03-Feb-2012	christos	branches: 1.38.6; ansi prototypes
1.37	06-Dec-2009	dyoung	branches: 1.37.12; 1.37.16; Simplify device-activation hooks.
1.36	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.35	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.34	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.31	28-Apr-2008	martin	branches: 1.31.8; 1.31.14; Remove clause 3 and 4 from TNF licenses
1.30	10-Apr-2008	cegger	branches: 1.30.2; 1.30.4; use aprint_*_dev and device_xname
1.29	19-Oct-2007	ad	branches: 1.29.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.28	09-Jul-2007	ad	branches: 1.28.6; 1.28.8; 1.28.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.27	16-Nov-2006	christos	branches: 1.27.8; 1.27.10; __unused removal on arguments; approved by core.
1.26	16-Oct-2006	pooka	* __unused games + other relevant fixes * give me ANSI C or give me death (at least where it could be easily spotted)
1.25	11-Dec-2005	christos	branches: 1.25.20; 1.25.22; merge ktrace-lwp.
1.24	01-Jun-2005	drochner	branches: 1.24.2; quell cast-qual and shadow warnings
1.23	27-Feb-2005	perry	nuke trailing whitespace
1.22	04-Feb-2005	perry	de-__P
1.21	22-Jul-2004	drochner	branches: 1.21.4; 1.21.6; if the register space os the ISDN card can't be mapped, abort device attachment instead of dereferencing NULL and so (this needs a serious cleanup -- the framework was thought to be generally usable for ~all PCI ISDN cards; this didn't work out, so we are left with a lot of nonsense to support just 1 particular card type...)
1.20	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.19	03-Oct-2003	pooka	Ready the kernel side of i4b for primary rate interface support by removing assumptions that there are only two B channels and by adding support for a varying number of channels. Due to this, rename previously used isdn identified "bri" to "isdnif", which better describes the current situation.
1.18	02-Oct-2002	thorpej	branches: 1.18.6; Add trailing ; to CFATTACH_DECL.
1.17	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.16	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.15	19-Apr-2002	drochner	check interrupt status of the PCI interface chip (if possible) before calling the (slow) isic handler being here, remove #ifdefs for FreeBSD and old version support and add RCS ID tags
1.14	14-Apr-2002	martin	Fix copyright notice.
1.13	13-Apr-2002	martin	Defer initialization of the ISAC chip until we enable interrupts - it gets into a strange failure mode if we do it with disabled interrupt. When (re-)enabling interrupts reset transmitter and receiver and clear any pending state.
1.12	10-Apr-2002	martin	Remove some redundant tries to disable interupts. We already avoided enabling them. Fix a stupid typo when enabling/disabling interrupts: we clear pending interrupts when enabling, not disbling isac interrupts.
1.11	08-Apr-2002	martin	Combine sc_enabled and sc_dying into one field sc_intr_valid, they are used for the same purpose (ignoring invalid interrupts). For cards that are not able to stop all interrupts (or we don't know a way to do that in software, at least) run the clearirq callback even when ignoring an interrupt because we are not enabled. Otherwise the card would stop interrupting. Reserve a driver specific callout handle and an int value in the generic isic_softc to allow card drivers to implement fancy blinkenlights.
1.10	30-Mar-2002	martin	Be a nice citizen, enable interrupts only when we are ready to talk to the ISDN (i.e. when userland has /dev/isdn open).
1.9	27-Mar-2002	martin	After Jason Thorpe explained to me how it is supposed to work, implement (de)activate for pcmcia cards. Implement detach/(de)activate for PCI cards. Clean up internal state (free call-descriptors) if a controller is detached while it has open connections.
1.8	25-Mar-2002	martin	Now that we have all pieces in place (and enough granularity to specify B-channel and D-channel drivers separately) split the Fritz!PCI card driver out of the isic driver. The new device is called "ifpci" and uses the same D-channel driver as the isic devices, but has it's own B-channel driver.
1.7	24-Mar-2002	martin	Now that we have all the pieces of the puzzle available start to unriddle and move them in their proper places. Move the BRI registry from layer 2 (duh!) to layer 4, so active cards (which don't have layer 3 or layer 2 in their driver). Remove all remaining hard coded controller and driver types. Remove any arbitrary hard coded limits, at least those that show up in the internal API. This fixes PR 15950.
1.6	10-Feb-2002	wiz	pcmcia -> pci in comment.
1.5	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.4	13-Nov-2001	lukem	add RCSID
1.3	24-Mar-2001	martin	branches: 1.3.2; First step to cleanup the hardware driver <-> upper layers interface. This now provides slightly more functionality than the FreeBSD layer1-newbus interface. It was meant to be a simple change to one header and a few c files, but the change rippled all through various stuff. To prevent a change to the kernel<->userland interface right now the kernel is now lying about card types to userland (but who cares). This will be fixed when the userland interface changes, after layer 3 <-> layer 4 has been fixed. Functional changes: Provide a clean interface for hardware drivers to attach to the upper layers. This will need another small change in the B-channel handling when a similar change to the layer 3 <-> layer 4 interface happens. Avoid passing indices into global arrays of pointers around, instead pass the pointers itself. Don't code hardware driver types by predefined magic numbers (think LKM). Prepare for detachable drivers (think pcmcia). While there remove some sets of function pointers always pointing to the same function (meant to be the configurable set of D channel protocol handlers). It is unlikely another supported D-channel protocol will fit into that (maximal layer interface) abstraction. When we get support for another protocol, we will need to come up with a workable interface. Besides, the old implementation was, uhm, strange.
1.2	20-Feb-2001	martin	branches: 1.2.2; 1.2.4; Filename cleanup: remove i4b_ prefixes outside sys/netisdn, last round. Renamed in sys/dev/ic: i4b_hscx.c -> hscx.c i4b_hscx.h -> hscx.h i4b_ipac.h -> ipac.h i4b_isac.c -> isac.c i4b_isac.h -> isac.h i4b_isic.c -> isic.c i4b_isicbchan.c -> isic_bchan.c i4b_isicl1.c -> isic_l1.c i4b_isicl1.h -> isic_l1.h i4b_isicl1fsm.c -> isic_l1fsm.c
1.1	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.2.4.3	27-Mar-2001	bouyer	Sync with HEAD.
1.2.4.2	12-Mar-2001	bouyer	Sync with HEAD.
1.2.4.1	20-Feb-2001	bouyer	file isic_pci.c was added on branch thorpej_scsipi on 2001-03-12 13:31:08 +0000
1.2.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.2.2.7	20-Jun-2002	nathanw	Catch up to -current.
1.2.2.6	17-Apr-2002	nathanw	Catch up to -current.
1.2.2.5	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.2.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.2.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.3.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.2.2	16-Mar-2002	jdolecek	Catch up with -current.
1.3.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.6.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.6.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.18.6.2	04-Feb-2005	skrll	Sync with HEAD.
1.18.6.1	03-Aug-2004	skrll	Sync with HEAD
1.21.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.6.1	12-Feb-2005	yamt	sync with head.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.24.2.3	27-Oct-2007	yamt	sync with head.
1.24.2.2	03-Sep-2007	yamt	sync with head.
1.24.2.1	30-Dec-2006	yamt	sync with head.
1.25.22.2	10-Dec-2006	yamt	sync with head.
1.25.22.1	22-Oct-2006	yamt	sync with head
1.25.20.1	18-Nov-2006	ad	Sync with head.
1.27.10.1	11-Jul-2007	mjf	Sync with head.
1.27.8.2	23-Oct-2007	ad	Sync with head.
1.27.8.1	01-Jul-2007	ad	Adapt to callout API change.
1.28.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.28.8.1	06-Nov-2007	matt	sync with HEAD
1.28.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.29.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.30.4.4	11-Mar-2010	yamt	sync with head
1.30.4.3	16-May-2009	yamt	sync with head
1.30.4.2	04-May-2009	yamt	sync with head.
1.30.4.1	16-May-2008	yamt	sync with head.
1.30.2.1	18-May-2008	yamt	sync with head.
1.31.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.31.8.1	28-Apr-2009	skrll	Sync with HEAD.
1.37.16.1	18-Feb-2012	mrg	merge to -current.
1.37.12.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.12.2	30-Oct-2012	yamt	sync with head
1.37.12.1	17-Apr-2012	yamt	sync with head
1.38.6.3	03-Dec-2017	jdolecek	update from HEAD
1.38.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.38.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.39.2.1	18-May-2014	rmind	sync with head
1.40.6.1	05-Oct-2016	skrll	Sync with HEAD
1.41.18.1	10-Jun-2019	christos	Sync with HEAD
1.41.16.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.12	22-Sep-2018	maxv	Remove isic(4). It is part of ISDN, which we are now retiring.
1.11	28-Apr-2008	martin	branches: 1.11.86; 1.11.88; Remove clause 3 and 4 from TNF licenses
1.10	11-Dec-2005	christos	branches: 1.10.70; 1.10.72; 1.10.74; merge ktrace-lwp.
1.9	04-Feb-2005	perry	de-__P
1.8	22-Jul-2004	drochner	branches: 1.8.4; 1.8.6; if the register space os the ISDN card can't be mapped, abort device attachment instead of dereferencing NULL and so (this needs a serious cleanup -- the framework was thought to be generally usable for ~all PCI ISDN cards; this didn't work out, so we are left with a lot of nonsense to support just 1 particular card type...)
1.7	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	03-May-2002	drochner	branches: 1.6.10; make the LEDs work with both channels active, some cleanup and simplification
1.5	19-Apr-2002	drochner	check interrupt status of the PCI interface chip (if possible) before calling the (slow) isic handler being here, remove #ifdefs for FreeBSD and old version support and add RCS ID tags
1.4	14-Apr-2002	martin	Fix copyright notice.
1.3	27-Mar-2002	martin	After Jason Thorpe explained to me how it is supposed to work, implement (de)activate for pcmcia cards. Implement detach/(de)activate for PCI cards. Clean up internal state (free call-descriptors) if a controller is detached while it has open connections.
1.2	24-Mar-2002	martin	Now that we have all the pieces of the puzzle available start to unriddle and move them in their proper places. Move the BRI registry from layer 2 (duh!) to layer 4, so active cards (which don't have layer 3 or layer 2 in their driver). Remove all remaining hard coded controller and driver types. Remove any arbitrary hard coded limits, at least those that show up in the internal API. This fixes PR 15950.
1.1	18-Feb-2001	martin	branches: 1.1.2; 1.1.4; 1.1.6; Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.1.6.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.4.2	12-Mar-2001	bouyer	Sync with HEAD.
1.1.4.1	18-Feb-2001	bouyer	file isic_pci.h was added on branch thorpej_scsipi on 2001-03-12 13:31:08 +0000
1.1.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.1.2.2	17-Apr-2002	nathanw	Catch up to -current.
1.1.2.1	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.6.10.2	04-Feb-2005	skrll	Sync with HEAD.
1.6.10.1	03-Aug-2004	skrll	Sync with HEAD
1.8.6.1	12-Feb-2005	yamt	sync with head.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.74.1	16-May-2008	yamt	sync with head.
1.10.72.1	18-May-2008	yamt	sync with head.
1.10.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.11.88.1	10-Jun-2019	christos	Sync with HEAD
1.11.86.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.9	25-Mar-2002	martin	Now that we have all pieces in place (and enough granularity to specify B-channel and D-channel drivers separately) split the Fritz!PCI card driver out of the isic driver. The new device is called "ifpci" and uses the same D-channel driver as the isic devices, but has it's own B-channel driver.
1.8	25-Mar-2002	martin	Remove a leftover from the BRI registry at layer 2. Initialize layer 2's idea of it's BRI id after attaching.
1.7	24-Mar-2002	martin	Now that we have all the pieces of the puzzle available start to unriddle and move them in their proper places. Move the BRI registry from layer 2 (duh!) to layer 4, so active cards (which don't have layer 3 or layer 2 in their driver). Remove all remaining hard coded controller and driver types. Remove any arbitrary hard coded limits, at least those that show up in the internal API. This fixes PR 15950.
1.6	17-Mar-2002	martin	Remove all knowledge about specific application (layer 4) drivers from the generic layer 4 and layer 3 management system. This should make the layer 4 driver API LKM clean - finaly. Make the Fritz!PCI driver work again after resent changes (oops!), noted by Frank Kardel (PR 15948) and Matthias Scheeler.
1.5	13-Nov-2001	lukem	add RCSID
1.4	07-Jul-2001	thorpej	branches: 1.4.2; bcopy -> memcpy
1.3	24-Mar-2001	martin	First step to cleanup the hardware driver <-> upper layers interface. This now provides slightly more functionality than the FreeBSD layer1-newbus interface. It was meant to be a simple change to one header and a few c files, but the change rippled all through various stuff. To prevent a change to the kernel<->userland interface right now the kernel is now lying about card types to userland (but who cares). This will be fixed when the userland interface changes, after layer 3 <-> layer 4 has been fixed. Functional changes: Provide a clean interface for hardware drivers to attach to the upper layers. This will need another small change in the B-channel handling when a similar change to the layer 3 <-> layer 4 interface happens. Avoid passing indices into global arrays of pointers around, instead pass the pointers itself. Don't code hardware driver types by predefined magic numbers (think LKM). Prepare for detachable drivers (think pcmcia). While there remove some sets of function pointers always pointing to the same function (meant to be the configurable set of D channel protocol handlers). It is unlikely another supported D-channel protocol will fit into that (maximal layer interface) abstraction. When we get support for another protocol, we will need to come up with a workable interface. Besides, the old implementation was, uhm, strange.
1.2	20-Feb-2001	martin	branches: 1.2.2; 1.2.4; Filename cleanup: remove i4b_ prefixes outside sys/netisdn, last round. Renamed in sys/dev/ic: i4b_hscx.c -> hscx.c i4b_hscx.h -> hscx.h i4b_ipac.h -> ipac.h i4b_isac.c -> isac.c i4b_isac.h -> isac.h i4b_isic.c -> isic.c i4b_isicbchan.c -> isic_bchan.c i4b_isicl1.c -> isic_l1.c i4b_isicl1.h -> isic_l1.h i4b_isicl1fsm.c -> isic_l1fsm.c
1.1	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.2.4.3	27-Mar-2001	bouyer	Sync with HEAD.
1.2.4.2	12-Mar-2001	bouyer	Sync with HEAD.
1.2.4.1	20-Feb-2001	bouyer	file isic_pci_avm_fritz_pci.c was added on branch thorpej_scsipi on 2001-03-12 13:31:08 +0000
1.2.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.2.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.4.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.4.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.21	22-Sep-2018	maxv	Remove isic(4). It is part of ISDN, which we are now retiring.
1.20	27-Oct-2012	chs	branches: 1.20.36; 1.20.38; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.19	14-Mar-2009	dsl	branches: 1.19.12; 1.19.22; Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.18	10-Apr-2008	cegger	branches: 1.18.4; 1.18.12; 1.18.18; use aprint_*_dev and device_xname
1.17	19-Oct-2007	ad	branches: 1.17.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	09-Jul-2007	ad	branches: 1.16.6; 1.16.8; 1.16.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.15	11-Dec-2005	christos	branches: 1.15.30; 1.15.32; merge ktrace-lwp.
1.14	01-Jun-2005	drochner	branches: 1.14.2; quell cast-qual and shadow warnings
1.13	27-Feb-2005	perry	nuke trailing whitespace
1.12	22-Jul-2004	drochner	branches: 1.12.4; 1.12.6; if the register space os the ISDN card can't be mapped, abort device attachment instead of dereferencing NULL and so (this needs a serious cleanup -- the framework was thought to be generally usable for ~all PCI ISDN cards; this didn't work out, so we are left with a lot of nonsense to support just 1 particular card type...)
1.11	08-May-2003	martin	branches: 1.11.2; We are getting passed an integer value via a void* arg (in an ioctl() like function) - cast it to an intptr_t. Addresses parts of PR kern/21448 by Frank Kardel.
1.10	03-May-2002	drochner	make the LEDs work with both channels active, some cleanup and simplification
1.9	19-Apr-2002	drochner	check interrupt status of the PCI interface chip (if possible) before calling the (slow) isic handler being here, remove #ifdefs for FreeBSD and old version support and add RCS ID tags
1.8	18-Apr-2002	martin	Duh, 0 bits enable interrupts in the IPAC! So to disable them all better use 0xff instead of 0x00 as mask. Noted by Matthias Drochner. Move some initialization unrelated to interrupts back to its place at attach time.
1.7	17-Apr-2002	martin	Be carefull not to cause interrupts while /dev/isdn is not open. Make the LEDs work.
1.6	27-Mar-2002	martin	After Jason Thorpe explained to me how it is supposed to work, implement (de)activate for pcmcia cards. Implement detach/(de)activate for PCI cards. Clean up internal state (free call-descriptors) if a controller is detached while it has open connections.
1.5	25-Mar-2002	martin	Now that we have all pieces in place (and enough granularity to specify B-channel and D-channel drivers separately) split the Fritz!PCI card driver out of the isic driver. The new device is called "ifpci" and uses the same D-channel driver as the isic devices, but has it's own B-channel driver.
1.4	24-Mar-2002	martin	Now that we have all the pieces of the puzzle available start to unriddle and move them in their proper places. Move the BRI registry from layer 2 (duh!) to layer 4, so active cards (which don't have layer 3 or layer 2 in their driver). Remove all remaining hard coded controller and driver types. Remove any arbitrary hard coded limits, at least those that show up in the internal API. This fixes PR 15950.
1.3	13-Nov-2001	lukem	add RCSID
1.2	20-Feb-2001	martin	branches: 1.2.2; 1.2.4; 1.2.6; Filename cleanup: remove i4b_ prefixes outside sys/netisdn, last round. Renamed in sys/dev/ic: i4b_hscx.c -> hscx.c i4b_hscx.h -> hscx.h i4b_ipac.h -> ipac.h i4b_isac.c -> isac.c i4b_isac.h -> isac.h i4b_isic.c -> isic.c i4b_isicbchan.c -> isic_bchan.c i4b_isicl1.c -> isic_l1.c i4b_isicl1.h -> isic_l1.h i4b_isicl1fsm.c -> isic_l1fsm.c
1.1	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.2.6.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.4.2	12-Mar-2001	bouyer	Sync with HEAD.
1.2.4.1	20-Feb-2001	bouyer	file isic_pci_elsa_qs1p.c was added on branch thorpej_scsipi on 2001-03-12 13:31:09 +0000
1.2.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.2.2.2	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.11.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.11.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.11.2.1	03-Aug-2004	skrll	Sync with HEAD
1.12.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.12.4.1	29-Apr-2005	kent	sync with -current
1.14.2.2	27-Oct-2007	yamt	sync with head.
1.14.2.1	03-Sep-2007	yamt	sync with head.
1.15.32.1	11-Jul-2007	mjf	Sync with head.
1.15.30.2	23-Oct-2007	ad	Sync with head.
1.15.30.1	01-Jul-2007	ad	Adapt to callout API change.
1.16.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.8.1	06-Nov-2007	matt	sync with HEAD
1.16.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.17.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.18.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.18.4.1	04-May-2009	yamt	sync with head.
1.19.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.19.12.1	30-Oct-2012	yamt	sync with head
1.20.38.1	10-Jun-2019	christos	Sync with HEAD
1.20.36.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.12	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.11	07-Aug-2023	msaitoh	ichsmb(4),ismt(4): Add Snow Ridge support.
1.10	10-May-2023	riastradh	ismt(4): Use config_detach_children.
1.9	07-Aug-2021	thorpej	branches: 1.9.6; Merge thorpej-cfargs2.
1.8	24-Apr-2021	thorpej	branches: 1.8.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.7	22-Dec-2019	thorpej	branches: 1.7.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.6	17-Aug-2017	msaitoh	branches: 1.6.2; 1.6.6; 1.6.12; Add C3000 devices.
1.5	27-Apr-2017	msaitoh	branches: 1.5.2; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.4	11-Jul-2016	knakahara	branches: 1.4.8; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.3	01-Feb-2016	msaitoh	branches: 1.3.2; Add parenthesis for the addr argument of ISMT_DESC_ADDR_RW(addr, is_read). It was no real bug.
1.2	06-Jan-2016	christos	fix 32 bit compilation
1.1	05-Jan-2016	msaitoh	Port FreeBSD's ismt(4) driver. ismt(4) supports Intel Chipset internal SMBus 2.0 controller with DMA. It's different from ichsmb(4). Supported chipsets are S1200 and C2000.
1.3.2.4	28-Aug-2017	skrll	Sync with HEAD
1.3.2.3	05-Oct-2016	skrll	Sync with HEAD
1.3.2.2	19-Mar-2016	skrll	Sync with HEAD
1.3.2.1	01-Feb-2016	skrll	file ismt.c was added on branch nick-nhusb on 2016-03-19 11:30:11 +0000
1.4.8.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.5.2.2	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1891): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.5.2.1	26-Feb-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #565): sys/dev/pci/ichsmb.c: 1.51 sys/dev/pci/ismt.c: 1.6 Add C3000 devices.
1.6.12.1	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1723): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.6.6.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.6.2.2	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	17-Aug-2017	jdolecek	file ismt.c was added on branch tls-maxphys on 2017-12-03 11:37:08 +0000
1.7.10.2	03-Apr-2021	thorpej	ismt_config_interrupts(): remove a now-unused variable.
1.7.10.1	24-Mar-2021	thorpej	Don't filter interface attributes in rescan functions for devices that carry only a single interface attribute. The autoconfiguration machinery already considers interface attributes when searching for possible parents.
1.8.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.9.6.1	23-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #338): sys/dev/pci/ichsmb.c: revision 1.85 sys/dev/pci/ismt.c: revision 1.11 ichsmb(4),ismt(4): Add Snow Ridge support.
1.122	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.121	15-Oct-2019	christos	remove conditional portion that was wrong and always false.
1.120	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.119	14-Jul-2016	msaitoh	branches: 1.119.16; 1.119.18; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.118	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.117	29-Mar-2014	christos	branches: 1.117.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.116	14-Sep-2013	martin	Remove unused variable
1.115	24-May-2011	joerg	branches: 1.115.4; 1.115.14; 1.115.18; Use proper format string
1.114	28-Feb-2011	mjacob	Update isp driver to be in sync with other platforms. Mostly very minor changes with effectively zero impact on NetBSD.
1.113	13-Nov-2010	uebayasi	branches: 1.113.2; 1.113.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.112	12-Apr-2010	mjacob	Oops. Don't forget to set the isp_port value based upon PCI function code. Otherwise you end up with both ports using the same WWPN. Bad. Discovered by Havard. Pointy hat to me.
1.111	26-Mar-2010	mjacob	Synchronize with other platforms: Strongly architect handles so we can more easily detect bogus handles. This switches us to a full 32 bits for all handles. Handle the case of FC disks disappearing and then reappearing- at least at the FC transport level. Some better and finer control of debug and non-debug printouts.
1.110	07-Sep-2009	tsutsui	branches: 1.110.2; 1.110.4; Split device_t/softc. Tested QLogic 1020 Fast Wide SCSI HBA at PCI. Sbus attachment is untested, but not so much quirks in it.
1.109	25-Jun-2009	mjacob	Update ISP driver to latest and greatest. Includes support for the 8Gb part.
1.108	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.107	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.106	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.105	07-Apr-2008	cegger	branches: 1.105.4; 1.105.18; use aprint_*_dev and device_xname and get rid of isp_name
1.104	11-Mar-2008	mjacob	Checkpoint of some fairly major isp(4) rototilling. Reintroduce more of a 'channel' concept in preparation for NP-IV support. This gets rid of the chanA/chanB concept as the 2400 can have up to 128 virtual channels. Actually, with MID firmware you can also have the 2200 and 2300 support 'channels, but they do it with an FL-Port topology. Because FC cards can now have 'channels', just about every support function for fibre channel had to be redone to have a channel index as well. Rototill isp_ioctl.h for channel stuff as well. Pick up a lot of work about fabric management (hopefully better) and keep work in place that will allow for dynamic attachment/detachment of devices (if I can figure out how to make the midlayer support it). Merge the target code with external trees. Eventually it might even be sorted out on NetBSD. Update some firmware stuff.
1.103	26-Jun-2007	mjacob	branches: 1.103.8; 1.103.24; 1.103.28; If we have 2K Login firmware for 23XX cards, remember to put out the 'extended' structure- not the older one. Matters for the swozzling that goes on for big endian machines.
1.102	25-May-2007	mjacob	Whoops- I almost forgot. Thank MetrumRG Associates for material support in making 4Gb NetBSD support possible.
1.101	24-May-2007	mjacob	Major update to isp(4) driver to bring it in line with external sources. The major changes are: + 4Gb (24XX) card support + Rewritten fabric and loop evaluation code + New f/w sets The 4Gb changes required major rototilling, which caused a rewrite of fabric and loop eval code. The latter can now be set up to tune for dynamic device arrival/departure if the framework is set up for it, or to be firm about waiting for devices. Testing has been principally on amd64, i386 and sparc64 and seems to not have broken things for me.
1.100	04-Mar-2007	christos	branches: 1.100.2; 1.100.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.99	13-Jan-2007	cube	branches: 1.99.2; Make the logic of handling the options right.
1.98	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.97	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.96	02-Sep-2006	christos	branches: 1.96.2; 1.96.4; add missing initializers
1.95	29-Mar-2006	thorpej	Use device_cfdata().
1.94	11-Dec-2005	christos	branches: 1.94.4; 1.94.6; 1.94.8; 1.94.10; 1.94.12; merge ktrace-lwp.
1.93	30-May-2005	christos	branches: 1.93.2; - const poisoning - avoid variable shadowing.
1.92	27-Feb-2005	perry	nuke trailing whitespace
1.91	10-Mar-2004	matt	branches: 1.91.8; 1.91.10; Fixes for that only support for the 2030 can be compiled in.
1.90	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.89	07-Aug-2003	mjacob	Add in changes from Matt Thomas to check for 64 bit PCI stuff. Note that a Qlogic 1020 is Fast Wide, not Ultra Fast Wide.
1.88	03-May-2003	wiz	branches: 1.88.2; DMA, not dma nor Dma.
1.87	25-Nov-2002	thorpej	Add ULL to 64-bit integer constants.
1.86	18-Oct-2002	mjacob	Put in, sight unseen, changes which should enable 10160 support. Interestingly enough, Sun actually made a product out of the 10160, which is a single channel variant of the 12160 (Ultra3) card.
1.85	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.84	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.83	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.82	12-Aug-2002	mjacob	Various changes to allow for overriding just loopid and/or iid. Fix a random 23XX ISR reading bug. Be chattier about HBA_BOTCH errors. More f/w cashdump layering.
1.81	15-Jun-2002	mjacob	Make all 23XX cards 'touched'. We seem to, too often, hang when trying to run ABOUT FIRMWARE at the first isp_reset call. We do check for the registers being set with values which would tell us if there's firmware running or not- but this seems to not always work. It's not essential, so move on.
1.80	17-May-2002	mjacob	Slight touchup on last change- make sure HBA_BOTCH is set (just in case somebody at a different level does the wrong thing) on ENOMEM/EAGAIN returns from loading DMA.
1.79	12-May-2002	matt	branches: 1.79.2; If a dma map can't be loaded due to resource shortage, make we report that as a resource shortage instead of a driver stuffup.
1.78	28-Apr-2002	mjacob	suggestion from Jason: have (in non-verbose boot case) card type printed out on same line as attachment line.
1.77	22-Mar-2002	mjacob	Handle Port Unavailable the same as Port Logout (i.e., send a LIP). Unconst pointer to f/w in the ispdv structure. Too many compilers get unhappy over our walking the array. Make casts as appropriate so that initialization in structure is still happy. Limit length of fabric to 256. This will all go away soon. Do a cleaner case of keeping multiple CPUs/threads from reading the same response queue entries.
1.76	21-Feb-2002	mjacob	Implement F/W crashdumps for 2200 && 2300 cards (kernel compile option). Distinguish between 2312 and 2300 cards (they are different). Enable RIO (Reduced Interrupt Operation) for the LVD cards (hey- I've seen batched completions of the 30 commands at a time with this,....)... If we get a Port Logout on local loop topologies, we have to force the f/w to log back in. The easiest way (for us) to do this is to force a LIP. This also will wake up the disk that probably just had a f/w crash. Implement mailbox 'continuations'- this allows interrupts to re-drive a mailbox command if it's one that just essentially repeats the previous mailbox command (e.g., f/w download). This saves a boatload of sleep/wakeup twitches. If we're not a 2300 and we're about to return with a 'bogus interrupt'- check the semaphore register to be non-zero at all and outgoing mailbox 0- this seems to be where some of the lost ISP1080 commands came from.
1.75	14-Dec-2001	mjacob	Major restructuring for swizzling to the request queue and unswizzling from the response queue. Instead of the ad hoc ISP_SWIZZLE_REQUEST, we now have a complete set of inline functions in isp_inline.h. Each platform is responsible for providing just one of a set of ISP_IOX_{GET,PUT}{8,16,32} macros. The reason this needs to be done is that we need to have a single set of functions that will work correctly on multiple architectures for both little and big endian machines. It also needs to work correctly in the case that we have the request or response queues in memory that has to be treated specially (e.g., have ddi_dma_sync called on it for Solaris after we update it or before we read from it). One thing that falls out of this is that we no longer build requests in the request queue itself. Instead, we build the request locally (e.g., on the stack) and then as part of the swizzling operation, copy it to the request queue entry we've allocated. I thought long and hard about whether this was too expensive a change to make as it in a lot of cases requires an extra copy. On balance, the flexbility is worth it. With any luck, the entry that we build locally stays in a processor writeback cache (after all, it's only 64 bytes) so that the cost of actually flushing it to the memory area that is the shared queue with the PCI device is not all that expensive. We may examine this again and try to get clever in the future to try and avoid copies. Another change that falls out of this is that MEMORYBARRIER should be taken a lot more seriously. The macro ISP_ADD_REQUEST does a MEMORYBARRIER on the entry being added. But there had been many other places this had been missing. It's now very important that it be done. For NetBSD, it does a ddi_dmamap_sync as appropriate. This gets us out of the explicit ddi_dmamap_sync on the whole response queue that we did for SBus cards at each interrupt. Set things up so that platforms that cannot have an SBus don't get a lot of the SBus code checks (dead coded out). Additional changes: Fix a longstanding buglet of sorts. When we get an entry via isp_getrqentry, the iptr value that gets returned is the value we intend to eventually plug into the ISP registers as the entry one past the last one we've written- not the current entry we're updating. All along we've been calling sync functions on the wrong index value. Argh. The 'fix' here is to rename all 'iptr' variables as 'nxti' to remember that this is the 'next' pointer- not the current pointer. Devote a single bit to mboxbsy- and set aside bits for output mbox registers that we need to pick up- we can have at least one command which does not have any defined output registers (MBOX_EXECUTE_FIRMWARE). Explicitly decode GetAllNext SNS Response back as a GetAllNext response. Otherwise, we won't unswizzle it correctly. Nuke some additional __P macros.
1.74	13-Nov-2001	lukem	add RCSID
1.73	06-Oct-2001	mjacob	Respect QLogic's errata- read BIU_ISR even on the 2300 to see if there's an interrupt (avoids PCI parity errors which can occur on the 2312 if you access some registers from the host at the same time the RISC on the 2312 is accessing them).
1.72	01-Sep-2001	mjacob	branches: 1.72.2; Add support for 2 Gigabit cards (2300/2312). This necessitated a change in how interrupts are down- the 23XX has not only a different place to check for an interrupt, but unlike all other QLogic cards, you have to read the status as a 32 bit word- not 16 bit words. Rather than have device specific functions as called from the core module (in isp_intr), it makes more sense to have the platform/bus modules do the gruntwork of splitting out the isr, semaphore register and the first outgoing mailbox register (if needed) prior to calling isp_intr (if calling isp_intr is necessary at all).
1.71	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.70	07-Jul-2001	thorpej	branches: 1.70.2; bzero -> memset
1.69	14-Jun-2001	mjacob	Ooops- we need to allocate an array of pointers to scsipi_xfer structures isp_maxcmds in length- not an array of scsipi_xfer structures isp_maxcmds in length.
1.68	14-Mar-2001	mjacob	Clean up licence a bit. ANSIfy.
1.67	07-Mar-2001	thorpej	Add BUS_DMA_STREAMING to data transfer maps.
1.66	12-Feb-2001	mjacob	branches: 1.66.2; set default HBA role
1.65	30-Dec-2000	mjacob	Use bootverbose to now set whether or not we add ISP_LOGCONFIG and ISP_LOGINFO messages into the default logging mask.
1.64	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.63	28-Dec-2000	mjacob	Apply Bill Sommerfeld's -Wformat changes.
1.62	23-Dec-2000	wiz	Fix pathnames in comment.
1.61	09-Dec-2000	mjacob	Finally fix this driver to be sensible about the ENDIAN dance. It's not quite simply a question of the Qlogic being little endian and having to have stuff swapped on big endian machines- it also has to do with the fact that the SBus and PCI DMA layouts are wierd with respect to this. At any rate, now finally fixed- works on Mac G4, tested it on a SS10 for sparc, checked on alpha to see if I've broken anything, and as soon as I get another spare afternoon I'll finally install a sparc64 version which should just work (as it'll be like the Mac).
1.60	06-Dec-2000	mjacob	Correct omission by the previous committer.
1.59	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.58	17-Oct-2000	mjacob	quiet chatty boot messages
1.57	16-Oct-2000	mjacob	Remember to establish default PORT WWN. Add in some LOGDEBUG2 messages for the dma mapping code.
1.56	14-Aug-2000	mjacob	Add a maintenance note. Minor change in some logic in the DMA mapping.
1.55	03-Aug-2000	mjacob	More compilation breakage in the non-DEBUG case (from Castor Fu <castor@geocast.com>)
1.54	02-Aug-2000	mjacob	Fix bonehead bug for compiles w/o DEBUG set (thanks to KANETA Shin'ichi <kaneta@cr.chiba-u.ac.jp>)
1.53	01-Aug-2000	mjacob	Core version 2.0 (platform version 1.0) rewrite of ISP driver. Some interace cleanups, some new common functions. The major impact that will be noticeable right away is that if you boot with not Fibre connected to the FC cards, you no longer hang indefinitely.
1.52	05-Jul-2000	mjacob	Use new isp_handle_index function. Redo how firmware is checked for and loaded. Remember to enable interrupts after isp_reset but before isp_attach. Return CMD_EAGAIN on request queue overflow so we can retry the command when there's more queue space.
1.51	19-Feb-2000	mjacob	branches: 1.51.4; Support for > 12 byte commands for parallel SCSI.
1.50	12-Feb-2000	mjacob	add (untested in NetBSD yet) Qlogic 12160 (Ultra3) suppot
1.49	20-Dec-1999	mjacob	get back to looking at config flags
1.48	16-Dec-1999	mjacob	Dual LVD (1280) support.
1.47	04-Dec-1999	mjacob	request/response queues now indexed via u_int16_t. Add back in f/w inclusion and clean up it's compilation. get chip rev out of pci header.
1.46	28-Oct-1999	mjacob	Remove reference to loadable f/w.
1.45	17-Oct-1999	mjacob	branches: 1.45.2; 1.45.4; Oops- missed a MAXISPREQUEST sized thing- these are all gone now. It's amazing this didn't eat me.
1.44	17-Oct-1999	mjacob	stylistic cleanups
1.43	14-Oct-1999	mjacob	F/W version no longer relevant. Only do WWN seed for FC cards. Now we malloc xflist and dma maps in mailbox setup routines. We also do the appropriate endian swizzling at the end of a dma map routine.
1.42	30-Sep-1999	thorpej	branches: 1.42.2; Update for SCSPI changes.
1.41	05-Jul-1999	mjacob	Change to approved NASA/Ames copyright. Add in Qlogic 2200 support.
1.40	12-May-1999	mjacob	Do a fairly large internal restructuring to accomodate dual-bus host adapters (e.g., the 1240). Include the new 1080/1240 NVRAM layout reading code. Some moderately significant mailbox changes were necessary also to accomodate a second channel.
1.39	04-Apr-1999	mjacob	Read the board revision and trim cache line size back from 16 to 1 for early revision 2100 boards. Make sure to turn ROM off for these boards.
1.38	27-Mar-1999	mjacob	branches: 1.38.2; 1.38.4; Oops- if you enable 1080 support, remember to load the f/w
1.37	26-Mar-1999	mjacob	add isp1080 support and some basic PDB change stuff
1.36	17-Mar-1999	mjacob	prototype 1080/1240 support
1.35	09-Feb-1999	mjacob	clean up pci configuration a bit
1.34	30-Jan-1999	mjacob	roll internal release tag
1.33	10-Jan-1999	mjacob	oops. Forgot to turn on PCI burst.
1.32	28-Dec-1998	mjacob	I mean really clean up headers
1.31	28-Dec-1998	mjacob	clean up headers; remove early revision 2100 workarounds
1.30	17-Sep-1998	mjacob	dma setup now returns CMD_QUEUED or CMD_COMPLETE. Clarify a ternary operation that needed some parentheses.
1.29	08-Sep-1998	mjacob	Remove clocks as common code now can determine what clock rates to use. Remove other than 8 byte bursts- until we clarify whether or not buggy PCI implementatiosn really cause problems.
1.28	12-Aug-1998	mjacob	minor change to make the message less linux like
1.27	31-Jul-1998	thorpej	Back out part of previous; it is perfectly reasonable, and desirable, to print out debugging info ifdef DEBUG.
1.26	31-Jul-1998	mjacob	Part A (Cosmetic nonsense): >We're not Linux. If we still want driver version strings to be displayed >at boot, then I suppose DEBUG is a reasonable compromise. Makes the whole concept useless. This is for default printouts. If you can build a debug kernel, you know what version you have. This was under the concept of 'RAS' so that hapless users could tell you microversion things. But I guess this isn't the right way according to our local Jesuits. Oh, well. I'll think up something different and hopefully less objectionable. And yes, NetBSD isn't linux. The developers seem to be equally bad tempered, but linux is more successful. Part B: Field interrupts in OS layer so that (in this OS) bus_dmamap_sync(POSTREAD) can (formally) ensure that the result queue is stable wrt to buffering and that for sending a command a bus_dmamap_sync(PREWRITE) is done to ensure that the device gets a good view of what the mailbox contents should be.
1.25	20-Jul-1998	thorpej	branches: 1.25.2; Oops, unused variable warning.
1.24	20-Jul-1998	thorpej	We're not Linux. If we still want driver version strings to be displayed at boot, then I suppose DEBUG is a reasonable compromise.
1.23	18-Jul-1998	mjacob	fix bogus initialization
1.22	18-Jul-1998	mjacob	locking now out of main framework for isp_reset/isp_init
1.21	15-Jul-1998	mjacob	no framework
1.20	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.19	04-Feb-1998	thorpej	Add offset and length parameters to bus_dmamap_sync(), used for specifiying partial syncs of a DMA mapping.
1.18	04-Feb-1998	thorpej	Add dm_mapsize to bus_dmamap_t and rename BUS_DMAMEM_NOSYNC to BUS_DMA_COHERENT.
1.17	10-Sep-1997	mjacob	branches: 1.17.2; For the 2100- the flags weren't quite right- but things worked anyway. Wierd.
1.16	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.15	16-Aug-1997	mjacob	Major rework to support Qlogic ISP2100 (Fibre Channel PCI card). Also now able to read SXP registers (instead of panicing). Probably have done a bit of setup overkill- we now alloc some scratch memory for the 2100 that I thought I would need for the Port database, but so far I haven't needed to retrieve the port database. Well, early days yet.
1.14	08-Jun-1997	thorpej	branches: 1.14.2; 1.14.4; Fix pasto: sbus -> pci
1.13	08-Jun-1997	thorpej	- Update for changes to core ISP driver. - Convert to use bus dma, greatly simplifying DMA setup.
1.12	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.11	13-Apr-1997	cgd	use pci_map_register(). map both mem and I/O spaces, preferring mem space if it's usable, but falling back to I/O space if mem isn't usable.
1.10	05-Apr-1997	mjacob	Add 60 (MHz) value for PCI clock frequency value. Also enable DMA bursting.
1.9	28-Mar-1997	cgd	clean up (unify) the register mapping code a bit
1.8	28-Mar-1997	cgd	punt pci_mem_find's &cacheable arg, and pass NULL instead. We never want to map the memory register cacheable.
1.7	28-Mar-1997	cgd	clean up isp_pci_prefer_io comment
1.6	16-Mar-1997	cgd	map either via memory or I/O space, depending on the setting of a patchable kernel variable. eventually, it might make sense to turn this into a driver flag or something.
1.5	13-Mar-1997	cgd	change microcode array definitions from "unsigned short" to the more correct u_int16_t, and remove bogus casts that the old definition required.
1.4	13-Mar-1997	mycroft	Make the microcode table const.
1.3	13-Mar-1997	cgd	include <vm/vm.h> for vtophys() on x86. kill extra <sys/kernel.h> incl.
1.2	12-Mar-1997	cgd	NetBSD RCS ID tweaks, a few comment block tweaks. Also, make copyright notices consistent (per Matt Jacob).
1.1	12-Mar-1997	cgd	branches: 1.1.1; Initial revision
1.1.1.1	12-Mar-1997	cgd	ISP 10x0 driver from Matthew Jacob of NASA Ames Research Center. (March 12, 1997 version).
1.14.4.3	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.4.2	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.4.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.2.2	26-Aug-1997	bouyer	Update from trunk.
1.14.2.1	01-Jul-1997	bouyer	Updates for new scsipi subsystem. Actally known to work on i386 and sparc.
1.17.2.1	07-Nov-1998	cgd	pull up revs 1.18-1.30 from trunk, and patch to make functional in 1.3.x. (mjacob)
1.25.2.1	08-Aug-1998	eeh	Revert cdevsw mmap routines to return int.
1.38.4.2	02-Aug-1999	thorpej	Update from trunk.
1.38.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.38.2.2	13-May-2000	he	Apply patch (requested by Matthew Jacob via he): Correctly account for F-port vs. F-port (no FLOGI_ACC) topologies. Make sure we get a port database entry for the fabric name server. Preserve fabric logins if the device didn't change across fabric or port database changes, or the device has already logged into us (e.g., for target/initiator dual role devices like Veritas SANbox). Propagate class 3 service parameter changes where devices can change roles. Fix all occurrences of setting a sendmarker so that setting it for one bus on dual bus cards doesn't wipe a pending sendmarker for other busses on the same card. Comments added and clarifications made in some of the target mode code. Add support for > 12 byte CDBs. Split out nvram reading into per-card functions. Add proper Ultra2/Ultra3 support. Upgrade firmware.
1.38.2.1	08-Jan-2000	he	Pull up revisions 1.31-1.41,1.43-1.49 (via patch, requested by mjacob): Untangle Qlogic firmware copyright problems and update firmware. Untangle some MD support issues for said firmware. Add 1280 (Dual LVD), 1240 (Dual Ultra), 2200 (2nd Generation Qlogic FC chipset). Fix some synchronous negotiation problems for parallel SCSI. Firm up some Fabric Support issues.
1.42.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.45.4.1	15-Nov-1999	fvdl	Sync with -current
1.45.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.45.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.45.2.6	05-Jan-2001	bouyer	Sync with HEAD
1.45.2.5	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.45.2.4	08-Dec-2000	bouyer	Sync with HEAD.
1.45.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.45.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.45.2.1	20-Oct-1999	thorpej	Eliminate sleeping from adapter routines.
1.51.4.4	16-Mar-2001	he	Pull up revisions 1.62-1.63,1.66,1.68 (requested by mjacob): Fix pathnames in comment. Apply -Wformat changes. Set default HBA role (to ISP_ROLE_INITIATOR).
1.51.4.3	26-Jan-2001	jhawk	Back out revision 1.60 (requested by mjacob), erroneously pulled up.
1.51.4.2	25-Jan-2001	jhawk	Pull up revisions 1.57-1.58, 1.60-1.61 (requested by mjacob): Add in correct SBus bursting; upgrade to 2.01.26 firmware; pull to latest initiator mode level; make changes to WWN default handling; quiet chatty boot messages; fix endian code so MacPPC works; fix bug in lock recursion counter; fix bug which excluded all but NL-ports from being logged into on a fabric.
1.51.4.1	28-Aug-2000	mjacob	Complete rewrite of internals for isp core version 2.0. Pullup to netbsd-1-5 approved by thorpej@netbsd.org.
1.66.2.14	11-Dec-2002	thorpej	Sync with HEAD.
1.66.2.13	11-Nov-2002	nathanw	Catch up to -current
1.66.2.12	18-Oct-2002	nathanw	Catch up to -current.
1.66.2.11	13-Aug-2002	nathanw	Catch up to -current.
1.66.2.10	20-Jun-2002	nathanw	Catch up to -current.
1.66.2.9	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.66.2.8	28-Feb-2002	nathanw	Catch up to -current.
1.66.2.7	08-Jan-2002	nathanw	Catch up to -current.
1.66.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.66.2.5	22-Oct-2001	nathanw	Catch up to -current.
1.66.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.66.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.66.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.66.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.70.2.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.70.2.6	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.70.2.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.70.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.70.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.70.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.70.2.1	03-Aug-2001	lukem	update to -current
1.72.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.79.2.3	29-Aug-2002	gehenna	catch up with -current.
1.79.2.2	20-Jun-2002	gehenna	catch up with -current.
1.79.2.1	30-May-2002	gehenna	Catch up with -current.
1.88.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.88.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.88.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.88.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.88.2.1	03-Aug-2004	skrll	Sync with HEAD
1.91.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.91.8.1	29-Apr-2005	kent	sync with -current
1.93.2.5	17-Mar-2008	yamt	sync with head.
1.93.2.4	03-Sep-2007	yamt	sync with head.
1.93.2.3	26-Feb-2007	yamt	sync with head.
1.93.2.2	30-Dec-2006	yamt	sync with head.
1.93.2.1	21-Jun-2006	yamt	sync with head.
1.94.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.94.10.1	19-Apr-2006	elad	sync with head.
1.94.8.2	03-Sep-2006	yamt	sync with head.
1.94.8.1	01-Apr-2006	yamt	sync with head.
1.94.6.1	22-Apr-2006	simonb	Sync with head.
1.94.4.1	09-Sep-2006	rpaulo	sync with head
1.96.4.2	10-Dec-2006	yamt	sync with head.
1.96.4.1	22-Oct-2006	yamt	sync with head
1.96.2.2	01-Feb-2007	ad	Sync with head.
1.96.2.1	18-Nov-2006	ad	Sync with head.
1.99.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.100.4.1	11-Jul-2007	mjf	Sync with head.
1.100.2.2	15-Jul-2007	ad	Sync with head.
1.100.2.1	27-May-2007	ad	Sync with head.
1.103.28.2	02-Jun-2008	mjf	Sync with HEAD.
1.103.28.1	03-Apr-2008	mjf	Sync with HEAD.
1.103.24.1	24-Mar-2008	keiichi	sync with head.
1.103.8.1	23-Mar-2008	matt	sync with HEAD
1.105.18.2	23-Jul-2009	jym	Sync with HEAD.
1.105.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.105.4.4	11-Aug-2010	yamt	sync with head.
1.105.4.3	16-Sep-2009	yamt	sync with head
1.105.4.2	18-Jul-2009	yamt	sync with head.
1.105.4.1	16-May-2009	yamt	sync with head
1.110.4.3	31-May-2011	rmind	sync with head
1.110.4.2	05-Mar-2011	rmind	sync with head
1.110.4.1	30-May-2010	rmind	sync with head
1.110.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.113.4.1	05-Mar-2011	bouyer	Sync with HEAD
1.113.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.115.18.1	18-May-2014	rmind	sync with head
1.115.14.2	03-Dec-2017	jdolecek	update from HEAD
1.115.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.115.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.117.6.2	05-Oct-2016	skrll	Sync with HEAD
1.117.6.1	09-Jul-2016	skrll	Sync with HEAD
1.119.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.119.18.1	10-Jun-2019	christos	Sync with HEAD
1.119.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.19	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.18	27-Oct-2012	chs	branches: 1.18.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.17	31-Jul-2012	bouyer	branches: 1.17.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.16	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.15	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.14	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.13	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.12	04-Apr-2011	dyoung	branches: 1.12.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.11	05-Nov-2010	jakllsch	branches: 1.11.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.10	10-Apr-2008	cegger	branches: 1.10.24; 1.10.26; use aprint_*_dev and device_xname
1.9	18-Mar-2008	cube	Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.8	09-Feb-2007	ad	branches: 1.8.20; 1.8.36; 1.8.40; Merge newlock2 to head.
1.7	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.6	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.5	30-Jun-2006	xtraeme	branches: 1.5.4; 1.5.6; Match IT8211 IDE Controller.
1.4	11-Dec-2005	christos	branches: 1.4.4; 1.4.8; 1.4.16; merge ktrace-lwp.
1.3	24-May-2005	lukem	branches: 1.3.2; add __KERNEL_RCSID
1.2	27-Feb-2005	perry	nuke trailing whitespace
1.1	01-Dec-2004	grant	branches: 1.1.4; 1.1.6; 1.1.8; add iteide(4) driver for ITE8212-based IDE controllers. from OpenBSD, ported to NetBSD by me. ok'd by bouyer@, thorpej@.
1.1.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	29-Apr-2005	kent	sync with -current
1.1.4.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.4.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.2	18-Dec-2004	skrll	Sync with HEAD.
1.1.4.1	01-Dec-2004	skrll	file iteide.c was added on branch ktrace-lwp on 2004-12-18 09:32:08 +0000
1.3.2.3	24-Mar-2008	yamt	sync with head.
1.3.2.2	26-Feb-2007	yamt	sync with head.
1.3.2.1	30-Dec-2006	yamt	sync with head.
1.4.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.4.8.1	11-Aug-2006	yamt	sync with head
1.4.4.1	09-Sep-2006	rpaulo	sync with head
1.5.6.2	10-Dec-2006	yamt	sync with head.
1.5.6.1	22-Oct-2006	yamt	sync with head
1.5.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.5.4.1	18-Nov-2006	ad	Sync with head.
1.8.40.2	02-Jun-2008	mjf	Sync with HEAD.
1.8.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.8.36.1	24-Mar-2008	keiichi	sync with head.
1.8.20.1	23-Mar-2008	matt	sync with HEAD
1.10.26.2	21-Apr-2011	rmind	sync with head
1.10.26.1	05-Mar-2011	rmind	sync with head
1.10.24.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.11.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.12.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.4.1	30-Oct-2012	yamt	sync with head
1.17.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.17.2.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.17.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.18.2.1	18-May-2014	rmind	sync with head
1.11	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.10	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.9	23-Mar-2014	christos	branches: 1.9.28; 1.9.30; remove unused
1.8	27-Oct-2012	chs	branches: 1.8.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.7	10-Apr-2008	cegger	branches: 1.7.38; 1.7.48; use aprint_*_dev and device_xname
1.6	19-Oct-2007	ad	branches: 1.6.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	11-Dec-2005	christos	branches: 1.5.30; 1.5.44; 1.5.46; 1.5.50; merge ktrace-lwp.
1.4	09-Jul-2005	christos	make the isdn drivers compile again with -Wconst, -Wshadow
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	27-Sep-2002	provos	branches: 1.2.2; 1.2.4; 1.2.10; 1.2.18; 1.2.20; remove trailing \n in panic(). approved perry.
1.1	24-Sep-2002	pooka	Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.2.20.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.18.1	29-Apr-2005	kent	sync with -current
1.2.10.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.10.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.4.1	27-Sep-2002	nathanw	file iwic_bchan.c was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.2.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.1	27-Sep-2002	jdolecek	file iwic_bchan.c was added on branch kqueue on 2002-10-10 18:40:53 +0000
1.3.4.2	27-Oct-2007	yamt	sync with head.
1.3.4.1	21-Jun-2006	yamt	sync with head.
1.5.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.46.1	06-Nov-2007	matt	sync with HEAD
1.5.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.30.1	23-Oct-2007	ad	Sync with head.
1.6.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.7.48.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.7.48.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.7.38.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.38.1	30-Oct-2012	yamt	sync with head
1.8.2.1	18-May-2014	rmind	sync with head
1.9.30.1	10-Jun-2019	christos	Sync with HEAD
1.9.28.2	30-Sep-2018	pgoyette	Ssync with HEAD
1.9.28.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.11	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.10	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.9	23-Mar-2014	christos	branches: 1.9.28; 1.9.30; remove unused
1.8	23-Mar-2014	christos	remove unused
1.7	27-Oct-2012	chs	branches: 1.7.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.6	10-Apr-2008	cegger	branches: 1.6.38; 1.6.48; use aprint_*_dev and device_xname
1.5	19-Oct-2007	ad	branches: 1.5.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.4	11-Dec-2005	christos	branches: 1.4.30; 1.4.44; 1.4.46; 1.4.50; merge ktrace-lwp.
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	27-Sep-2002	provos	branches: 1.2.2; 1.2.4; 1.2.10; 1.2.18; 1.2.20; remove trailing \n in panic(). approved perry.
1.1	24-Sep-2002	pooka	Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.2.20.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.18.1	29-Apr-2005	kent	sync with -current
1.2.10.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.4.1	27-Sep-2002	nathanw	file iwic_dchan.c was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.2.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.1	27-Sep-2002	jdolecek	file iwic_dchan.c was added on branch kqueue on 2002-10-10 18:40:53 +0000
1.3.4.1	27-Oct-2007	yamt	sync with head.
1.4.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.4.46.1	06-Nov-2007	matt	sync with HEAD
1.4.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.4.30.1	23-Oct-2007	ad	Sync with head.
1.5.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.6.48.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.48.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.6.38.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6.38.1	30-Oct-2012	yamt	sync with head
1.7.2.1	18-May-2014	rmind	sync with head
1.9.30.1	10-Jun-2019	christos	Sync with HEAD
1.9.28.2	30-Sep-2018	pgoyette	Ssync with HEAD
1.9.28.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.5	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.4	19-Oct-2007	ad	branches: 1.4.106; 1.4.108; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.3	11-Dec-2005	christos	branches: 1.3.30; 1.3.44; 1.3.46; 1.3.50; merge ktrace-lwp.
1.2	09-Jul-2005	christos	make the isdn drivers compile again with -Wconst, -Wshadow
1.1	24-Sep-2002	pooka	branches: 1.1.2; 1.1.4; 1.1.10; 1.1.26; Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.1.26.2	27-Oct-2007	yamt	sync with head.
1.1.26.1	21-Jun-2006	yamt	sync with head.
1.1.10.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.1	24-Sep-2002	nathanw	file iwic_fsm.c was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.1.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.1	24-Sep-2002	jdolecek	file iwic_fsm.c was added on branch kqueue on 2002-10-10 18:40:53 +0000
1.3.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.3.46.1	06-Nov-2007	matt	sync with HEAD
1.3.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.30.1	23-Oct-2007	ad	Sync with head.
1.4.108.1	10-Jun-2019	christos	Sync with HEAD
1.4.106.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.21	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.20	14-Jul-2016	msaitoh	branches: 1.20.16; 1.20.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.19	29-Mar-2014	christos	branches: 1.19.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	27-Oct-2012	chs	branches: 1.18.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.17	22-May-2011	mrg	branches: 1.17.4; 1.17.14; add some fujitsu ethernet id's and a new vendor id. partly from Michael Moll <kvedulv@kvedulv.de>, taken from openbsd. XXX: move the #if 0'd usage of PCI_VENDOR_CITICORP inside the #if 0 code XXX: region, as this id conflits with the (4th) fijutsi id.
1.16	26-Nov-2009	njoly	branches: 1.16.4; 1.16.6; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.15	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.14	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.13	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.12	10-Apr-2008	cegger	branches: 1.12.4; 1.12.18; use aprint_*_dev and device_xname
1.11	19-Oct-2007	ad	branches: 1.11.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.10	16-Nov-2006	christos	branches: 1.10.8; 1.10.22; 1.10.24; 1.10.28; __unused removal on arguments; approved by core.
1.9	16-Oct-2006	pooka	* __unused games + other relevant fixes * give me ANSI C or give me death (at least where it could be easily spotted)
1.8	11-Dec-2005	christos	branches: 1.8.20; 1.8.22; merge ktrace-lwp.
1.7	03-Oct-2003	pooka	branches: 1.7.16; Ready the kernel side of i4b for primary rate interface support by removing assumptions that there are only two B channels and by adding support for a varying number of channels. Due to this, rename previously used isdn identified "bri" to "isdnif", which better describes the current situation.
1.6	08-May-2003	martin	branches: 1.6.2; We are getting passed an integer value via a void* arg (in an ioctl() like function) - cast it to an intptr_t. Addresses parts of PR kern/21448 by Frank Kardel.
1.5	23-Oct-2002	pooka	avoid namespace collision with IMASK
1.4	02-Oct-2002	thorpej	branches: 1.4.2; 1.4.4; Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	24-Sep-2002	pooka	Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.4.4.3	11-Nov-2002	nathanw	Catch up to -current
1.4.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.4.4.1	02-Oct-2002	nathanw	file iwic_pci.c was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.4.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.4.2.1	02-Oct-2002	jdolecek	file iwic_pci.c was added on branch kqueue on 2002-10-10 18:40:54 +0000
1.6.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.7.16.2	27-Oct-2007	yamt	sync with head.
1.7.16.1	30-Dec-2006	yamt	sync with head.
1.8.22.2	10-Dec-2006	yamt	sync with head.
1.8.22.1	22-Oct-2006	yamt	sync with head
1.8.20.1	18-Nov-2006	ad	Sync with head.
1.10.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.10.24.1	06-Nov-2007	matt	sync with HEAD
1.10.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.10.8.1	23-Oct-2007	ad	Sync with head.
1.11.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.12.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.12.4.2	11-Mar-2010	yamt	sync with head
1.12.4.1	16-May-2009	yamt	sync with head
1.16.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.16.4.1	31-May-2011	rmind	sync with head
1.17.14.3	03-Dec-2017	jdolecek	update from HEAD
1.17.14.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.17.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.17.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.17.4.1	30-Oct-2012	yamt	sync with head
1.18.2.1	18-May-2014	rmind	sync with head
1.19.6.1	05-Oct-2016	skrll	Sync with HEAD
1.20.18.1	10-Jun-2019	christos	Sync with HEAD
1.20.16.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.3	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.2	23-Oct-2002	pooka	branches: 1.2.194; 1.2.196; avoid namespace collision with IMASK
1.1	24-Sep-2002	pooka	branches: 1.1.2; 1.1.4; Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.1.4.3	11-Nov-2002	nathanw	Catch up to -current
1.1.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.1	24-Sep-2002	nathanw	file iwicreg.h was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.1.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.1	24-Sep-2002	jdolecek	file iwicreg.h was added on branch kqueue on 2002-10-10 18:40:54 +0000
1.2.196.1	10-Jun-2019	christos	Sync with HEAD
1.2.194.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.7	22-Sep-2018	maxv	Remove iwic(4). It is part of ISDN, which we are now retiring. This driver was still marked as experimental (its man page dates back to 2002).
1.6	27-Oct-2012	chs	branches: 1.6.36; 1.6.38; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.5	19-Oct-2007	ad	branches: 1.5.54; 1.5.64; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.4	11-Dec-2005	christos	branches: 1.4.30; 1.4.44; 1.4.46; 1.4.50; merge ktrace-lwp.
1.3	09-Jul-2005	christos	make the isdn drivers compile again with -Wconst, -Wshadow
1.2	08-Jul-2003	itojun	branches: 1.2.16; function prototype must not have variable name
1.1	24-Sep-2002	pooka	branches: 1.1.2; 1.1.4; 1.1.10; Driver for Winbond W6692 passive ISDN cards. Ported from the FreeBSD driver by Ilpo Ruotsalainen <lonewolf@cubical.fi>, and reviewed by Martin.
1.1.10.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.10.1	03-Aug-2004	skrll	Sync with HEAD
1.1.4.2	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.1	24-Sep-2002	nathanw	file iwicvar.h was added on branch nathanw_sa on 2002-10-18 02:43:10 +0000
1.1.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.1	24-Sep-2002	jdolecek	file iwicvar.h was added on branch kqueue on 2002-10-10 18:40:54 +0000
1.2.16.2	27-Oct-2007	yamt	sync with head.
1.2.16.1	21-Jun-2006	yamt	sync with head.
1.4.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.4.46.1	06-Nov-2007	matt	sync with HEAD
1.4.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.4.30.1	23-Oct-2007	ad	Sync with head.
1.5.64.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.5.54.1	30-Oct-2012	yamt	sync with head
1.6.38.1	10-Jun-2019	christos	Sync with HEAD
1.6.36.1	30-Sep-2018	pgoyette	Ssync with HEAD
1.25	14-Jul-2016	msaitoh	KNF. No functional change.
1.24	31-Jul-2012	bouyer	branches: 1.24.2; 1.24.16; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.23	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.22	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.21	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.20	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.19	04-Apr-2011	dyoung	branches: 1.19.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.18	13-Feb-2011	jakllsch	Turn on detachment for ixpide(4).
1.17	05-Nov-2010	jakllsch	branches: 1.17.2; 1.17.4; Pave the way for detachment of pciide(4)-family ATA controllers.
1.16	24-Feb-2010	dyoung	branches: 1.16.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.15	08-Jan-2010	dyoung	branches: 1.15.2; Expand PMF_FN_* macros.
1.14	04-Nov-2008	reinoud	Implement powersave for ixpide(4). Tested it on an Asus IXP250 with IGP. It seems to sleep and awake fine now without errors or warnings.
1.13	06-Sep-2008	rmind	branches: 1.13.2; 1.13.4; PR/39442: Jeff Thieleke: Support for ATI SB700/SB800 controllers.
1.12	29-Apr-2008	martin	branches: 1.12.2; 1.12.6; Convert to new 2 clause license
1.11	18-Mar-2008	cube	branches: 1.11.2; 1.11.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.10	10-Sep-2007	cube	branches: 1.10.16; 1.10.20; Remove 3rd clause and my name from all the licences which were only in my name.
1.9	09-Feb-2007	ad	branches: 1.9.6; 1.9.18; 1.9.20; Merge newlock2 to head.
1.8	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.7	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.6	30-Jun-2006	xtraeme	branches: 1.6.4; 1.6.6; Match ATI IXP SB600 IDE/SATA.
1.5	11-Dec-2005	christos	branches: 1.5.4; 1.5.8; 1.5.16; merge ktrace-lwp.
1.4	04-Jul-2005	augustss	branches: 1.4.2; The SB400 SATA controllers seem to work fine for me with this driver.
1.3	07-Mar-2005	christos	PR/29614: Andrzej Kukula: Missing PCI IDs of ATI IXP 300 and 400 ATA controller
1.2	27-Feb-2005	perry	nuke trailing whitespace
1.1	06-Dec-2004	cube	branches: 1.1.4; 1.1.6; 1.1.8; Add ixpide(4), a driver for the IDE controller found in the ATI IXP series of integrated chipsets. Documented from the Linux driver. Tested by reinoud@, ok'd by bouyer@.
1.1.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	29-Apr-2005	kent	sync with -current
1.1.4.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.4.4	08-Mar-2005	skrll	Sync with HEAD.
1.1.4.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.2	18-Dec-2004	skrll	Sync with HEAD.
1.1.4.1	06-Dec-2004	skrll	file ixpide.c was added on branch ktrace-lwp on 2004-12-18 09:32:08 +0000
1.4.2.4	24-Mar-2008	yamt	sync with head.
1.4.2.3	27-Oct-2007	yamt	sync with head.
1.4.2.2	26-Feb-2007	yamt	sync with head.
1.4.2.1	30-Dec-2006	yamt	sync with head.
1.5.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.5.8.1	11-Aug-2006	yamt	sync with head
1.5.4.1	09-Sep-2006	rpaulo	sync with head
1.6.6.2	10-Dec-2006	yamt	sync with head.
1.6.6.1	22-Oct-2006	yamt	sync with head
1.6.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.6.4.1	18-Nov-2006	ad	Sync with head.
1.9.20.2	23-Mar-2008	matt	sync with HEAD
1.9.20.1	06-Nov-2007	matt	sync with HEAD
1.9.18.1	02-Oct-2007	joerg	Sync with HEAD.
1.9.6.1	09-Oct-2007	ad	Sync with head.
1.10.20.4	17-Jan-2009	mjf	Sync with HEAD.
1.10.20.3	28-Sep-2008	mjf	Sync with HEAD.
1.10.20.2	02-Jun-2008	mjf	Sync with HEAD.
1.10.20.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.16.1	24-Mar-2008	keiichi	sync with head.
1.11.4.3	11-Mar-2010	yamt	sync with head
1.11.4.2	04-May-2009	yamt	sync with head.
1.11.4.1	16-May-2008	yamt	sync with head.
1.11.2.1	18-May-2008	yamt	sync with head.
1.12.6.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.12.6.1	19-Oct-2008	haad	Sync with HEAD.
1.12.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.13.4.1	12-Nov-2008	snj	Pull up following revision(s) (requested by reinoud in ticket #42): sys/dev/pci/ixpide.c: revision 1.14 Implement powersave for ixpide(4). Tested it on an Asus IXP250 with IGP. It seems to sleep and awake fine now without errors or warnings.
1.13.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.15.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.15.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.16.2.2	21-Apr-2011	rmind	sync with head
1.16.2.1	05-Mar-2011	rmind	sync with head
1.17.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.17.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.19.4.1	30-Oct-2012	yamt	sync with head
1.24.16.1	05-Oct-2016	skrll	Sync with HEAD
1.24.2.2	03-Dec-2017	jdolecek	update from HEAD
1.24.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.25	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.24	24-Apr-2021	thorpej	branches: 1.24.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.23	10-Nov-2019	chs	branches: 1.23.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.22	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.21	14-Jul-2016	msaitoh	branches: 1.21.16; 1.21.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.20	14-Jul-2016	msaitoh	KNF. No functional change.
1.19	29-Mar-2014	christos	branches: 1.19.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	31-Jul-2012	bouyer	branches: 1.18.2; 1.18.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.17	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.16	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.15	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.14	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.13	29-Mar-2012	pgoyette	Add match for JMB362 (with two SATA ports).
1.12	30-Jan-2012	drochner	branches: 1.12.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.11	13-Aug-2011	jakllsch	branches: 1.11.2; 1.11.6; Make correct spelling in comments a higher priority.
1.10	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.9	05-Nov-2010	jakllsch	branches: 1.9.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.8	27-Jul-2010	jakllsch	Support detachment of ahcisata(4). Use use 64-bit DMA tag (where available) for ahcisata(4) at jmide(4). Beginnings of detach/resume support for jmide(4). Sprinkle static. Misc. little changes.
1.7	19-Oct-2009	bouyer	branches: 1.7.2; 1.7.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.6	05-Jun-2008	bouyer	Call ahci_intr() with the device's softc, not the device_t. Should fix kern/38742
1.5	21-Mar-2008	xtraeme	branches: 1.5.2; 1.5.4; 1.5.6; Don't forget to set the atac dev pointer in the softc; ahcisata at jmide works again after that. Also use CFATTACH_DECL_NEW, device_t and related changes for the NJMAHCI block.
1.4	18-Mar-2008	cube	Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.3	11-Feb-2008	xtraeme	branches: 1.3.2; 1.3.6; Attach ahcisata(4) to any RAID controller that supports AHCI, and make it pass the ATAC_CAP_RAID capability for them. ok'ed by bouyer@.
1.2	31-May-2007	bouyer	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.14; 1.2.16; 1.2.18; 1.2.24; Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes.
1.1	15-May-2007	bouyer	branches: 1.1.2; add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work.
1.1.2.2	17-May-2007	yamt	sync with head.
1.1.2.1	15-May-2007	yamt	file jmide.c was added on branch yamt-idlelwp on 2007-05-17 13:41:31 +0000
1.2.24.1	18-Feb-2008	mjf	Sync with HEAD.
1.2.18.4	24-Mar-2008	yamt	sync with head.
1.2.18.3	11-Feb-2008	yamt	sync with head.
1.2.18.2	03-Sep-2007	yamt	sync with head.
1.2.18.1	31-May-2007	yamt	file jmide.c was added on branch yamt-lazymbuf on 2007-09-03 14:37:10 +0000
1.2.16.2	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.2.16.1	31-May-2007	wrstuden	file jmide.c was added on branch wrstuden-fixsa on 2007-09-03 07:04:35 +0000
1.2.14.2	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.2.14.1	31-May-2007	pavel	file jmide.c was added on branch netbsd-4 on 2007-08-31 20:09:25 +0000
1.2.12.1	23-Mar-2008	matt	sync with HEAD
1.2.4.2	11-Jul-2007	mjf	Sync with head.
1.2.4.1	31-May-2007	mjf	file jmide.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:51 +0000
1.2.2.3	09-Jun-2007	ad	Sync with head.
1.2.2.2	09-Jun-2007	ad	Sync with head.
1.2.2.1	31-May-2007	ad	file jmide.c was added on branch vmlocking on 2007-06-09 21:37:24 +0000
1.3.6.2	05-Jun-2008	mjf	Sync with HEAD. Also fix build.
1.3.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.3.2.1	24-Mar-2008	keiichi	sync with head.
1.5.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.5.4.3	11-Aug-2010	yamt	sync with head.
1.5.4.2	11-Mar-2010	yamt	sync with head
1.5.4.1	04-May-2009	yamt	sync with head.
1.5.2.1	17-Jun-2008	yamt	sync with head.
1.7.4.2	21-Apr-2011	rmind	sync with head
1.7.4.1	05-Mar-2011	rmind	sync with head
1.7.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.7.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.9.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.11.6.2	05-Apr-2012	mrg	sync to latest -current.
1.11.6.1	18-Feb-2012	mrg	merge to -current.
1.11.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.11.2.2	30-Oct-2012	yamt	sync with head
1.11.2.1	17-Apr-2012	yamt	sync with head
1.12.2.1	03-Apr-2012	riz	Pull up following revision(s) (requested by pgoyette in ticket #149): sys/dev/pci/pcidevs: revision 1.1108 sys/dev/pci/jmide.c: revision 1.13 Add JMB 362 (as found on ASUS M5A99X motherboard) Add match for JMB362 (with two SATA ports).
1.18.4.1	18-May-2014	rmind	sync with head
1.18.2.3	03-Dec-2017	jdolecek	update from HEAD
1.18.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.18.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.19.6.1	05-Oct-2016	skrll	Sync with HEAD
1.21.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.21.18.1	10-Jun-2019	christos	Sync with HEAD
1.21.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.23.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.24.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.4	24-Oct-2011	njoly	Remove unused local defines for PCI_COMMAND_STATUS_REG bits (already available from pcireg.h).
1.3	24-Jul-2010	jakllsch	correct transposition in comment
1.2	19-Oct-2009	bouyer	branches: 1.2.2; 1.2.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.1	15-May-2007	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.16; 1.1.18; 1.1.20; 1.1.44; add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work.
1.1.44.2	11-Aug-2010	yamt	sync with head.
1.1.44.1	11-Mar-2010	yamt	sync with head
1.1.20.2	03-Sep-2007	yamt	sync with head.
1.1.20.1	15-May-2007	yamt	file jmide_reg.h was added on branch yamt-lazymbuf on 2007-09-03 14:37:10 +0000
1.1.18.2	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.1.18.1	15-May-2007	wrstuden	file jmide_reg.h was added on branch wrstuden-fixsa on 2007-09-03 07:04:36 +0000
1.1.16.2	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.1.16.1	15-May-2007	pavel	file jmide_reg.h was added on branch netbsd-4 on 2007-08-31 20:09:25 +0000
1.1.6.2	11-Jul-2007	mjf	Sync with head.
1.1.6.1	15-May-2007	mjf	file jmide_reg.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:52 +0000
1.1.4.2	09-Jun-2007	ad	Sync with head.
1.1.4.1	15-May-2007	ad	file jmide_reg.h was added on branch vmlocking on 2007-06-09 21:37:24 +0000
1.1.2.2	17-May-2007	yamt	sync with head.
1.1.2.1	15-May-2007	yamt	file jmide_reg.h was added on branch yamt-idlelwp on 2007-05-17 13:41:32 +0000
1.2.4.1	05-Mar-2011	rmind	sync with head
1.2.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.17	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.16	26-Apr-2021	thorpej	branches: 1.16.6; Be explicit about the "eap" interface attribute when attaching a "joy" instance.
1.15	24-Apr-2021	thorpej	branches: 1.15.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.14	08-May-2019	isaki	branches: 1.14.12; Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.13	23-Nov-2011	jmcneill	branches: 1.13.50; 1.13.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.12	07-May-2009	cegger	branches: 1.12.12; 1.12.14; use device_private()
1.11	06-May-2009	cegger	use device_private()
1.10	10-Apr-2008	cegger	branches: 1.10.4; 1.10.16; 1.10.18; use aprint_*_dev and device_xname
1.9	26-Mar-2008	xtraeme	Split device_t/softc for joy(4) and all its attachments. Also use device_lookup_private() and related cosmetic changes. Reviewed by cube@.
1.8	11-Dec-2007	lukem	branches: 1.8.8; use __KERNEL_RCSID()
1.7	19-Oct-2007	ad	branches: 1.7.4; 1.7.6; 1.7.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.6	16-Nov-2006	christos	branches: 1.6.8; 1.6.22; 1.6.24; 1.6.28; __unused removal on arguments; approved by core.
1.5	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.4	11-Dec-2005	christos	branches: 1.4.20; 1.4.22; merge ktrace-lwp.
1.3	28-Jun-2005	thorpej	branches: 1.3.2; Use ANSI function decls and static.
1.2	03-Aug-2004	drochner	use joydetach() now that we have it
1.1	08-Jul-2004	drochner	branches: 1.1.2; support the game port on eap1371 This is only usable with some caution because these soundcards only allow to map IO port 0x20x for this, thus bypassing PCI address management. Very likely this will only work on primary PCI buses, and there is some potential for conflicts with ISA devices as well. (XXX cannot be detached because the "joy" driver doesn't support it yet)
1.1.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.1.2.2	03-Aug-2004	skrll	Sync with HEAD
1.1.2.1	08-Jul-2004	skrll	file joy_eap.c was added on branch ktrace-lwp on 2004-08-03 10:49:09 +0000
1.3.2.3	21-Jan-2008	yamt	sync with head
1.3.2.2	27-Oct-2007	yamt	sync with head.
1.3.2.1	30-Dec-2006	yamt	sync with head.
1.4.22.2	10-Dec-2006	yamt	sync with head.
1.4.22.1	22-Oct-2006	yamt	sync with head
1.4.20.1	18-Nov-2006	ad	Sync with head.
1.6.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.6.24.2	09-Jan-2008	matt	sync with HEAD
1.6.24.1	06-Nov-2007	matt	sync with HEAD
1.6.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.6.8.1	23-Oct-2007	ad	Sync with head.
1.7.8.2	02-Jan-2008	bouyer	Sync with HEAD
1.7.8.1	13-Dec-2007	bouyer	Sync with HEAD
1.7.6.1	11-Dec-2007	yamt	sync with head.
1.7.4.1	26-Dec-2007	ad	Sync with head.
1.8.8.2	02-Jun-2008	mjf	Sync with HEAD.
1.8.8.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.10.16.1	12-Dec-2008	ad	Checkpoint work in progress.
1.10.4.1	16-May-2009	yamt	sync with head
1.12.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.12.12.1	17-Apr-2012	yamt	sync with head
1.13.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.13.50.1	10-Jun-2019	christos	Sync with HEAD
1.14.12.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.15.2.1	13-May-2021	thorpej	Sync with HEAD.
1.16.6.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.17	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.16	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.15	23-Nov-2011	jmcneill	branches: 1.15.50; 1.15.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.14	28-Apr-2008	martin	branches: 1.14.12; 1.14.34; 1.14.36; Remove clause 3 and 4 from TNF licenses
1.13	26-Mar-2008	xtraeme	branches: 1.13.2; 1.13.4; Split device_t/softc for joy(4) and all its attachments. Also use device_lookup_private() and related cosmetic changes. Reviewed by cube@.
1.12	19-Oct-2007	ad	branches: 1.12.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	16-Nov-2006	christos	branches: 1.11.8; 1.11.22; 1.11.24; 1.11.28; __unused removal on arguments; approved by core.
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	11-Dec-2005	christos	branches: 1.9.20; 1.9.22; merge ktrace-lwp.
1.8	28-Jun-2005	thorpej	branches: 1.8.2; Use ANSI function decls and static.
1.7	04-Feb-2005	perry	de-__P
1.6	04-Dec-2003	keihan	branches: 1.6.8; 1.6.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.4	02-Oct-2002	thorpej	branches: 1.4.6; Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	25-Apr-2002	kleink	branches: 1.1.6; 1.1.8; Add a joystick attachment to the ESS Solo-1 driver.
1.1.8.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.8.1	25-Apr-2002	jdolecek	file joy_eso.c was added on branch kqueue on 2002-06-23 17:47:50 +0000
1.1.6.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.1.6.1	25-Apr-2002	nathanw	file joy_eso.c was added on branch nathanw_sa on 2002-06-20 03:45:35 +0000
1.4.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.4.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.6.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	12-Feb-2005	yamt	sync with head.
1.6.8.1	29-Apr-2005	kent	sync with -current
1.8.2.2	27-Oct-2007	yamt	sync with head.
1.8.2.1	30-Dec-2006	yamt	sync with head.
1.9.22.2	10-Dec-2006	yamt	sync with head.
1.9.22.1	22-Oct-2006	yamt	sync with head
1.9.20.1	18-Nov-2006	ad	Sync with head.
1.11.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.24.1	06-Nov-2007	matt	sync with HEAD
1.11.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.8.1	23-Oct-2007	ad	Sync with head.
1.12.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.4.1	16-May-2008	yamt	sync with head.
1.13.2.1	18-May-2008	yamt	sync with head.
1.14.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.14.34.1	17-Apr-2012	yamt	sync with head
1.14.12.1	12-Dec-2008	ad	Checkpoint work in progress.
1.15.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.15.50.1	10-Jun-2019	christos	Sync with HEAD
1.21	07-May-2014	joerg	Fix impossible error condition.
1.20	30-Jan-2012	drochner	branches: 1.20.6; 1.20.10; 1.20.20; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.19	05-Dec-2011	christos	make joy code consistent (avoid games with pointers), and fix the isa softc declaration
1.18	23-Nov-2011	jmcneill	branches: 1.18.2; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.17	28-Apr-2008	martin	branches: 1.17.12; 1.17.34; 1.17.36; Remove clause 3 and 4 from TNF licenses
1.16	26-Mar-2008	xtraeme	branches: 1.16.2; 1.16.4; Split device_t/softc for joy(4) and all its attachments. Also use device_lookup_private() and related cosmetic changes. Reviewed by cube@.
1.15	19-Oct-2007	ad	branches: 1.15.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.14	16-Nov-2006	christos	branches: 1.14.8; 1.14.22; 1.14.24; 1.14.28; __unused removal on arguments; approved by core.
1.13	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.12	11-Dec-2005	christos	branches: 1.12.20; 1.12.22; merge ktrace-lwp.
1.11	28-Jun-2005	thorpej	branches: 1.11.2; Use ANSI function decls and static.
1.10	04-Feb-2005	perry	de-__P
1.9	23-Apr-2004	itojun	branches: 1.9.4; 1.9.6; pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.8	03-Apr-2004	martti	Added another SB Live! 5.1 soundcard and joystick version
1.7	04-Dec-2003	keihan	branches: 1.7.2; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	16-Oct-2002	jdolecek	branches: 1.6.6; match the joystick device on Creative SBLive! card; note this only works when emuxki(4) driver is compiled into the kernel, it has to enable the joystick port
1.5	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	14-Apr-2002	martin	Fix copyright notice.
1.1	02-Feb-2002	jdolecek	branches: 1.1.2; 1.1.4; Reduce the massive code duplication regarding joy(4). Split it into MI and MD parts, and make ISA/ISAPNP/PCI joy(4) attachments MI.
1.1.4.5	11-Nov-2002	nathanw	Catch up to -current
1.1.4.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.3	17-Apr-2002	nathanw	Catch up to -current.
1.1.4.2	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.1	02-Feb-2002	nathanw	file joy_pci.c was added on branch nathanw_sa on 2002-02-28 04:14:03 +0000
1.1.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.1	02-Feb-2002	jdolecek	file joy_pci.c was added on branch kqueue on 2002-02-11 20:10:01 +0000
1.6.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.6.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.6.1	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	05-Apr-2004	tron	Pull up revision 1.8 (requested by martti in ticket #54): Added another SB Live! 5.1 soundcard and joystick version
1.9.6.1	12-Feb-2005	yamt	sync with head.
1.9.4.1	29-Apr-2005	kent	sync with -current
1.11.2.2	27-Oct-2007	yamt	sync with head.
1.11.2.1	30-Dec-2006	yamt	sync with head.
1.12.22.2	10-Dec-2006	yamt	sync with head.
1.12.22.1	22-Oct-2006	yamt	sync with head
1.12.20.1	18-Nov-2006	ad	Sync with head.
1.14.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.14.24.1	06-Nov-2007	matt	sync with HEAD
1.14.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.14.8.1	23-Oct-2007	ad	Sync with head.
1.15.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.15.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.16.4.1	16-May-2008	yamt	sync with head.
1.16.2.1	18-May-2008	yamt	sync with head.
1.17.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.17.34.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.17.34.1	17-Apr-2012	yamt	sync with head
1.17.12.1	12-Dec-2008	ad	Checkpoint work in progress.
1.18.2.1	18-Feb-2012	mrg	merge to -current.
1.20.20.1	10-Aug-2014	tls	Rebase.
1.20.10.1	18-May-2014	rmind	sync with head
1.20.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.26	13-Apr-2025	rin	ld(4): Convert blkno argument for sc_dump() to daddr_t PR kern/59153 (1) For backends that accept 64-bit block address, i.e., nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4), this should enable to dump beyond 2Gi blocks. (2) sdmmc(4) backend allows to dump up to the last block. (3) For other backends, block address is handled as `int`. Some of them may support blocks up to 4Gi, but I do not have enough time to examine datasheets. So, continue to reject >2Gi blocks as before. XXX This is KABI change, and cannot be pulled up into netbsd-{10,9}. XXX Compile-test only (for amd64/ALL) due to lack of large SSDs ;) Thanks mlelstv@ for discussion and careful review!!
1.25	27-Sep-2016	pgoyette	branches: 1.25.54; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.24	16-Sep-2016	jdolecek	modify ldattach() to have default strategy as a parameter
1.23	13-Apr-2015	riastradh	branches: 1.23.2; Convert sys/dev to use <sys/rndsource.h>.
1.22	02-Mar-2015	christos	From Tom Ivar Helbekkmo: - Use mutexes and condvars instead of spl() calls. - Add a couple of bugfixes from FreeBSD (the easily observed ones). - Add a simple watchdog thread for the newer versions of the hardware, that tickles the controller at task submission time if it seems not ready (from FreeBSD). - Add a buffer allocation size hack to the ioctl handler, to work around a firmware bug in some versions of the controller (also from FreeBSD). - Rearrange the bus_dmamap_sync() calls to make the use match the documentation.
1.21	02-Feb-2012	tls	branches: 1.21.6; 1.21.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.20	13-Nov-2010	uebayasi	branches: 1.20.8; 1.20.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.19	12-May-2009	cegger	branches: 1.19.4; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.18	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.17	09-Sep-2008	tron	branches: 1.17.8; Complete device_t/softc split for the ld(4) attachments. This should prevent crashes while attaching a drive. Patch supplied by Juan RP in PR kern/39468.
1.16	28-Apr-2008	martin	branches: 1.16.2; 1.16.6; Remove clause 3 and 4 from TNF licenses
1.15	10-Apr-2008	cegger	branches: 1.15.2; 1.15.4; use aprint_*_dev and device_xname
1.14	19-Oct-2007	ad	branches: 1.14.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	29-Jul-2007	ad	branches: 1.13.4; 1.13.6; 1.13.10; 1.13.12; It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.18; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	23-Jul-2006	bouyer	branches: 1.10.4; 1.10.6; Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives). From FreeBSD, with some adjustements by Andrew Doran and me.
1.9	25-Mar-2006	thorpej	Use device_parent().
1.8	11-Dec-2005	christos	branches: 1.8.4; 1.8.6; 1.8.8; 1.8.10; 1.8.12; merge ktrace-lwp.
1.7	28-Jun-2005	thorpej	branches: 1.7.2; Use ANSI function decls and static.
1.6	28-Oct-2004	yamt	branches: 1.6.10; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.5	04-May-2003	ad	branches: 1.5.2; - Fix the mailbox protocol. Not tested with newer 'Quartz' controllers. - Handle failure during initalisation more gracefully. - Create a watchdog thread to poke the board and check for status changes. - Make ready for > 64kB transfers.
1.4	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	30-Jan-2002	ad	branches: 1.1.2; 1.1.4; Bare-bones driver for AMI RAID. Parts taken from FreeBSD. This was tried a good while ago and it had problems under load. Changes were made to address that, but I don't have the ability to test them. So, I'm committing it before it rots.
1.1.4.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.2	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.1	30-Jan-2002	nathanw	file ld_amr.c was added on branch nathanw_sa on 2002-02-28 04:14:03 +0000
1.1.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.1	30-Jan-2002	jdolecek	file ld_amr.c was added on branch kqueue on 2002-02-11 20:10:01 +0000
1.5.2.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.2.1	02-Nov-2004	skrll	Sync with HEAD.
1.6.10.1	30-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1431): - Add a /dev/amr* control file for amr(4) devices, which allows sending raw commands to the controller. - Add a amrctl(8) control tool, which for now only allows to get status from the adapter (status of adapter, logical volumes and and individual drives).
1.7.2.4	27-Oct-2007	yamt	sync with head.
1.7.2.3	03-Sep-2007	yamt	sync with head.
1.7.2.2	30-Dec-2006	yamt	sync with head.
1.7.2.1	21-Jun-2006	yamt	sync with head.
1.8.12.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.8.10.1	19-Apr-2006	elad	sync with head.
1.8.8.2	11-Aug-2006	yamt	sync with head
1.8.8.1	01-Apr-2006	yamt	sync with head.
1.8.6.1	22-Apr-2006	simonb	Sync with head.
1.8.4.1	09-Sep-2006	rpaulo	sync with head
1.10.6.2	10-Dec-2006	yamt	sync with head.
1.10.6.1	22-Oct-2006	yamt	sync with head
1.10.4.1	18-Nov-2006	ad	Sync with head.
1.12.18.1	15-Aug-2007	skrll	Sync with HEAD.
1.12.8.3	23-Oct-2007	ad	Sync with head.
1.12.8.2	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.12.8.1	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.13.12.2	29-Jul-2007	ad	It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.13.12.1	29-Jul-2007	ad	file ld_amr.c was added on branch matt-mips64 on 2007-07-29 12:50:23 +0000
1.13.10.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.6.1	06-Nov-2007	matt	sync with HEAD
1.13.4.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.14.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.14.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.15.4.3	16-May-2009	yamt	sync with head
1.15.4.2	04-May-2009	yamt	sync with head.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.6.1	19-Oct-2008	haad	Sync with HEAD.
1.16.2.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.17.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.19.4.1	05-Mar-2011	rmind	sync with head
1.20.12.1	18-Feb-2012	mrg	merge to -current.
1.20.8.1	17-Apr-2012	yamt	sync with head
1.21.24.3	05-Oct-2016	skrll	Sync with HEAD
1.21.24.2	06-Jun-2015	skrll	Sync with HEAD
1.21.24.1	06-Apr-2015	skrll	Sync with HEAD
1.21.6.1	03-Dec-2017	jdolecek	update from HEAD
1.23.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.25.54.1	02-Aug-2025	perseant	Sync with HEAD
1.21	13-Apr-2025	rin	ld(4): Convert blkno argument for sc_dump() to daddr_t PR kern/59153 (1) For backends that accept 64-bit block address, i.e., nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4), this should enable to dump beyond 2Gi blocks. (2) sdmmc(4) backend allows to dump up to the last block. (3) For other backends, block address is handled as `int`. Some of them may support blocks up to 4Gi, but I do not have enough time to examine datasheets. So, continue to reject >2Gi blocks as before. XXX This is KABI change, and cannot be pulled up into netbsd-{10,9}. XXX Compile-test only (for amd64/ALL) due to lack of large SSDs ;) Thanks mlelstv@ for discussion and careful review!!
1.20	27-Feb-2017	jdolecek	branches: 1.20.52; refactor the ld(4) DIOCCACHESYNC hook into general ioctl hook, so that attachments would be able to implement arbitrary other ioctls
1.19	27-Sep-2016	pgoyette	branches: 1.19.2; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.18	16-Sep-2016	jdolecek	modify ldattach() to have default strategy as a parameter
1.17	13-Apr-2015	riastradh	branches: 1.17.2; Convert sys/dev to use <sys/rndsource.h>.
1.16	17-Oct-2013	christos	branches: 1.16.6; - remove unused variables - move debugging code inside debugging sections
1.15	02-Feb-2012	tls	branches: 1.15.6; 1.15.10; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.14	13-Nov-2010	uebayasi	branches: 1.14.8; 1.14.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.13	09-Sep-2008	tron	branches: 1.13.16; Complete device_t/softc split for the ld(4) attachments. This should prevent crashes while attaching a drive. Patch supplied by Juan RP in PR kern/39468.
1.12	11-Aug-2008	simonb	Add a flags argument to the ld(4) flush handlers, and call these with a "poll" flag when called from ld_shutdown(). This is the infrastructure part of kern/38655 - in itself it doesn't fix the panic referenced in that PR. XXX: At least ld_twa.c and perhaps ld_iop.c and ld_icp.c need to check for this new poll flag and do something useful.
1.11	07-May-2008	joerg	branches: 1.11.2; 1.11.4; 1.11.6; First part of twa(4) improvements: - more/better diagnostic messages - dynamically allocate units in preparation for 9650+ support - better distribute queue slots over disks according to size - drain large request queue on 9550 - keep track of resets, will be needed for 9650. This changes have been contributed by Wasabi Systems.
1.10	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.9	19-Oct-2007	ad	branches: 1.9.16; 1.9.18; 1.9.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.8	07-Aug-2007	simonb	branches: 1.8.2; 1.8.6; Unwrap not-too-long line.
1.7	29-Jul-2007	ad	branches: 1.7.4; 1.7.6; It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.6	21-Jul-2007	ad	Don't depend on uvm_extern.h pulling in proc.h.
1.5	08-Feb-2007	ad	branches: 1.5.6; 1.5.14; Add a quick hack to allow addressing over 2G sectors. XXX This driver needs to be rewritten to use the scsipi layer. Additionally: - Missing copyright years: 2000, 2001, 2002, 2003 - Missing attribution: Andrew Doran and Jason R. Thorpe
1.4	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.3	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.2	25-May-2006	wrstuden	branches: 1.2.2; 1.2.6; 1.2.8; 1.2.10; 1.2.12; 1.2.16; 1.2.18; 1.2.20; Add $NetBSD: ids, which work better than $netbsd: ones.
1.1	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.2.20.2	10-Dec-2006	yamt	sync with head.
1.2.20.1	22-Oct-2006	yamt	sync with head
1.2.18.2	09-Sep-2006	rpaulo	sync with head
1.2.18.1	25-May-2006	rpaulo	file ld_twa.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:18 +0000
1.2.16.2	09-Feb-2007	ad	Sync with HEAD.
1.2.16.1	18-Nov-2006	ad	Sync with head.
1.2.12.2	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.2.12.1	25-May-2006	tron	file ld_twa.c was added on branch netbsd-3 on 2006-07-31 12:35:22 +0000
1.2.10.2	26-Jun-2006	yamt	sync with head.
1.2.10.1	25-May-2006	yamt	file ld_twa.c was added on branch yamt-pdpolicy on 2006-06-26 12:51:22 +0000
1.2.8.6	27-Oct-2007	yamt	sync with head.
1.2.8.5	03-Sep-2007	yamt	sync with head.
1.2.8.4	26-Feb-2007	yamt	sync with head.
1.2.8.3	30-Dec-2006	yamt	sync with head.
1.2.8.2	21-Jun-2006	yamt	sync with head.
1.2.8.1	25-May-2006	yamt	file ld_twa.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:05 +0000
1.2.6.2	19-Jun-2006	chap	Sync with head.
1.2.6.1	25-May-2006	chap	file ld_twa.c was added on branch chap-midi on 2006-06-19 04:01:36 +0000
1.2.2.2	01-Jun-2006	kardel	Sync with head.
1.2.2.1	25-May-2006	kardel	file ld_twa.c was added on branch simonb-timecounters on 2006-06-01 22:36:46 +0000
1.5.14.1	15-Aug-2007	skrll	Sync with HEAD.
1.5.6.5	23-Oct-2007	ad	Sync with head.
1.5.6.4	09-Oct-2007	ad	Sync with head.
1.5.6.3	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.5.6.2	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.5.6.1	05-Apr-2007	ad	Compile fixes.
1.7.6.2	29-Jul-2007	ad	It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.7.6.1	29-Jul-2007	ad	file ld_twa.c was added on branch matt-mips64 on 2007-07-29 12:50:23 +0000
1.7.4.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.7.4.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.8.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.8.2.1	06-Nov-2007	matt	sync with HEAD
1.9.20.2	04-May-2009	yamt	sync with head.
1.9.20.1	16-May-2008	yamt	sync with head.
1.9.18.1	18-May-2008	yamt	sync with head.
1.9.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.9.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.11.6.1	19-Oct-2008	haad	Sync with HEAD.
1.11.4.2	11-Jun-2008	simonb	Revert my suggested fix for kern/38655 - that shouldn't have snuck in to this branch.
1.11.4.1	10-Jun-2008	simonb	Initial commit of Wasabi System's WAPBL (Write Ahead Physical Block Logging) journaling code. Originally written by Darrin B. Jewell while at Wasabi and updated to -current by Antti Kantee, Andy Doran, Greg Oster and Simon Burge. Still a number of issues - look in doc/BRANCHES for "simonb-wapbl" for more info.
1.11.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.11.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.13.16.1	05-Mar-2011	rmind	sync with head
1.14.12.1	18-Feb-2012	mrg	merge to -current.
1.14.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.14.8.1	17-Apr-2012	yamt	sync with head
1.15.10.1	18-May-2014	rmind	sync with head
1.15.6.2	03-Dec-2017	jdolecek	update from HEAD
1.15.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.16.6.3	28-Aug-2017	skrll	Sync with HEAD
1.16.6.2	05-Oct-2016	skrll	Sync with HEAD
1.16.6.1	06-Jun-2015	skrll	Sync with HEAD
1.17.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.17.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.19.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.20.52.1	02-Aug-2025	perseant	Sync with HEAD
1.41	13-Apr-2025	rin	ld(4): Convert blkno argument for sc_dump() to daddr_t PR kern/59153 (1) For backends that accept 64-bit block address, i.e., nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4), this should enable to dump beyond 2Gi blocks. (2) sdmmc(4) backend allows to dump up to the last block. (3) For other backends, block address is handled as `int`. Some of them may support blocks up to 4Gi, but I do not have enough time to examine datasheets. So, continue to reject >2Gi blocks as before. XXX This is KABI change, and cannot be pulled up into netbsd-{10,9}. XXX Compile-test only (for amd64/ALL) due to lack of large SSDs ;) Thanks mlelstv@ for discussion and careful review!!
1.40	27-Feb-2017	jdolecek	branches: 1.40.52; refactor the ld(4) DIOCCACHESYNC hook into general ioctl hook, so that attachments would be able to implement arbitrary other ioctls
1.39	27-Sep-2016	pgoyette	branches: 1.39.2; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.38	16-Sep-2016	jdolecek	modify ldattach() to have default strategy as a parameter
1.37	13-Apr-2015	riastradh	branches: 1.37.2; Convert sys/dev to use <sys/rndsource.h>.
1.36	02-Feb-2012	tls	branches: 1.36.6; 1.36.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.35	13-Nov-2010	uebayasi	branches: 1.35.8; 1.35.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.34	12-May-2009	cegger	branches: 1.34.4; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.33	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.32	09-Sep-2008	tron	branches: 1.32.8; Complete device_t/softc split for the ld(4) attachments. This should prevent crashes while attaching a drive. Patch supplied by Juan RP in PR kern/39468.
1.31	11-Aug-2008	simonb	Add polling support to ld_twe_flush(). Fixes the "cpu_switcho: switching above IPL_SCHED" panic for ld@twe disks on shutdown in kern/38655.
1.30	28-Apr-2008	martin	branches: 1.30.2; 1.30.4; 1.30.6; Remove clause 3 and 4 from TNF licenses
1.29	19-Oct-2007	ad	branches: 1.29.16; 1.29.18; 1.29.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.28	29-Jul-2007	ad	branches: 1.28.4; 1.28.6; 1.28.10; 1.28.12; It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.27	21-Jul-2007	ad	Don't depend on uvm_extern.h pulling in proc.h.
1.26	16-Nov-2006	christos	branches: 1.26.8; 1.26.18; __unused removal on arguments; approved by core.
1.25	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.24	25-Mar-2006	thorpej	branches: 1.24.8; 1.24.10; Use device_parent().
1.23	11-Dec-2005	christos	branches: 1.23.4; 1.23.6; 1.23.8; 1.23.10; 1.23.12; merge ktrace-lwp.
1.22	08-Feb-2005	briggs	branches: 1.22.6; Use aprint_*() instead of printf() at attach time.
1.21	28-Oct-2004	yamt	branches: 1.21.4; 1.21.6; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.20	27-May-2004	thorpej	Add a "flush" operation so the array's write cache is flushed on close and at shutdown.
1.19	22-Apr-2004	itojun	sprintf -> snprintf
1.18	26-Sep-2003	thorpej	branches: 1.18.2; Confirmed; don't need to query stripe size on TwinStor.
1.17	23-Sep-2003	thorpej	- Make CCB allocation slightly more efficient by changing how the CCB is returned to the caller. - Make code paths that can use twe_ccb_alloc_wait() use it, and assert that a CCB is always returned from that function. - Assert that a CCB is always returned when allocating the reserved CCB for an AEN fetch.
1.16	22-Sep-2003	thorpej	Add support for dynamically attaching and detaching RAID array units.
1.15	21-Sep-2003	thorpej	Report the status of the logical drive (normal, degraded, etc.) at attach time.
1.14	21-Sep-2003	thorpej	- Record more information about the array unit, including array type and stripe depth. - Report array type and stripe depth when attaching the logical drive.
1.13	02-Oct-2002	thorpej	branches: 1.13.6; Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	24-May-2002	christos	NBPG is not constant on the sparc, so don't use CPP tricks.
1.9	18-May-2002	ad	branches: 1.9.2; - Use write barriers. - NBPG -> PAGE_SIZE.
1.8	13-Nov-2001	lukem	branches: 1.8.8; add RCSID
1.7	10-Jun-2001	ad	branches: 1.7.2; 1.7.8; Centralise the geometry fudge.
1.6	26-Jan-2001	ad	branches: 1.6.2; No need to submit before poll any more.
1.5	22-Jan-2001	ad	Nuke useless argument to ld_twe_dobio().
1.4	22-Jan-2001	ad	Minor optimisation.
1.3	22-Jan-2001	ad	opt_twe.h isn't needed any more.
1.2	22-Jan-2001	ad	Don't artificially clamp the maximum number of outstanding commands.
1.1	26-Nov-2000	ad	branches: 1.1.2; lsu -> ld, by popular request.
1.1.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.1.2.2	08-Dec-2000	bouyer	Sync with HEAD.
1.1.2.1	26-Nov-2000	bouyer	file ld_twe.c was added on branch thorpej_scsipi on 2000-12-08 09:12:32 +0000
1.6.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.6.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.6.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.6.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.7.8.2	25-Oct-2001	he	Pull up revisions 1.1-1.7 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.7.8.1	10-Jun-2001	he	file ld_twe.c was added on branch netbsd-1-5 on 2001-10-25 17:54:07 +0000
1.7.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.7.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.8.1	30-May-2002	gehenna	Catch up with -current.
1.9.2.1	26-May-2002	perry	partially resolve 1.6 pullup ticket #29 from grant pulls up revision 1.10 Original commit message (by christos): > NBPG is not constant on the sparc, so don't use CPP tricks.
1.13.6.5	09-Feb-2005	skrll	Sync with HEAD.
1.13.6.4	02-Nov-2004	skrll	Sync with HEAD.
1.13.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.6.1	03-Aug-2004	skrll	Sync with HEAD
1.18.2.1	30-May-2004	tron	Pull up revision 1.20 (requested by thorpej in ticket #411): Add a "flush" operation so the array's write cache is flushed on close and at shutdown.
1.21.6.1	12-Feb-2005	yamt	sync with head.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.22.6.4	27-Oct-2007	yamt	sync with head.
1.22.6.3	03-Sep-2007	yamt	sync with head.
1.22.6.2	30-Dec-2006	yamt	sync with head.
1.22.6.1	21-Jun-2006	yamt	sync with head.
1.23.12.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.23.10.1	19-Apr-2006	elad	sync with head.
1.23.8.1	01-Apr-2006	yamt	sync with head.
1.23.6.1	22-Apr-2006	simonb	Sync with head.
1.23.4.1	09-Sep-2006	rpaulo	sync with head
1.24.10.2	10-Dec-2006	yamt	sync with head.
1.24.10.1	22-Oct-2006	yamt	sync with head
1.24.8.1	18-Nov-2006	ad	Sync with head.
1.26.18.1	15-Aug-2007	skrll	Sync with HEAD.
1.26.8.4	23-Oct-2007	ad	Sync with head.
1.26.8.3	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.26.8.2	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.26.8.1	05-Apr-2007	ad	Compile fixes.
1.28.12.2	29-Jul-2007	ad	It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.28.12.1	29-Jul-2007	ad	file ld_twe.c was added on branch matt-mips64 on 2007-07-29 12:50:23 +0000
1.28.10.1	25-Oct-2007	bouyer	Sync with HEAD.
1.28.6.1	06-Nov-2007	matt	sync with HEAD
1.28.4.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.29.20.3	16-May-2009	yamt	sync with head
1.29.20.2	04-May-2009	yamt	sync with head.
1.29.20.1	16-May-2008	yamt	sync with head.
1.29.18.1	18-May-2008	yamt	sync with head.
1.29.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.29.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.30.6.1	19-Oct-2008	haad	Sync with HEAD.
1.30.4.2	11-Jun-2008	simonb	Revert my suggested fix for kern/38655 - that shouldn't have snuck in to this branch.
1.30.4.1	10-Jun-2008	simonb	Initial commit of Wasabi System's WAPBL (Write Ahead Physical Block Logging) journaling code. Originally written by Darrin B. Jewell while at Wasabi and updated to -current by Antti Kantee, Andy Doran, Greg Oster and Simon Burge. Still a number of issues - look in doc/BRANCHES for "simonb-wapbl" for more info.
1.30.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.30.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.32.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.34.4.1	05-Mar-2011	rmind	sync with head
1.35.12.1	18-Feb-2012	mrg	merge to -current.
1.35.8.1	17-Apr-2012	yamt	sync with head
1.36.24.3	28-Aug-2017	skrll	Sync with HEAD
1.36.24.2	05-Oct-2016	skrll	Sync with HEAD
1.36.24.1	06-Jun-2015	skrll	Sync with HEAD
1.36.6.1	03-Dec-2017	jdolecek	update from HEAD
1.37.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.37.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.39.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.40.52.1	02-Aug-2025	perseant	Sync with HEAD
1.46	21-Oct-2025	pgoyette	When building the module we actually need the ioconf.c
1.45	27-Aug-2025	isaki	Fix string length in T_GET_ID response.
1.44	05-Jul-2025	mlelstv	Ignore empty id_data and keep default string.
1.43	13-Apr-2025	rin	ld(4): Convert blkno argument for sc_dump() to daddr_t PR kern/59153 (1) For backends that accept 64-bit block address, i.e., nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4), this should enable to dump beyond 2Gi blocks. (2) sdmmc(4) backend allows to dump up to the last block. (3) For other backends, block address is handled as `int`. Some of them may support blocks up to 4Gi, but I do not have enough time to examine datasheets. So, continue to reject >2Gi blocks as before. XXX This is KABI change, and cannot be pulled up into netbsd-{10,9}. XXX Compile-test only (for amd64/ALL) due to lack of large SSDs ;) Thanks mlelstv@ for discussion and careful review!!
1.42	27-Feb-2025	jakllsch	Implement VIRTIO_BLK_F_TOPOLOGY support for physical sector size information
1.41	23-Feb-2025	mlelstv	Need to allocate DMA buffer.
1.40	22-Feb-2025	mlelstv	For fetching id data: - actually enter synchronous mode. - add and use polling support, which is necessary during attach. - exit synchronous mode on errors.
1.39	22-Feb-2025	mlelstv	Query device for id string.
1.38	22-Feb-2025	mlelstv	Remove module attachment hack, use more of standard boilerplate for modules.
1.37	20-Feb-2025	jakllsch	Remove trailing whitespace
1.36	06-Feb-2025	jakllsch	Add discard support to ld@virtio.
1.35	12-Jun-2024	riastradh	branches: 1.35.2; ld@virtio(4): Fix maximum size parameters. - SEG_MAX is the maximum number of segments. - SIZE_MAX is the maximum number of bytes in a single segment. The maximum transfer size is, therefore, SEG_MAX * SIZE_MAX. => Don't add two extra segments in the dmamap vr_payload for the header and status -- we already have a separate dmamap vr_cmdsts for that. => Don't recalculate payload dmamap parameters based on division by NBPG, just use the ones specified by the host. => Allow SIZE_MAX below MAXPHYS as long as SIZE_MAX*SEG_MAX >= MAXPHYS. Even though ldattach clamps ld->sc_maxxfer to MAXPHYS, make sure to clamp it in ld_virtio_attach before ld_virtio_alloc_reqs since it determines the dmamap sizes and bounce buffer allocation and there's no sense in allocating those larger than ld will use anyway. PR kern/58338
1.34	09-Mar-2024	isaki	Modify a confused expression in ld_virtio_attach(). VIRTIO_BLK_MIN_SEGMENTS should be the total number of non-data segments, so I rename it to VIRTIO_BLK_CTRL_SEGMENTS. PR kern/57981.
1.33	12-Feb-2024	isaki	Fix typo in error message.
1.32	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.31	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.30	13-Apr-2022	uwe	branches: 1.30.4; virtio: use the new syntax for snprintb(3) format strings. The old syntax is limited to 32 bits only (and has 1-based bit numbers which is rather incovenient too).
1.29	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.28	24-Oct-2020	skrll	branches: 1.28.2; Use BUS_DMA_WAITOK in ld_virtio_alloc_reqs which is called at attached time...
1.27	17-Jan-2019	hannken	Use cv_broadcast() on "sc_sync_wait" to prevent deadlock when more than one thread try to "DIOCCACHESYNC" at the same time.
1.26	23-Nov-2018	jmcneill	Use MSIX when available
1.25	12-Jul-2018	jakllsch	revert previous, unintended, commit
1.24	12-Jul-2018	jakllsch	<sys/pmc.h> is gone; dont
1.23	21-Jun-2018	jakllsch	branches: 1.23.2; destroy 'sc_sync_wait' condvar and mutex upon detach
1.22	10-Jun-2018	jakllsch	remove irrelevant pci(9) #includes from virtio child drivers
1.21	10-Jun-2018	jakllsch	The virtio block device capacity config item is expressed always in request protocol sector units of 512 bytes. Also, add and use a symbolic constant to refer to request protocol sector units.
1.20	07-Jun-2018	jakllsch	Make ld_virtio aware of a possible device-side write cache.
1.19	03-Jun-2018	jakllsch	unload payload dma map upon command completion
1.18	03-Jun-2018	jakllsch	add feature/register definitions from virtio-v1.0-cs04
1.17	03-Jun-2018	jakllsch	remove trailing whitespace
1.16	09-Aug-2017	mlelstv	branches: 1.16.2; While ld(4) is MP safe, many backends are not. Add a flag for backends that are MP safe. Take KERNEL_LOCK when calling into a backend that doesn't have the flag set. Do the same for the discard routine. Fixes PR 52462.
1.15	25-Mar-2017	jdolecek	branches: 1.15.6; reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.14	27-Feb-2017	jdolecek	refactor the ld(4) DIOCCACHESYNC hook into general ioctl hook, so that attachments would be able to implement arbitrary other ioctls
1.13	30-Nov-2016	christos	branches: 1.13.2; Don't call virtio_enqueue_abort when virtio_enqueue_reserve fails. Pointed out by uwe@
1.12	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.11	16-Sep-2016	jdolecek	modify ldattach() to have default strategy as a parameter
1.10	07-Jul-2016	msaitoh	branches: 1.10.2; KNF. Remove extra spaces. No functional change.
1.9	27-Oct-2015	christos	- Print the negotiated feature bits - Define a constant for the minimum number of segments and use that. - Remove unused sc_lock - Use single variables instead of arrays of 1 element - To the transfer size and segment size negotiation more correctly - Make vr_bp NULL when unused. - Add missing abort calls
1.8	05-May-2015	ozaki-r	Use NULL for initialization of sc_config_change
1.7	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.6	22-Jul-2014	ozaki-r	branches: 1.6.4; Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio It is set by a child driver, e.g., if_vioif. If set, virtio sets PCI_INTR_MPSAFE for pci_intr_establish.
1.5	02-Feb-2012	tls	branches: 1.5.6; 1.5.20; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.4	03-Dec-2011	hannken	branches: 1.4.2; 1.4.4; Don't try to set ld->sc_maxxfer below MAXPHYS. At least genfs_io assumes 'ld->sc_maxxfer == MAXPHYS'.
1.3	23-Nov-2011	hannken	According to "Virtio PCI Card Specification v0.9.2 DRAFT" there is no feature named VIRTIO_BLK_F_SECTOR_MAX so remove it. Linux seems to use this feature bit as VIRTIO_BLK_F_TOPOLOGY. Use VIRTIO_BLK_F_BLK_SIZE * VIRTIO_BLK_F_SEG_MAX as the drivers maxxfer and reorder so sc_secsize gets set before use. As maxxfer may not be a multiple of page size add one more segment to the dma maps. Tested on Linux 3.1.1 host by Guillaume Lasmayous.
1.2	02-Nov-2011	hannken	branches: 1.2.2; Make sure we allocate enough segments for the request dma maps.
1.1	30-Oct-2011	hannken	Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.2.2.1	17-Apr-2012	yamt	sync with head
1.4.4.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.4.4.1	03-Dec-2011	riz	file ld_virtio.c was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.4.2.1	18-Feb-2012	mrg	merge to -current.
1.5.20.1	10-Aug-2014	tls	Rebase.
1.5.6.2	03-Dec-2017	jdolecek	update from HEAD
1.5.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.4.6	28-Aug-2017	skrll	Sync with HEAD
1.6.4.5	05-Dec-2016	skrll	Sync with HEAD
1.6.4.4	05-Oct-2016	skrll	Sync with HEAD
1.6.4.3	09-Jul-2016	skrll	Sync with HEAD
1.6.4.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.6.4.1	06-Jun-2015	skrll	Sync with HEAD
1.10.2.4	26-Apr-2017	pgoyette	Sync with HEAD
1.10.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.10.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.10.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.13.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.15.6.2	22-Jun-2018	martin	Pull up following revision(s) (requested by jakllsch in ticket #888): sys/dev/pci/ld_virtio.c: revision 1.19-1.21,1.23 remove trailing whitespace add feature/register definitions from virtio-v1.0-cs04 unload payload dma map upon command completion Make ld_virtio aware of a possible device-side write cache. The virtio block device capacity config item is expressed always in request protocol sector units of 512 bytes. Also, add and use a symbolic constant to refer to request protocol sector units. destroy 'sc_sync_wait' condvar and mutex upon detach
1.15.6.1	01-Sep-2017	martin	Pull up following revision(s) (requested by mlelstv in ticket #261): sys/dev/sdmmc/ld_sdmmc.c: revision 1.32 sys/dev/sdmmc/ld_sdmmc.c: revision 1.33 sys/dev/sdmmc/ld_sdmmc.c: revision 1.34 sys/dev/sdmmc/sdmmc_mem.c: revision 1.62 sys/dev/i2o/ld_iop.c: revision 1.39 sys/dev/ld.c: revision 1.102 sys/dev/ld.c: revision 1.103 sys/dev/dksubr.c: revision 1.98 sys/dev/dksubr.c: revision 1.99 sys/dev/sdmmc/sdmmcvar.h: revision 1.29 sys/dev/ic/ld_nvme.c: revision 1.17 sys/dev/ldvar.h: revision 1.31 sys/dev/ldvar.h: revision 1.32 sys/dev/ic/ld_cac.c: revision 1.31 sys/dev/pci/ld_virtio.c: revision 1.16 While ld(4) is MP safe, many backends are not. Add a flag for backends that are MP safe. Take KERNEL_LOCK when calling into a backend that doesn't have the flag set. Do the same for the discard routine. Fixes PR 52462. Defer sdmmc discard operations to the sdmmc task queue. Fixes a panic introduced by ld.c r1.102. validate length for discard operation and split operation when byte length doesn't fit into 'int'. make the sc_discard interface for the ld backend asynchronous and signal completion through new callback lddiscardend. Use a standard struct buf to pass disk address and range instead of two off_t values. make lddiscard synchronous again. This is a requirement of the current ffs discard code. Initialize error also in the case where len=0, which just succeeds. while here, assert that the len is indeed non-negative. this is already confirmed by sys_fdiscard, but let's be sure. reported by: GCC, but with different compile flags
1.16.2.4	18-Jan-2019	pgoyette	Synch with HEAD
1.16.2.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.16.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.16.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.23.2.1	10-Jun-2019	christos	Sync with HEAD
1.28.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.30.4.2	02-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #914): sys/dev/pci/virtio_pci.c: revision 1.50 sys/dev/pci/virtio_pci.c: revision 1.51 sys/dev/pci/virtio_pci.c: revision 1.52 sys/dev/pci/virtio_pci.c: revision 1.53 sys/dev/pci/virtio_pci.c: revision 1.54 sys/dev/pci/vioscsi.c: revision 1.35 sys/dev/pci/ld_virtio.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.34 sys/dev/virtio/viocon.c: revision 1.9 sys/dev/pci/ld_virtio.c: revision 1.35 sys/dev/pci/virtio.c: revision 1.80 sys/dev/pci/virtio.c: revision 1.81 sys/dev/pci/virtio.c: revision 1.82 sys/dev/pci/virtio_pci.c: revision 1.45 sys/dev/pci/virtio_pci.c: revision 1.46 sys/dev/pci/virtio_pci.c: revision 1.47 sys/dev/pci/virtio_pci.c: revision 1.48 sys/dev/pci/virtio_pci.c: revision 1.49 sys/dev/pci/virtio.c: revision 1.79 A single SCSI request may require multiple slots in the virtio queue but the queue isn't sized for the theoretical maximum. So just use XS_BUSY to pace the scsipi layer. viocon(4): Show error in assertion failure. virtio_read_device_config_le_[24]: Fix for {aarch64,arm}eb Stop byte-swapping for big-endian aarch64 and arm to fix corrupted read for, e.g., sc_taglen for vio9p(4). As described as comments in virtio_pci.c, big-endian aarch64 and armv7 (armeb) are somewhat special. At the moment, all supported virtual/real machines are configured as little-endian, and only CPU cores are switched to big-endian mode during early boot stage. Most peripheral buses for armeb are LSB as a result, and their default bus_space(9) functions swap byte-order. Therefore, PIOed data from memory-mapped devices, as well as pci(4) ones, are actually LSB, but observed as MSB for armeb. Therefore, we should not swap byte-order further in virtio_read_device_config_le_[24] on armeb. See virtio_pci.c and sys/dev/fdt/virtio_mmio_fdt.c also. fix spelling mistakes, mainly in comments and log messages. Split KASSERT(A && B) into KASSERT(A); KASSERT(B). Fix typo in error message. Modify a confused expression in ld_virtio_attach(). VIRTIO_BLK_MIN_SEGMENTS should be the total number of non-data segments, so I rename it to VIRTIO_BLK_CTRL_SEGMENTS. PR kern/57981. ld@virtio(4): Fix maximum size parameters. - SEG_MAX is the maximum number of segments. - SIZE_MAX is the maximum number of bytes in a single segment. The maximum transfer size is, therefore, SEG_MAX * SIZE_MAX. => Don't add two extra segments in the dmamap vr_payload for the header and status -- we already have a separate dmamap vr_cmdsts for that. => Don't recalculate payload dmamap parameters based on division by NBPG, just use the ones specified by the host. => Allow SIZE_MAX below MAXPHYS as long as SIZE_MAX*SEG_MAX >= MAXPHYS. Even though ldattach clamps ld->sc_maxxfer to MAXPHYS, make sure to clamp it in ld_virtio_attach before ld_virtio_alloc_reqs since it determines the dmamap sizes and bounce buffer allocation and there's no sense in allocating those larger than ld will use anyway. PR kern/58338 virtio_pci.c: Nix trailing whitespace. virtio@pci: Unmap the BARs we actually mapped on >=1.0 detach. PR kern/58335 virtio@pci: Nix useless sc_mapped_iosize. This appears to have been a mistaken attempt to pass the subregion length through to bus_space_unmap, which doesn't make any sense -- bus_space_subregion does not allocate anything and the resulting handle must not be passed to bus_space_unmap. Once we pass the size that was actually returned by pci_mapreg_map to bus_space_unmap, which matches the size passed internally by pci_mapreg_map to bus_space_map, sc_mapped_iosize is no longer used and can be flushed. Prompted by PR kern/58335. virtio@pci: Don't detach on shutdown. The virtio bus does not keep state in memory that needs to flushed to persistent storage. Typically DVF_DETACH_SHUTDOWN is only needed by, e.g., disk drivers to flush disk caches (although really, vfs sync should do that anyway, so maybe we should just nix the option altogether). Prompted by PR kern/58335. virtio@pci: Zero sc_bars_iosize[i] on unmap. No functional change intended, but this makes the code less brittle. Prompted by PR kern/58335. virtio_pci.c: Sort includes. No functional change intended. virtio_pci.c: Omit needless blank lines and commented code. No functional change intended. virtio_pci.c: Sprinkle KNF. - No parameter names in function prototypes. - Fix indentation. - Break overlong lines. - Blank line between declarations and rest of block. virtio_pci.c: Sprinkle const. No functional change intended. virtio_pci.c: Use container_of, not cast. virtio(4): Assert child attach finished before we allow queueing. Should help catch mistakes like: PR kern/58555: Kernel panic during boot when using viocon with PCIe
1.30.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.35.2.1	02-Aug-2025	perseant	Sync with HEAD
1.19	30-Nov-2018	jmcneill	Add support for polling com devices when no interrupt is available.
1.18	27-Apr-2017	msaitoh	branches: 1.18.8; 1.18.10; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.17	29-Mar-2014	christos	branches: 1.17.6; 1.17.20; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.16	31-Jul-2013	soren	On x86, print the I/O address of lpt_puc as for com_puc so that hacks using iopl/ioperm can be told where to find the parallel port. Closes PR kern/43683.
1.15	22-Jul-2013	soren	Since puc(4) serial ports are typically not identified in the BIOS COM[1234] table, the I/O address must be manually set using installboot(8) in order to enable a serial console. Print the address here so the user doesn't have to dig through PCI configuration space to find it. Enable suspend/resume hooks.
1.14	07-Mar-2008	cube	branches: 1.14.38; 1.14.48; 1.14.54; 1.14.62; Split the softc from the device_t for all lpt(4) variants and attachments except the ppbus stuff (which doesn't compile) and ulpt(4) which is unrelated and can be dealt with separately. As usual, it comes with related cosmetic changes.
1.13	19-Oct-2007	ad	branches: 1.13.12; 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	27-Feb-2005	perry	nuke trailing whitespace
1.7	04-Feb-2005	perry	de-__P
1.6	02-Oct-2002	thorpej	branches: 1.6.6; 1.6.14; 1.6.16; Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.2	13-Nov-2001	lukem	add RCSID
1.1	26-Jun-1998	cgd	branches: 1.1.26; 1.1.28; add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.1.28.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.28.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.26.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.26.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.26.1	14-Nov-2001	nathanw	Catch up to -current.
1.6.16.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.6.16.1	12-Feb-2005	yamt	sync with head.
1.6.14.1	29-Apr-2005	kent	sync with -current
1.6.6.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.6.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.6.1	04-Feb-2005	skrll	Sync with HEAD.
1.9.2.3	17-Mar-2008	yamt	sync with head.
1.9.2.2	27-Oct-2007	yamt	sync with head.
1.9.2.1	30-Dec-2006	yamt	sync with head.
1.10.22.2	10-Dec-2006	yamt	sync with head.
1.10.22.1	22-Oct-2006	yamt	sync with head
1.10.20.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.2	23-Mar-2008	matt	sync with HEAD
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.12.1	24-Mar-2008	keiichi	sync with head.
1.14.62.1	23-Jul-2013	riastradh	sync with HEAD
1.14.54.2	18-May-2014	rmind	sync with head
1.14.54.1	28-Aug-2013	rmind	sync with head
1.14.48.2	03-Dec-2017	jdolecek	update from HEAD
1.14.48.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.14.38.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.17.20.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.17.6.1	28-Aug-2017	skrll	Sync with HEAD
1.18.10.1	10-Jun-2019	christos	Sync with HEAD
1.18.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.3	26-Nov-2000	ad	lsu -> ld, by popular request.
1.2	14-Nov-2000	itojun	branches: 1.2.2; need uvm/uvm_extern.h for PAGE_SIZE
1.1	08-Nov-2000	ad	Move lsu stuff to more logical locations, and push queueing into the lsu driver.
1.2.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.2.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.2.2.1	14-Nov-2000	bouyer	file lsu_twe.c was added on branch thorpej_scsipi on 2000-11-22 16:04:09 +0000
1.7	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.6	24-Apr-2021	thorpej	branches: 1.6.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.5	10-Jul-2016	jmcneill	branches: 1.5.34; fix inverted screen blanking logic
1.4	14-Mar-2012	nonaka	branches: 1.4.2; 1.4.4; 1.4.16; 1.4.18; lynxfb too. > Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with > something meaningful. All relevant documentation has been updated or > written.
1.3	11-Mar-2012	nonaka	fix mmaped offset. X works on Yeeloong Notebook now.
1.2	11-Mar-2012	nonaka	path correct device to wsdisplayio_busid_pci().
1.1	02-Mar-2012	nonaka	branches: 1.1.2; Added Lemote Yeeloong Notebook support.
1.1.2.5	05-Apr-2012	mrg	sync to latest -current.
1.1.2.4	06-Mar-2012	mrg	sync to -current
1.1.2.3	06-Mar-2012	mrg	sync to -current
1.1.2.2	04-Mar-2012	mrg	sync to latest -current.
1.1.2.1	02-Mar-2012	mrg	file lynxfb.c was added on branch jmcneill-usbmp on 2012-03-04 00:46:20 +0000
1.4.18.1	05-Oct-2016	skrll	Sync with HEAD
1.4.16.1	11-Jul-2016	martin	Pull up following revision(s) (requested by jmcneill in ticket #1203): sys/dev/pci/lynxfb.c: revision 1.5 fix inverted screen blanking logic
1.4.4.1	03-Dec-2017	jdolecek	update from HEAD
1.4.2.2	17-Apr-2012	yamt	sync with head
1.4.2.1	14-Mar-2012	yamt	file lynxfb.c was added on branch yamt-pagecache on 2012-04-17 00:07:51 +0000
1.5.34.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.6.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	11-Mar-2012	nonaka	branches: 1.2.2; fix mmaped offset. X works on Yeeloong Notebook now.
1.1	02-Mar-2012	nonaka	branches: 1.1.2; Added Lemote Yeeloong Notebook support.
1.1.2.5	05-Apr-2012	mrg	sync to latest -current.
1.1.2.4	06-Mar-2012	mrg	sync to -current
1.1.2.3	06-Mar-2012	mrg	sync to -current
1.1.2.2	04-Mar-2012	mrg	sync to latest -current.
1.1.2.1	02-Mar-2012	mrg	file lynxfbreg.h was added on branch jmcneill-usbmp on 2012-03-04 00:46:20 +0000
1.2.2.2	17-Apr-2012	yamt	sync with head
1.2.2.1	11-Mar-2012	yamt	file lynxfbreg.h was added on branch yamt-pagecache on 2012-04-17 00:07:51 +0000
1.1	02-Mar-2012	nonaka	branches: 1.1.2; 1.1.4; Added Lemote Yeeloong Notebook support.
1.1.4.2	17-Apr-2012	yamt	sync with head
1.1.4.1	02-Mar-2012	yamt	file lynxfbvar.h was added on branch yamt-pagecache on 2012-04-17 00:07:51 +0000
1.1.2.4	06-Mar-2012	mrg	sync to -current
1.1.2.3	06-Mar-2012	mrg	sync to -current
1.1.2.2	04-Mar-2012	mrg	sync to latest -current.
1.1.2.1	02-Mar-2012	mrg	file lynxfbvar.h was added on branch jmcneill-usbmp on 2012-03-04 00:46:20 +0000
1.107	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.106	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.105	21-Jun-2021	christos	branches: 1.105.2; fix proplib deprecation
1.104	24-Apr-2021	thorpej	branches: 1.104.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.103	10-Oct-2020	jdc	branches: 1.103.4; Don't change a videomode that's been setup in the firmware. Extend the register debugging output (and hopefully make it easier to read).
1.102	07-Aug-2020	macallan	the drawing engine needs a stride that's a multiple of 8 in pixels, so let's actually distinguish between width and stride, and always round the stride up to the next multiple of 8 now modes with odd widths work
1.101	07-Aug-2020	jdc	Also set the bits for horizontal sync start delay. This makes machfb work correctly for some modes (e.g. 1152x720). Modify mach64_get_mode() to get the mode from the chip registers (only used with debug on). Whilst here tidy up some other debug output.
1.100	30-Jul-2020	macallan	if we find composity sync enabled, leave it on instead of unconditionally enabling it on sparc*
1.99	05-Jul-2020	martin	Adapt to proplib interface changes
1.98	07-Dec-2019	macallan	clean up the video mode selection logic, switch modes only when actually necessary while there make some debug output optional
1.97	05-Feb-2019	mrg	branches: 1.97.4; add or adjust fallthru comments.
1.96	03-Feb-2019	tnn	machfb: guard against absent mode information from firmware
1.95	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.94	02-Jun-2017	macallan	branches: 1.94.8; 1.94.10; enable font loading and screen resizing, now that the glyph cache can be reconfigured whenever needed
1.93	28-May-2017	macallan	g/c some unused / #if 0-ed goop, support firmware-provided EDID, get rid of private list of video modes tested on macppc, should Just Work(tm) on sparc64
1.92	11-Jul-2016	msaitoh	KNF. No functional change.
1.91	18-Dec-2013	macallan	branches: 1.91.6; get rid of /dev/fb support which has never been used for anything other than some highly experimental code in xf86/Xsun
1.90	06-Nov-2013	macallan	Don't turn off aperture registers - some firmwares don't know how to turn them back on. Also distinguish between WSDISPLAYIO_MODE_MAPPED and _DUMBFB - we only need to turn the registers off in _DUMBFB mode when mapping all 8MB of video memory.
1.89	09-Oct-2013	macallan	support WSDISPLAYIO_GET_FBINFO
1.88	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.87	28-May-2013	macallan	branches: 1.87.2; support RI_FULLCLEAR
1.86	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.85	05-Oct-2012	macallan	gc_attr is no more
1.84	01-Oct-2012	jdc	Match the Rage Mobility M1 used in the Tadpole SPARCle.
1.83	16-Aug-2012	macallan	branches: 1.83.2; fix a bunch of typos, thinkos and cargocultos
1.82	15-Aug-2012	macallan	don't forget to clear the screen and wipe the glyph cache when re-entering terminal emulation mode
1.81	15-Aug-2012	macallan	some cleanup: - get rid of some redundant variables ( sc_phys and sc_base were essentially the same ) - we don't access video memory directly anymore, so: * no more need for byte order tests * don't map video memory if we have a MMIO register aperture * get rid of various bits of support code dealing with video memory access - use MMIO register aperture if available, turn off register blocks in the main aperture if not in use, to avoid overlap with video memory on 8MB cards - don't scissor off access to off-screen memory. How the hell did this work on Rage II? Apparently Rage XL treats the scissor registers slightly different. Now this works properly on 8MB and Rage XL boards like Sun's PGX64.
1.80	03-Aug-2012	macallan	machfb's memsize is in MB, glyphcache_init expects bytes. doh.
1.79	02-Aug-2012	macallan	add support for anti-aliased fonts
1.78	14-Jun-2012	macallan	more preparation for alpha support: - use R3G3B2 colour map - consistently use ri_devcmap[] - reshuffle cmap setup, screen cleaning, msg buffer replay etc. to deal with the above
1.77	14-Jun-2012	macallan	do some minimal setup if we're not the console
1.76	14-Jun-2012	macallan	cosmetics: - fix/remove outdated comments - printf() -> aprint_*() - #if 1 removal
1.75	08-May-2012	macallan	don't enable IO access - we don't use it and at least on some macppc machines the IO BAR contains garbage should probably be pulled into 6.0
1.74	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.73	30-Jan-2012	drochner	branches: 1.73.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.72	11-Jan-2012	macallan	use rasops_init(0, 0)
1.71	04-Aug-2011	jakllsch	branches: 1.71.2; 1.71.6; Fix machfb build on ports not based on powerpc or sparc.
1.70	03-Aug-2011	macallan	don't try to a 'better' vertical frequency, it's almost certainly wrong with flat panels
1.69	01-Aug-2011	mbalmer	According to the Oxford Dictionary of Etymology, a wether is a male sheep or ram.
1.68	29-Jun-2011	macallan	support WSDISPLAYIO_GET_EDID
1.67	01-Jun-2011	macallan	move wsdisplay_accessops into the softc so multiple instances have a chance to work without stepping on each other's toes
1.66	16-May-2011	macallan	don't leave a mess on screen when attaching, while there use VCONS_DONT_READ if VCONS_DRAW_INTR is set
1.65	10-May-2011	dyoung	Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.64	04-May-2011	macallan	add support for (ancient) Mach64 CX and GX, while there make sure the visible vram area starts at the same offset as the area we're going to draw into - zero that is. from David Riley
1.63	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.62	16-Dec-2010	cegger	branches: 1.62.2; 1.62.4; ioctl: KNF switch-case, remove useless brackets
1.61	02-Oct-2010	macallan	cleanup debug code
1.60	02-Oct-2010	macallan	fix a typo
1.59	04-May-2010	macallan	this has been sitting in my tree too long: - device_t-ify - sanitize debugging #ifdefs - deal with firmwares setting up modes using other clocks than #0 - get rid of sparc and macppc-specific code, use device properties instead - while there, use alternate box drawing font if needed
1.58	11-Mar-2010	mrg	branches: 1.58.2; various aprint_* fixes.
1.57	06-May-2009	elad	branches: 1.57.2; Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.56	03-Jan-2009	yamt	branches: 1.56.2; remove extra semicolons.
1.55	01-Jul-2008	dyoung	branches: 1.55.4; Comment out some device IDs that do not actually belong to mach64 devices. This stops machfb0 from trying and failing to attach as console on my G4 PowerBook. Now, genfb0 will attach as console, instead. Apparently, Michael Lorenz made a similar change on the netbsd-4 branch, but it was never committed to -current.
1.54	01-Jul-2008	dyoung	Cosmetic: use device_t, cfdata_t, __arraycount(), device_private().
1.53	12-Jun-2008	cegger	use device_lookup_private to get softc
1.52	10-Apr-2008	cegger	branches: 1.52.2; 1.52.4; 1.52.6; 1.52.8; use aprint_*_dev and device_xname
1.51	24-Mar-2008	elad	Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. This and similar requests will be replaced Soon, but these two in particular should be pulled up to netbsd-4, and to allow doing that smoothly, we're first introducing the "issuser" version. Discussed with and okay macallan@.
1.50	15-Jan-2008	jmcneill	branches: 1.50.6; For WSDISPLAY_GINFO, depth should be sc->bits_per_pixel, not sc->color_depth. While we're here, implement WSDISPLAYIO_LINEBYTES.
1.49	04-Mar-2007	christos	branches: 1.49.16; 1.49.22; 1.49.28; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.48	16-Nov-2006	christos	branches: 1.48.2; 1.48.4; 1.48.8; __unused removal on arguments; approved by core.
1.47	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.46	02-Sep-2006	christos	branches: 1.46.2; 1.46.4; add missing initializers
1.45	10-May-2006	mrg	quell GCC 4.1 uninitialised variable warnings. XXX: we should audit the tree for which old ones are no longer needed after getting the older compilers out of the tree..
1.44	19-Apr-2006	macallan	fix some ugly mixup of virtual and physical addresses
1.43	16-Apr-2006	macallan	correct a cast to avoid a warning. Pointed out by he.
1.42	15-Apr-2006	jmmv	Remove the getwschar and putwschar accessops from wsdisplay drivers as requested by uwe@. These were wrong because they were receiving an emulcookie yet they were accessops (thus having to receive an accesscookie). Instead, just handle the WSDISPLAYIO_{GET,PUT}WSCHAR ioctls from the driver's ioctl accessop. As this reduces the amount of code needed to handle these operations to two small functions in each driver, remove the WSDISPLAY_CHARFUNCS kernel option. Reviewed by, at least, uwe@ and macallan@. No objections in tech-kern@.
1.41	12-Apr-2006	macallan	convert to vcons while there, attach an fb device when running on sparc
1.40	12-Apr-2006	jmmv	Add an extra cookie to the ioctl and mmap wsdisplay accessops that points to the screen on which they are being called. The driver cannot guess this by itself but it is needed to implement, at least, the getwschar and putwschar functions in the correct place. There are no functional changes yet. Tested on i386 (vga, vga_raster, machfb, vesafb), macppc and sparc64. Suggested and reviewed by macallan@.
1.39	05-Apr-2006	drochner	update for accessops change (get/serborder removed)
1.38	18-Dec-2005	macallan	branches: 1.38.4; 1.38.6; 1.38.8; 1.38.10; 1.38.12; Add devname: to printf()s From rivo nurges
1.37	12-Dec-2005	christos	welcome to the new lwp world.
1.36	11-Dec-2005	christos	merge ktrace-lwp.
1.35	03-Oct-2005	macallan	Build mach64_get_mode only on SPARC and PowerPC to shut up a compiler warning.
1.34	01-Oct-2005	macallan	Cleanup, remove some duplicate code, behave better with more than one instance present.
1.33	21-Sep-2005	macallan	- avoid using global variables unless necessary, this should fix weird problems when more than one mach64 is present - check memory BARs in mach64_mmap() and adjust allowed ranges in case something ( XFree86 for instance ) changed them - disable 'standard' framebuffer mapping at offset 0 on sparc64 because some Sun/ATI firmware likes to map PCI resources there. May be necessary on other 64bit architectures as well.
1.32	02-Aug-2005	macallan	fix a stupid typo
1.31	02-Aug-2005	macallan	disable mmap()ing the framebuffer at offset 0 on sparc64 because some ATI firmware likes to put PCI memory resources into this range, notably a Rage IIc which puts the 2nd register aperture to 0x2000. This should allow a few graphics chips to work with XFree86 which previously failed with something like this: (WW) ATI: PCI/AGP Mach64 in slot 2:5:0 could not be detected! No devices to configure. Configuration failed. Thanks to Florian Stoehr for doing most of the work tracking this down.
1.30	07-Jul-2005	thorpej	Apply const and static where appropriate.
1.29	31-May-2005	christos	branches: 1.29.2; Always call mach64_init_screen, otherwise we'll panic later if DIAGNOSTIC when we call switch screen and the machine has a serial console. Thanks to martin for the fix.
1.28	31-May-2005	christos	minor knf. no functional change.
1.27	30-May-2005	christos	add const.
1.26	02-May-2005	macallan	more formatting fixes
1.25	02-May-2005	macallan	fixed formatting
1.24	02-May-2005	macallan	- made blitter operationd asynchronous - removed some debug code - prevent blitter operations when not in WSDISPLAYIO_MODE_EMUL (approved by martin)
1.23	07-Mar-2005	martin	branches: 1.23.2; Make it compile for 32bit sparc kernels (bus_addr_t is larger than void* there)
1.22	27-Feb-2005	perry	nuke trailing whitespace
1.21	25-Feb-2005	martin	From Michael Lorenz: more cleanup and various bugfixes, support for WSDISPLAYIO_GETWSCHAR and WSDISPLAYIO_PUTWSCHAR ioctls.
1.20	17-Jan-2005	martin	branches: 1.20.2; More changes from Michael Lorenz: - fix a panic in mach64_alloc_screen() - some cleanup - restrict mach64_mmap() to addresses which belong to it - mach64_attach now prints bus addresses instead of kernel vm addresses - initial support for macppc
1.19	09-Jan-2005	martin	branches: 1.19.2; Bugfixes and acceleration support. From Michael Lorenz.
1.18	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.17	22-Mar-2004	pk	sparc: OF_instance_to_package => prom_instance_to_package
1.16	22-Mar-2004	martin	Adapt to promlib changes on sparc.
1.15	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.14	29-Jun-2003	fvdl	branches: 1.14.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.13	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.12	20-Jan-2003	martin	Grrr, add another set of casts working around sc_aperbase having the wrong type. We should fix this! This makes it compile for 32bit sparc64 kernels.
1.11	15-Jan-2003	martin	Fix typo - make it compile with DIAGNOSTIC.
1.10	11-Dec-2002	junyoung	'.' was missing in license notice.
1.9	25-Nov-2002	martin	Remove assumptions about endianess of bus_space_vaddr mappings, replace with an explicit probe.
1.8	02-Nov-2002	martin	branches: 1.8.2; Need to cast a bus_space_addr_t to (u_long) before gcc allows us to cast it to a pointer on 32bit spacr64 kernels. Sigh.
1.7	31-Oct-2002	martin	Rearange struct mach64screen to be derived from struct rasops_info, move a few bits around and make adding screens after attach time actually work. When not booting as console, try to properly set up the hardware to get a display nevertheless (XXX - does not yet work on my U5). #if 0 some unused functions planned for future extensions (to make clear they are unused now)
1.6	29-Oct-2002	junyoung	Reduce the number of #ifdef __sparc__ from 5 to 3.
1.5	29-Oct-2002	junyoung	Rename sparc_screen_is_console() to mach64_is_console() and turn it into a MI interface.
1.4	25-Oct-2002	junyoung	Register aperture is now mapped as a subregion of the framebuffer aperture using bus_space_subregion(). This makes reg*() look better.
1.3	25-Oct-2002	martin	Sprinkle a few bus_space_read/bus_space_write calls and simplify register offset calculation. Mostly from Bang Jun Young. Don't call wsdisplay_cnattach unconditionally. On sparc use OF to decide whether we are console output. This makes it actually work on my U5 - if only we had a keyboard driver to produce wskbd events (coming soon).
1.2	24-Oct-2002	martin	A few 64bit nits.
1.1	24-Oct-2002	junyoung	Add machfb, ATI Mach64/Rage framebuffer display driver. XXX this version is not fully functional yet. More to come shortly.
1.8.2.5	17-Jan-2003	thorpej	Sync with HEAD.
1.8.2.4	19-Dec-2002	thorpej	Sync with HEAD.
1.8.2.3	11-Dec-2002	thorpej	Sync with HEAD.
1.8.2.2	11-Nov-2002	nathanw	Catch up to -current
1.8.2.1	02-Nov-2002	nathanw	file machfb.c was added on branch nathanw_sa on 2002-11-11 22:11:19 +0000
1.14.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.8	08-Mar-2005	skrll	Sync with HEAD.
1.14.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.6	24-Jan-2005	skrll	Sync with HEAD.
1.14.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.14.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.2	03-Aug-2004	skrll	Sync with HEAD
1.14.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.19.2.1	29-Apr-2005	kent	sync with -current
1.20.2.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.23.2.5	08-Jun-2005	tron	Pull up revision 1.29 (requested by martin in ticket #377): Always call mach64_init_screen, otherwise we'll panic later if DIAGNOSTIC when we call switch screen and the machine has a serial console. Thanks to martin for the fix.
1.23.2.4	08-Jun-2005	tron	Pull up revision 1.28 (requested by martin in ticket #377): minor knf. no functional change.
1.23.2.3	08-Jun-2005	tron	Pull up revision 1.26 (requested by martin in ticket #377): more formatting fixes
1.23.2.2	08-Jun-2005	tron	Pull up revision 1.25 (requested by martin in ticket #377): fixed formatting
1.23.2.1	08-Jun-2005	tron	Pull up revision 1.24 (requested by martin in ticket #377): - made blitter operationd asynchronous - removed some debug code - prevent blitter operations when not in WSDISPLAYIO_MODE_EMUL (approved by martin)
1.29.2.4	21-Jan-2008	yamt	sync with head
1.29.2.3	03-Sep-2007	yamt	sync with head.
1.29.2.2	30-Dec-2006	yamt	sync with head.
1.29.2.1	21-Jun-2006	yamt	sync with head.
1.38.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.38.10.2	11-May-2006	elad	sync with head
1.38.10.1	19-Apr-2006	elad	sync with head.
1.38.8.3	03-Sep-2006	yamt	sync with head.
1.38.8.2	24-May-2006	yamt	sync with head.
1.38.8.1	11-Apr-2006	yamt	sync with head
1.38.6.2	01-Jun-2006	kardel	Sync with head.
1.38.6.1	22-Apr-2006	simonb	Sync with head.
1.38.4.1	09-Sep-2006	rpaulo	sync with head
1.46.4.2	10-Dec-2006	yamt	sync with head.
1.46.4.1	22-Oct-2006	yamt	sync with head
1.46.2.1	18-Nov-2006	ad	Sync with head.
1.48.8.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.48.8.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.48.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.48.2.2	24-Mar-2008	bouyer	Pull up following revision(s) (requested by elad in ticket #1107): sys/dev/pci/machfb.c: revision 1.51 sys/dev/pci/voodoofb.c: revision 1.14 Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. Discussed with and okay macallan@.
1.48.2.1	21-Jul-2007	liamjfoy	branches: 1.48.2.1.4; Pull up following revision(s) (requested by macallan in ticket #781): This keeps machfb from matching Rage Mobility chips which we don't support properly, mainly for lack of hardware in the right hands. Provided by a patch.
1.48.2.1.4.1	24-Mar-2008	bouyer	Pull up following revision(s) (requested by elad in ticket #1107): sys/dev/pci/machfb.c: revision 1.51 sys/dev/pci/voodoofb.c: revision 1.14 Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. Discussed with and okay macallan@.
1.49.28.1	19-Jan-2008	bouyer	Sync with HEAD
1.49.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.49.16.1	23-Mar-2008	matt	sync with HEAD
1.50.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.50.6.4	02-Jul-2008	mjf	Sync with HEAD.
1.50.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.50.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.50.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.52.8.2	03-Jul-2008	simonb	Sync with head.
1.52.8.1	18-Jun-2008	simonb	Sync with head.
1.52.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.52.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.52.4.4	09-Oct-2010	yamt	sync with head
1.52.4.3	11-Aug-2010	yamt	sync with head.
1.52.4.2	16-May-2009	yamt	sync with head
1.52.4.1	04-May-2009	yamt	sync with head.
1.52.2.1	17-Jun-2008	yamt	sync with head.
1.55.4.1	19-Jan-2009	skrll	Sync with HEAD.
1.56.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.57.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.57.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.57.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.58.2.4	12-Jun-2011	rmind	sync with head
1.58.2.3	31-May-2011	rmind	sync with head
1.58.2.2	05-Mar-2011	rmind	sync with head
1.58.2.1	30-May-2010	rmind	sync with head
1.62.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.62.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.71.6.3	02-Jun-2012	mrg	sync to latest -current.
1.71.6.2	05-Apr-2012	mrg	sync to latest -current.
1.71.6.1	18-Feb-2012	mrg	merge to -current.
1.71.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.71.2.3	30-Oct-2012	yamt	sync with head
1.71.2.2	23-May-2012	yamt	sync with head.
1.71.2.1	17-Apr-2012	yamt	sync with head
1.73.2.1	12-Jun-2012	riz	Pull up following revision(s) (requested by macallan in ticket #322): sys/dev/pci/machfb.c: revision 1.75 don't enable IO access - we don't use it and at least on some macppc machines the IO BAR contains garbage should probably be pulled into 6.0
1.83.2.4	03-Dec-2017	jdolecek	update from HEAD
1.83.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.83.2.2	23-Jun-2013	tls	resync from head
1.83.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.87.2.2	18-May-2014	rmind	sync with head
1.87.2.1	28-Aug-2013	rmind	sync with head
1.91.6.2	28-Aug-2017	skrll	Sync with HEAD
1.91.6.1	05-Oct-2016	skrll	Sync with HEAD
1.94.10.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.94.10.1	10-Jun-2019	christos	Sync with HEAD
1.94.8.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.97.4.1	16-Apr-2020	martin	Pull up following revision(s) (requested by jdc in ticket #836): sys/dev/pci/machfb.c: revision 1.98 clean up the video mode selection logic, switch modes only when actually necessary while there make some debug output optional
1.103.4.3	24-Apr-2021	thorpej	Make sure to explcitly specify the "wsemuldisplaydev" interface attribute when attaching the wscons display instance, because these devices can also attach children using the "drm" interface attribute.
1.103.4.2	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.103.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.104.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.105.2.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	15-Aug-2012	macallan	add a few register definitions
1.4	04-May-2010	macallan	branches: 1.4.8; add some PLL-specific register definitions
1.3	11-Dec-2005	christos	branches: 1.3.74; 1.3.96; 1.3.98; merge ktrace-lwp.
1.2	03-Oct-2005	macallan	Cleanup, remove some duplicate code, behave better with more than one instance present. ( for machfb.c 1.34+ )
1.1	24-Oct-2002	junyoung	branches: 1.1.2; 1.1.8; 1.1.24; Add machfb, ATI Mach64/Rage framebuffer display driver. XXX this version is not fully functional yet. More to come shortly.
1.1.24.1	21-Jun-2006	yamt	sync with head.
1.1.8.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.2	11-Nov-2002	nathanw	Catch up to -current
1.1.2.1	24-Oct-2002	nathanw	file machfbreg.h was added on branch nathanw_sa on 2002-11-11 22:11:20 +0000
1.3.98.1	30-May-2010	rmind	sync with head
1.3.96.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.3.74.1	11-Aug-2010	yamt	sync with head.
1.4.8.1	30-Oct-2012	yamt	sync with head
1.2	03-Sep-2002	augustss	Replace the mr driver with the gtp driver. From OpenBSD.
1.1	01-Jan-2002	augustss	branches: 1.1.2; 1.1.4; Add support for radio cards. Written by Maxim Tsyplakov and Vladimir Popov for OpenBSD, from where it was imported.
1.1.4.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.1.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.1	01-Jan-2002	thorpej	file maxiradio.c was added on branch kqueue on 2002-01-10 19:56:50 +0000
1.1.2.3	17-Sep-2002	nathanw	Catch up to -current.
1.1.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.1	01-Jan-2002	nathanw	file maxiradio.c was added on branch nathanw_sa on 2002-01-08 00:31:08 +0000
1.22	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.21	05-May-2022	msaitoh	Add some code for the SKINNY variant to make Dell PERC H310 work.
1.20	09-Dec-2018	jdolecek	branches: 1.20.4; use pci_intr_establish_xname() everywhere
1.19	14-Jul-2016	msaitoh	branches: 1.19.10; 1.19.16; 1.19.18; - Use aprint*() instead of printf() in xxx_attach(). - Add missing aprint_naive("\n"); - KNF
1.18	29-Mar-2014	christos	branches: 1.18.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.17	31-Aug-2012	jnemeth	branches: 1.17.2; 1.17.4; PR/46877 - Wolfgang Stukenbrock -- missing definition for Symbios SAS 9261-8i
1.16	23-Aug-2012	bouyer	Add support newer LSI RAID controllers based on the SAS2208 chip, codenamed "ThunderBolt". Add tagged queuing support for all adapters supported by mfi(4). Tested with a MegaRAID SAS 9265-8i adapter, and an older Dell PERC 5/i.
1.15	05-Aug-2012	bouyer	MFI_IOP_SKINNY is an enum not a single bit value, so it can't be used as a bit flag. As sc_flags is really used to hold an enum mfi_iop value, change it to enum mfi_iop and rename to sc_ioptype. While there init it in mfi_attach() instead of mfi_pci_attach().
1.14	22-Mar-2012	sborrill	Add IBM ServeRAID M5014 as subtype
1.13	21-Mar-2012	sborrill	Add support for skinny variants (e.g. IBM ServeRAID M1015). Based on OpenBSD changes with some improvements. Tested on IBM x3550M3 with RAID0 and RAID1 volumes including bioctl(8) operation.
1.12	09-Feb-2010	msaitoh	branches: 1.12.10; 1.12.14; 1.12.16; Add newer cars supports. Tested on MegaRAID SAS 9260-8i. - Add MFI gen2 support from OpenBSD. - Add entry for MegaRAID SAS 9260-8i
1.11	16-Jul-2009	dyoung	branches: 1.11.2; Let us detach & re-attach children of mfi0. Detach mfi0 at shutdown. Detachment may fail after freeing some but not all resources, so take care not to re-release any resource during detachment. Tested on a Dell PowerEdge 1950.
1.10	16-Jul-2009	dyoung	device_t/softc split. Tested and shown to work on a Dell PowerEdge 1950.
1.9	16-Jul-2009	dyoung	Add a rudimentary detachment hook for mfi(4).
1.8	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.7	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.6	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.5	26-Feb-2008	xtraeme	branches: 1.5.4; 1.5.14; 1.5.18; 1.5.22; Make it follow the standard convention of autoconf printing: mfi0 at pci0 ... mfi0: interrupting at ... and only print the model string if it was matched. Add the Dell PERC 5/i integraded subproduct ID.
1.4	25-Feb-2008	xtraeme	Added support for the mfi(4) controllers with powerpc IOPs such as LSI SAS1078 or Dell PERC 6, from OpenBSD. Tested by Akira Kato on current-users@.
1.3	19-Oct-2007	ad	branches: 1.3.12; 1.3.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.2	20-Dec-2006	bouyer	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.14; 1.2.26; 1.2.28; 1.2.32; Add missing $NetBSD: $
1.1	17-Dec-2006	bouyer	branches: 1.1.2; Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64.
1.1.2.3	21-Dec-2006	yamt	sync with head.
1.1.2.2	18-Dec-2006	yamt	sync with head.
1.1.2.1	17-Dec-2006	yamt	file mfi_pci.c was added on branch yamt-splraiseipl on 2006-12-18 11:42:13 +0000
1.2.32.1	25-Oct-2007	bouyer	Sync with HEAD.
1.2.28.2	23-Mar-2008	matt	sync with HEAD
1.2.28.1	06-Nov-2007	matt	sync with HEAD
1.2.26.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.2.14.1	23-Oct-2007	ad	Sync with head.
1.2.8.2	12-Jan-2007	ad	Sync with head.
1.2.8.1	20-Dec-2006	ad	file mfi_pci.c was added on branch newlock2 on 2007-01-12 00:57:41 +0000
1.2.6.4	27-Feb-2008	yamt	sync with head.
1.2.6.3	27-Oct-2007	yamt	sync with head.
1.2.6.2	30-Dec-2006	yamt	sync with head.
1.2.6.1	20-Dec-2006	yamt	file mfi_pci.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:46 +0000
1.2.4.2	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1609): sys/arch/i386/conf/INSTALL: revision 1.298 sys/arch/i386/conf/GENERIC: revision 1.802 sys/dev/pci/files.pci: revision 1.274 sys/arch/amd64/conf/GENERIC: revision 1.122 sys/dev/pci/mfi_pci.c: revision 1.1 via patch distrib/sets/lists/man/mi: revision 1.958 sys/arch/amd64/conf/INSTALL: revision 1.61 sys/dev/ic/mfireg.h: revision 1.1 share/man/man4/Makefile: revision 1.416 via patch sys/dev/ic/mfi.c: revision 1.1 sys/arch/i386/conf/XEN2_DOM0: revision 1.19 sys/dev/ic/mfivar.h: revision 1.1 sys/conf/files: revision 1.821 share/man/man4/mfi.4: revision 1.1 share/man/man4/mfi.4: revision 1.2 sys/dev/pci/pcidevs via patch Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64. Punctuation issues.
1.2.4.1	20-Dec-2006	ghen	file mfi_pci.c was added on branch netbsd-3 on 2006-12-23 16:43:18 +0000
1.2.2.3	21-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #286): sys/dev/pci/mfi_pci.c: revision 1.2 sys/dev/ic/mfireg.h: revision 1.2 sys/dev/ic/mfi.c: revision 1.2 sys/dev/ic/mfivar.h: revision 1.2 Add missing $NetBSD: $
1.2.2.2	21-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #286): sys/arch/i386/conf/INSTALL: revision 1.298 sys/arch/i386/conf/GENERIC: revision 1.802 sys/dev/pci/files.pci: revision 1.274 sys/arch/amd64/conf/GENERIC: revision 1.122 sys/dev/pci/mfi_pci.c: revision 1.1 distrib/sets/lists/man/mi: revision 1.958 sys/arch/amd64/conf/INSTALL: revision 1.61 sys/arch/i386/conf/ALL: revision 1.75 sys/dev/ic/mfireg.h: revision 1.1 share/man/man4/Makefile: revision 1.416 sys/dev/ic/mfi.c: revision 1.1 sys/arch/i386/conf/XEN2_DOM0: revision 1.19 sys/dev/ic/mfivar.h: revision 1.1 sys/conf/files: revision 1.821 share/man/man4/mfi.4: revision 1.1 Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64.
1.2.2.1	20-Dec-2006	tron	file mfi_pci.c was added on branch netbsd-4 on 2006-12-21 12:46:21 +0000
1.3.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.3.12.1	24-Mar-2008	keiichi	sync with head.
1.5.22.1	21-Apr-2010	matt	sync to netbsd-5
1.5.18.2	23-Jul-2009	jym	Sync with HEAD.
1.5.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.5.14.2	30-Sep-2012	bouyer	Pull up following revision(s) (requested by jnemeth in ticket #1795): sys/dev/pci/mfi_pci.c: revision 1.17 via patch PR/46877 - Wolfgang Stukenbrock -- missing definition for Symbios SAS 9261-8i
1.5.14.1	28-Mar-2010	snj	Apply patch (requested by msaitoh in ticket #1326): Add support for MFI gen2 devices. Mention newer devices, RAID 6 and RAID60 in mfi.4.
1.5.4.3	11-Mar-2010	yamt	sync with head
1.5.4.2	18-Jul-2009	yamt	sync with head.
1.5.4.1	16-May-2009	yamt	sync with head
1.11.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.12.16.4	24-Oct-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #628): sys/dev/ic/mfi.c: revision 1.40 sys/dev/ic/mfi.c: revision 1.41 sys/dev/ic/mfi.c: revision 1.42 sys/dev/ic/mfi.c: revision 1.43 sys/dev/ic/mfi.c: revision 1.44 sys/dev/ic/mfi.c: revision 1.45 sys/dev/ic/mfi.c: revision 1.46 sys/dev/pci/pcidevs: revision 1.1128 sys/dev/pci/mfi_pci.c: revision 1.15 sys/dev/pci/mfi_pci.c: revision 1.16 sys/dev/ic/mfi.c: revision 1.39 sys/dev/ic/mfivar.h: revision 1.16 sys/dev/ic/mfireg.h: revision 1.6 sys/dev/ic/mfivar.h: revision 1.17 sys/dev/ic/mfireg.h: revision 1.7 sys/dev/ic/mfivar.h: revision 1.18 sys/dev/ic/mfivar.h: revision 1.19 Add some support for 64bit DMA but stick to 32bit DMA for now. From OpenBSD mfi.c rev 1.119. Add LSI MegaRAID SAS2208 MFI_IOP_SKINNY is an enum not a single bit value, so it can't be used as a bit flag. As sc_flags is really used to hold an enum mfi_iop value, change it to enum mfi_iop and rename to sc_ioptype. While there init it in mfi_attach() instead of mfi_pci_attach(). sg64.len is 32bits, so use htole32() Really init sc_ioptype in mfi_attach() Add support newer LSI RAID controllers based on the SAS2208 chip, codenamed "ThunderBolt". Add tagged queuing support for all adapters supported by mfi(4). Tested with a MegaRAID SAS 9265-8i adapter, and an older Dell PERC 5/i. Avoid "unused variable" warning for non-DIAGNOSTIC kernels. Pointed out by Havard Eidnes After discussion on tech-kern@, fix performance issue related to cache flush commands from WAPBL, by skipping the cache flush if the BBU is present and considered good. Users which still want the write back cache with a non-working BBU can set vfs.wapbl.flush_disk_cache to 0. - add commands to monitor the BBU state. Add a boolean BBU sensor to monitor the BBU state via sysmon_envsys(9). - if the BBU is considered good, turn SCSI_SYNCHRONIZE_CACHE_10 and SCSI_SYNCHRONIZE_CACHE_16 commands from upper layer into NOOPs. While there, handle SCSI_SYNCHRONIZE_CACHE_16 in addition to SCSI_SYNCHRONIZE_CACHE_10. - Add a shutdown pmf(9) handler, which flushes the cache and shutdown the firmware - on detach, also flush cache and shutdown firmware. - on attach, print the firmware-provided name, and the BBU state Tested on a LSI MegaRAID SAS 9265-8i and a PERC 5/i Integrated Make MFI_DEBUG build on i386. Report BBU state changes with aprint_normal(), it seems that sysmon_envsys() doens't report changes for ENVSYS_INDICATOR as it does for ENVSYS_DRIVE.
1.12.16.3	03-Sep-2012	riz	Pull up following revision(s) (requested by jnemeth in ticket #536): sys/dev/pci/mfi_pci.c: revision 1.17 PR/46877 - Wolfgang Stukenbrock -- missing definition for Symbios SAS 9261-8i
1.12.16.2	19-May-2012	riz	branches: 1.12.16.2.2; Pull up following revision(s) (requested by sborrill in ticket #261): sys/dev/pci/mfi_pci.c: revision 1.14 share/man/man4/mfi.4: revision 1.11 Add IBM ServeRAID M5014 as subtype Add IBM ServeRAID M1015 and M5014
1.12.16.1	22-Mar-2012	riz	Pull up following revision(s) (requested by sborrill in ticket #138): sys/dev/pci/mfi_pci.c: revision 1.13 sys/dev/ic/mfi.c: revision 1.38 sys/dev/ic/mfivar.h: revision 1.15 sys/dev/ic/mfireg.h: revision 1.5 Add support for skinny variants (e.g. IBM ServeRAID M1015). Based on OpenBSD changes with some improvements. Tested on IBM x3550M3 with RAID0 and RAID1 volumes including bioctl(8) operation.
1.12.16.2.2.1	01-Nov-2012	matt	sync with netbsd-6-0-RELEASE.
1.12.14.1	05-Apr-2012	mrg	sync to latest -current.
1.12.10.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.10.2	30-Oct-2012	yamt	sync with head
1.12.10.1	17-Apr-2012	yamt	sync with head
1.17.4.1	18-May-2014	rmind	sync with head
1.17.2.2	03-Dec-2017	jdolecek	update from HEAD
1.17.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.18.6.1	05-Oct-2016	skrll	Sync with HEAD
1.19.18.1	10-Jun-2019	christos	Sync with HEAD
1.19.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.19.10.1	17-May-2022	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1454): sys/dev/ic/mfi.c: revisions 1.63, 1.66-77 sys/dev/ic/mfireg.h: revisions 1.11-1.20 via patch sys/dev/pci/mfi_pci.c: revision 1.21 sys/dev/pci/mfii.c: revisions 1.6-1.7, 1.10-1.15 share/man/man4/mfi.4: revision 1.13 Improve mfi(4) and mfii(4): - Set 'ld_sync' to NULL as part of 'again', to prevent use-after-free. - Add some code for the SKINNY variant to make Dell PERC H310 work. - Print the percentage correctly when the background initialization is running. - Clear mailbox to not to pass garbage data. - Use union mbox instead of unit8_t xxx[] to avoid unaligned access. - Set stripe size for BIOCVOL to show the size correctly in bioctl. - Add support for iBBU-09 to show BBU voltage, current and temperature correctly. - Fix typos in comments. - Sprinkle static. - Improve debug printf()s. - KNF. Remove extra semicolon. Whitespace fixes.
1.20.4.1	17-May-2022	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1454): sys/dev/ic/mfi.c: revisions 1.63, 1.66-77 sys/dev/ic/mfireg.h: revisions 1.11-1.20 via patch sys/dev/pci/mfi_pci.c: revision 1.21 sys/dev/pci/mfii.c: revisions 1.6-1.7, 1.10-1.15 share/man/man4/mfi.4: revision 1.13 Improve mfi(4) and mfii(4): - Set 'ld_sync' to NULL as part of 'again', to prevent use-after-free. - Add some code for the SKINNY variant to make Dell PERC H310 work. - Print the percentage correctly when the background initialization is running. - Clear mailbox to not to pass garbage data. - Use union mbox instead of unit8_t xxx[] to avoid unaligned access. - Set stripe size for BIOCVOL to show the size correctly in bioctl. - Add support for iBBU-09 to show BBU voltage, current and temperature correctly. - Fix typos in comments. - Sprinkle static. - Improve debug printf()s. - KNF. Remove extra semicolon. Whitespace fixes.
1.32	13-Feb-2024	msaitoh	mfii(4): Apply two changes from OpenBSD to fix an unknown firmware state. My own MegaRAID 946N-8i 2G", firmware 50.5.0-2594 failed to attach. mfii0: unknown firmware state 1879048192 1879048192 equals to 0x70000000(== MFI_STATE_FW_INIT_2). Apply following two OpenBSD commits to resolve this problem.
1.31	05-Oct-2023	christos	provide a separate pointer that points to the drive sensors (Edgar Fuss)
1.30	23-Sep-2023	christos	When attaching and detaching disks, use the proper index, skipping the first 4, which are used for the controller itself (Edgar Fuss)
1.29	22-Sep-2023	christos	update present variable when disks are added or removed (Edgar Fuss)
1.28	29-Sep-2022	bouyer	Convert from pci_intr_map() to pci_intr_alloc(); makes the driver use MSI/MSI-x when available. Tested by Edgar Fu�.
1.27	14-Sep-2022	msaitoh	mfii: Don't panic on detach. Call sysmon_envsys_destroy() and clear sc->sc_sme to NULL if sysmon_envsys_register failed.
1.26	16-Jul-2022	msaitoh	Enable Aero.
1.25	16-Jul-2022	msaitoh	It seems that a logical volume number and the target ID is different on newer systems. It's observed on an Aero. Add sc->sc_ld[X].ld_target_id to keep the target ID and use it for scsipi layer. Now sdX is attached on Aero system. Tested by Mark Davies.
1.24	16-Jul-2022	msaitoh	Add max 256 VD support.
1.23	09-Jul-2022	msaitoh	Add preliminary support for Aero. - Aero supports 32bit descriptor. - Even if the Aero card supports 32bit descriptor, 64bit descriptor access is required for MFI_CMD_INIT. - Add quirk for Aero to retry register read. - The PCI ID table for Aero is disabled. Even if it's enabled, any logical volumes are not found. Perhaps the reason is that our mfii driver is currently not supports more than 64 logical volumes and MR_DCMD_LD_GET_LIST returns incorrect value. Tested by Mark Daies' H750.
1.22	09-Jul-2022	msaitoh	Use bus_space_write_8() if available.
1.21	09-Jul-2022	msaitoh	Modify strange bus_space_barrier().
1.20	29-Jun-2022	msaitoh	Add CVPM02 BBU support.
1.19	27-Jun-2022	msaitoh	KNF a little. No functional change.
1.18	27-Jun-2022	msaitoh	Add some MegaRAID devices (except Aero series). Not tested.
1.17	27-Jun-2022	msaitoh	Add comment. Sort entries. No functional change.
1.16	22-Jun-2022	msaitoh	Fix wrong access in mfii_start(). I hope this fixes kern/55192 and kern/56669.
1.15	13-May-2022	msaitoh	Fix compile error when MFII_DEBUG is set. Whitespace fix.
1.14	12-May-2022	msaitoh	Use IBBU's data structure for iBBU-09.
1.13	11-May-2022	msaitoh	Add support for iBBU-09 to show BBU voltage, current and temp correctly.
1.12	11-May-2022	msaitoh	mfii(4): Set stripe size correctly for BIOCVOL. Same as mfi.c rev. 1.75.
1.11	07-May-2022	msaitoh	Use mlp_bgi for MFI_LD_PROG_BGI. Same as OpenBSD.
1.10	05-May-2022	msaitoh	Whitespace.
1.9	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.8	24-Apr-2021	thorpej	branches: 1.8.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.7	14-May-2020	msaitoh	branches: 1.7.4; Remove extra semicolon.
1.6	27-Nov-2019	maxv	localify
1.5	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.4	24-Apr-2019	bouyer	branches: 1.4.2; 1.4.4; Clean up timeout logic. Now the system survives to a ccb timeout; before the filesystem would stall because one I/O did never completes.
1.3	03-Dec-2018	bouyer	branches: 1.3.2; Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0
1.2	24-Nov-2018	bouyer	branches: 1.2.2; Add my name in copyright list
1.1	24-Nov-2018	bouyer	Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6.
1.2.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.2.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.2.2.1	24-Nov-2018	pgoyette	file mfii.c was added on branch pgoyette-compat on 2018-11-26 01:52:32 +0000
1.3.2.7	29-Sep-2022	snj	Pull up following revision(s) (requested by bouyer in ticket #1768): sys/dev/pci/mfii.c: 1.28 Convert from pci_intr_map() to pci_intr_alloc(); makes the driver use MSI/MSI-x when available. Tested by Edgar Fu?.
1.3.2.6	16-Sep-2022	martin	Pull up the following, requested by msaitoh in ticket #1764: sys/dev/ic/mfireg.h 1.22 sys/dev/pci/mfii.c 1.16-1.22,1.27 via patch share/man/man4/mfii.4 1.3 via patch - Fix wrong access in mfii_start(). - Add SAS3216,SAS3224,SAS3316 and SAS3324 support. - Add CVPM02 BBU support. - Use bus_space_write_8() if available. - Don't panic on detach if no any sensor device. - Add comment. Sort entries. KNF.
1.3.2.5	17-May-2022	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1454): sys/dev/ic/mfi.c: revisions 1.63, 1.66-77 sys/dev/ic/mfireg.h: revisions 1.11-1.20 via patch sys/dev/pci/mfi_pci.c: revision 1.21 sys/dev/pci/mfii.c: revisions 1.6-1.7, 1.10-1.15 share/man/man4/mfi.4: revision 1.13 Improve mfi(4) and mfii(4): - Set 'ld_sync' to NULL as part of 'again', to prevent use-after-free. - Add some code for the SKINNY variant to make Dell PERC H310 work. - Print the percentage correctly when the background initialization is running. - Clear mailbox to not to pass garbage data. - Use union mbox instead of unit8_t xxx[] to avoid unaligned access. - Set stripe size for BIOCVOL to show the size correctly in bioctl. - Add support for iBBU-09 to show BBU voltage, current and temperature correctly. - Fix typos in comments. - Sprinkle static. - Improve debug printf()s. - KNF. Remove extra semicolon. Whitespace fixes.
1.3.2.4	26-Apr-2019	martin	Pull up following revision(s) (requested by bouyer in ticket #1247): sys/dev/pci/mfii.c: revision 1.4 Clean up timeout logic. Now the system survives to a ccb timeout; before the filesystem would stall because one I/O did never completes.
1.3.2.3	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1131): sys/dev/pci/mpiireg.h: revision 1.2 sys/dev/pci/mfii.c: revision 1.3 share/man/man4/mpii.4: revision 1.3 share/man/man4/mpii.4: revision 1.4 sys/dev/pci/mpii.c: revision 1.15 sys/dev/pci/mpii.c: revision 1.16 sys/dev/pci/mpii.c: revision 1.17 sys/dev/pci/mpii.c: revision 1.18 Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0 Update for the new mpii(4) driver, mostly from OpenBSD. Also tell the scsi layer that we are MPSAFE Remove trailing whitespace NULL-terminate mpii_devices[], this is the end condition of the loop in mpii_match(). Noticed by Mike Pumford There's no sensors for physical disks so don't try to detach them.
1.3.2.2	07-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1129): sys/arch/amd64/conf/XEN3_DOM0: revision 1.159 sys/arch/amd64/conf/GENERIC: revision 1.508 sys/arch/i386/conf/ALL: revision 1.458 sys/dev/ic/mfi.c: revision 1.60 sys/dev/pci/mpiireg.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.1628 sys/dev/pci/mfii.c: revision 1.1,1.2 (adapted) sys/arch/i386/conf/GENERIC: revision 1.1194 sys/dev/pci/mpii.c: revision 1.13 sys/dev/ic/mfireg.h: revision 1.9 share/man/man4/mfii.4: revision 1.1 share/man/man4/Makefile: revision 1.673 (patch) sys/dev/pci/files.pci: revision 1.410 share/man/man4/mfii.4: revision 1.2 sys/arch/amd64/conf/ALL: revision 1.108 sys/arch/i386/conf/XEN3PAE_DOM0: revision 1.8 (patch, in XEN3_DOM0) Add some definitions from OpenBSD, needed by the upcoming mfii driver. No functionnal change. - Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change. - Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6. - Add my name in copyright list - add a man page for the new mpii, mostly from OpenBSD. - Sort SEE ALSO. Fix date. Whitespace fixes.
1.3.2.1	03-Dec-2018	martin	file mfii.c was added on branch netbsd-8 on 2018-12-07 17:11:37 +0000
1.4.4.3	29-Sep-2022	martin	Pull up following revision(s) (requested by bouyer in ticket #1534): sys/dev/pci/mfii.c: revision 1.28 Convert from pci_intr_map() to pci_intr_alloc(); makes the driver use MSI/MSI-x when available. Tested by Edgar Fu?.
1.4.4.2	16-Sep-2022	martin	Pull up the following, requested by msaitoh in ticket #1524: sys/dev/ic/mfireg.h 1.22 sys/dev/pci/mfii.c 1.16-1.22,1.27 via patch share/man/man4/mfii.4 1.3 via patch - Fix wrong access in mfii_start(). - Add SAS3216,SAS3224,SAS3316 and SAS3324 support. - Add CVPM02 BBU support. - Use bus_space_write_8() if available. - Don't panic on detach if no any sensor device. - Add comment. Sort entries. KNF.
1.4.4.1	17-May-2022	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1454): sys/dev/ic/mfi.c: revisions 1.63, 1.66-77 sys/dev/ic/mfireg.h: revisions 1.11-1.20 via patch sys/dev/pci/mfi_pci.c: revision 1.21 sys/dev/pci/mfii.c: revisions 1.6-1.7, 1.10-1.15 share/man/man4/mfi.4: revision 1.13 Improve mfi(4) and mfii(4): - Set 'ld_sync' to NULL as part of 'again', to prevent use-after-free. - Add some code for the SKINNY variant to make Dell PERC H310 work. - Print the percentage correctly when the background initialization is running. - Clear mailbox to not to pass garbage data. - Use union mbox instead of unit8_t xxx[] to avoid unaligned access. - Set stripe size for BIOCVOL to show the size correctly in bioctl. - Add support for iBBU-09 to show BBU voltage, current and temperature correctly. - Fix typos in comments. - Sprinkle static. - Improve debug printf()s. - KNF. Remove extra semicolon. Whitespace fixes.
1.4.2.4	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.4.2.3	08-Apr-2020	martin	Merge changes from current as of 20200406
1.4.2.2	10-Jun-2019	christos	Sync with HEAD
1.4.2.1	24-Apr-2019	christos	file mfii.c was added on branch phil-wifi on 2019-06-10 22:07:17 +0000
1.7.4.3	28-Mar-2021	thorpej	No need to pass interface attribute or locators to config_found().
1.7.4.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.7.4.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.8.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.28	24-Apr-2021	thorpej	Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.27	09-Dec-2018	jdolecek	branches: 1.27.14; use pci_intr_establish_xname() everywhere
1.26	27-Sep-2016	pgoyette	branches: 1.26.14; 1.26.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.25	29-Mar-2014	christos	branches: 1.25.6; 1.25.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.24	27-Oct-2012	chs	branches: 1.24.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.23	26-Nov-2009	njoly	branches: 1.23.12; 1.23.22; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.22	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.21	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.20	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.19	28-Apr-2008	martin	branches: 1.19.14; Remove clause 3 and 4 from TNF licenses
1.18	10-Apr-2008	cegger	branches: 1.18.2; 1.18.4; use aprint_*_dev and device_xname
1.17	19-Oct-2007	ad	branches: 1.17.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	16-Nov-2006	christos	branches: 1.16.8; 1.16.22; 1.16.24; 1.16.28; __unused removal on arguments; approved by core.
1.15	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.14	02-Sep-2006	christos	branches: 1.14.2; 1.14.4; static goes first
1.13	11-Dec-2005	christos	branches: 1.13.4; 1.13.8; merge ktrace-lwp.
1.12	27-Feb-2005	perry	branches: 1.12.4; nuke trailing whitespace
1.11	02-Oct-2002	thorpej	branches: 1.11.6; 1.11.14; 1.11.16; Add trailing ; to CFATTACH_DECL.
1.10	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.9	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.8	31-Aug-2002	ad	Nuke incorrect mailbox length defines and just use 13 everywhere. May help with PR/14266.
1.7	26-Aug-2002	ad	Handle aincent firmware as found on EISA/MCA boards. Untested.
1.6	13-Nov-2001	lukem	branches: 1.6.8; 1.6.10; add RCSID
1.5	26-Jul-2001	ad	branches: 1.5.6; Map both memory and IO space, but prefer using memory space.
1.4	26-Jul-2001	ad	Don't map I/O space if we have already have a good memory mapping.
1.3	15-May-2001	ad	branches: 1.3.2; Correct arguments to bus_dmamap_create(), and add a reset routine for the v3 interface.
1.2	18-Feb-2001	mjacob	branches: 1.2.2; Make it compile (for alpha). The porting author might want to look at this more closely than I have.
1.1	04-Feb-2001	ad	branches: 1.1.2; Add a driver for the Mylex DAC960 family (including DEC SWXCR).
1.1.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.1.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.2.1	04-Feb-2001	bouyer	file mlx_pci.c was added on branch thorpej_scsipi on 2001-02-11 19:15:58 +0000
1.2.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.2.2.5	17-Sep-2002	nathanw	Catch up to -current.
1.2.2.4	27-Aug-2002	nathanw	Catch up to -current.
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.2.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.3.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	03-Aug-2001	lukem	update to -current
1.5.6.2	25-Oct-2001	he	Pull up revisions 1.1-1.5 (new, via patch, requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.5.6.1	26-Jul-2001	he	file mlx_pci.c was added on branch netbsd-1-5 on 2001-10-25 18:02:41 +0000
1.6.10.1	28-Jul-2003	he	Pull up revisions 1.7-1.8 (requested by ad in ticket #1128): Improve stability of the mlx driver. o Better handling of old firmware o Use correct mailbox length, fixes PR#14266
1.6.8.2	31-Aug-2002	gehenna	catch up with -current.
1.6.8.1	29-Aug-2002	gehenna	catch up with -current.
1.11.16.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.14.1	29-Apr-2005	kent	sync with -current
1.11.6.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.4.2	27-Oct-2007	yamt	sync with head.
1.12.4.1	30-Dec-2006	yamt	sync with head.
1.13.8.1	03-Sep-2006	yamt	sync with head.
1.13.4.1	09-Sep-2006	rpaulo	sync with head
1.14.4.2	10-Dec-2006	yamt	sync with head.
1.14.4.1	22-Oct-2006	yamt	sync with head
1.14.2.1	18-Nov-2006	ad	Sync with head.
1.16.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.24.1	06-Nov-2007	matt	sync with HEAD
1.16.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.8.1	23-Oct-2007	ad	Sync with head.
1.17.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.18.4.3	11-Mar-2010	yamt	sync with head
1.18.4.2	16-May-2009	yamt	sync with head
1.18.4.1	16-May-2008	yamt	sync with head.
1.18.2.1	18-May-2008	yamt	sync with head.
1.19.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.23.22.3	03-Dec-2017	jdolecek	update from HEAD
1.23.22.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.23.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.23.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.23.12.1	30-Oct-2012	yamt	sync with head
1.24.2.1	18-May-2014	rmind	sync with head
1.25.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.25.6.1	05-Oct-2016	skrll	Sync with HEAD
1.26.16.1	10-Jun-2019	christos	Sync with HEAD
1.26.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.27.14.1	28-Mar-2021	thorpej	Correct the names of the arguments passed to the rescan function.
1.56	03-Sep-2021	andvar	s/existant/existent/ in comments and messages, plus few more similar fixes.
1.55	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.54	24-Apr-2021	thorpej	branches: 1.54.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.53	10-Nov-2019	chs	branches: 1.53.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.52	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.51	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.50	07-Jul-2016	msaitoh	branches: 1.50.16; 1.50.18; KNF. Remove extra spaces. No functional change.
1.49	25-Jul-2014	dholland	branches: 1.49.4; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.48	29-Mar-2014	christos	branches: 1.48.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.47	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.46	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.45	27-Oct-2012	chs	branches: 1.45.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.44	13-Nov-2010	uebayasi	branches: 1.44.8; 1.44.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.43	26-Nov-2009	njoly	branches: 1.43.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.42	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.41	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.40	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.39	08-Jun-2008	tsutsui	branches: 1.39.12; Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.38	28-Apr-2008	martin	branches: 1.38.2; Remove clause 3 and 4 from TNF licenses
1.37	10-Apr-2008	cegger	branches: 1.37.2; 1.37.4; use aprint_*_dev and device_xname
1.36	19-Oct-2007	ad	branches: 1.36.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.35	09-Jul-2007	ad	branches: 1.35.6; 1.35.8; 1.35.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.34	04-Mar-2007	christos	branches: 1.34.2; 1.34.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.33	02-Dec-2006	elad	branches: 1.33.2; Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.32	16-Nov-2006	christos	branches: 1.32.2; __unused removal on arguments; approved by core.
1.31	08-Nov-2006	elad	Replace securelevel checks with kauth(9) calls.
1.30	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.29	02-Sep-2006	christos	branches: 1.29.2; 1.29.4; - fix incomplete initializer - static goes first
1.28	23-Aug-2006	christos	Acting as Elad's keyboard: Place the securelevel checks in their logical locations. This will be clearer in the future when code changes to use kauth(9) calls. input and okay ad@
1.27	14-May-2006	elad	integrate kauth.
1.26	14-Apr-2006	christos	Coverity CID 1103: Avoid NULL pointer deref.
1.25	14-Apr-2006	christos	Coverity CID 2439: Remove bogus test.
1.24	11-Dec-2005	christos	branches: 1.24.4; 1.24.6; 1.24.8; 1.24.10; 1.24.12; merge ktrace-lwp.
1.23	27-Feb-2005	perry	branches: 1.23.4; nuke trailing whitespace
1.22	21-Feb-2005	thorpej	Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name everything "scsi_*", since we really are talking about the SCSI command set, ATAPI transport not withstanding. Improve the names of many structures, and prepend "SCSI_" onto all SCSI command opcodes. Place items described by the SCSI Primary Commands document into scsi_spc.h.
1.21	29-Jun-2003	fvdl	branches: 1.21.2; 1.21.10; 1.21.12; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.20	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.19	01-Apr-2003	thorpej	Use PAGE_SIZE rather than NBPG.
1.18	31-Mar-2003	perry	For consistency, initialise->initialize. (No it isn't anti-UK bias -- just trying to keep the whole system to one consistent spelling.)
1.17	06-Jan-2003	wiz	interrupt with two rs.
1.16	25-Nov-2002	thorpej	Avoid strict-alias warnings.
1.15	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.14	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.13	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.12	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.11	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.10	01-Jun-2002	lukem	SIMPLEQ rototill: - implement SIMPLEQ_REMOVE(head, elm, type, field). whilst it's O(n), this mirrors the functionality of SLIST_REMOVE() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE() - remove the unnecessary elm arg from SIMPLEQ_REMOVE_HEAD(). this mirrors the functionality of SLIST_REMOVE_HEAD() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE_HEAD() - remove notes about SIMPLEQ not supporting arbitrary element removal - use SIMPLEQ_FOREACH() instead of home-grown for loops - use SIMPLEQ_EMPTY() appropriately - use SIMPLEQ_*() instead of accessing sqh_first,sqh_last,sqe_next directly - reorder manual page; be consistent about how the types are listed - other minor cleanups
1.9	14-Jan-2002	tsutsui	branches: 1.9.8; Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.8	13-Nov-2001	lukem	add RCSID
1.7	03-Aug-2001	ad	branches: 1.7.2; 1.7.4; Remove unnecessary test.
1.6	01-Aug-2001	ad	branches: 1.6.2; Hook in the mly control interface on i386.
1.5	31-Jul-2001	ad	Oops, put back the initalization of "done" in mly_thread().
1.4	30-Jul-2001	ad	Assume that me->lun contains the device number if this event is for a logical device.
1.3	30-Jul-2001	ad	- Forget to report events from logical devices. - Cosmetic changes.
1.2	30-Jul-2001	ad	Don't match boards with a PCI class code of `I2O'.
1.1	30-Jul-2001	ad	Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.6.2.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.6.2.6	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.6.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.6.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.3	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.6.2.2	03-Aug-2001	lukem	update to -current
1.6.2.1	01-Aug-2001	lukem	file mly.c was added on branch kqueue on 2001-08-03 04:13:18 +0000
1.7.4.3	01-Oct-2001	fvdl	Catch up with -current.
1.7.4.2	26-Sep-2001	fvdl	* add a VCLONED vnode flag that indicates a vnode representing a cloned device. * rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass to VOP_REVOKE * the revoke system call will revoke all aliases, as before, but not the clones * vdevgone is called when detaching a device, so make it use REVOKECLONE to get rid of all clones as well * clean up all uses of VOP_OPEN wrt. locking. * add a few VOPS to spec_vnops that need to do something when it's a clone vnode (access and getattr) * add a copy of the vnode vattr structure of the original 'master' vnode to the specinfo of a cloned vnode. could possibly redirect getattr to the 'master' vnode, but this has issues with revoke * add a vdev_reassignvp function that disassociates a vnode from its original device, and reassociates it with the specified dev_t. to be used by cloning devices only, in case a new minor is allocated. * change all direct references in drivers to v_devcookie and v_rdev to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes when debugging race conditions that still exist wrt. locking and revoking vnodes. * make the locking state of a vnode consistent when passed to d_open and d_close (unlocked). locked would be better, but has some deadlock issues
1.7.4.1	07-Sep-2001	thorpej	Commit my "devvp" changes to the thorpej-devvp branch. This replaces the use of dev_t in most places with a struct vnode *. This will form the basic infrastructure for real cloning device support (besides being architecurally cleaner -- it'll be good to get away from using numbers to represent objects).
1.7.2.10	07-Jan-2003	thorpej	Sync with HEAD.
1.7.2.9	11-Dec-2002	thorpej	Sync with HEAD.
1.7.2.8	11-Nov-2002	nathanw	Catch up to -current
1.7.2.7	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.6	17-Sep-2002	nathanw	Catch up to -current.
1.7.2.5	20-Jun-2002	nathanw	Catch up to -current.
1.7.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.7.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.7.2.1	03-Aug-2001	nathanw	file mly.c was added on branch nathanw_sa on 2001-08-24 00:10:13 +0000
1.9.8.2	20-Jun-2002	gehenna	catch up with -current.
1.9.8.1	16-May-2002	gehenna	Add the character device switch.
1.21.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.10.1	29-Apr-2005	kent	sync with -current
1.21.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.21.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.21.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.21.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.23.4.4	27-Oct-2007	yamt	sync with head.
1.23.4.3	03-Sep-2007	yamt	sync with head.
1.23.4.2	30-Dec-2006	yamt	sync with head.
1.23.4.1	21-Jun-2006	yamt	sync with head.
1.24.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.24.10.1	19-Apr-2006	elad	sync with head.
1.24.8.2	03-Sep-2006	yamt	sync with head.
1.24.8.1	24-May-2006	yamt	sync with head.
1.24.6.1	22-Apr-2006	simonb	Sync with head.
1.24.4.1	09-Sep-2006	rpaulo	sync with head
1.29.4.2	10-Dec-2006	yamt	sync with head.
1.29.4.1	22-Oct-2006	yamt	sync with head
1.29.2.2	12-Jan-2007	ad	Sync with head.
1.29.2.1	18-Nov-2006	ad	Sync with head.
1.32.2.1	04-Dec-2006	tron	Pull up following revision(s) (requested by elad in ticket #247): sys/dev/ic/dpt.c: revision 1.55 sys/dev/pci/amr.c: revision 1.43 sys/secmodel/bsd44/secmodel_bsd44_securelevel.c: revision 1.19 sys/dev/pci/mly.c: revision 1.33 share/man/man9/kauth.9: revision 1.37 sys/dev/ic/mlx.c: revision 1.49 sys/dev/ic/icp_ioctl.c: revision 1.14 sys/dev/i2o/iop.c: revision 1.62 sys/dev/pci/twe.c: revision 1.82 sys/sys/kauth.h: revision 1.25 sys/dev/i2o/dpti.c: revision 1.31 sys/kern/kern_auth.c: revision 1.33 sys/dev/tc/stic.c: revision 1.37 Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.33.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.34.4.1	11-Jul-2007	mjf	Sync with head.
1.34.2.4	23-Oct-2007	ad	Sync with head.
1.34.2.3	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.34.2.2	10-Apr-2007	ad	Nuke the deferred kthread creation stuff, as it's no longer needed. Pointed out by thorpej@.
1.34.2.1	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.35.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.35.8.1	06-Nov-2007	matt	sync with HEAD
1.35.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.36.16.3	29-Jun-2008	mjf	Sync with HEAD.
1.36.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.36.16.1	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.37.4.4	11-Mar-2010	yamt	sync with head
1.37.4.3	16-May-2009	yamt	sync with head
1.37.4.2	04-May-2009	yamt	sync with head.
1.37.4.1	16-May-2008	yamt	sync with head.
1.37.2.2	17-Jun-2008	yamt	sync with head.
1.37.2.1	18-May-2008	yamt	sync with head.
1.38.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.39.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.43.4.1	05-Mar-2011	rmind	sync with head
1.44.18.3	03-Dec-2017	jdolecek	update from HEAD
1.44.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.44.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.44.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.44.8.1	30-Oct-2012	yamt	sync with head
1.45.2.1	18-May-2014	rmind	sync with head
1.48.2.1	10-Aug-2014	tls	Rebase.
1.49.4.1	09-Jul-2016	skrll	Sync with HEAD
1.50.18.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.50.18.1	10-Jun-2019	christos	Sync with HEAD
1.50.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.50.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.53.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.54.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.6	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.5	11-Dec-2005	christos	branches: 1.5.70; 1.5.72; 1.5.74; merge ktrace-lwp.
1.4	27-Feb-2005	perry	nuke trailing whitespace
1.3	31-Mar-2003	perry	branches: 1.3.2; 1.3.10; 1.3.12; revert for now -- cancelled seems to be used more often than canceled
1.2	31-Mar-2003	perry	again, for consistency, cancelled->canceled (I fear the flames)
1.1	30-Jul-2001	ad	branches: 1.1.2; 1.1.4; Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	30-Jul-2001	nathanw	file mly_tables.h was added on branch nathanw_sa on 2001-08-24 00:10:13 +0000
1.1.2.2	03-Aug-2001	lukem	update to -current
1.1.2.1	30-Jul-2001	lukem	file mly_tables.h was added on branch kqueue on 2001-08-03 04:13:18 +0000
1.3.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.3.10.1	29-Apr-2005	kent	sync with -current
1.3.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.5.74.1	16-May-2008	yamt	sync with head.
1.5.72.1	18-May-2008	yamt	sync with head.
1.5.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.5	06-Sep-2015	dholland	More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.4	28-Apr-2008	martin	branches: 1.4.44; 1.4.64; Remove clause 3 and 4 from TNF licenses
1.3	11-Dec-2005	christos	branches: 1.3.70; 1.3.72; 1.3.74; merge ktrace-lwp.
1.2	27-Feb-2005	perry	nuke trailing whitespace
1.1	30-Jul-2001	ad	branches: 1.1.2; 1.1.4; 1.1.24; 1.1.32; 1.1.34; Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.1.34.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.32.1	29-Apr-2005	kent	sync with -current
1.1.24.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	30-Jul-2001	nathanw	file mlyio.h was added on branch nathanw_sa on 2001-08-24 00:10:14 +0000
1.1.2.2	03-Aug-2001	lukem	update to -current
1.1.2.1	30-Jul-2001	lukem	file mlyio.h was added on branch kqueue on 2001-08-03 04:13:18 +0000
1.3.74.1	16-May-2008	yamt	sync with head.
1.3.72.1	18-May-2008	yamt	sync with head.
1.3.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.4.64.1	22-Sep-2015	skrll	Sync with HEAD
1.4.44.1	03-Dec-2017	jdolecek	update from HEAD
1.9	10-Feb-2024	andvar	fix various typos in comments and log messages.
1.8	24-Oct-2021	andvar	fix various typos in comments, mainly copypasta.
1.7	28-Dec-2019	gutteridge	Fix typo in comment.
1.6	21-Nov-2017	maxv	branches: 1.6.4; Mmh, surprising bug. It's __packed, not __packed__. Here the structure is not packed for real, but instead a global __packed__ symbol is declared.
1.5	08-Sep-2008	gmcgarry	branches: 1.5.38; 1.5.74; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.4	28-Apr-2008	martin	branches: 1.4.2; 1.4.6; Remove clause 3 and 4 from TNF licenses
1.3	11-Dec-2005	christos	branches: 1.3.70; 1.3.72; 1.3.74; merge ktrace-lwp.
1.2	27-Feb-2005	perry	nuke trailing whitespace
1.1	30-Jul-2001	ad	branches: 1.1.2; 1.1.4; 1.1.24; 1.1.32; 1.1.34; Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.1.34.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.32.1	29-Apr-2005	kent	sync with -current
1.1.24.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	30-Jul-2001	nathanw	file mlyreg.h was added on branch nathanw_sa on 2001-08-24 00:10:14 +0000
1.1.2.2	03-Aug-2001	lukem	update to -current
1.1.2.1	30-Jul-2001	lukem	file mlyreg.h was added on branch kqueue on 2001-08-03 04:13:18 +0000
1.3.74.2	04-May-2009	yamt	sync with head.
1.3.74.1	16-May-2008	yamt	sync with head.
1.3.72.1	18-May-2008	yamt	sync with head.
1.3.70.2	28-Sep-2008	mjf	Sync with HEAD.
1.3.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.4.6.1	19-Oct-2008	haad	Sync with HEAD.
1.4.2.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.5.74.1	27-Nov-2017	martin	Pull up following revision(s) (requested by maxv in ticket #399): sys/dev/pci/mlyreg.h: revision 1.6 Mmh, surprising bug. It's __packed, not __packed__. Here the structure is not packed for real, but instead a global __packed__ symbol is declared.
1.5.38.1	03-Dec-2017	jdolecek	update from HEAD
1.6.4.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.6	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.5	28-Apr-2008	martin	branches: 1.5.34; 1.5.44; Remove clause 3 and 4 from TNF licenses
1.4	09-Jul-2007	ad	branches: 1.4.28; 1.4.30; 1.4.32; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.3	16-Feb-2006	perry	branches: 1.3.24; 1.3.26; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.2	24-Dec-2005	perry	branches: 1.2.2; 1.2.4; 1.2.6; __inline__ -> inline
1.1	30-Jul-2001	ad	branches: 1.1.2; 1.1.4; 1.1.40; Add a driver for Mylex AcceleRAID and eXtremeRAID controllers with v6 firmware. Based off the FreeBSD driver, and re-worked by tls, erh and I.
1.1.40.2	03-Sep-2007	yamt	sync with head.
1.1.40.1	21-Jun-2006	yamt	sync with head.
1.1.4.2	24-Aug-2001	nathanw	Catch up with -current.
1.1.4.1	30-Jul-2001	nathanw	file mlyvar.h was added on branch nathanw_sa on 2001-08-24 00:10:14 +0000
1.1.2.2	03-Aug-2001	lukem	update to -current
1.1.2.1	30-Jul-2001	lukem	file mlyvar.h was added on branch kqueue on 2001-08-03 04:13:18 +0000
1.2.6.1	22-Apr-2006	simonb	Sync with head.
1.2.4.1	09-Sep-2006	rpaulo	sync with head
1.2.2.1	18-Feb-2006	yamt	sync with head.
1.3.26.1	11-Jul-2007	mjf	Sync with head.
1.3.24.1	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.4.32.1	16-May-2008	yamt	sync with head.
1.4.30.1	18-May-2008	yamt	sync with head.
1.4.28.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.44.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.5.44.1	12-Sep-2012	tls	Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for physio (I/O to raw devices); needs more doing to get it going with the filesystems, but it shouldn't damage data. All work's been done on amd64 so far. Not hard to add support to other ports. If others want to pitch in, one very helpful thing would be to sort out when and how IDE disks can do 128K or larger transfers, and adjust the various PCI IDE (or at least ahcisata) drivers and wd.c accordingly -- it would make testing much easier. Another very helpful thing would be to implement a smart minphys() for RAIDframe along the lines detailed in the MAXPHYS-NOTES file.
1.5.34.1	30-Oct-2012	yamt	sync with head
1.31	04-Feb-2024	andvar	s/substract/subtract/ in comments.
1.30	25-Oct-2023	buhrow	Fixes for PR kern/57133: I can now explain why this assert is firing and have a fix for it. It is a regression introduced in R1.22 of mpii.c. If a request comes in and the IOC returns a MPII_SCSIIO_STATUS_CHECK_COND condition, after a successful transfer, or one that is a recovered error, mpii(4) correctly sets the xs->error to XS_SENSE, but incorrectly sets xs->resid to 0 before returning the xfer to the upper scsi layers. Once the upper layers get it, they notice the XS_SENSE check condition and because it's a retryable error, they increment xs_requeuecnt, set ERESTART and send the xfer request down to the mpii(4) layer again for a retry. What they do not do is reset xs->resid equal to xs->datalen. When the xfer comes down to mpii(4) again, the assert happens. The fix is for the mpii(4) driver to leave xs->resid alone when it encounters a MPII_SCSIIO_STATUS_CHECK_COND condition. This bug affects NetBSD-10, netbsd-9 and netbsd-8.
1.29	07-Aug-2021	thorpej	branches: 1.29.6; Merge thorpej-cfargs2.
1.28	24-Apr-2021	thorpej	branches: 1.28.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.27	23-Feb-2021	skrll	branches: 1.27.2; Whitespace (mostly trailing)
1.26	23-Feb-2021	skrll	Use the PCI_MAPREG_TYPE macro. No binary change.
1.25	08-Aug-2020	jnemeth	branches: 1.25.2; make this compile without bio(4)
1.24	28-Nov-2019	maxv	localify
1.23	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.22	11-Mar-2019	kardel	branches: 1.22.4; PR/54045 fix mpii to adhere to physio diagnostic invariant that fully processed data must not post an error: 1) verify expected scspi state via KASSERT() instead of just setting the variables. 2) set xs->resid only in known good conditions 3) insure setting errors in all error paths and refrain from clearing xs->resid in error paths. While there do some cosmectic clean up: 1) extend and relocate some debug output 2) mpii HBAs can also manage non-disk devices like tapes etc, so log that physical "devices" instead of physical "disks" are attached or detached. Tested with NEOSeries FlexStor II and luckily a broken tape drive 8-( mpii0 at pci1 dev 0 function 0: vendor 1000 product 00ab (rev. 0x01) mpii0: interrupting at irq 11 mpii0: HBA 9400-8i8e, firmware 3.0.4.0, MPI 2.6 mpii0: physical device inserted in slot 9 mpii0: physical device inserted in slot 13 mpii0: physical device inserted in slot 16 st0 at scsibus0 target 9 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st0: density code 92, variable blocks, write-enabled ch0 at scsibus0 target 9 lun 1: <BDT, FlexStor II, 5.50> changer removable ch0: 23 slots, 2 drives, 1 picker, 1 portal st0: tagged queueing ch0: tagged queueing st1 at scsibus0 target 13 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st1: density code 92, variable blocks, write-enabled st1: tagged queueing ses0 at scsibus0 target 16 lun 0: <LSI, VirtualSES, 01> enclosure services fixed Note: pullup-8
1.21	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.20	27-Jan-2019	pgoyette	Merge the [pgoyette-compat] branch
1.19	28-Dec-2018	kardel	fix LUN handling (byte order issue) now ch0 on a NEOSeries FlexStor II is detected again on LUN 1 this is a regression introduced in 1.15 update and 1.8.10.3 pullup Tested with NEOSeries FlexStor II: mpii0 at pci1 dev 0 function 0: vendor 1000 product 00ab (rev. 0x01) mpii0: interrupting at irq 11 mpii0: HBA 9400-8i8e, firmware 3.0.4.0, MPI 2.6 mpii0: physical disk inserted in slot 9 mpii0: physical disk inserted in slot 13 mpii0: physical disk inserted in slot 16 st0 at scsibus0 target 9 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st0: drive empty st0: tagged queueing st1 at scsibus0 target 13 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st1: drive empty ch0 at scsibus0 target 13 lun 1: <BDT, FlexStor II, 5.50> changer removable ch0: 23 slots, 2 drives, 1 picker, 1 portal st1: tagged queueing ch0: tagged queueing ses0 at scsibus0 target 16 lun 0: <LSI, VirtualSES, 01> enclosure services fixed ses0: SCSI-3 SES Device
1.18	05-Dec-2018	bouyer	There's no sensors for physical disks so don't try to detach them.
1.17	04-Dec-2018	bouyer	NULL-terminate mpii_devices[], this is the end condition of the loop in mpii_match(). Noticed by Mike Pumford
1.16	03-Dec-2018	bouyer	Also tell the scsi layer that we are MPSAFE
1.15	03-Dec-2018	bouyer	Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0
1.14	02-Dec-2018	jdolecek	use pci_intr_establish_xname()
1.13	24-Nov-2018	bouyer	Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change.
1.12	14-Oct-2018	jdolecek	remove M_CANFAIL flag for malloc(9) - it was completely ignored, so had actually no effect
1.11	15-Jan-2018	maya	branches: 1.11.2; 1.11.4; indent to magic logic
1.10	09-Oct-2017	maya	Reorder to avoid null deref before null test
1.9	25-Jun-2017	christos	PR/52330: Jia-Ju Bai: mpii driver: a sleep-in-interrupt bug in mpii_intr Since the enclosing routime mpii_event_raid already calls malloc with M_NOWAIT, fix the cache routine to do the same. While there check the result of the cache routine and change some error prints to aprint.
1.8	02-May-2016	christos	branches: 1.8.10; move scsipi_strvis -> libkern:strnvisx() change the prototype to match userland fix sizes of strings passed to it
1.7	08-Feb-2016	christos	PR/50785: David Binderman: Remove dead code.
1.6	12-Mar-2015	christos	Dedup the conversion of bioc_disk and bioc_vol to envsys_data_t
1.5	29-Mar-2014	christos	branches: 1.5.4; 1.5.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.4	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.3	09-Aug-2013	kardel	fix issues when reading variable block sized tapes. symptoms: generic HBA error on console when reading with a larger blocksize. blocks read are padded to requested block size with a 5a... a5... pattern. problems fixed: - controller scsi_status values did not match the ones used by the spsipi layer. a mapping function was introduced. - when experiencing an underrun (read 64k and get a 63k block) the controller posted not a SUCCESS status but CHECK status. handle that like SUCCESS adjusting xs->resid and set XS_SENSE. now the correct data amount is returned and nothing is 'added' and no 'generic HBA error' occurs. - make decisions using variables and constants from the controller domain.
1.2	08-Aug-2013	kardel	Allow 8 luns instead of 1. This enables access to the changer device on a Dell PV-124T: ch0 at scsibus0 target 9 lun 1: <DELL, PV-124T, 0086> changer removable ch0: 16 slots, 1 drive, 1 picker, 0 portals
1.1	19-Apr-2012	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; Add mpii(4), a driver for LSI Logic Fusion-MPT Message Passing Interface II SAS controllers. Ported from OpenBSD.
1.1.10.2	18-May-2014	rmind	sync with head
1.1.10.1	28-Aug-2013	rmind	sync with head
1.1.8.2	03-Dec-2017	jdolecek	update from HEAD
1.1.8.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.6.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.2	23-May-2012	yamt	sync with head.
1.1.6.1	19-Apr-2012	yamt	file mpii.c was added on branch yamt-pagecache on 2012-05-23 10:07:58 +0000
1.1.4.2	29-Apr-2012	mrg	sync to latest -current.
1.1.4.1	19-Apr-2012	mrg	file mpii.c was added on branch jmcneill-usbmp on 2012-04-29 23:04:51 +0000
1.1.2.3	07-Sep-2013	bouyer	Pull up following revision(s) (requested by kardel in ticket #932): sys/dev/pci/mpii.c: revision 1.2 sys/dev/pci/mpii.c: revision 1.3 Allow 8 luns instead of 1. This enables access to the changer device on a Dell PV-124T: ch0 at scsibus0 target 9 lun 1: <DELL, PV-124T, 0086> changer removable ch0: 16 slots, 1 drive, 1 picker, 0 portals fix issues when reading variable block sized tapes. symptoms: generic HBA error on console when reading with a larger blocksize. blocks read are padded to requested block size with a 5a... a5... pattern. problems fixed: - controller scsi_status values did not match the ones used by the spsipi layer. a mapping function was introduced. - when experiencing an underrun (read 64k and get a 63k block) the controller posted not a SUCCESS status but CHECK status. handle that like SUCCESS adjusting xs->resid and set XS_SENSE. now the correct data amount is returned and nothing is 'added' and no 'generic HBA error' occurs. - make decisions using variables and constants from the controller domain.
1.1.2.2	23-Apr-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #193): sys/arch/i386/conf/GENERIC: revision 1.1072 sys/dev/pci/mpii.c: revision 1.1 sys/arch/i386/conf/XEN3_DOM0: revision 1.66 sys/dev/pci/files.pci: revision 1.357 share/man/man4/Makefile: revision 1.584 distrib/sets/lists/man/mi: revision 1.1387 share/man/man4/mpii.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.354 sys/arch/amd64/conf/XEN3_DOM0: revision 1.83 sys/arch/i386/conf/ALL: revision 1.337 Add mpii(4), a driver for LSI Logic Fusion-MPT Message Passing Interface II SAS controllers. Ported from OpenBSD.
1.1.2.1	19-Apr-2012	riz	file mpii.c was added on branch netbsd-6 on 2012-04-23 16:31:36 +0000
1.5.6.4	28-Aug-2017	skrll	Sync with HEAD
1.5.6.3	29-May-2016	skrll	Sync with HEAD
1.5.6.2	19-Mar-2016	skrll	Sync with HEAD
1.5.6.1	06-Apr-2015	skrll	Sync with HEAD
1.5.4.1	06-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #1614): sys/dev/pci/mpii.c: revision 1.9 PR/52330: Jia-Ju Bai: mpii driver: a sleep-in-interrupt bug in mpii_intr Since the enclosing routime mpii_event_raid already calls malloc with M_NOWAIT, fix the cache routine to do the same. While there check the result of the cache routine and change some error prints to aprint.
1.8.10.8	26-Oct-2023	martin	Pull up following revision(s) (requested by buhrow in ticket #1916): sys/dev/pci/mpii.c: revision 1.30 Fixes for PR kern/57133: I can now explain why this assert is firing and have a fix for it. It is a regression introduced in R1.22 of mpii.c. If a request comes in and the IOC returns a MPII_SCSIIO_STATUS_CHECK_COND condition, after a successful transfer, or one that is a recovered error, mpii(4) correctly sets the xs->error to XS_SENSE, but incorrectly sets xs->resid to 0 before returning the xfer to the upper scsi layers. Once the upper layers get it, they notice the XS_SENSE check condition and because it's a retryable error, they increment xs_requeuecnt, set ERESTART and send the xfer request down to the mpii(4) layer again for a retry. What they do not do is reset xs->resid equal to xs->datalen. When the xfer comes down to mpii(4) again, the assert happens. The fix is for the mpii(4) driver to leave xs->resid alone when it encounters a MPII_SCSIIO_STATUS_CHECK_COND condition. This bug affects NetBSD-10, netbsd-9 and netbsd-8.
1.8.10.7	29-Sep-2022	snj	Apply patch (requested by bouyer in ticket #1767): Convert from pci_intr_map() to pci_intr_alloc(); makes the driver use MSI/MSI-x when available.
1.8.10.6	09-Aug-2020	martin	Pull up following revision(s) (requested by jnemeth in ticket #1596): sys/dev/pci/mpii.c: revision 1.25 make this compile without bio(4)
1.8.10.5	15-Mar-2019	martin	Pull up following revision(s) (requested by kardel in ticket #1217): sys/dev/pci/mpii.c: revision 1.22 PR/54045 fix mpii to adhere to physio diagnostic invariant that fully processed data must not post an error: 1) verify expected scspi state via KASSERT() instead of just setting the variables. 2) set xs->resid only in known good conditions 3) insure setting errors in all error paths and refrain from clearing xs->resid in error paths. While there do some cosmectic clean up: 1) extend and relocate some debug output 2) mpii HBAs can also manage non-disk devices like tapes etc, so log that physical "devices" instead of physical "disks" are attached or detached. Tested with NEOSeries FlexStor II and luckily a broken tape drive 8-( mpii0 at pci1 dev 0 function 0: vendor 1000 product 00ab (rev. 0x01) mpii0: interrupting at irq 11 mpii0: HBA 9400-8i8e, firmware 3.0.4.0, MPI 2.6 mpii0: physical device inserted in slot 9 mpii0: physical device inserted in slot 13 mpii0: physical device inserted in slot 16 st0 at scsibus0 target 9 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st0: density code 92, variable blocks, write-enabled ch0 at scsibus0 target 9 lun 1: <BDT, FlexStor II, 5.50> changer removable ch0: 23 slots, 2 drives, 1 picker, 1 portal st0: tagged queueing ch0: tagged queueing st1 at scsibus0 target 13 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st1: density code 92, variable blocks, write-enabled st1: tagged queueing ses0 at scsibus0 target 16 lun 0: <LSI, VirtualSES, 01> enclosure services fixed Note: pullup-8
1.8.10.4	07-Jan-2019	martin	Pull up following revision(s) (requested by kardel in ticket #1157): sys/dev/pci/mpii.c: revision 1.19 fix LUN handling (byte order issue) now ch0 on a NEOSeries FlexStor II is detected again on LUN 1 this is a regression introduced in 1.15 update and 1.8.10.3 pullup Tested with NEOSeries FlexStor II: mpii0 at pci1 dev 0 function 0: vendor 1000 product 00ab (rev. 0x01) mpii0: interrupting at irq 11 mpii0: HBA 9400-8i8e, firmware 3.0.4.0, MPI 2.6 mpii0: physical disk inserted in slot 9 mpii0: physical disk inserted in slot 13 mpii0: physical disk inserted in slot 16 st0 at scsibus0 target 9 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st0: drive empty st0: tagged queueing st1 at scsibus0 target 13 lun 0: <IBM, ULTRIUM-HH7, J4D1> tape removable st1: drive empty ch0 at scsibus0 target 13 lun 1: <BDT, FlexStor II, 5.50> changer removable ch0: 23 slots, 2 drives, 1 picker, 1 portal st1: tagged queueing ch0: tagged queueing ses0 at scsibus0 target 16 lun 0: <LSI, VirtualSES, 01> enclosure services fixed ses0: SCSI-3 SES Device
1.8.10.3	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1131): sys/dev/pci/mpiireg.h: revision 1.2 sys/dev/pci/mfii.c: revision 1.3 share/man/man4/mpii.4: revision 1.3 share/man/man4/mpii.4: revision 1.4 sys/dev/pci/mpii.c: revision 1.15 sys/dev/pci/mpii.c: revision 1.16 sys/dev/pci/mpii.c: revision 1.17 sys/dev/pci/mpii.c: revision 1.18 Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0 Update for the new mpii(4) driver, mostly from OpenBSD. Also tell the scsi layer that we are MPSAFE Remove trailing whitespace NULL-terminate mpii_devices[], this is the end condition of the loop in mpii_match(). Noticed by Mike Pumford There's no sensors for physical disks so don't try to detach them.
1.8.10.2	07-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1129): sys/arch/amd64/conf/XEN3_DOM0: revision 1.159 sys/arch/amd64/conf/GENERIC: revision 1.508 sys/arch/i386/conf/ALL: revision 1.458 sys/dev/ic/mfi.c: revision 1.60 sys/dev/pci/mpiireg.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.1628 sys/dev/pci/mfii.c: revision 1.1,1.2 (adapted) sys/arch/i386/conf/GENERIC: revision 1.1194 sys/dev/pci/mpii.c: revision 1.13 sys/dev/ic/mfireg.h: revision 1.9 share/man/man4/mfii.4: revision 1.1 share/man/man4/Makefile: revision 1.673 (patch) sys/dev/pci/files.pci: revision 1.410 share/man/man4/mfii.4: revision 1.2 sys/arch/amd64/conf/ALL: revision 1.108 sys/arch/i386/conf/XEN3PAE_DOM0: revision 1.8 (patch, in XEN3_DOM0) Add some definitions from OpenBSD, needed by the upcoming mfii driver. No functionnal change. - Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change. - Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6. - Add my name in copyright list - add a man page for the new mpii, mostly from OpenBSD. - Sort SEE ALSO. Fix date. Whitespace fixes.
1.8.10.1	09-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #857): sys/dev/pci/mpii.c: revision 1.9 PR/52330: Jia-Ju Bai: mpii driver: a sleep-in-interrupt bug in mpii_intr Since the enclosing routime mpii_event_raid already calls malloc with M_NOWAIT, fix the cache routine to do the same. While there check the result of the cache routine and change some error prints to aprint.
1.11.4.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.11.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.11.4.1	10-Jun-2019	christos	Sync with HEAD
1.11.2.4	18-Jan-2019	pgoyette	Synch with HEAD
1.11.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.11.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.11.2.1	20-Oct-2018	pgoyette	Sync with head
1.22.4.2	26-Oct-2023	martin	Pull up following revision(s) (requested by buhrow in ticket #1756): sys/dev/pci/mpii.c: revision 1.30 Fixes for PR kern/57133: I can now explain why this assert is firing and have a fix for it. It is a regression introduced in R1.22 of mpii.c. If a request comes in and the IOC returns a MPII_SCSIIO_STATUS_CHECK_COND condition, after a successful transfer, or one that is a recovered error, mpii(4) correctly sets the xs->error to XS_SENSE, but incorrectly sets xs->resid to 0 before returning the xfer to the upper scsi layers. Once the upper layers get it, they notice the XS_SENSE check condition and because it's a retryable error, they increment xs_requeuecnt, set ERESTART and send the xfer request down to the mpii(4) layer again for a retry. What they do not do is reset xs->resid equal to xs->datalen. When the xfer comes down to mpii(4) again, the assert happens. The fix is for the mpii(4) driver to leave xs->resid alone when it encounters a MPII_SCSIIO_STATUS_CHECK_COND condition. This bug affects NetBSD-10, netbsd-9 and netbsd-8.
1.22.4.1	09-Aug-2020	martin	Pull up following revision(s) (requested by jnemeth in ticket #1044): sys/dev/pci/mpii.c: revision 1.25 make this compile without bio(4)
1.25.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.27.2.3	28-Mar-2021	thorpej	No need to pass interface attribute or locators to config_found().
1.27.2.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.27.2.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.28.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.29.6.1	26-Oct-2023	martin	Pull up following revision(s) (requested by buhrow in ticket #435): sys/dev/pci/mpii.c: revision 1.30 Fixes for PR kern/57133: I can now explain why this assert is firing and have a fix for it. It is a regression introduced in R1.22 of mpii.c. If a request comes in and the IOC returns a MPII_SCSIIO_STATUS_CHECK_COND condition, after a successful transfer, or one that is a recovered error, mpii(4) correctly sets the xs->error to XS_SENSE, but incorrectly sets xs->resid to 0 before returning the xfer to the upper scsi layers. Once the upper layers get it, they notice the XS_SENSE check condition and because it's a retryable error, they increment xs_requeuecnt, set ERESTART and send the xfer request down to the mpii(4) layer again for a retry. What they do not do is reset xs->resid equal to xs->datalen. When the xfer comes down to mpii(4) again, the assert happens. The fix is for the mpii(4) driver to leave xs->resid alone when it encounters a MPII_SCSIIO_STATUS_CHECK_COND condition. This bug affects NetBSD-10, netbsd-9 and netbsd-8.
1.2	03-Dec-2018	bouyer	branches: 1.2.2; 1.2.6; Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0
1.1	24-Nov-2018	bouyer	branches: 1.1.2; Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change.
1.1.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.1.2.1	24-Nov-2018	pgoyette	file mpiireg.h was added on branch pgoyette-compat on 2018-11-26 01:52:32 +0000
1.2.6.2	10-Jun-2019	christos	Sync with HEAD
1.2.6.1	03-Dec-2018	christos	file mpiireg.h was added on branch phil-wifi on 2019-06-10 22:07:17 +0000
1.2.2.3	08-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1131): sys/dev/pci/mpiireg.h: revision 1.2 sys/dev/pci/mfii.c: revision 1.3 share/man/man4/mpii.4: revision 1.3 share/man/man4/mpii.4: revision 1.4 sys/dev/pci/mpii.c: revision 1.15 sys/dev/pci/mpii.c: revision 1.16 sys/dev/pci/mpii.c: revision 1.17 sys/dev/pci/mpii.c: revision 1.18 Update the mpii(4) driver to the latest OpenBSD version. This adds support for the SAS3xxx LSI controllers, and this also makes the driver MP-safe. adjust mfii.c for changes in mpiireg.h Tested with a mpii0: SMC2008-IR, firmware 9.0.0.0 IR, MPI 2.0 Update for the new mpii(4) driver, mostly from OpenBSD. Also tell the scsi layer that we are MPSAFE Remove trailing whitespace NULL-terminate mpii_devices[], this is the end condition of the loop in mpii_match(). Noticed by Mike Pumford There's no sensors for physical disks so don't try to detach them.
1.2.2.2	07-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1129): sys/arch/amd64/conf/XEN3_DOM0: revision 1.159 sys/arch/amd64/conf/GENERIC: revision 1.508 sys/arch/i386/conf/ALL: revision 1.458 sys/dev/ic/mfi.c: revision 1.60 sys/dev/pci/mpiireg.h: revision 1.1 distrib/sets/lists/man/mi: revision 1.1628 sys/dev/pci/mfii.c: revision 1.1,1.2 (adapted) sys/arch/i386/conf/GENERIC: revision 1.1194 sys/dev/pci/mpii.c: revision 1.13 sys/dev/ic/mfireg.h: revision 1.9 share/man/man4/mfii.4: revision 1.1 share/man/man4/Makefile: revision 1.673 (patch) sys/dev/pci/files.pci: revision 1.410 share/man/man4/mfii.4: revision 1.2 sys/arch/amd64/conf/ALL: revision 1.108 sys/arch/i386/conf/XEN3PAE_DOM0: revision 1.8 (patch, in XEN3_DOM0) Add some definitions from OpenBSD, needed by the upcoming mfii driver. No functionnal change. - Move registers definitions to a separate file, needed for the upcomning mpii driver. No functionnal change. - Add mpii(4), a driver for LSI Megaraid Fusion controllers. Ported from OpenBSD. This driver is MP-safe. Note that the earlier fusion controllers (Megaraid 2208, codenamed Thunderbold) are also supported by mfi(4). mpii will take precedence if both drivers are enabled. Tested on a mfii0 at pci6 dev 0 function 0: "PERC H740P Adapter ", firmware 50.3.0-1512, 819 2MB cache mfii0: interrupting at ioapic2 pin 2 scsibus0 at mfii0: 64 targets, 8 luns per target scsibus0: waiting 2 seconds for devices to settle... sd0 at scsibus0 target 0 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd0: fabricating a geometry sd0: 99 GB, 102399 cyl, 64 head, 32 sec, 512 bytes/sect x 209714688 sectors sd0: tagged queueing sd1 at scsibus0 target 1 lun 0: <DELL, PERC H740P Adp, 5.03> disk fixed sd1: fabricating a geometry sd1: 22254 GB, 22788608 cyl, 64 head, 32 sec, 512 bytes/sect x 46671069696 sectors sd1: fabricating a geometry It supports bioctl(8) ioctls, as well as sensors for the BBU and logical drives. Sponsored by LIP6. - Add my name in copyright list - add a man page for the new mpii, mostly from OpenBSD. - Sort SEE ALSO. Fix date. Whitespace fixes.
1.2.2.1	03-Dec-2018	martin	file mpiireg.h was added on branch netbsd-8 on 2018-12-07 17:11:37 +0000
1.26	02-Dec-2018	jdolecek	use pci_intr_establish_xname()
1.25	14-Jul-2016	msaitoh	branches: 1.25.16; 1.25.18; KNF. No functional change.
1.24	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.23	29-Mar-2014	christos	branches: 1.23.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.22	18-Mar-2012	martin	branches: 1.22.2; 1.22.4; Split device_t
1.21	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.20	24-Nov-2009	njoly	branches: 1.20.12; 1.20.16; Do not use aprint_normal for error messages.
1.19	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.18	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.17	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.16	10-Apr-2008	cegger	branches: 1.16.4; 1.16.18; use aprint_*_dev and device_xname
1.15	25-Feb-2008	xtraeme	Do not attach on Symbios SAS1078, doesn't work and it will be matched soon by the mfi(4) driver.
1.14	04-Feb-2008	joerg	branches: 1.14.2; 1.14.6; Fix quirk to clean only PCI register 0x6a and 0x6b, which is likely what the original code intended. This avoids the unaligned access. According to jmmv@'s testing, this seems to work.
1.13	03-Jan-2008	tron	Fix perfomance problems for old revisions of the Symbios 53c1030. Patch supplied by Patrick Welche in PR kern/30531.
1.12	17-Sep-2007	msaitoh	branches: 1.12.6; 1.12.12; add SAS1068_2, SAS1064E_2 and SAS1068E_2
1.11	27-Jul-2007	tron	branches: 1.11.4; 1.11.6; 1.11.8; Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.10	16-Nov-2006	christos	branches: 1.10.2; 1.10.8; 1.10.16; 1.10.18; __unused removal on arguments; approved by core.
1.9	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.8	29-Mar-2006	thorpej	branches: 1.8.8; 1.8.10; mpt_pci_link_peer(): Rather than testing if the candidate device is attached with mpt_pci (by comparing the dv_cfattach pointer), check to see if it has the same parent. If so, then it is guaranteed to have the same attachment (which is mpt_pci).
1.7	28-Mar-2006	thorpej	Use device_unit().
1.6	11-Dec-2005	christos	branches: 1.6.4; 1.6.6; 1.6.8; 1.6.10; 1.6.12; merge ktrace-lwp.
1.5	27-Feb-2005	perry	branches: 1.5.2; 1.5.4; 1.5.12; 1.5.14; nuke trailing whitespace
1.4	06-Jan-2005	he	branches: 1.4.2; 1.4.4; Add recognition of LSI Logic FC919X chips, as found on LSI Logic LSI7102XP-LC controllers. Document FC919X and FC929X with separate entries in the man page.
1.3	18-Sep-2004	martin	Add support for PCI-X version of LSI FC929; from Bert Kiers in PR kern/26836.
1.2	14-Jul-2003	lukem	branches: 1.2.2; add missing __KERNEL_RCSID()
1.1	16-Apr-2003	thorpej	branches: 1.1.2; Device driver for the LSI Logic Fusion-MPT based SCSI and Fibre Channel adapters. Currently supports: * LSI 53c1030 Ultra320 SCSI * LSI FC909, FC909A, FC919, and FC929 Fibre Channel Ported from the FreeBSD "mpt" driver, written by Greg Ansley. Thanks to Frank van der Linden for testing and some bug finding. This work was sponsored by Wasabi Systems, Inc.
1.1.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.1.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.1	03-Aug-2004	skrll	Sync with HEAD
1.2.2.1	18-Sep-2004	he	Pull up revision 1.3 (requested by martin in ticket #856): Add support for PCI-X version of LSI FC929, from PR#26836.
1.4.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.2.1	29-Apr-2005	kent	sync with -current
1.5.14.1	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #1822): sys/dev/ic/mpt_netbsd.c: revision 1.12 via patch sys/dev/ic/mpt_netbsd.h: revision 1.6 via patch sys/dev/ic/mpt_mpilib.h: revision 1.3 via patch sys/dev/ic/mpt.h: revision 1.6 via patch sys/dev/ic/mpt.c: revisions 1.9, 1.10 via patch sys/dev/ic/mpt_debug.c: patch sys/dev/pci/mpt_pci.c: revision 1.11 via patch sys/dev/pci/pcidevs: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/pcidevs_data.h: patch Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.5.12.1	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #1822): sys/dev/ic/mpt_netbsd.c: revision 1.12 via patch sys/dev/ic/mpt_netbsd.h: revision 1.6 via patch sys/dev/ic/mpt_mpilib.h: revision 1.3 via patch sys/dev/ic/mpt.h: revision 1.6 via patch sys/dev/ic/mpt.c: revisions 1.9, 1.10 via patch sys/dev/ic/mpt_debug.c: patch sys/dev/pci/mpt_pci.c: revision 1.11 via patch sys/dev/pci/pcidevs: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/pcidevs_data.h: patch Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.5.4.7	27-Feb-2008	yamt	sync with head.
1.5.4.6	11-Feb-2008	yamt	sync with head.
1.5.4.5	21-Jan-2008	yamt	sync with head
1.5.4.4	27-Oct-2007	yamt	sync with head.
1.5.4.3	03-Sep-2007	yamt	sync with head.
1.5.4.2	30-Dec-2006	yamt	sync with head.
1.5.4.1	21-Jun-2006	yamt	sync with head.
1.5.2.1	13-Sep-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #1822): sys/dev/ic/mpt_netbsd.c: revision 1.12 via patch sys/dev/ic/mpt_netbsd.h: revision 1.6 via patch sys/dev/ic/mpt_mpilib.h: revision 1.3 via patch sys/dev/ic/mpt.h: revision 1.6 via patch sys/dev/ic/mpt.c: revisions 1.9, 1.10 via patch sys/dev/ic/mpt_debug.c: patch sys/dev/pci/mpt_pci.c: revision 1.11 via patch sys/dev/pci/pcidevs: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/pcidevs_data.h: patch Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.6.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.6.10.1	19-Apr-2006	elad	sync with head.
1.6.8.1	01-Apr-2006	yamt	sync with head.
1.6.6.1	22-Apr-2006	simonb	Sync with head.
1.6.4.1	09-Sep-2006	rpaulo	sync with head
1.8.10.2	10-Dec-2006	yamt	sync with head.
1.8.10.1	22-Oct-2006	yamt	sync with head
1.8.8.1	18-Nov-2006	ad	Sync with head.
1.10.18.1	15-Aug-2007	skrll	Sync with HEAD.
1.10.16.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.10.8.2	09-Oct-2007	ad	Sync with head.
1.10.8.1	20-Aug-2007	ad	Sync with HEAD.
1.10.2.2	29-Sep-2008	jdc	Pull up revision 1.12 (requested by bouyer in ticket #1192). add SAS1068_2, SAS1064E_2 and SAS1068E_2
1.10.2.1	30-Jul-2007	liamjfoy	Pull up following revision(s) (requested by tron in ticket #797): sys/dev/ic/mpt_netbsd.c: revision 1.12 sys/dev/ic/mpt_mpilib.h: revision 1.3 sys/dev/ic/mpt.h: revision 1.6 sys/dev/ic/mpt.c: revision 1.9 sys/dev/ic/mpt_netbsd.h: revision 1.6 sys/dev/pci/mpt_pci.c: revision 1.11 doc/CHANGES: revision 1.893 Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG. Mention support for new hardware in mpt(4).
1.11.8.2	27-Jul-2007	tron	Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.11.8.1	27-Jul-2007	tron	file mpt_pci.c was added on branch matt-mips64 on 2007-07-27 13:06:52 +0000
1.11.6.3	23-Mar-2008	matt	sync with HEAD
1.11.6.2	09-Jan-2008	matt	sync with HEAD
1.11.6.1	06-Nov-2007	matt	sync with HEAD
1.11.4.1	02-Oct-2007	joerg	Sync with HEAD.
1.12.12.1	08-Jan-2008	bouyer	Sync with HEAD
1.12.6.1	18-Feb-2008	mjf	Sync with HEAD.
1.14.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.14.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.14.2.1	24-Mar-2008	keiichi	sync with head.
1.16.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.4.2	11-Mar-2010	yamt	sync with head
1.16.4.1	16-May-2009	yamt	sync with head
1.20.16.2	05-Apr-2012	mrg	sync to latest -current.
1.20.16.1	18-Feb-2012	mrg	merge to -current.
1.20.12.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.20.12.1	17-Apr-2012	yamt	sync with head
1.22.4.1	18-May-2014	rmind	sync with head
1.22.2.3	03-Dec-2017	jdolecek	update from HEAD
1.22.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.22.2.1	12-Sep-2012	tls	Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for physio (I/O to raw devices); needs more doing to get it going with the filesystems, but it shouldn't damage data. All work's been done on amd64 so far. Not hard to add support to other ports. If others want to pitch in, one very helpful thing would be to sort out when and how IDE disks can do 128K or larger transfers, and adjust the various PCI IDE (or at least ahcisata) drivers and wd.c accordingly -- it would make testing much easier. Another very helpful thing would be to implement a smart minphys() for RAIDframe along the lines detailed in the MAXPHYS-NOTES file.
1.23.6.2	05-Oct-2016	skrll	Sync with HEAD
1.23.6.1	09-Jul-2016	skrll	Sync with HEAD
1.25.18.1	10-Jun-2019	christos	Sync with HEAD
1.25.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.17	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	23-Nov-2011	jmcneill	branches: 1.16.50; 1.16.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.15	28-Apr-2008	martin	branches: 1.15.12; 1.15.34; 1.15.36; Remove clause 3 and 4 from TNF licenses
1.14	27-Mar-2008	xtraeme	branches: 1.14.2; 1.14.4; Split device_t/softc for mpu(4) and its attachments, plus other related cosmetic changes. Reviewed by cube@.
1.13	19-Oct-2007	ad	branches: 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	04-Feb-2005	perry	de-__P
1.7	02-Dec-2004	xtraeme	branches: 1.7.4; 1.7.6; Add ACPI mpu(4) attachment, based on lpt_acpi.c by jmcneill@, reviewed by christos@, cube@ and martin@. While here, remove a printf() in mpu_attach() (ic/mpu.c) to remove an empty line in some frontends (mpu_isa.c, etc).
1.6	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	02-Oct-2002	thorpej	branches: 1.5.6; Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	13-Nov-2001	lukem	add RCSID
1.1	04-Sep-2001	itohy	branches: 1.1.4; 1.1.6; MPU-401 compatible MIDI UART driver for C-Media CMI8x38
1.1.6.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.6.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.6.1	04-Sep-2001	nathanw	file mpu_cmpci.c was added on branch nathanw_sa on 2001-09-21 22:35:57 +0000
1.1.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.4.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.4.1	04-Sep-2001	thorpej	file mpu_cmpci.c was added on branch kqueue on 2001-09-13 01:15:55 +0000
1.5.6.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.6.5	04-Feb-2005	skrll	Sync with HEAD.
1.5.6.4	18-Dec-2004	skrll	Sync with HEAD.
1.5.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.6.1	03-Aug-2004	skrll	Sync with HEAD
1.7.6.1	12-Feb-2005	yamt	sync with head.
1.7.4.1	29-Apr-2005	kent	sync with -current
1.9.2.1	27-Oct-2007	yamt	sync with head.
1.10.22.2	10-Dec-2006	yamt	sync with head.
1.10.22.1	22-Oct-2006	yamt	sync with head
1.10.20.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.15.34.1	17-Apr-2012	yamt	sync with head
1.15.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.16.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.16.50.1	10-Jun-2019	christos	Sync with HEAD
1.18	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.17	23-Nov-2011	jmcneill	branches: 1.17.50; 1.17.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.16	28-Apr-2008	martin	branches: 1.16.12; 1.16.34; 1.16.36; Remove clause 3 and 4 from TNF licenses
1.15	27-Mar-2008	xtraeme	branches: 1.15.2; 1.15.4; Split device_t/softc for mpu(4) and its attachments, plus other related cosmetic changes. Reviewed by cube@.
1.14	19-Oct-2007	ad	branches: 1.14.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	28-Jun-2005	thorpej	branches: 1.10.2; Use ANSI function decls and static.
1.9	04-Feb-2005	perry	de-__P
1.8	02-Dec-2004	xtraeme	branches: 1.8.4; 1.8.6; Add ACPI mpu(4) attachment, based on lpt_acpi.c by jmcneill@, reviewed by christos@, cube@ and martin@. While here, remove a printf() in mpu_attach() (ic/mpu.c) to remove an empty line in some frontends (mpu_isa.c, etc).
1.7	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	02-Oct-2002	thorpej	branches: 1.6.6; Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	13-Nov-2001	lukem	add RCSID
1.2	14-Aug-1999	kleink	branches: 1.2.14; 1.2.16; Add missing MPU interrupt glue.
1.1	03-Aug-1999	augustss	Add the file I forgot at the previous MPU changes.
1.2.16.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.16.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.14.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.14.1	14-Nov-2001	nathanw	Catch up to -current.
1.6.6.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.6.5	04-Feb-2005	skrll	Sync with HEAD.
1.6.6.4	18-Dec-2004	skrll	Sync with HEAD.
1.6.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.6.1	03-Aug-2004	skrll	Sync with HEAD
1.8.6.1	12-Feb-2005	yamt	sync with head.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.2.1	27-Oct-2007	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.14.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.14.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.16.34.1	17-Apr-2012	yamt	sync with head
1.16.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.17.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.17.50.1	10-Jun-2019	christos	Sync with HEAD
1.18	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.17	23-Nov-2011	jmcneill	branches: 1.17.50; 1.17.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.16	28-Apr-2008	martin	branches: 1.16.12; 1.16.34; 1.16.36; Remove clause 3 and 4 from TNF licenses
1.15	27-Mar-2008	xtraeme	branches: 1.15.2; 1.15.4; Split device_t/softc for mpu(4) and its attachments, plus other related cosmetic changes. Reviewed by cube@.
1.14	19-Oct-2007	ad	branches: 1.14.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	28-Jun-2005	thorpej	branches: 1.10.2; Use ANSI function decls and static.
1.9	04-Feb-2005	perry	de-__P
1.8	02-Dec-2004	xtraeme	branches: 1.8.4; 1.8.6; Add ACPI mpu(4) attachment, based on lpt_acpi.c by jmcneill@, reviewed by christos@, cube@ and martin@. While here, remove a printf() in mpu_attach() (ic/mpu.c) to remove an empty line in some frontends (mpu_isa.c, etc).
1.7	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	02-Oct-2002	thorpej	branches: 1.6.6; Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	13-Nov-2001	lukem	add RCSID
1.2	15-May-2000	thorpej	branches: 1.2.6; 1.2.8; 1.2.10; ac97.h -> ac97var.h
1.1	01-Nov-1999	augustss	branches: 1.1.4; 1.1.6; Add FM801 audio device driver. From Witold J. Wnuk ww181302@zodiac.mimuw.edu.pl Closes PR kern/8729.
1.1.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.4.2	15-Nov-1999	fvdl	Sync with -current
1.1.4.1	01-Nov-1999	fvdl	file mpu_fms.c was added on branch fvdl-softdep on 1999-11-15 00:41:04 +0000
1.2.10.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.10.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.8.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.8.1	14-Nov-2001	nathanw	Catch up to -current.
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	15-May-2000	bouyer	file mpu_fms.c was added on branch thorpej_scsipi on 2000-11-20 11:42:25 +0000
1.6.6.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.6.5	04-Feb-2005	skrll	Sync with HEAD.
1.6.6.4	18-Dec-2004	skrll	Sync with HEAD.
1.6.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.6.1	03-Aug-2004	skrll	Sync with HEAD
1.8.6.1	12-Feb-2005	yamt	sync with head.
1.8.4.1	29-Apr-2005	kent	sync with -current
1.10.2.1	27-Oct-2007	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.14.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.14.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.16.34.1	17-Apr-2012	yamt	sync with head
1.16.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.17.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.17.50.1	10-Jun-2019	christos	Sync with HEAD
1.17	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	23-Nov-2011	jmcneill	branches: 1.16.50; 1.16.54; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.15	28-Apr-2008	martin	branches: 1.15.12; 1.15.34; 1.15.36; Remove clause 3 and 4 from TNF licenses
1.14	27-Mar-2008	xtraeme	branches: 1.14.2; 1.14.4; Split device_t/softc for mpu(4) and its attachments, plus other related cosmetic changes. Reviewed by cube@.
1.13	19-Oct-2007	ad	branches: 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	02-Dec-2004	xtraeme	Add ACPI mpu(4) attachment, based on lpt_acpi.c by jmcneill@, reviewed by christos@, cube@ and martin@. While here, remove a printf() in mpu_attach() (ic/mpu.c) to remove an empty line in some frontends (mpu_isa.c, etc).
1.7	08-Nov-2004	kent	ANSIfy and style fixes
1.6	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	02-Oct-2002	thorpej	branches: 1.5.6; Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	13-Nov-2001	lukem	add RCSID
1.1	30-Mar-2001	minoura	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.1.8.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.6.2	01-May-2001	he	Pull up revision 1.1 (new, requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.1.6.1	30-Mar-2001	he	file mpu_yds.c was added on branch netbsd-1-5 on 2001-05-01 11:45:26 +0000
1.1.4.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.1	30-Mar-2001	bouyer	file mpu_yds.c was added on branch thorpej_scsipi on 2001-04-21 17:49:16 +0000
1.1.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.1.2.1	30-Mar-2001	nathanw	file mpu_yds.c was added on branch nathanw_sa on 2001-04-09 01:57:03 +0000
1.5.6.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.6.5	18-Dec-2004	skrll	Sync with HEAD.
1.5.6.4	14-Nov-2004	skrll	Sync with HEAD.
1.5.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.6.1	03-Aug-2004	skrll	Sync with HEAD
1.9.2.1	27-Oct-2007	yamt	sync with head.
1.10.22.2	10-Dec-2006	yamt	sync with head.
1.10.22.1	22-Oct-2006	yamt	sync with head
1.10.20.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.15.34.1	17-Apr-2012	yamt	sync with head
1.15.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.16.54.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.16.50.1	10-Jun-2019	christos	Sync with HEAD
1.10	31-Aug-2018	jdolecek	constify mvsata_pci_products[]
1.9	21-Jun-2017	jdolecek	branches: 1.9.4; 1.9.6; switch to pci_intr_establish_xname()
1.8	29-Mar-2014	christos	branches: 1.8.6; 1.8.22; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.7	30-Jan-2012	drochner	branches: 1.7.6; 1.7.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.6	31-Jan-2011	jakllsch	branches: 1.6.4; 1.6.8; Make the Adaptec 1420SA entry match the Marvell 88SX6041 entry. According to http://lists.freebsd.org/pipermail/freebsd-current/2010-April/016973.html the Adaptec 1420SA is a 88SX6041 (and the 1430SA is a 88SX7042). This should fix the problem described in http://mail-index.netbsd.org/current-users/2010/10/14/msg014526.html
1.5	01-Aug-2010	kiyohara	branches: 1.5.2; 1.5.4; Remove mvsata_pci_products's terminator(-1). Instead of it use __arraycount().
1.4	13-Jul-2010	kiyohara	Move to pci/mvsata_pci.c (marvell/mvsata_mv.c) splitting mvsata_products[]. mvsata(4) uses these to match.
1.3	24-Feb-2010	dyoung	branches: 1.3.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.2	08-Jan-2010	dyoung	branches: 1.2.2; Expand PMF_FN_* macros.
1.1	27-Jul-2009	kiyohara	branches: 1.1.2; Support Marvell Hercules-I/II SATA Controllers.
1.1.2.4	11-Aug-2010	yamt	sync with head.
1.1.2.3	11-Mar-2010	yamt	sync with head
1.1.2.2	19-Aug-2009	yamt	sync with head.
1.1.2.1	27-Jul-2009	yamt	file mvsata_pci.c was added on branch yamt-nfs-mp on 2009-08-19 18:47:12 +0000
1.2.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.2.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.3.2.1	05-Mar-2011	rmind	sync with head
1.5.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.5.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.6.8.1	18-Feb-2012	mrg	merge to -current.
1.6.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6.4.1	17-Apr-2012	yamt	sync with head
1.7.10.1	18-May-2014	rmind	sync with head
1.7.6.2	03-Dec-2017	jdolecek	update from HEAD
1.7.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.8.22.1	25-Jun-2017	snj	Pull up following revision(s) (requested by jdolecek in ticket #56): sys/dev/pci/mvsata_pci.c: revision 1.9 sys/dev/pci/siisata_pci.c: revision 1.16 switch to pci_intr_establish_xname()
1.8.6.1	28-Aug-2017	skrll	Sync with HEAD
1.9.6.1	10-Jun-2019	christos	Sync with HEAD
1.9.4.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.2	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.1	01-Apr-2010	jakllsch	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.14; 1.1.18; Add support for Domex 536 PCI SCSI controller to nca(4). This truly remarkable chip is found on the Domex DMX-3191D SCSI adapter.
1.1.18.1	18-Feb-2012	mrg	merge to -current.
1.1.14.1	17-Apr-2012	yamt	sync with head
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	01-Apr-2010	yamt	file nca_pci.c was added on branch yamt-nfs-mp on 2010-08-11 22:53:50 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	01-Apr-2010	rmind	file nca_pci.c was added on branch rmind-uvmplock on 2010-05-30 05:17:36 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	01-Apr-2010	uebayasi	file nca_pci.c was added on branch uebayasi-xip on 2010-04-30 14:43:39 +0000
1.107	07-Jul-2001	thorpej	Remove the old NCR 53c8xx driver. It no longer compiles (it was not updated for the thorpej_scsipi branch), and has been replaced by Manuel's MUCH BETTER `siop' driver. The code can still be referenced by using cvsweb, checking out an old revision, whatever. But it's no longer useful to have it appear in a checked out source tree.
1.106	25-Apr-2001	bouyer	Pull up the thorpej_scsipi branch to main branch. This is a completely rewritten scsipi_xfer execution engine, and the associated changes to HBA drivers. Overview of changes & features: - All xfers are queued in the mid-layer, rather than doing so in an ad-hoc fashion in individual adapter drivers. - Adapter/channel resource management in the mid-layer, avoids even trying to start running an xfer if the adapter/channel doesn't have the resources. - Better communication between the mid-layer and the adapters. - Asynchronous event notification mechanism from adapter to mid-layer and peripherals. - Better peripheral queue management: freeze/thaw, sorted requeueing during recovery, etc. - Clean separation of peripherals, adapters, and adapter channels (no more scsipi_link). - Kernel thread for each scsipi_channel makes error recovery much easier (no more dealing with interrupt context when recovering from an error). - Mid-layer support for tagged queueing: commands can have the tag type set explicitly, tag IDs are allocated in the mid-layer (thus eliminating the need to use buggy tag ID allocation schemes in many adapter drivers). - support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command will be requeued, or a REQUEST SENSE will be sent as appropriate. Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
1.105	28-Dec-2000	sommerfeld	branches: 1.105.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.104	24-Sep-2000	jdolecek	use new generic bootverbose instead of local definition
1.103	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.102	26-Jun-2000	mrg	remove redundant vm includes
1.101	08-Apr-2000	tsutsui	Remove #ifdef __NetBSD__ code inside !__NetBSD__ part.
1.100	08-Apr-2000	tsutsui	Add vaddr_t cast to vtophys() arg. (for powerpc)
1.99	06-Apr-2000	tsutsui	Fix bogus byte-swap code in ncr_snooptest(). XXX This snoop test is not needed if ncr had real bus_dma(9) support...
1.98	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.97	18-Mar-2000	cgd	* rework the quirk table ('device_tab'): * QUIRK_NOMSG only has any meaning when NCR_GETCC_WITHMSG is defined. Therefore, there's no harm in using it when NCR_GETCC_WITHMSG is not defined. so, simplify the table by removing the #ifdef. * there's really no point in having table entries after an entry which will match everything. * add some comments, clean up spacing. * add an entry for "QUANTUM"/"ATLAS IV" drives with flags QUIRK_NOTAGS\|QUIRK_NOMSG. (I included the latter flag only because everything else had it before! ... which means that all the functionality added with the NCR_GETCC_WITHMSG define would never get used! sigh) The latter fixes the problems I was having on an Atlas, and should fix the problems mentioned by Hans Hoppe <hopha@casema.net> in comments on PR#7694.
1.96	18-Mar-2000	cgd	undo the non-script changes from rev 1.87. They caused problems in 1.4.x and i have concerns (but no concrete proof) they will cause/have caused problems in -current as well. Really, the right way to fix this is to rewrite the driver, and push up tagged queueing handling into a common middle layer that'll do it right in a low-level-driver-independent manner. I'll fix my particular issues by using the ncr driver quirk mechanism.
1.95	17-Mar-2000	soren	Use a kvtophys() hack for mips like for alpha. No idea if it actually works..
1.94	20-Feb-2000	sommerfeld	deal with a DELAY vs. delay issue
1.93	26-Dec-1999	tsutsui	Use htole32() and le32toh() instead of SCR_BO(). XXX Should we still keep non-NetBSD part?
1.92	20-Dec-1999	tsutsui	Add more SCR_BO(). NCR53c875 now works on macppc. XXX All SCR_BO() macros should be replaced with htole32() or le32toh().
1.91	05-Dec-1999	thorpej	Mark exactly 3 uses of the SCR_BO() macro as being bogus; these 3 uses already involve byte swapping on big-endian systems due to bus_space_*(). However, the use is self-consistent, and the value is not interpreted by the chip, so it probably does not matter. Leave them in for now; we can always look at their removal later.
1.90	05-Dec-1999	thorpej	Slight change to previous, to make it more obvious as to what's going on.
1.89	05-Dec-1999	thorpej	I have the feeling that OpenBSD has never tried the 53c825a, 53c875, 53c875j, 53c885, 53c895, or 53c896 on big-endian systems.
1.88	05-Dec-1999	thorpej	Changes from the OpenBSD `ncr' driver to byte-swap the script on big-endian systems. Should make this go on e.g. the Apple Network Server.
1.87	08-Oct-1999	cgd	branches: 1.87.2; Don't burst into flame when a QUEUE FULL message is received. Also, do a gross hack which allows seemingly-broken quantum drives to function with this driver. The gross hack is to disable tagged queueing completely when QUEUE FULL is received. That costs performance on drives which do tagged queueing properly and which return QUEUE FULL, but given the way this driver works it's seems to be the only thing short of significant recoding which will make it function with the quantum drives in question.
1.86	05-Oct-1999	thorpej	branches: 1.86.4; Have I mentioned how much I love this driver? Fix the code in the tagged queueing support that decreases the number of openings on a device; it previously assumed that a scsipi_link's `openings' were descreased as commands were issued, which is not longer the case (`active' is increased).
1.85	05-Oct-1999	thorpej	Fix a botch when updating for the SCSIPI changes.
1.84	30-Sep-1999	thorpej	Update for SCSPI changes.
1.83	19-Aug-1999	matt	after bus_dmamem creating some memory, zero it out. This has stopped my alpha from getting COMMAND FAILED on random boots.
1.82	04-Aug-1999	tron	Add missing prototype for read_tekram_eeprom() to make driver compile if "options NCR_TEKRAM_EEPROM" is specified in the kernel configuration file. Patch supplied by Hans Hoppe in PR kern/8141.
1.81	29-Jul-1999	bouyer	Bump MAX_START to 256, so that the ncr driver can work with more than 5 devices on the same bus. See PR kern/6347 for details.
1.80	13-Dec-1998	thorpej	branches: 1.80.2; 1.80.4; Don't DMA the scsi command from or the sense data to the scsipi_xfer directly. That would require that we map the scsipi_xfer into DMA space. Instead, copy to/from the NCR CCB, which the script already has to DMA to/from. These copies are small, and don't seem to affect performance.
1.79	12-Dec-1998	thorpej	Step 3: Use DMA maps for the actual data transfers.
1.78	12-Dec-1998	thorpej	Step two: Use DMA-safe memory for the script instances. The most important thing here is that is guarantees that the script is contiguous in DMA space.
1.77	11-Dec-1998	thorpej	Step one of bus_dma'ing the PCI NCR driver: Separate the ncb (i.e. softc) members that are accessed by the script into a separate structure. Allocate one of these structures in DMA safe memory using bus_dma, and change RELOC_SOFTC to use the DMA address of this structure.
1.76	05-Dec-1998	mjacob	Update HBAs to incorporate the new max_lun property.
1.75	19-Nov-1998	thorpej	Adapt to the new scsipi_adapter interface.
1.74	10-Oct-1998	thorpej	Garbage-collect the open_target_lu and close_target_lu entry points from struct scsipi_adapter; they were not used. Add a scsipi_ioctl entry point to struct scsipi_adapter. This will be used to issue ioctl commands to the host adapters. Inspired by PR #6090, from Matt Jacob.
1.73	21-Sep-1998	mjacob	Roll back to 1.71 version- too many unexplainable problems with 1.72.
1.72	08-Sep-1998	mjacob	add more general probe to regular more operation stuff
1.71	20-Aug-1998	veego	branches: 1.71.2; Add some braces to stop the new egcs warnings.
1.70	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.69	10-Aug-1998	ross	Fix a race involving determination of pollmode. This is not known to be associated with any observed lossage...it was just noticed while reading ncr_start(). g/c some now-unreachable code produced by the earlier race condition fix.
1.68	08-Aug-1998	ross	1. Untangle preprocessor conditionals for sanity and {} balance 2. Fix a race between the first callout and the first non-polled ncr_start()
1.67	08-Jun-1998	thorpej	branches: 1.67.2; Nuke __BROKEN_INDIRECT_CONFIG.
1.66	28-May-1998	matt	Use PCI_REVISION and eliminate some pci_conf_reads
1.65	28-Jan-1998	thorpej	Use offsetof() from libkern.h
1.64	12-Jan-1998	thorpej	Update for config changes.
1.63	23-Sep-1997	perry	ncr_reg.h -> ncrreg.h per Dave Huang
1.62	23-Sep-1997	perry	updates from FreeBSD via Dave Huang <khym@bga.com>
1.61	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.60	01-Jul-1997	cjs	branches: 1.60.2; 1.60.4; Add QUIRK_NOTAGS quirk to disable tagged command queuing for particular drives. Add a quirk entry for HP C372x drives with QUIRK_NOTAGS. Also, print quirks being used if quirks other than QUIRK_NOSYNC are being used.
1.59	17-Apr-1997	mikel	end line after printing WIDE SCSI status; fixes PR kern 3505.
1.58	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.57	13-Apr-1997	cgd	use pci_map_register(). map both mem and I/O spaces, preferring mem space by default if it's usable, but falling back to I/O space if mem isn't usable. If NCR_IOMAPPED is defined (default on the x86), prefer I/O space then fall back to mem. Also, clean up the various memory consistency checks so that they can deal with run-time determination of whether or not the device is to be memory- or I/O-mapped.
1.56	04-Mar-1997	mycroft	Fix a couple of splhigh()s that should be splbio().
1.55	10-Jan-1997	perry	branches: 1.55.4; Commit latest fixes/upgrades to NCR SCSI driver. From FreeBSD via Dave Huang <khym@bga.com> Tested on alpha by cgd, tested on several i386 boxes. Certainly causes no harm to the goddamned mess, but the NCR driver only works when you perform voodoo rituals on it anyway. This is what Dave said (in email) has been added to the driver: ---------------------------------------------------------------------- This seems to be the most significant change: General cleanup and new features for 53c875 based cards, especially the Tekram DC390W/U/F, whose config EEPROM can now be dumped, if the kernel is built with option NCR_TEKRAM_EEPROM. Other changes: - add brackets to expansion of OUTB/W/L macro arguments. - remove unused NCB structure element ns_async - support sync. SCSI offset of 16 (instead of only 8) on 825A and 875 - correctly identify 53c810A and 53c825A chips - preserve SCSI BIOS settings of PCI performance options - remove (already disabled) support for NCR reset because of command timeout - reverse order of reading of SCSI and DMA specific interrupt cause registers - add definition of Tekram config EEPROM contents (not currently used) ----------------------------------------------------------------------
1.54	10-Dec-1996	thorpej	Fill in sc_link.max_target
1.53	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.52	23-Nov-1996	cgd	cast pointers to long, not int, and fix the resulting printf format goof.
1.51	20-Nov-1996	perry	Incorporate updated version of fixes from FreeBSD submitted in PR 2841, plus some fixes to make the patches work on the Alpha. Seems to improve the NCR driver a lot. We probably should try to incorporate any updates that have happened since, too.
1.50	14-Nov-1996	cgd	print ptrdiff_t's by casting to long and printing with %l<whatever>. (This fixes problems with the printf format fixes i checked in yesterday. ptrdiff_t is an 'int' on the i386 but a 'long' on the alpha, so the cast really is necessary... sigh)
1.49	13-Nov-1996	cgd	fix a bunch of printf format botches
1.48	25-Oct-1996	cgd	update alpha_XXX_dmamap() invocations to match prototype.
1.47	25-Oct-1996	cgd	make script_kvars an array of unsigned longs, and cast when assigning the elements, to avoid warnings (with -Wcast-qual) about 'volatile' qualifiers being discarded.
1.46	23-Oct-1996	cgd	update drivers to use the new NetBSD/alpha DMA mapping hack, since the old one is not workable with the new bus.h.
1.45	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.44	13-Oct-1996	christos	backout kprintf changes
1.43	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.42	20-Sep-1996	cgd	put set of 'channel' field of scsi_link in the right #ifdef
1.41	28-Aug-1996	cgd	(1) set scsi_link channel to either the appropriate channel (if a multi-channel driver), or to SCSI_CHANNEL_ONLY_ONE if a single-channel driver. (2) use scsiprint() rather than a locally-defined autoconfig print function, and kill any locally-defined print function.
1.40	27-Aug-1996	cgd	change cfprint_t type definition to take a const char , rather than a char , because that's what was really intended, and because if the print function modifies the string, various things could become unhappy (so the string should _not_ be modified).
1.39	27-Aug-1996	cgd	#undef assert before defining it, so the new libkern definition doesn't conflict with this one.
1.38	10-Aug-1996	mycroft	Eliminate reference to stddef.h.
1.37	15-Jul-1996	cgd	before #defining the vtophys() hack for the alpha, #undef vtophys since it may be a macro.
1.36	03-Jun-1996	cgd	convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make the code more correct.
1.35	13-May-1996	mycroft	branches: 1.35.4; Use intr.h.
1.34	03-May-1996	christos	- fix printf format strings - #ifdef out unused variables/functions - parenthesize
1.33	03-Apr-1996	mycroft	Implement ncr_print().
1.32	30-Mar-1996	cgd	make sure that chipset tags are used properly (because the alpha cares), and include <machine/intr.h> on the alpha, for interrupt information. (i386 will probably switch to this eventually.)
1.31	27-Mar-1996	cgd	modify these to use the new PCI interfaces: (1) use pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register, and bus_{io,mem}_map() to map it. (2) use pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (5) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.30	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.29	14-Mar-1996	cgd	fix types on a bunch of variables (were u_long's, now are U_INT32s or ncrcmds, and a few other fixes). print pointers with %p, rather than casting to (unsigned), which breaks on the Alpha.
1.28	14-Mar-1996	cgd	include pcireg before pcivar.h
1.27	11-Mar-1996	cgd	kill extraneous empty line in comment block at start of file
1.26	11-Mar-1996	cgd	mostly convert to use interfaces provided by <machine/bus.h>. (For now, assumes that pci_map_mem/pci_map_io provide interfaces which utilitize bus_{io,mem}_handle_t's, or types which are compatible. This works on the i386, and will change eventually anyway.)
1.25	11-Mar-1996	cgd	Move INT32, U_INT32, TIMEOUT definitions to ncr_reg.h, and add definitions for U_INT8, INT16, U_INT16 definitions. Convert structs and definitions in ncr_reg.h (e.g. ncrcmd, the chip register layout, etc.) to use these definitions. Add INB_OFF, INL_OFF, and OUTL_OFF macros to access specified offsets into I/O or memory space. Convert register dumps (etc.), and cache snoop test to use these new macros, so that nothing accesses the device I/O or memory space directly. (Register dumps now come from I/O space if NCR_IOMAPPED. They used to bogusly use memory space.) Add a new relocation type for script entries, RELOC_KVAR. Allow scripts access to mono_time.tv_sec, mono_time, and ncr_cache via this mechanism, and convert scripts to use it. An ncrcmd is only 32 bits wide, and KVAs may be > 32 bits wide (e.g. on Alpha), leading to linker problems. This is a safer way to do this anyway; relocation is more deterministic this way, and doesn't rely on KVAs not looking like other relocation types. Panic if an unmatched relocation other than 'zero' is specified. That's now a script bug. (This used to be used to convert KVAs of kernel variables referenced in the script to PCI bus physical addresses, and that is now handled by RELOC_KVAR relocations.) Figure out and print the model of chip.
1.24	24-Feb-1996	cgd	use mono_time, rather than time, to keep track of timeouts. time can change unpredictably.
1.23	28-Dec-1995	thorpej	Squish a couple of rogue old-style RCS ids.
1.22	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.21	30-Nov-1995	jtc	merge in changes from 1.1 release branch
1.20	23-Oct-1995	pk	DV_NULL should be DV_DULL.
1.19	22-Oct-1995	pk	DV_DISK => DV_NULL (from Mike Long; PR#1607).
1.18	02-Oct-1995	mycroft	branches: 1.18.2; Fix an oversight in the minphys routine.
1.17	01-Oct-1995	mycroft	Incorporate recent changes from Wolfgang and Stefan.
1.16	12-Aug-1995	mycroft	minphys() functions really should return void.
1.15	24-Jul-1995	cgd	update SCSI minphys routines' definitions to match standard minphys() definition and usage.
1.14	17-Apr-1995	cgd	clean up several ISA device interfaces: autoconfiguration, header inclusion, and interrupt configuration. more work still needs to be done, but it's getting better...
1.13	28-Mar-1995	jtc	KERNEL -> _KERNEL
1.12	01-Feb-1995	mycroft	Patch from Wolfgang to fix a missing spl.
1.11	27-Jan-1995	cgd	include files from the correct places.
1.10	16-Jan-1995	mycroft	Incorporate recent changes from Wolfgang. Recognize the 815.
1.9	04-Jan-1995	mycroft	Remove many of the `static's, to facilitate debugging.
1.8	28-Dec-1994	mycroft	Numerous changes. Many bugs fixed, better autoconfig, a few new features.
1.7	04-Nov-1994	mycroft	Make a wrapper match function to check the bus and device numbers, rather than insisting that every driver do it.
1.6	03-Nov-1994	mycroft	Always use direct configuration.
1.5	27-Oct-1994	cgd	new RCS ID format.
1.4	27-Oct-1994	andrew	Default to decent performance.
1.3	27-Oct-1994	andrew	Merge changes from Wolfgang Stanglmeier's 941012 release.
1.2	01-Oct-1994	mycroft	Update from Wolfgang Stanglemeier.
1.1	09-Aug-1994	mycroft	branches: 1.1.1; Add NCR 53c8XX driver, originally from Wolfgang Stanglmeier.
1.1.1.1	29-Aug-1998	ross	Import current FreeBSD ncr.c, but without the id strings.
1.18.2.2	23-Nov-1995	perry	Changed to use the hacked up version Thor got from the original maintainers. It is uglier than mycroft's version, but it works on a lot more machines than his does. Hopefully this can be fixed "for real" after the release. Note that although it works on a superset of the machines that mycroft's did, it is still broken on some slower machines with fast disks. The thing is full of timing bugs. Sigh.
1.18.2.1	23-Oct-1995	pk	From trunk: DV_DISK -> DV_DULL
1.35.4.2	10-Dec-1996	mycroft	Eliminate references to user-level include files.
1.35.4.1	03-Jun-1996	cgd	pull up from trunk: >convert alpha vtophys() #defines to use __alpha_bus_XXX_dmamap(), to >get rid of the nasty (vtophys(va) \| 0x40000000) expressions, and to make >the code more correct.
1.55.4.1	12-Mar-1997	is	Merge in changes from The Trunk
1.60.4.2	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.60.4.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.60.2.2	01-Jul-1997	bouyer	Updates for new scsipi subsystem. Actally known to work on i386 and sparc.
1.60.2.1	01-Jul-1997	bouyer	file ncr.c was added on branch bouyer-scsipi on 1997-07-01 17:35:37 +0000
1.67.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.71.2.1	30-Aug-1998	ross	This is a merge on a branch of all NetBSD ncr.c changes into the FreeBSD-current (1.123) ncr.c. It didn't actually solve my '875 problems; we will decide whether to actually use this work later. ross-ncr-fb1-123-base the branch point ross-ncr-fb1-123-merge the branch
1.80.4.1	02-Aug-1999	thorpej	Update from trunk.
1.80.2.6	02-Apr-2000	he	Pull up revisions 1.87 and 1.96-1.97 (requested by cgd): Fix script bug in processing QUEUE FULL response (would always trigger an assertion failure), and add Quantum Atlas IV drives to the quirk table so that they don't use tagged queueing. Fixes problems reported in PR#7694.
1.80.2.5	08-Nov-1999	cgd	Back out pullup of rev. 1.87. It seems a bit broken, and the known evil is better than the unknown evil.
1.80.2.4	18-Oct-1999	cgd	pull up rev 1.87 from trunk (requested by cgd): Fix bugs in handling QUEUE FULL messages. Also, when a QUEUE FULL message is seen, disable tagged queueing completely for the target which generated it. This is (much) more conservative than necessary for normal operation, but allows the driver to function properly even with some broken drives at the cost of some performance on drives that have a very low tagged command limit.
1.80.2.3	18-Sep-1999	cgd	pull up rev 1.83 from trunk (cgd): Zero out DMA memory regions to avoid use of uninitialized memory (and resutling "COMMAND FAILED" errors) on systems which don't clear memory on boot. (cgd)
1.80.2.2	18-Sep-1999	cgd	pull up rev 1.82 (tron): Make driver build with Tekram EEPROM support enabled. Fixes PR#8141. (tron)
1.80.2.1	18-Sep-1999	cgd	pull up rev 1.81 (bouyer): Make the ncr driver work with more than 5 devices on the same bus. Fixes PR#6347. (bouyer)
1.86.4.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.87.2.6	15-Jan-2001	bouyer	req_sense_length is dead.
1.87.2.5	05-Jan-2001	bouyer	Sync with HEAD
1.87.2.4	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.87.2.3	26-Oct-1999	thorpej	Update for: Deal a little more gracefully with the fact that xfer mode parameters are for the I_T Nexus, and make all xfer mode updates `async events'.
1.87.2.2	20-Oct-1999	thorpej	Eliminate sleeping from adapter routines.
1.87.2.1	19-Oct-1999	thorpej	Adapt to scsipi API changes.
1.105.2.2	28-Feb-2002	nathanw	Catch up to -current.
1.105.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file ncr.c.old was initially added on branch thorpej_scsipi.
1.1.2.2	22-Nov-2000	bouyer	Should not be there
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file ncr.c.orig was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file ncr.c.rej was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.15	23-Sep-1997	perry	ncr_reg.h -> ncrreg.h per Dave Huang
1.14	23-Sep-1997	perry	updates from FreeBSD via Dave Huang <khym@bga.com>
1.13	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.12	23-Jul-1997	bouyer	branches: 1.12.2; Back out previous change. This was for bouyer-scsipi, and not for the trunk.
1.11	23-Jul-1997	bouyer	Forgot a scsi->dev/scsipi.
1.10	10-Jan-1997	perry	branches: 1.10.8; Commit latest fixes/upgrades to NCR SCSI driver. From FreeBSD via Dave Huang <khym@bga.com> Tested on alpha by cgd, tested on several i386 boxes. Certainly causes no harm to the goddamned mess, but the NCR driver only works when you perform voodoo rituals on it anyway. This is what Dave said (in email) has been added to the driver: ---------------------------------------------------------------------- This seems to be the most significant change: General cleanup and new features for 53c875 based cards, especially the Tekram DC390W/U/F, whose config EEPROM can now be dumped, if the kernel is built with option NCR_TEKRAM_EEPROM. Other changes: - add brackets to expansion of OUTB/W/L macro arguments. - remove unused NCB structure element ns_async - support sync. SCSI offset of 16 (instead of only 8) on 825A and 875 - correctly identify 53c810A and 53c825A chips - preserve SCSI BIOS settings of PCI performance options - remove (already disabled) support for NCR reset because of command timeout - reverse order of reading of SCSI and DMA specific interrupt cause registers - add definition of Tekram config EEPROM contents (not currently used) ----------------------------------------------------------------------
1.9	20-Nov-1996	perry	Incorporate updated version of fixes from FreeBSD submitted in PR 2841, plus some fixes to make the patches work on the Alpha. Seems to improve the NCR driver a lot. We probably should try to incorporate any updates that have happened since, too.
1.8	11-Mar-1996	cgd	RCS ID police; get rid of Id keyword.
1.7	11-Mar-1996	cgd	Move INT32, U_INT32, TIMEOUT definitions to ncr_reg.h, and add definitions for U_INT8, INT16, U_INT16 definitions. Convert structs and definitions in ncr_reg.h (e.g. ncrcmd, the chip register layout, etc.) to use these definitions. Add INB_OFF, INL_OFF, and OUTL_OFF macros to access specified offsets into I/O or memory space. Convert register dumps (etc.), and cache snoop test to use these new macros, so that nothing accesses the device I/O or memory space directly. (Register dumps now come from I/O space if NCR_IOMAPPED. They used to bogusly use memory space.) Add a new relocation type for script entries, RELOC_KVAR. Allow scripts access to mono_time.tv_sec, mono_time, and ncr_cache via this mechanism, and convert scripts to use it. An ncrcmd is only 32 bits wide, and KVAs may be > 32 bits wide (e.g. on Alpha), leading to linker problems. This is a safer way to do this anyway; relocation is more deterministic this way, and doesn't rely on KVAs not looking like other relocation types. Panic if an unmatched relocation other than 'zero' is specified. That's now a script bug. (This used to be used to convert KVAs of kernel variables referenced in the script to PCI bus physical addresses, and that is now handled by RELOC_KVAR relocations.) Figure out and print the model of chip.
1.6	30-Nov-1995	jtc	merge in changes from 1.1 release branch
1.5	01-Oct-1995	mycroft	branches: 1.5.2; Incorporate recent changes from Wolfgang and Stefan.
1.4	28-Dec-1994	mycroft	Numerous changes. Many bugs fixed, better autoconfig, a few new features.
1.3	27-Oct-1994	cgd	new RCS ID format.
1.2	01-Oct-1994	mycroft	Update from Wolfgang Stanglemeier.
1.1	09-Aug-1994	mycroft	Add NCR 53c8XX driver, originally from Wolfgang Stanglmeier.
1.5.2.1	23-Nov-1995	perry	Changed to use the hacked up version Thor got from the original maintainers. It is uglier than mycroft's version, but it works on a lot more machines than his does. Hopefully this can be fixed "for real" after the release. Note that although it works on a superset of the machines that mycroft's did, it is still broken on some slower machines with fast disks. The thing is full of timing bugs. Sigh.
1.10.8.1	23-Jul-1997	bouyer	scsi/->dev/scsipi/
1.12.2.2	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.12.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.18	07-Jul-2001	thorpej	Remove the old NCR 53c8xx driver. It no longer compiles (it was not updated for the thorpej_scsipi branch), and has been replaced by Manuel's MUCH BETTER `siop' driver. The code can still be referenced by using cvsweb, checking out an old revision, whatever. But it's no longer useful to have it appear in a checked out source tree.
1.17	26-Dec-1999	tsutsui	branches: 1.17.6; Use htole32() and le32toh() instead of SCR_BO(). XXX Should we still keep non-NetBSD part?
1.16	18-Dec-1999	tsutsui	Use bswap32() as SCR_BO() macro to avoid unexpected side effects. SCR_BO(*src++) caused some trouble.
1.15	05-Dec-1999	thorpej	Changes from the OpenBSD `ncr' driver to byte-swap the script on big-endian systems. Should make this go on e.g. the Apple Network Server.
1.14	23-Sep-1997	perry	branches: 1.14.2; 1.14.18; 1.14.24; updates from FreeBSD via Dave Huang <khym@bga.com>
1.13	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.12	23-Jul-1997	bouyer	branches: 1.12.2; Back out previous change. This was for bouyer-scsipi, and not for the trunk.
1.11	23-Jul-1997	bouyer	Forgot a scsi->dev/scsipi.
1.10	10-Jan-1997	perry	branches: 1.10.8; Commit latest fixes/upgrades to NCR SCSI driver. From FreeBSD via Dave Huang <khym@bga.com> Tested on alpha by cgd, tested on several i386 boxes. Certainly causes no harm to the goddamned mess, but the NCR driver only works when you perform voodoo rituals on it anyway. This is what Dave said (in email) has been added to the driver: ---------------------------------------------------------------------- This seems to be the most significant change: General cleanup and new features for 53c875 based cards, especially the Tekram DC390W/U/F, whose config EEPROM can now be dumped, if the kernel is built with option NCR_TEKRAM_EEPROM. Other changes: - add brackets to expansion of OUTB/W/L macro arguments. - remove unused NCB structure element ns_async - support sync. SCSI offset of 16 (instead of only 8) on 825A and 875 - correctly identify 53c810A and 53c825A chips - preserve SCSI BIOS settings of PCI performance options - remove (already disabled) support for NCR reset because of command timeout - reverse order of reading of SCSI and DMA specific interrupt cause registers - add definition of Tekram config EEPROM contents (not currently used) ----------------------------------------------------------------------
1.9	20-Nov-1996	perry	Incorporate updated version of fixes from FreeBSD submitted in PR 2841, plus some fixes to make the patches work on the Alpha. Seems to improve the NCR driver a lot. We probably should try to incorporate any updates that have happened since, too.
1.8	11-Mar-1996	cgd	RCS ID police; get rid of Id keyword.
1.7	11-Mar-1996	cgd	Move INT32, U_INT32, TIMEOUT definitions to ncr_reg.h, and add definitions for U_INT8, INT16, U_INT16 definitions. Convert structs and definitions in ncr_reg.h (e.g. ncrcmd, the chip register layout, etc.) to use these definitions. Add INB_OFF, INL_OFF, and OUTL_OFF macros to access specified offsets into I/O or memory space. Convert register dumps (etc.), and cache snoop test to use these new macros, so that nothing accesses the device I/O or memory space directly. (Register dumps now come from I/O space if NCR_IOMAPPED. They used to bogusly use memory space.) Add a new relocation type for script entries, RELOC_KVAR. Allow scripts access to mono_time.tv_sec, mono_time, and ncr_cache via this mechanism, and convert scripts to use it. An ncrcmd is only 32 bits wide, and KVAs may be > 32 bits wide (e.g. on Alpha), leading to linker problems. This is a safer way to do this anyway; relocation is more deterministic this way, and doesn't rely on KVAs not looking like other relocation types. Panic if an unmatched relocation other than 'zero' is specified. That's now a script bug. (This used to be used to convert KVAs of kernel variables referenced in the script to PCI bus physical addresses, and that is now handled by RELOC_KVAR relocations.) Figure out and print the model of chip.
1.6	30-Nov-1995	jtc	merge in changes from 1.1 release branch
1.5	01-Oct-1995	mycroft	branches: 1.5.2; Incorporate recent changes from Wolfgang and Stefan.
1.4	28-Dec-1994	mycroft	Numerous changes. Many bugs fixed, better autoconfig, a few new features.
1.3	27-Oct-1994	cgd	new RCS ID format.
1.2	01-Oct-1994	mycroft	Update from Wolfgang Stanglemeier.
1.1	09-Aug-1994	mycroft	Add NCR 53c8XX driver, originally from Wolfgang Stanglmeier.
1.5.2.1	23-Nov-1995	perry	Changed to use the hacked up version Thor got from the original maintainers. It is uglier than mycroft's version, but it works on a lot more machines than his does. Hopefully this can be fixed "for real" after the release. Note that although it works on a superset of the machines that mycroft's did, it is still broken on some slower machines with fast disks. The thing is full of timing bugs. Sigh.
1.10.8.1	23-Jul-1997	bouyer	scsi/->dev/scsipi/
1.12.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.24.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.14.18.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.14.2.2	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.14.2.1	23-Sep-1997	thorpej	file ncrreg.h was added on branch marc-pcmcia on 1997-09-29 07:21:05 +0000
1.17.6.1	28-Feb-2002	nathanw	Catch up to -current.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file ncrreg.h.old was initially added on branch thorpej_scsipi.
1.1.2.2	22-Nov-2000	bouyer	Should not be there
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8	07-Jul-2001	thorpej	Remove the old NCR 53c8xx driver. It no longer compiles (it was not updated for the thorpej_scsipi branch), and has been replaced by Manuel's MUCH BETTER `siop' driver. The code can still be referenced by using cvsweb, checking out an old revision, whatever. But it's no longer useful to have it appear in a checked out source tree.
1.7	17-Mar-1996	thorpej	branches: 1.7.42; New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.6	27-Jan-1995	cgd	include files from the correct places.
1.5	18-Nov-1994	mycroft	Convert port, IRQ, and DRQ numbers to ints.
1.4	27-Oct-1994	cgd	new RCS ID format.
1.3	01-Oct-1994	mycroft	Update from Wolfgang Stanglemeier.
1.2	11-Aug-1994	mycroft	Make this work.
1.1	09-Aug-1994	mycroft	branches: 1.1.2; Add NCR 53c8XX driver, originally from Wolfgang Stanglmeier.
1.1.2.1	11-Aug-1994	mycroft	update from trunk
1.7.42.1	28-Feb-2002	nathanw	Catch up to -current.
1.2	02-Nov-2002	perry	change array to unsigned char. fixes 23430 lint warnings. "really!"
1.1	05-Nov-2000	thorpej	branches: 1.1.2; 1.1.4; 1.1.6; Driver for NeoMagic 256 audio. Based on the OpenBSD driver, which is in turn based on the FreeBSD driver, ported to NetBSD and somewhat modified by me. Thanks to Matt Thomas for testing.
1.1.6.2	06-May-2001	he	Pull up revision 1.1 (new, requested by sommerfeld): Add a driver for the NeoMagic 256 AC'97 chip.
1.1.6.1	05-Nov-2000	he	file neo-coeff.h was added on branch netbsd-1-5 on 2001-05-06 14:39:09 +0000
1.1.4.1	11-Nov-2002	nathanw	Catch up to -current
1.1.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.1	05-Nov-2000	bouyer	file neo-coeff.h was added on branch thorpej_scsipi on 2000-11-22 16:04:10 +0000
1.56	06-Sep-2021	andvar	fix various typos in comments.
1.55	12-Nov-2019	msaitoh	Add missing initialization of sc_dev.
1.54	08-Jun-2019	isaki	branches: 1.54.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.53	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.52	16-Mar-2019	isaki	branches: 1.52.2; Use C99 style struct initializer to audio_hw_if.
1.51	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.50	29-Mar-2014	christos	branches: 1.50.22; 1.50.28; 1.50.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.49	27-Oct-2012	chs	branches: 1.49.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.48	24-Aug-2012	msaitoh	branches: 1.48.2; Fix typos
1.47	24-Nov-2011	mrg	put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.46	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.45	24-Feb-2010	dyoung	branches: 1.45.10; 1.45.12; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.44	08-Jan-2010	dyoung	branches: 1.44.2; Expand PMF_FN_* macros.
1.43	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.42	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.41	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.40	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.39	10-Apr-2008	cegger	branches: 1.39.4; 1.39.16; 1.39.18; use aprint_*_dev and device_xname
1.38	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.37	09-Dec-2007	jmcneill	branches: 1.37.6; 1.37.10; Merge jmcneill-pm branch.
1.36	19-Oct-2007	ad	branches: 1.36.4; 1.36.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.35	16-Nov-2006	christos	branches: 1.35.6; 1.35.8; 1.35.22; 1.35.24; 1.35.28; __unused removal on arguments; approved by core.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.32	02-Sep-2006	christos	branches: 1.32.2; 1.32.4; add missing initializer
1.31	25-Feb-2006	wiz	branches: 1.31.2; Fix some typos.
1.30	11-Dec-2005	christos	branches: 1.30.2; 1.30.4; 1.30.6; merge ktrace-lwp.
1.29	28-Jun-2005	thorpej	branches: 1.29.2; Use ANSI function decls and static.
1.28	08-Apr-2005	kent	* neo_intr() update the watermark for the recording buffer The fix was provided in PR#26706 by Chapman Flack.
1.27	27-Feb-2005	perry	branches: 1.27.2; nuke trailing whitespace
1.26	15-Jan-2005	kent	branches: 1.26.2; 1.26.4; ansify and KNF
1.25	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.24	29-Oct-2004	yamt	branches: 1.24.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.23	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.22	09-Jul-2004	mycroft	Clean up. Attempt to make these safer against stray interrupts.
1.21	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.20	30-Oct-2003	simonb	branches: 1.20.2; Remove some assigned-to but otherwise unused variables.
1.19	03-May-2003	wiz	branches: 1.19.2; DMA, not dma nor Dma.
1.18	22-Feb-2003	tsutsui	hz -> Hz
1.17	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	14-Jun-2002	thorpej	Map the device's memory space with BUS_SPACE_MAP_LINEAR, since we need to be able to access it as such. Fixes kern/17114.
1.12	15-Apr-2002	simonb	branches: 1.12.2; 1.12.4; Fix a couple of tyops.
1.11	13-Nov-2001	lukem	add RCSID
1.10	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.9	04-Sep-2001	thorpej	branches: 1.9.2; Use bus_space_mmap().
1.8	14-Mar-2001	drochner	branches: 1.8.2; 1.8.4; this driver doesn't work on my thinkpad, so disable the match for the IBM/0x00dd subsystem ID too
1.7	28-Dec-2000	sommerfeld	branches: 1.7.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.6	27-Nov-2000	thorpej	Power management (well, suspend/resume) is now implemented. Remove it from the TODO list.
1.5	27-Nov-2000	sommerfeld	Add a powerhook to reload AC97 registers after a resume. Sound now works after suspend/resume on my Sony PCG-Z505S
1.4	05-Nov-2000	thorpej	branches: 1.4.2; Two things off the TODO list.
1.3	05-Nov-2000	thorpej	Add mmap support (i386-specific for the moment).
1.2	05-Nov-2000	thorpej	Weed out the Dell and HP versions which are known to work in WSS emulation mode (i.e. won't work in AC'97 mode).
1.1	05-Nov-2000	thorpej	Driver for NeoMagic 256 audio. Based on the OpenBSD driver, which is in turn based on the FreeBSD driver, ported to NetBSD and somewhat modified by me. Thanks to Matt Thomas for testing.
1.4.2.5	27-Mar-2001	bouyer	Sync with HEAD.
1.4.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.4.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.4.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.4.2.1	05-Nov-2000	bouyer	file neo.c was added on branch thorpej_scsipi on 2000-11-22 16:04:11 +0000
1.7.2.7	18-Oct-2002	nathanw	Catch up to -current.
1.7.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.7.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.7.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.7.2.3	08-Oct-2001	nathanw	Catch up to -current.
1.7.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.7.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.8.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.8.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.8.2.2	06-May-2001	he	Pull up revisions 1.1-1.6, 1.8 (new, requested by sommerfeld): Add a driver for the NeoMagic 256 AC'97 chip.
1.8.2.1	14-Mar-2001	he	file neo.c was added on branch netbsd-1-5 on 2001-05-06 14:39:36 +0000
1.9.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.12.4.1	13-Apr-2005	tron	Pull up revision 1.28 (requested by kent in ticket #5743): * neo_intr() update the watermark for the recording buffer The fix was provided in PR#26706 by Chapman Flack.
1.12.2.1	20-Jun-2002	gehenna	catch up with -current.
1.19.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.19.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.6	17-Jan-2005	skrll	Sync with HEAD.
1.19.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.19.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.19.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.19.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.19.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.2.1	22-Sep-2004	jmc	branches: 1.20.2.1.2; Pullup rev 1.23 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.20.2.1.2.1	11-May-2005	riz	Pull up revision 1.28 (requested by kent in ticket #1413): * neo_intr() update the watermark for the recording buffer The fix was provided in PR#26706 by Chapman Flack.
1.24.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.24.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.26.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.26.2.1	29-Apr-2005	kent	sync with -current
1.27.2.1	08-Apr-2005	he	Pull up revision 1.28 (requested by kent in ticket #127): In neo_intr(), update the watermark for the recording buffer. Fixes PR#26706.
1.29.2.5	17-Mar-2008	yamt	sync with head.
1.29.2.4	21-Jan-2008	yamt	sync with head
1.29.2.3	27-Oct-2007	yamt	sync with head.
1.29.2.2	30-Dec-2006	yamt	sync with head.
1.29.2.1	21-Jun-2006	yamt	sync with head.
1.30.6.1	22-Apr-2006	simonb	Sync with head.
1.30.4.1	09-Sep-2006	rpaulo	sync with head
1.30.2.1	01-Mar-2006	yamt	sync with head.
1.31.2.1	03-Sep-2006	yamt	sync with head.
1.32.4.2	10-Dec-2006	yamt	sync with head.
1.32.4.1	22-Oct-2006	yamt	sync with head
1.32.2.1	18-Nov-2006	ad	Sync with head.
1.35.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.35.24.3	23-Mar-2008	matt	sync with HEAD
1.35.24.2	09-Jan-2008	matt	sync with HEAD
1.35.24.1	06-Nov-2007	matt	sync with HEAD
1.35.22.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.35.22.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.35.22.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.35.22.1	05-Oct-2007	joerg	Convert neo(4) to PNP power management.
1.35.8.1	23-Oct-2007	ad	Sync with head.
1.35.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.36.6.1	11-Dec-2007	yamt	sync with head.
1.36.4.1	26-Dec-2007	ad	Sync with head.
1.37.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.37.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.37.6.1	24-Mar-2008	keiichi	sync with head.
1.39.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.39.16.2	12-Dec-2008	ad	Checkpoint work in progress.
1.39.16.1	11-Dec-2008	ad	Checkpoint work in progress.
1.39.4.2	11-Mar-2010	yamt	sync with head
1.39.4.1	16-May-2009	yamt	sync with head
1.44.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.45.12.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.45.12.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.45.10.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.45.10.2	30-Oct-2012	yamt	sync with head
1.45.10.1	17-Apr-2012	yamt	sync with head
1.48.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.48.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.49.2.1	18-May-2014	rmind	sync with head
1.50.30.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.50.30.1	10-Jun-2019	christos	Sync with HEAD
1.50.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.50.22.1	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1439): sys/dev/sbus/sio16.c: revision 1.25 sys/dev/gpib/mt.c: revision 1.33 sys/arch/mvme68k/dev/wdsc.c: revision 1.33 sys/dev/gpib/ct.c: revision 1.31 sys/dev/isa/mcd.c: revision 1.119 sys/dev/pci/if_stge.c: revision 1.73 sys/dev/gpib/ppi.c: revision 1.25 sys/dev/pcmcia/if_tr_pcmcia.c: revision 1.29 sys/arch/amiga/dev/zssc.c: revision 1.46 sys/arch/mac68k/obio/iwm_fd.c: revision 1.57 sys/arch/amiga/dev/if_qn.c: revision 1.48 sys/dev/isa/tcic2_isa.c: revision 1.28 sys/dev/isa/uha_isa.c: revision 1.42 sys/dev/pci/neo.c: revision 1.55 sys/arch/next68k/dev/nextdisplay.c: revision 1.22 sys/dev/isa/if_iy.c: revision 1.111 Add missing initialization of sc_dev. Initialize sc_dev correctly to avoid null pointer dereference when bus_space_map() failed. Add missing initialization of sc_dev.
1.52.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.52.2.2	28-Apr-2019	isaki	Adapt to audio2. - XXX In audio2, number of blocks must be 3 or greater, so modify round_blocksize() to return bufsize / 4. - XXX Doesn't neo_trigger_input() need to subtract ssz from buffer end like trigger_output()?
1.52.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.54.2.1	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #425): sys/dev/sbus/sio16.c: revision 1.25 sys/dev/gpib/mt.c: revision 1.33 sys/arch/mvme68k/dev/wdsc.c: revision 1.33 sys/dev/gpib/ct.c: revision 1.31 sys/dev/isa/mcd.c: revision 1.119 sys/dev/pci/if_stge.c: revision 1.73 sys/dev/gpib/ppi.c: revision 1.25 sys/dev/pcmcia/if_tr_pcmcia.c: revision 1.29 sys/arch/amiga/dev/zssc.c: revision 1.46 sys/arch/mac68k/obio/iwm_fd.c: revision 1.57 sys/arch/amiga/dev/if_qn.c: revision 1.48 sys/dev/isa/tcic2_isa.c: revision 1.28 sys/dev/isa/uha_isa.c: revision 1.42 sys/dev/pci/neo.c: revision 1.55 sys/arch/next68k/dev/nextdisplay.c: revision 1.22 sys/dev/isa/if_iy.c: revision 1.111 Add missing initialization of sc_dev. Initialize sc_dev correctly to avoid null pointer dereference when bus_space_map() failed. Add missing initialization of sc_dev.
1.1	05-Nov-2000	thorpej	branches: 1.1.2; 1.1.6; Driver for NeoMagic 256 audio. Based on the OpenBSD driver, which is in turn based on the FreeBSD driver, ported to NetBSD and somewhat modified by me. Thanks to Matt Thomas for testing.
1.1.6.2	06-May-2001	he	Pull up revision 1.1 (new, requested by sommerfeld): Add a driver for the NeoMagic 256 AC'97 chip.
1.1.6.1	05-Nov-2000	he	file neoreg.h was added on branch netbsd-1-5 on 2001-05-06 14:39:10 +0000
1.1.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.1	05-Nov-2000	bouyer	file neoreg.h was added on branch thorpej_scsipi on 2000-11-22 16:04:12 +0000
1.29	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.28	03-Sep-2021	andvar	fix typos in comments, mainly s/extention/extension/ and s/sufficent/sufficient/
1.27	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.26	24-Apr-2021	thorpej	branches: 1.26.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.25	22-Dec-2019	thorpej	branches: 1.25.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.24	14-Feb-2016	chs	branches: 1.24.18; zero the i2c_attach_args structure before filling it in. fixes occasional crashes in iic_attach().
1.23	14-Feb-2012	pgoyette	branches: 1.23.6; 1.23.24; Replace the xxx_acquire()/xxx_release() rwlocks with mutexes. There are only RW_WRITERs for these, and no RW_READERs, so no need to incur the extra overhead of allowing for both. As discussed on tech-kern. For piixpm and ichsmb, the acquire/release protocol needs to be used, even if the request is I2C_F_POLL'd (or if the device supports only polled mode). Otherwise multiple requests can be running at the same time, and they stomp on each other and create anomolous results. Part 2 addresses my PR kern/45889 3 ACKs from releng
1.22	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.21	08-May-2010	pgoyette	branches: 1.21.8; 1.21.12; Recognize additional SMBus controllers. From FreeBSD.
1.20	09-May-2009	pgoyette	branches: 1.20.2; 1.20.4; For multi-byte transfers, the bytes need to be written to or read from consecutive data registers. Discovered while debugging sdtemp(4) driver which does _only_ 16-bit transfers!
1.19	09-May-2009	pgoyette	You can't use a BYTE protocol to read 16-bits worth of data from the i2c bus. Discovered while debugging my sdtemp(4) driver.
1.18	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.17	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.16	03-Feb-2009	pgoyette	branches: 1.16.2; Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.15	15-Oct-2008	pgoyette	branches: 1.15.2; If we're doing a 2-byte read, return the entire 16-bit result.
1.14	25-May-2008	cegger	branches: 1.14.4; add comment why NULL handlers are sufficient. requested and ok'd by jmcneill
1.13	25-May-2008	cegger	add power management support
1.12	27-Mar-2008	kiyohara	branches: 1.12.2; 1.12.4; 1.12.6; Split device_t/softc.
1.11	14-Jan-2008	xtraeme	branches: 1.11.6; Fix the BARs on nForce 430, which failed to attach. Reported by atsushi fukumoto in PR kern/37756.
1.10	04-Dec-2007	xtraeme	branches: 1.10.4; Newer nForce MCP SMBus devices use different PCI BARs than the old ones (nForce 2/3/4). nfsmb(4) is now properly working on newer MCPs, tested by wiz@ on MCP55: nfsmbc0 at pci0 dev 1 function 1: NVIDIA nForce MCP55 SMBus Controller (rev. 0xa2) nfsmb0 at nfsmbc0 SMBus 1 iic0 at nfsmb0: I2C bus spdmem0 at iic0 addr 0x51 spdmem0: DDR2 SDRAM memory, data ECC, 1024MB, 667MHz, PC2-5300 spdmem0: 14 rows, 10 cols, 2 ranks, 4 banks/chip, 3.00ns cycle time spdmem0: voltage SSTL 1.8V, refresh time 7.8us (self-refreshing) nfsmb1 at nfsmbc0 SMBus 2 iic1 at nfsmb1: I2C bus
1.9	14-Nov-2007	xtraeme	branches: 1.9.2; Attach to NVIDIA MCP73 now that we have the PCI id.
1.8	14-Nov-2007	xtraeme	Attach to more NVIDIA SMBus controllers.
1.7	19-Oct-2007	ad	branches: 1.7.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.6	01-Sep-2007	riz	branches: 1.6.2; 1.6.6; For SMBus, add the ability to enumerate devices on the bus. This does NOT identify the devices, merely indicates the presence of devices at certain addresses. Tested on ichsmb and nfsmb - other SMBus devices will need to ensure the proper bus type is set. (I2C_TYPE_SMBUS) From Nicolas Joly, via Paul Goyette, in PR#36744.
1.5	27-Aug-2007	xtraeme	branches: 1.5.2; Use rwlock(9) rather than lockmgr(9).
1.4	14-Aug-2007	kiyohara	Remove and insert null-line. No parentheses are needed around the return value.
1.3	28-Jul-2007	kiyohara	branches: 1.3.4; 1.3.6; s/uint32_t reg/pcireg_t reg/. Check condition `!= 0'.
1.2	21-Jul-2007	kiyohara	We wait for the clearness of the NFORCE_SMB_PROTOCOL as transfer completion. And, the maximum is assumed to be 10msec. I think that the thing for which it waits any further is dangerous for the system.
1.1	11-Jul-2007	kiyohara	branches: 1.1.2; 1.1.4; 1.1.6; Add support for NVIDIA nForce 2/3/4 SMBus controller and SMBus driver.
1.1.6.2	03-Sep-2007	skrll	Sync with HEAD.
1.1.6.1	15-Aug-2007	skrll	Sync with HEAD.
1.1.4.5	23-Oct-2007	ad	Sync with head.
1.1.4.4	09-Oct-2007	ad	Sync with head.
1.1.4.3	20-Aug-2007	ad	Sync with HEAD.
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	11-Jul-2007	ad	file nfsmb.c was added on branch vmlocking on 2007-07-15 13:21:37 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	11-Jul-2007	mjf	file nfsmb.c was added on branch mjf-ufs-trans on 2007-07-11 20:07:53 +0000
1.3.6.2	28-Jul-2007	kiyohara	s/uint32_t reg/pcireg_t reg/. Check condition `!= 0'.
1.3.6.1	28-Jul-2007	kiyohara	file nfsmb.c was added on branch matt-mips64 on 2007-07-28 12:31:51 +0000
1.3.4.5	09-Dec-2007	jmcneill	Sync with HEAD.
1.3.4.4	14-Nov-2007	joerg	Sync with HEAD.
1.3.4.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.4.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.3.4.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.5.2.3	23-Mar-2008	matt	sync with HEAD
1.5.2.2	09-Jan-2008	matt	sync with HEAD
1.5.2.1	06-Nov-2007	matt	sync with HEAD
1.6.6.2	18-Nov-2007	bouyer	Sync with HEAD
1.6.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.6.2.6	21-Jan-2008	yamt	sync with head
1.6.2.5	07-Dec-2007	yamt	sync with head
1.6.2.4	15-Nov-2007	yamt	sync with head.
1.6.2.3	27-Oct-2007	yamt	sync with head.
1.6.2.2	03-Sep-2007	yamt	sync with head.
1.6.2.1	01-Sep-2007	yamt	file nfsmb.c was added on branch yamt-lazymbuf on 2007-09-03 14:37:13 +0000
1.7.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.7.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.7.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.9.2.1	08-Dec-2007	ad	Sync with head.
1.10.4.1	19-Jan-2008	bouyer	Sync with HEAD
1.11.6.3	17-Jan-2009	mjf	Sync with HEAD.
1.11.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.11.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.12.4.3	11-Aug-2010	yamt	sync with head.
1.12.4.2	16-May-2009	yamt	sync with head
1.12.4.1	04-May-2009	yamt	sync with head.
1.12.2.1	04-Jun-2008	yamt	sync with head
1.14.4.1	19-Oct-2008	haad	Sync with HEAD.
1.15.2.1	03-Mar-2009	skrll	Sync with HEAD.
1.16.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.20.4.1	30-May-2010	rmind	sync with head
1.20.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.21.12.1	18-Feb-2012	mrg	merge to -current.
1.21.8.1	17-Apr-2012	yamt	sync with head
1.23.24.1	19-Mar-2016	skrll	Sync with HEAD
1.23.6.1	03-Dec-2017	jdolecek	update from HEAD
1.24.18.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.25.10.2	23-Mar-2021	thorpej	Convert config_found_ia() call sites where the device only carries a single interface attribute to bare config_found() calls.
1.25.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.26.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	04-Dec-2007	xtraeme	Newer nForce MCP SMBus devices use different PCI BARs than the old ones (nForce 2/3/4). nfsmb(4) is now properly working on newer MCPs, tested by wiz@ on MCP55: nfsmbc0 at pci0 dev 1 function 1: NVIDIA nForce MCP55 SMBus Controller (rev. 0xa2) nfsmb0 at nfsmbc0 SMBus 1 iic0 at nfsmb0: I2C bus spdmem0 at iic0 addr 0x51 spdmem0: DDR2 SDRAM memory, data ECC, 1024MB, 667MHz, PC2-5300 spdmem0: 14 rows, 10 cols, 2 ranks, 4 banks/chip, 3.00ns cycle time spdmem0: voltage SSTL 1.8V, refresh time 7.8us (self-refreshing) nfsmb1 at nfsmbc0 SMBus 2 iic1 at nfsmb1: I2C bus
1.1	11-Jul-2007	kiyohara	branches: 1.1.2; 1.1.4; 1.1.10; 1.1.12; 1.1.14; 1.1.20; 1.1.22; Add support for NVIDIA nForce 2/3/4 SMBus controller and SMBus driver.
1.1.22.1	08-Dec-2007	ad	Sync with head.
1.1.20.1	08-Dec-2007	mjf	Sync with HEAD.
1.1.14.3	07-Dec-2007	yamt	sync with head
1.1.14.2	03-Sep-2007	yamt	sync with head.
1.1.14.1	11-Jul-2007	yamt	file nfsmbreg.h was added on branch yamt-lazymbuf on 2007-09-03 14:37:14 +0000
1.1.12.1	09-Jan-2008	matt	sync with HEAD
1.1.10.1	09-Dec-2007	jmcneill	Sync with HEAD.
1.1.4.2	15-Jul-2007	ad	Sync with head.
1.1.4.1	11-Jul-2007	ad	file nfsmbreg.h was added on branch vmlocking on 2007-07-15 13:21:37 +0000
1.1.2.2	11-Jul-2007	mjf	Sync with head.
1.1.2.1	11-Jul-2007	mjf	file nfsmbreg.h was added on branch mjf-ufs-trans on 2007-07-11 20:07:54 +0000
1.13	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.12	14-Jul-2016	msaitoh	branches: 1.12.16; 1.12.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.11	11-Jul-2016	msaitoh	KNF. No functional change.
1.10	29-Mar-2014	christos	branches: 1.10.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.9	09-Jul-2008	joerg	branches: 1.9.28; 1.9.38; 1.9.44; - device/softc split
1.8	28-Apr-2008	martin	branches: 1.8.2; 1.8.4; 1.8.6; Remove clause 3 and 4 from TNF licenses
1.7	10-Apr-2008	cegger	branches: 1.7.2; 1.7.4; use aprint_*_dev and device_xname
1.6	19-Oct-2007	ad	branches: 1.6.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.5	16-Nov-2006	christos	branches: 1.5.8; 1.5.22; 1.5.24; 1.5.28; __unused removal on arguments; approved by core.
1.4	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.3	11-Dec-2005	christos	branches: 1.3.20; 1.3.22; merge ktrace-lwp.
1.2	26-Aug-2004	thorpej	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.18; Use ANSI function decls, more static.
1.1	26-Aug-2004	itohy	Workbit NinjaSCSI-32 PCI/CardBus SCSI driver (njs)
1.2.18.2	27-Oct-2007	yamt	sync with head.
1.2.18.1	30-Dec-2006	yamt	sync with head.
1.2.6.2	11-Nov-2004	he	Pull up revisions 1.1-1.2 (via patch, requested by itohy in ticket #1741): Add support for the Workbit NinjaSCSI-32 PCI/Cardbus SCSI driver, njs(4).
1.2.6.1	26-Aug-2004	he	file njs_pci.c was added on branch netbsd-1-6 on 2004-11-11 23:21:34 +0000
1.2.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.2.4.2	03-Sep-2004	skrll	Sync with HEAD
1.2.4.1	26-Aug-2004	skrll	file njs_pci.c was added on branch ktrace-lwp on 2004-09-03 12:45:28 +0000
1.2.2.3	30-Aug-2004	tron	Pull up revision 1.2 (requested by itohy in ticket #800): Use ANSI function decls, more static.
1.2.2.2	30-Aug-2004	tron	Pull up revision 1.1 (requested by itohy in ticket #800): Workbit NinjaSCSI-32 PCI/CardBus SCSI driver (njs)
1.2.2.1	26-Aug-2004	tron	file njs_pci.c was added on branch netbsd-2-0 on 2004-08-30 09:24:58 +0000
1.3.22.2	10-Dec-2006	yamt	sync with head.
1.3.22.1	22-Oct-2006	yamt	sync with head
1.3.20.1	18-Nov-2006	ad	Sync with head.
1.5.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.5.24.1	06-Nov-2007	matt	sync with HEAD
1.5.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.5.8.1	23-Oct-2007	ad	Sync with head.
1.6.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.6.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.7.4.2	04-May-2009	yamt	sync with head.
1.7.4.1	16-May-2008	yamt	sync with head.
1.7.2.1	18-May-2008	yamt	sync with head.
1.8.6.1	19-Oct-2008	haad	Sync with HEAD.
1.8.4.1	18-Jul-2008	simonb	Sync with head.
1.8.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.9.44.1	18-May-2014	rmind	sync with head
1.9.38.2	03-Dec-2017	jdolecek	update from HEAD
1.9.38.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.28.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.10.6.1	05-Oct-2016	skrll	Sync with HEAD
1.12.18.1	10-Jun-2019	christos	Sync with HEAD
1.12.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.10	04-Jan-2017	christos	PR/51774: David Binderman: Remove unused variable.
1.9	07-Oct-2013	jakllsch	branches: 1.9.6; 1.9.10; Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.8	07-Oct-2013	jakllsch	Remove unused structure.
1.7	31-Jul-2012	bouyer	branches: 1.7.2; 1.7.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.6	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.5	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.4	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.3	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.2	04-Apr-2011	dyoung	branches: 1.2.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.1	10-Nov-2010	skrll	branches: 1.1.2; 1.1.6; Add nside(4) - a driver for the National Semiconductor PC87415 IDE controller as found in many HP PA-RISC machines. From OpenBSD. Reviewed by Manuel Bouyer.
1.1.6.3	21-Apr-2011	rmind	sync with head
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	10-Nov-2010	rmind	file nside.c was added on branch rmind-uvmplock on 2011-03-05 20:53:47 +0000
1.1.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.2.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.1	30-Oct-2012	yamt	sync with head
1.7.4.1	18-May-2014	rmind	sync with head
1.7.2.3	03-Dec-2017	jdolecek	update from HEAD
1.7.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.7.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.9.10.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.9.6.1	05-Feb-2017	skrll	Sync with HEAD
1.37	15-Aug-2022	pgoyette	Remove stray debug line (already commented out)
1.36	09-Jul-2022	pgoyette	Remove redundant ``devsw_ok = false;'' Thanks mrg@
1.35	08-Jul-2022	pgoyette	devsw_ok needs to survive across invocations of nvme_modcmd() so allocate it statically. Should address remaining issues with kern/56914
1.34	08-Jul-2022	pgoyette	Initialize cmajor & bmajor so the devsw_attach() has a chance of succeeding. Record the success of devsw_attach(), and do not try later to devsw_detach() unless the attach succeeded. Partial fix for kern/56914
1.33	07-Jul-2022	pgoyette	Add errno tto message when fail to attach the devsw
1.32	31-Mar-2022	pgoyette	For device modules that provide both auto-config and /dev/xxx interfaces, make sure that initialization and destruction follow the proper sequence. This is triggered by the recent changes to the devsw stuff; per riastradh@ the required call sequence is: devsw_attach() config_init_component() or config_cf_attach() ... config_fini_component() or config_cf_detach() devsw_detach() While here, add a few missing calls to some of the detach routines. Testing of these changes has been limited to: 1. compile without build break 2. no related test failures from atf 3. modload/modunload work as well as before. No functional device testing done, since I don't have any of these devices. Let me know of any damage I might cause here! XXX Some of the modules affected by this commit are already XXX broken; see kern/56772. This commit does not break any additional modules (as far as I know).
1.31	10-Oct-2021	skrll	Catch up with pcireg.h change
1.30	29-May-2021	riastradh	nvme(4): Add suspend/resume, derived from OpenBSD.
1.29	29-Jul-2020	jdolecek	branches: 1.29.6; 1.29.8; rename label now that it has dual purpose; suggested by mrg@
1.28	28-Jul-2020	jdolecek	add a quirk to disable MSI, and enable it for Intel SSD DC P4500 this device seems to cause serious system responsiveness issues when configured to use MSI, while it works fine when configured for either INTx or MSI-X this is important so this works well under Xen Dom0, which doesn't support MSI-X yet fixes another issue reported as feedback for PR port-xen/55285 by Frank Kardel
1.27	02-Dec-2019	msaitoh	Use PCI_MSIX_"TBL"BIR_MASK instead of PCI_MSIX_"PBA"BIR_MASK for MSI-X table. This is not a real bug because both macros have the same value.
1.26	23-Jan-2019	msaitoh	branches: 1.26.4; Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them expilicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.25	07-Dec-2018	msaitoh	No functional change. OK'd by nonaka: - Use correct macro. - Don't use magic number.
1.24	01-Dec-2018	jdolecek	fix pasto (MSI -> MSIX)
1.23	01-Dec-2018	jdolecek	allow MSI-X even with just one interrupt, on some systems MSI/INTx might not be available (or working, for that matter)
1.22	01-Dec-2018	jdolecek	simplify the interrupt allocation - it's not necessary to do the explicit fallbacks, pci_intr_alloc() does this already internally
1.21	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.20	18-Apr-2018	nonaka	branches: 1.20.2; nvme(4): Added some delay before check RDY bit quirk when disabling device. Pick from FreeBSD nvme(4) r326937.
1.19	01-Jun-2017	chs	branches: 1.19.2; 1.19.8; 1.19.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.18	13-Feb-2017	nonaka	remove extra crlf.
1.17	25-Nov-2016	knakahara	branches: 1.17.2; "#ifdef __HAVE_PCI_MSI_MSIX" workaround is not required any more. see sys/dev/pci/pci_stub.c r1.6
1.16	19-Oct-2016	jdolecek	follow advice of spec and block interrupts via INTMS/INTMC for intx handler; this also makes it possible to offload the actual interrupt processing to softintr handler, similar as for MSI/MSI-X
1.15	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.14	18-Sep-2016	jdolecek	fix several bugs, make nvme(4) MPSAFE by default and also bump default number of ioq from 128 to 1024; tested with VirtualBox and QEMU * remove NVME_INTMC/NVME_INTMS writes in hw intr handler as this is not MPSAFE, fortunately they don't seem to be necessary; shaves two register writes * need to use full mutex_enter() in nvme_q_complete(), to avoid small race between one handler exiting the loop and another entering * for MSI, handover the command result processing to softintr; unfortunately can't easily do that for INTx interrupts as they require doorbell write to deassert * unlock/relock q->q_cq_mtx before calling ccb_done to avoid potential deadlocks * make sure to destroy queue mutexes when destroying the queue (LOCKDEBUG) * make ns ctx pool per-device, so that it's deallocated properly on module unload * handle ctx allocation failure in ld_nvme_dobio() * remove splbio() calls in ld_nvme_dobio() and sync, the paths are exercised only for dump/shutdown, and that already disables interrupts * free the ns ctx in ld_nvme_biodone() before calling lddone() to avoid memory starvation, as lddone() can trigger another i/o request * be more careful with using PR_WAITOK, the paths are called from interrupt context and there we can't wait
1.13	18-Sep-2016	jdolecek	attach also the nvme_cdevsw when loaded as module, so /dev/nvme* works
1.12	17-Sep-2016	jdolecek	remove dev code included in previous commit by mistake
1.11	17-Sep-2016	jdolecek	#ifdef out MSI related code on platforms not supporting MSI, so that it's possible to load the driver as module on those platforms (the weak symbols are not found by module load, claiming they don't exist); this makes it possible to load the driver on e.g. sparc64, which is supposed to be working under OpenBSD unfortunately QEMU sparc64 emulator starts causing data access errors on first device register read in nvme_attach(), so can't confirm the driver is actually working on sparc64; same happens in QEMU when booting OpenBSD image, so it seems to be emulator bug
1.10	17-Sep-2016	jdolecek	actually detach the ld {b,c}devsw on unload, and fail load if it's already present
1.9	17-Sep-2016	jdolecek	need to set PCI_COMMAND_MASTER_ENABLE flag; fixes the driver under QEMU
1.8	17-Sep-2016	pgoyette	Since the ld code included with the nvme module references stuff in the dk_subr module, add a dependency.
1.7	16-Sep-2016	pgoyette	We need sys/module.h for built-in modules as well as for the .kmod so #include it unconditionally.
1.6	16-Sep-2016	jdolecek	make it possible to load nvme(4) as module to ease testing; currently somewhat non-optimal, since it includes the ld(4) code also and hence requires the kernel config to have 'no ld'
1.5	16-Sep-2016	jdolecek	fix nvme_pci_detach() to not pci_intr_disestablish(); the interrupts are disestablished and handler pointers are set to NULL already by nvme_detach(), leading to panic here; this was happening when running on system without MSI/MSI-X change all the alloc/dealloc code for sc_ih to use psc->psc_nintrs, so that we avoid some duplicated logic
1.4	11-Jul-2016	knakahara	branches: 1.4.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.3	24-Jun-2016	nonaka	fix free memory size.
1.2	11-May-2016	nonaka	branches: 1.2.2; Don't establish interrupt in nvme_pci_attach() when multi queue isn't used.
1.1	01-May-2016	nonaka	Added nvme(4) for Non-Volatile Memory Host Controller Interface devices. Ported from OpenBSD.
1.2.2.6	28-Aug-2017	skrll	Sync with HEAD
1.2.2.5	05-Dec-2016	skrll	Sync with HEAD
1.2.2.4	05-Oct-2016	skrll	Sync with HEAD
1.2.2.3	09-Jul-2016	skrll	Sync with HEAD
1.2.2.2	29-May-2016	skrll	Sync with HEAD
1.2.2.1	11-May-2016	skrll	file nvme_pci.c was added on branch nick-nhusb on 2016-05-29 08:44:22 +0000
1.4.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.4.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.4.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.17.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.19.10.4	26-Jan-2019	pgoyette	Sync with HEAD
1.19.10.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.19.10.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.19.10.1	22-Apr-2018	pgoyette	Sync with HEAD
1.19.8.2	03-Dec-2017	jdolecek	update from HEAD
1.19.8.1	01-Jun-2017	jdolecek	file nvme_pci.c was added on branch tls-maxphys on 2017-12-03 11:37:08 +0000
1.19.2.3	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1714: sys/dev/pci/pcireg.h 1.148-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.217-1.222, 1.224, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158 sys/dev/pci/ppb.c 1.74 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. - Change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Remove unused shift and mask definitions. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.19.2.2	27-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1172): sys/dev/pci/nvme_pci.c: revision 1.26 sys/dev/pci/xhci_pci.c: revision 1.21 sys/dev/pci/ixgbe/ixv.c: revision 1.108 sys/dev/pci/ixgbe/ixgbe.c: revision 1.171 sys/dev/pci/if_fxp_pci.c: revision 1.84 sys/dev/pci/if_fxp_pci.c: revision 1.85 sys/dev/pci/xhci_pci.c: revision 1.16 remove #ifdef DEBUG printf, it seems to have outlived it's usefulness - KNF. No functional change. - Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them explicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.19.2.1	19-Apr-2018	martin	Pull up following revision(s) (requested by nonaka in ticket #781): sbin/nvmectl/Makefile 1.4 sbin/nvmectl/bignum.c 1.2 sbin/nvmectl/devlist.c 1.3-1.5 sbin/nvmectl/firmware.c 1.3,1.4 sbin/nvmectl/identify.c 1.3-1.5 sbin/nvmectl/logpage.c 1.5-1.7 sbin/nvmectl/nvme.h 1.3 sbin/nvmectl/nvmectl.8 1.5 sbin/nvmectl/nvmectl.c 1.5-1.7 sbin/nvmectl/nvmectl.h 1.5-1.8 sbin/nvmectl/perftest.c 1.3-1.5 sbin/nvmectl/power.c 1.3,1.4 sbin/nvmectl/reset.c 1.2,1.3 sbin/nvmectl/util.c 1.1,1.2 sbin/nvmectl/wdc.c 1.2-1.4 sys/dev/ic/ld_nvme.c 1.20 sys/dev/ic/nvme.c 1.38,1.39 sys/dev/ic/nvmeio.h 1.2 sys/dev/ic/nvmereg.h 1.10,1.11 sys/dev/ic/nvmevar.h 1.16 sys/dev/pci/nvme_pci.c 1.20 nvmectl(8): Sync with FreeBSD nvmecontrol(8) r328763. nvmectl(8): fix wdc command usage. nvme(4): Added some delay before check RDY bit quirk when disabling device. Pick from FreeBSD nvme(4) r326937. Add some new structure fileds, opcodes and statuses from NVMe 1.3a. nvmectl(8): Add big-endian support. from FreeBSD nvmecontolr(8) r329824. nvmectl(8): fix subcommand usage. nvmectl(8): Remove some wdc subcommands from man page. - wdc drive-log - wdc get-crash-dump - wdc purge - wdc purge-monitor Typos. use setprogname()/getprogname(), do not hardcode the prognam name in fixed strings
1.20.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.20.2.1	10-Jun-2019	christos	Sync with HEAD
1.26.4.2	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1384: sys/dev/pci/pcireg.h 1.152-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.222, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158, 1.163 sys/dev/pci/ppb.c 1.74 - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.26.4.1	21-Jun-2021	martin	Pull up following revision(s) (requested by riastradh in ticket #1305): sys/dev/ic/nvmevar.h: revision 1.22 sys/dev/ic/nvme.c: revision 1.56 sys/dev/ic/nvme.c: revision 1.57 sys/dev/pci/nvme_pci.c: revision 1.30 nvme(4): Add suspend/resume, derived from OpenBSD. nvme(4): Move disestablishment of admin q interrupt to nvme_detach. Nothing re-established this after suspend/resume, so attempting suspend/resume/suspend would crash, and presumably we would miss interrupts after resume. This keeps the establish/disestablish more symmetric in attach/detach.
1.29.8.1	31-May-2021	cjep	sync with head
1.29.6.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.51	26-Sep-2021	thorpej	- Use seltrue_filtops rather than rolling our own with filt_seltrue. - Remove sc_wsel completely; nothing actually uses it.
1.50	26-Sep-2021	thorpej	Change the kqueue filterops::f_isfd field to filterops::f_flags, and define a flag FILTEROP_ISFD that has the meaning of the prior f_isfd. Field and flag name aligned with OpenBSD. This does not constitute a functional or ABI change, as the field location and size, and the value placed in that field, are the same as the previous code, but we're bumping __NetBSD_Version__ so 3rd-party module source code can adapt, as needed. NetBSD 9.99.89
1.49	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.48	24-Apr-2021	thorpej	branches: 1.48.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.47	18-Dec-2020	thorpej	branches: 1.47.2; Use sel{record,remove}_knote().
1.46	09-Dec-2018	jdolecek	branches: 1.46.12; use pci_intr_establish_xname() everywhere
1.45	25-Oct-2017	maya	branches: 1.45.2; 1.45.4; Use C99 initializer for filterops Mostly done with spatch with touchups for indentation @@ expression a; identifier b,c,d; identifier p; @@ const struct filterops p = - { a, b, c, d + { + .f_isfd = a, + .f_attach = b, + .f_detach = c, + .f_event = d, };
1.44	20-Aug-2017	maxv	M_WAITOK cannot fail, so remove the test branches. Otherwise it looks like leak/uninitialized area.
1.43	24-Jul-2015	martin	Pass the device, not the struct softc to config_found(). Avoids a crash at attach time, PR port-i386/50076.
1.42	29-Mar-2014	christos	branches: 1.42.4; 1.42.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.41	20-Mar-2014	christos	fix format confusion
1.40	20-Mar-2014	christos	make it compile on amd64
1.39	27-Oct-2012	chs	branches: 1.39.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.38	13-Nov-2010	uebayasi	branches: 1.38.8; 1.38.18; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.37	26-Nov-2009	njoly	branches: 1.37.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.36	12-Nov-2009	dyoung	Remove superfluous activation hook.
1.35	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.34	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.33	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.32	28-Apr-2008	martin	branches: 1.32.14; Remove clause 3 and 4 from TNF licenses
1.31	10-Apr-2008	cegger	branches: 1.31.2; 1.31.4; use aprint_*_dev and device_xname
1.30	01-Mar-2008	rmind	Welcome to 4.99.55: - Add a lot of missing selinit() and seldestroy() calls. - Merge selwakeup() and selnotify() calls into a single selnotify(). - Add an additional 'events' argument to selnotify() call. It will indicate which event (POLL_IN, POLL_OUT, etc) happen. If unknown, zero may be used. Note: please pass appropriate value of 'events' where possible. Proposed on: <tech-kern>
1.29	15-Dec-2007	perry	branches: 1.29.2; 1.29.6; __FUNCTION__ -> __func__
1.28	05-Dec-2007	pooka	branches: 1.28.4; Do not "return 1" from kqfilter for errors. That value is passed directly to the userland caller and results in a mysterious EPERM. Instead, return EINVAL or something else sensible depending on the case.
1.27	19-Oct-2007	ad	branches: 1.27.2; 1.27.4; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.26	09-Jul-2007	ad	branches: 1.26.6; 1.26.8; 1.26.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.25	04-Mar-2007	christos	branches: 1.25.2; 1.25.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.24	16-Nov-2006	christos	branches: 1.24.4; __unused removal on arguments; approved by core.
1.23	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.22	17-Aug-2006	he	branches: 1.22.2; 1.22.4; Make this build again without OBOE_DEBUG defined (get rid of warnings about unused variables).
1.21	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.20	11-Dec-2005	christos	branches: 1.20.4; 1.20.8; merge ktrace-lwp.
1.19	28-Jun-2005	thorpej	branches: 1.19.2; Use ANSI function decls and static.
1.18	27-Feb-2005	perry	nuke trailing whitespace
1.17	28-Oct-2003	mycroft	branches: 1.17.8; 1.17.10; Remove unnecessary code.
1.16	25-Oct-2003	christos	Fix uninitialized variable warnings
1.15	21-Oct-2003	fvdl	And the wise man said: An u_int32_t is not a pointer, my young apprentice, so thou shalt not compare it to NULL.
1.14	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.13	29-Jun-2003	fvdl	branches: 1.13.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.12	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.11	26-Nov-2002	christos	si_ -> sel_
1.10	26-Nov-2002	fvdl	Add a comment marking this file as broken on most non-i386 -- vtophys usage.
1.9	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.8	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.7	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.6	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.5	11-Dec-2001	augustss	branches: 1.5.2; 1.5.4; Recognized Danube chip as well. Small cleanup.
1.4	05-Dec-2001	augustss	Do SIR framing on incoming frames.
1.3	04-Dec-2001	augustss	Simplify interface to irframe.
1.2	02-Dec-2001	augustss	Touch up a little. No functional change.
1.1	02-Dec-2001	augustss	Add driver for the Toshiba's Oboe IrDA chip. From Jan Sparud.
1.5.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.5.4.3	02-Oct-2002	jdolecek	do not need the (void *) cast for kn_hook anymore
1.5.4.2	10-Jan-2002	thorpej	Add kqueue support to the IrDA framework.
1.5.4.1	11-Dec-2001	thorpej	file oboe.c was added on branch kqueue on 2002-01-10 18:29:10 +0000
1.5.2.5	11-Dec-2002	thorpej	Sync with HEAD.
1.5.2.4	11-Nov-2002	nathanw	Catch up to -current
1.5.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.5.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.5.2.1	11-Dec-2001	nathanw	file oboe.c was added on branch nathanw_sa on 2002-01-08 00:31:08 +0000
1.13.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.13.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.13.2.2	03-Aug-2004	skrll	Sync with HEAD
1.13.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.17.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.17.8.1	29-Apr-2005	kent	sync with -current
1.19.2.7	17-Mar-2008	yamt	sync with head.
1.19.2.6	21-Jan-2008	yamt	sync with head
1.19.2.5	07-Dec-2007	yamt	sync with head
1.19.2.4	27-Oct-2007	yamt	sync with head.
1.19.2.3	03-Sep-2007	yamt	sync with head.
1.19.2.2	30-Dec-2006	yamt	sync with head.
1.19.2.1	21-Jun-2006	yamt	sync with head.
1.20.8.1	03-Sep-2006	yamt	sync with head.
1.20.4.1	09-Sep-2006	rpaulo	sync with head
1.22.4.2	10-Dec-2006	yamt	sync with head.
1.22.4.1	22-Oct-2006	yamt	sync with head
1.22.2.1	18-Nov-2006	ad	Sync with head.
1.24.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.25.4.1	11-Jul-2007	mjf	Sync with head.
1.25.2.2	23-Oct-2007	ad	Sync with head.
1.25.2.1	05-Apr-2007	ad	Compile fixes.
1.26.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.26.8.3	23-Mar-2008	matt	sync with HEAD
1.26.8.2	09-Jan-2008	matt	sync with HEAD
1.26.8.1	06-Nov-2007	matt	sync with HEAD
1.26.6.2	09-Dec-2007	jmcneill	Sync with HEAD.
1.26.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.27.4.2	26-Dec-2007	ad	Sync with head.
1.27.4.1	08-Dec-2007	ad	Sync with head.
1.27.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.27.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.28.4.1	02-Jan-2008	bouyer	Sync with HEAD
1.29.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.29.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.29.2.1	24-Mar-2008	keiichi	sync with head.
1.31.4.3	11-Mar-2010	yamt	sync with head
1.31.4.2	16-May-2009	yamt	sync with head
1.31.4.1	16-May-2008	yamt	sync with head.
1.31.2.1	18-May-2008	yamt	sync with head.
1.32.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.37.4.1	05-Mar-2011	rmind	sync with head
1.38.18.3	03-Dec-2017	jdolecek	update from HEAD
1.38.18.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.38.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.38.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.38.8.1	30-Oct-2012	yamt	sync with head
1.39.2.1	18-May-2014	rmind	sync with head
1.42.6.2	28-Aug-2017	skrll	Sync with HEAD
1.42.6.1	22-Sep-2015	skrll	Sync with HEAD
1.42.4.1	30-Jul-2015	snj	Pull up following revision(s) (requested by martin in ticket #887): sys/dev/pci/oboe.c: revision 1.43 Pass the device, not the struct softc to config_found(). Avoids a crash at attach time, PR port-i386/50076.
1.45.4.1	10-Jun-2019	christos	Sync with HEAD
1.45.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.46.12.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.47.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.48.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.1	02-Dec-2001	augustss	branches: 1.1.2; 1.1.4; 1.1.118; 1.1.120; 1.1.122; Add driver for the Toshiba's Oboe IrDA chip. From Jan Sparud.
1.1.122.1	16-May-2008	yamt	sync with head.
1.1.120.1	18-May-2008	yamt	sync with head.
1.1.118.1	02-Jun-2008	mjf	Sync with HEAD.
1.1.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.1	02-Dec-2001	thorpej	file oboereg.h was added on branch kqueue on 2002-01-10 19:56:51 +0000
1.1.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.1	02-Dec-2001	nathanw	file oboereg.h was added on branch nathanw_sa on 2002-01-08 00:31:08 +0000
1.61	31-Mar-2025	riastradh	ohci@pci: Don't detach on shutdown. If there's some USB device that requires detach on shutdown (which would be surprising, because USB devices should generally be safe to yank at any time, and _storage_ drivers such as ld(4) or sd(4) should already be synced at shutdown some other way) then that driver should have DVF_DETACH_SHUTDOWN -- not the entire HCI. PR port-amd64/59180: System reboots instead of shutting down PR kern/59179: DVF_DETACH_SHUTDOWN: audit drivers for unnecessary use
1.60	31-Mar-2025	riastradh	ohci(4): Rework detach logic and justify the ordering. Handle failed attach when we detach. This changes the signature of the ohci_detach function, but it is only ever used by statically linked ohci bus attachments, never by modules so far, so no kernel revbump. PR port-amd64/59180: System reboots instead of shutting down
1.59	07-Aug-2021	thorpej	branches: 1.59.12; Merge thorpej-cfargs2.
1.58	24-Apr-2021	thorpej	branches: 1.58.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.57	09-Apr-2018	jakllsch	branches: 1.57.18; Stop potential misuse of vendor names and USB vendor IDs in root hub device and string descriptors. Firstly: Few vendors have identical PCI-SIG vendor IDs and USB-IF vendor IDs. As such, using the PCI vendor ID as a USB vendor ID may trample on whomever is allocated that USB vendor ID. Secondly: The vendor of the host controller hardware implementation has little to nothing to do with our usbroothub implementation. Thus we should not potentially associate any problems therewith to such third party. This change will result in root hubs being identified by USB Vendor ID 0x0000. Root hub vendor string will now be "NetBSD" (or, specifically: ostype). Product ID (0x0000) and product strings remain unchanged.
1.56	10-May-2017	msaitoh	branches: 1.56.8; Use pci_intr_establish_xname().
1.55	23-Apr-2016	skrll	branches: 1.55.8; Merge nick-nhusb - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix
1.54	19-Aug-2015	skrll	Interrupt handlers aren't MP-safe yet so use IPL_USB which is IPL_VM which takes the kernel lock.
1.53	21-Sep-2014	christos	branches: 1.53.2; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.52	29-Mar-2014	christos	branches: 1.52.4; 1.52.8; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.51	28-Jan-2014	skrll	Support the OHCI in the NS "SUPERIO" chip as found in hppa machines by twiddling the same bits as Linux.
1.50	10-Jun-2012	mrg	branches: 1.50.2; 1.50.4; merge the jmcneill-usbmp branch. many thanks to jared for the initial work, and every one else who has tested things for me. this is largely my fault at this point :-) the main changes are something like: - usbd_bus_methods{} gains a get_lock() to enable the host controller to provide a lock for the USB code. if the lock isn't provided, old-style protection is (partially) applied. - ehci/ohci/uhci have been converted to the new interfaces, including mutex/cv/etc conversion. - usbdivar.h contains a discussion about locking and what locks are held for which method calls. more to come for usbdi(9) here. - audio drivers (uaudio, umidi, auvitek) have been properly SMPified now that USB is ready. - scsi drivers have been modified to take the kernel lock explicitly before calling into scsi code. - usb pipes are associated with a lock, that is the same as the controller lock. (this could be split up further in the future.) - several usbfoo_locked() or usbfoo_unlocked() functions have been added to the usbdi(9) to enable functionality with or without the USB lock (per controller) already being held. the TODO.usbmp file has specific details on what is left to do, including what device-specific changes should be done now that the whole framework is ready.
1.49	05-Apr-2012	macallan	abort attach if memory access is not enabled this avoids a panic later on when trying to attach to an unconfigured ohci found for example in later iBooks when DIAGNOSTIC is set
1.48	30-Jan-2012	drochner	branches: 1.48.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.47	04-Apr-2011	dyoung	branches: 1.47.4; 1.47.8; Deregister PMF handlers during detachment. Complete the host controller shutdown.
1.46	25-May-2010	pgoyette	branches: 1.46.2; Revert earlier changes to callers of pci_find{vendor,product} since the original routine entrypoint name has been restored. Requested by dyoung@ and mrg@
1.45	24-May-2010	pgoyette	Update all callers of the pci_find{vendor,product} routines to now call these routines through their global pointers.
1.44	11-Mar-2010	mrg	branches: 1.44.2; various aprint_* fixes.
1.43	26-Apr-2009	cegger	branches: 1.43.2; detach ohci on shutdown
1.42	26-Apr-2009	cegger	Fix error handling.
1.41	17-Apr-2009	dyoung	Use device_t instead of 'struct device ' or device_ptr_t. Use cfdata_t instead of 'struct cfdata '. Delete some gratuitous parentheses.
1.40	17-Apr-2009	dyoung	Disable interrupts on the h/w before disestablishing the PCI interrupt handler.
1.39	28-Apr-2008	martin	branches: 1.39.8; 1.39.14; Remove clause 3 and 4 from TNF licenses
1.38	29-Mar-2008	nakayama	branches: 1.38.2; 1.38.4; Fix ohci brokenness in previous commit.
1.37	28-Mar-2008	drochner	split device/softc for USB host controllers and the usb (control) device, this is hairy stuff, and I've only tested with uhci/ehci at pci, please test the rest and report problems
1.36	07-Mar-2008	dyoung	Use device_t and accessors. In ohci_{pci,cardbus}_attach, do not get the device_t by casting the softc to void *, but use `self' instead.
1.35	07-Mar-2008	dyoung	Register the shutdown hook using pmf_device_register1().
1.34	22-Feb-2008	dyoung	Add method to detach children. Temporary: use PMF_FN_PROTO, PMF_FN_ARGS.
1.33	09-Dec-2007	jmcneill	branches: 1.33.6; 1.33.10; Merge jmcneill-pm branch.
1.32	19-Oct-2007	ad	branches: 1.32.4; 1.32.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.31	16-Nov-2006	christos	branches: 1.31.8; 1.31.18; 1.31.24; 1.31.26; 1.31.30; __unused removal on arguments; approved by core.
1.30	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.29	11-Dec-2005	christos	branches: 1.29.20; 1.29.22; merge ktrace-lwp.
1.28	28-Jun-2005	thorpej	branches: 1.28.2; Use ANSI function decls and static.
1.27	27-Feb-2005	perry	nuke trailing whitespace
1.26	02-Aug-2004	mycroft	branches: 1.26.4; 1.26.6; For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.25	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.24	22-Apr-2004	itojun	sprintf -> snprintf
1.23	02-Oct-2002	thorpej	branches: 1.23.6; Add trailing ; to CFATTACH_DECL.
1.22	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.21	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.20	13-Nov-2001	lukem	add RCSID
1.19	06-Nov-2001	augustss	Add some data structures to keep enable finding companion controllers for EHCI.
1.18	28-Dec-2000	sommerfeld	branches: 1.18.2; 1.18.4; 1.18.8; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.17	06-Sep-2000	augustss	ANSIfy.
1.16	27-Apr-2000	augustss	Change my email address.
1.15	25-Apr-2000	augustss	Move the size of the mapped bus_space region into the bus independent softc.
1.14	12-Oct-1999	augustss	branches: 1.14.2; Deallocate resources (bus space and interrupts) when detaching from the bus.
1.13	15-Sep-1999	augustss	branches: 1.13.2; Add preliminary (untested) code for detaching the USB host controller (needed for CardBus based controllers).
1.12	14-Sep-1999	augustss	And commit two files I forgot in the rearrangement for testing for interrupt context.
1.11	04-Sep-1999	augustss	Make sure not to use the io handle before it's been initialized.
1.10	21-Aug-1999	augustss	Disable interrupts early.
1.9	20-May-1999	augustss	Keep track of vendor id and make sure the vendor string is NUL terminated.
1.8	13-May-1999	thorpej	Rework the way ukbd attaches itself as the console (again). We now allow the code to pick the first USB keyboard instance as the console, ignoring which USB controller it's on. Should eventually allow detaching of the console keyboard. From Jason Thorpe <thorpej@nas.nasa.gov>
1.7	06-May-1999	thorpej	Add a way for machine-dependent code to tell a USB controller that it has the console input device. The USB keyboard driver uses this to attach the first USB keyboard instance as the console keyboard. Unfortunately, this must still be deferred to autoconfiguration time, but there's not much we can do about that right now.
1.6	27-Dec-1998	augustss	branches: 1.6.2; Fold lines to 80 chars.
1.5	25-Nov-1998	augustss	Make the copyright header conform to the NetBSD template.
1.4	25-Jul-1998	augustss	Use PCI class/subclass/interface to match USB host controllers instead of vendor/product id. This makes the probe find more controllers without recompilation.
1.3	25-Jul-1998	augustss	I forgot to commit these when usb_mem was added.
1.2	22-Jul-1998	augustss	Add an Acer labs USB controller.
1.1	12-Jul-1998	augustss	Add USB support. Supported so far: * UHCI and OHCI host controllers on PCI * Hubs * HID devices withe special drivers for mouse and keyboard * Printers
1.6.2.2	28-Jun-1999	perry	pullup 1.7->1.8 (wrstuden)
1.6.2.1	06-May-1999	perry	branches: 1.6.2.1.2; 1.6.2.1.4; pullup 1.6->1.7 (thorpej)
1.6.2.1.4.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.6.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.13.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.14.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.14.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.18.8.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.18.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.18.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.2.2	18-Oct-2002	nathanw	Catch up to -current.
1.18.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.23.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.23.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.23.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.23.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.23.6.1	03-Aug-2004	skrll	Sync with HEAD
1.26.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.26.4.1	29-Apr-2005	kent	sync with -current
1.28.2.5	17-Mar-2008	yamt	sync with head.
1.28.2.4	27-Feb-2008	yamt	sync with head.
1.28.2.3	21-Jan-2008	yamt	sync with head
1.28.2.2	27-Oct-2007	yamt	sync with head.
1.28.2.1	30-Dec-2006	yamt	sync with head.
1.29.22.2	10-Dec-2006	yamt	sync with head.
1.29.22.1	22-Oct-2006	yamt	sync with head
1.29.20.1	18-Nov-2006	ad	Sync with head.
1.31.30.1	25-Oct-2007	bouyer	Sync with HEAD.
1.31.26.3	23-Mar-2008	matt	sync with HEAD
1.31.26.2	09-Jan-2008	matt	sync with HEAD
1.31.26.1	06-Nov-2007	matt	sync with HEAD
1.31.24.5	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.31.24.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.31.24.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.31.24.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.31.24.1	14-Aug-2007	jmcneill	Add ohci power management support, currently untested.
1.31.18.1	22-May-2007	itohy	Overhaul of USB stack, mostly DMA related This applies to NetBSD 4.99.13 (March 1, 2007) usbdi(9) interface is based on FreeBSD version, excluding - removal of portability code Patch most NetBSD changes, excluding - DMA memory "reserve", since we don't need contiguous buffers any longer - volatiles in DMA structure, since it should not be needed with proper bus_dmamap_sync(9)s DMA/non-DMA memory management overhaul - Move all DMA related code to usb_mem.[ch] (add usb_alloc_buffer_dma(), usb_free_buffer_dma(), etc.). XXX Should usb_mem.[ch] be renamed as usb_mem_dma.[ch] ? - Add corresponding non-DMA code to usb_mem_nodma.[ch] . Currently just use malloc(9). - Above files are conditionally used by config framework (added attributes to conf/files and dev/usb/files.usb). - Add diagnostic panics when resource allocation is requested on interrupt context. - Change memory allocations (that require context) from NOWAIT to WAITOK. Allocate DMA/non-DMA buffer per host interface, not globally. advantage: Buffers can be freed on detaching host interface. Activity of a host interface does not affect others. disadvantages: It possibly consumes more memory. API changes - usbd_alloc_xfer() is changed: old: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev); new: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev, usbd_pipe_handle pipe); - pipe argument of usbd_setup_xfer() are now unused XXX the pipe argument should be removed? - add mapping APIs - async request will be processed as a task (kernel thread context), and delayed to some extent - usbdivar.h: struct usbd_xfer: renamed a member "allocbuf" to "hcbuffer" (mapped/allocated/refered buffer for HCI driver) - usb_port.h: change usb_proc_ptr from struct ptoc to struct lwp * - usb_port.h: add usb_sigproc_ptr for psignal(9) (struct proc ) - usb.h: add UE_MAXPKTSZ(ep) and UE_MAXPKTSZ_MASK macros for USB 2.0 changes to USB device drivers - atu, aue, axe, cdce, cue, kue, rum, udav, upl, ural, url, uaudio, ubt, ucom, ugen, uhidev, uirda, ulpt, umidi, urio, uscanner, ustir, utoppy: catch up API change of usbd_alloc_xfer() - umass, usscanner: * catch up API change of usbd_alloc_xfer() * eliminate memory copy for large transfer ohci - free resources on detach - add lots of bus_dmamap_sync() operations - simplify the code of loading std chain - rewrite code of looking up TD/ITD from DMA addr by using allocation chunk - add workaround for CMD Tech 670 and 673 chipsets - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf slhci - allocate xfer and slhci_xfer at once, and simplify relevant code - add slhci_detach() - remove second arg of slhci_attach() since it is the same as the first arg. - add support for "mapping" (no, it doesn't map since it doesn't do DMA) buffer and mbuf - add pcmcia frontend - NOT TESTED, missing hardware ehci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf - done only simple test uhci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf To do - review, test, debug - rewrite network drivers to utilize usbd_map_buffer_mbuf() - rewrite uaudio(4) to eliminate memcpy - "pipe" argument of usbd_setup_*xfer() should eventually be removed
1.31.8.1	23-Oct-2007	ad	Sync with head.
1.32.6.1	11-Dec-2007	yamt	sync with head.
1.32.4.1	26-Dec-2007	ad	Sync with head.
1.33.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.33.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.33.6.1	24-Mar-2008	keiichi	sync with head.
1.38.4.3	11-Aug-2010	yamt	sync with head.
1.38.4.2	04-May-2009	yamt	sync with head.
1.38.4.1	16-May-2008	yamt	sync with head.
1.38.2.1	18-May-2008	yamt	sync with head.
1.39.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.39.8.1	28-Apr-2009	skrll	Sync with HEAD.
1.43.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.44.2.1	21-Apr-2011	rmind	sync with head
1.46.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.47.8.3	05-Apr-2012	mrg	sync to latest -current.
1.47.8.2	18-Feb-2012	mrg	merge to -current.
1.47.8.1	09-Dec-2011	mrg	- make pipe->close method take the thread lock - convert usb_taskq to use mutex/cv - convert needs_explore usage into a cv on the thread lock - remove KERNEL_*LOCK from uaudio and umidi, since we're supposedly MPSAFE here now - use IPL_SCHED instead of IPL_USB (aka biglocked) interrupts - drop the audio thread lock when calling into usb when it may sleep, avoiding a deadlock between audiowrite and audioioctl. this fixes mixerctl -a vs. playing hanging the system XXX probably need to check this in a bunch more places.
1.47.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.47.4.2	30-Oct-2012	yamt	sync with head
1.47.4.1	17-Apr-2012	yamt	sync with head
1.48.2.1	16-Apr-2012	riz	Pull up following revision(s) (requested by macallan in ticket #182): sys/dev/pci/ohci_pci.c: revision 1.49 abort attach if memory access is not enabled this avoids a panic later on when trying to attach to an unconfigured ohci found for example in later iBooks when DIAGNOSTIC is set
1.50.4.1	18-May-2014	rmind	sync with head
1.50.2.2	03-Dec-2017	jdolecek	update from HEAD
1.50.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.52.8.2	14-Sep-2016	skrll	Use IPL_USB for interrupt handler
1.52.8.1	06-Sep-2016	skrll	First pass at netbsd-7 updated with USB code from HEAD
1.52.4.1	05-Apr-2017	snj	Pull up following revision(s) (requested by skrll in ticket #1395): share/man/man4/axe.4: netbsd-7-nhusb share/man/man4/axen.4: netbsd-7-nhusb share/man/man4/cdce.4: netbsd-7-nhusb share/man/man4/uaudio.4: netbsd-7-nhusb share/man/man4/ucom.4: netbsd-7-nhusb share/man/man4/uep.4: netbsd-7-nhusb share/man/man4/urtw.4: netbsd-7-nhusb share/man/man4/usb.4: netbsd-7-nhusb share/man/man4/uyap.4: netbsd-7-nhusb share/man/man4/xhci.4: netbsd-7-nhusb share/man/man9/usbdi.9: netbsd-7-nhusb sys/arch/amd64/conf/ALL: netbsd-7-nhusb sys/arch/amd64/conf/GENERIC: netbsd-7-nhusb sys/arch/amiga/dev/slhci_zbus.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_otg.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_usb.c: netbsd-7-nhusb sys/arch/arm/amlogic/amlogic_dwctwo.c: netbsd-7-nhusb sys/arch/arm/at91/at91ohci.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm2835_dwctwo.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm53xx_usb.c: netbsd-7-nhusb sys/arch/arm/ep93xx/epohci.c: netbsd-7-nhusb sys/arch/arm/gemini/obio_ehci.c: netbsd-7-nhusb sys/arch/arm/imx/files.imx23: netbsd-7-nhusb sys/arch/arm/imx/imxusb.c: netbsd-7-nhusb sys/arch/arm/imx/imxusbreg.h: netbsd-7-nhusb sys/arch/arm/omap/obio_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/omap3_ehci.c: netbsd-7-nhusb sys/arch/arm/omap/omapl1x_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/tiotg.c: netbsd-7-nhusb sys/arch/arm/s3c2xx0/ohci_s3c24x0.c: netbsd-7-nhusb sys/arch/arm/samsung/exynos_usb.c: netbsd-7-nhusb sys/arch/arm/xscale/pxa2x0_ohci.c: netbsd-7-nhusb sys/arch/arm/zynq/zynq_usb.c: netbsd-7-nhusb sys/arch/hpcarm/dev/nbp_slhci.c: netbsd-7-nhusb sys/arch/hpcmips/dev/plumohci.c: netbsd-7-nhusb sys/arch/i386/conf/ALL: netbsd-7-nhusb sys/arch/i386/conf/GENERIC: netbsd-7-nhusb sys/arch/i386/pci/gcscehci.c: netbsd-7-nhusb sys/arch/luna68k/conf/GENERIC: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahci.c: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahcivar.h: netbsd-7-nhusb sys/arch/mips/alchemy/dev/ohci_aubus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ehci_arbus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ohci_arbus.c: netbsd-7-nhusb sys/arch/mips/conf/files.adm5120: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ehci.c: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ohci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ehci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ohci.c: netbsd-7-nhusb sys/arch/playstation2/dev/ohci_sbus.c: netbsd-7-nhusb sys/arch/powerpc/booke/dev/pq3ehci.c: netbsd-7-nhusb sys/arch/powerpc/ibm4xx/dev/dwctwo_plb.c: netbsd-7-nhusb sys/arch/x68k/dev/slhci_intio.c: netbsd-7-nhusb sys/conf/files: netbsd-7-nhusb sys/dev/cardbus/ehci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/ohci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/uhci_cardbus.c: netbsd-7-nhusb sys/dev/ic/sl811hs.c: netbsd-7-nhusb sys/dev/ic/sl811hsvar.h: netbsd-7-nhusb sys/dev/isa/slhci_isa.c: netbsd-7-nhusb sys/dev/marvell/ehci_mv.c: netbsd-7-nhusb sys/dev/pci/ehci_pci.c: netbsd-7-nhusb sys/dev/pci/ohci_pci.c: netbsd-7-nhusb sys/dev/pci/uhci_pci.c: netbsd-7-nhusb sys/dev/pci/xhci_pci.c: netbsd-7-nhusb sys/dev/pcmcia/slhci_pcmcia.c: netbsd-7-nhusb sys/dev/usb/Makefile.usbdevs: netbsd-7-nhusb sys/dev/usb/TODO: netbsd-7-nhusb sys/dev/usb/TODO.usbmp: netbsd-7-nhusb sys/dev/usb/aubtfwl.c: netbsd-7-nhusb sys/dev/usb/auvitek.c: netbsd-7-nhusb sys/dev/usb/auvitek_audio.c: netbsd-7-nhusb sys/dev/usb/auvitek_dtv.c: netbsd-7-nhusb sys/dev/usb/auvitek_i2c.c: netbsd-7-nhusb sys/dev/usb/auvitek_video.c: netbsd-7-nhusb sys/dev/usb/auvitekvar.h: netbsd-7-nhusb sys/dev/usb/ehci.c: netbsd-7-nhusb sys/dev/usb/ehcireg.h: netbsd-7-nhusb sys/dev/usb/ehcivar.h: netbsd-7-nhusb sys/dev/usb/emdtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_dtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_ir.c: netbsd-7-nhusb sys/dev/usb/emdtvvar.h: netbsd-7-nhusb sys/dev/usb/ezload.c: netbsd-7-nhusb sys/dev/usb/ezload.h: netbsd-7-nhusb sys/dev/usb/files.usb: netbsd-7-nhusb sys/dev/usb/hid.c: netbsd-7-nhusb sys/dev/usb/hid.h: netbsd-7-nhusb sys/dev/usb/if_athn_usb.c: netbsd-7-nhusb sys/dev/usb/if_athn_usb.h: netbsd-7-nhusb sys/dev/usb/if_atu.c: netbsd-7-nhusb sys/dev/usb/if_atureg.h: netbsd-7-nhusb sys/dev/usb/if_aue.c: netbsd-7-nhusb sys/dev/usb/if_auereg.h: netbsd-7-nhusb sys/dev/usb/if_axe.c: netbsd-7-nhusb sys/dev/usb/if_axen.c: netbsd-7-nhusb sys/dev/usb/if_axenreg.h: netbsd-7-nhusb sys/dev/usb/if_axereg.h: netbsd-7-nhusb sys/dev/usb/if_cdce.c: netbsd-7-nhusb sys/dev/usb/if_cdcereg.h: netbsd-7-nhusb sys/dev/usb/if_cue.c: netbsd-7-nhusb sys/dev/usb/if_cuereg.h: netbsd-7-nhusb sys/dev/usb/if_kue.c: netbsd-7-nhusb sys/dev/usb/if_kuereg.h: netbsd-7-nhusb sys/dev/usb/if_otus.c: netbsd-7-nhusb sys/dev/usb/if_otusvar.h: netbsd-7-nhusb sys/dev/usb/if_rum.c: netbsd-7-nhusb sys/dev/usb/if_rumreg.h: netbsd-7-nhusb sys/dev/usb/if_rumvar.h: netbsd-7-nhusb sys/dev/usb/if_run.c: netbsd-7-nhusb sys/dev/usb/if_runvar.h: netbsd-7-nhusb sys/dev/usb/if_smsc.c: netbsd-7-nhusb sys/dev/usb/if_smscreg.h: netbsd-7-nhusb sys/dev/usb/if_smscvar.h: netbsd-7-nhusb sys/dev/usb/if_udav.c: netbsd-7-nhusb sys/dev/usb/if_udavreg.h: netbsd-7-nhusb sys/dev/usb/if_upgt.c: netbsd-7-nhusb sys/dev/usb/if_upgtvar.h: netbsd-7-nhusb sys/dev/usb/if_upl.c: netbsd-7-nhusb sys/dev/usb/if_ural.c: netbsd-7-nhusb sys/dev/usb/if_uralreg.h: netbsd-7-nhusb sys/dev/usb/if_uralvar.h: netbsd-7-nhusb sys/dev/usb/if_url.c: netbsd-7-nhusb sys/dev/usb/if_urlreg.h: netbsd-7-nhusb sys/dev/usb/if_urndis.c: netbsd-7-nhusb sys/dev/usb/if_urndisreg.h: netbsd-7-nhusb sys/dev/usb/if_urtw.c: netbsd-7-nhusb sys/dev/usb/if_urtwn.c: netbsd-7-nhusb sys/dev/usb/if_urtwn_data.h: netbsd-7-nhusb sys/dev/usb/if_urtwnreg.h: netbsd-7-nhusb sys/dev/usb/if_urtwnvar.h: netbsd-7-nhusb sys/dev/usb/if_urtwreg.h: netbsd-7-nhusb sys/dev/usb/if_zyd.c: netbsd-7-nhusb sys/dev/usb/if_zydreg.h: netbsd-7-nhusb sys/dev/usb/irmce.c: netbsd-7-nhusb sys/dev/usb/moscom.c: netbsd-7-nhusb sys/dev/usb/motg.c: netbsd-7-nhusb sys/dev/usb/motgvar.h: netbsd-7-nhusb sys/dev/usb/ohci.c: netbsd-7-nhusb sys/dev/usb/ohcireg.h: netbsd-7-nhusb sys/dev/usb/ohcivar.h: netbsd-7-nhusb sys/dev/usb/pseye.c: netbsd-7-nhusb sys/dev/usb/slurm.c: netbsd-7-nhusb sys/dev/usb/stuirda.c: netbsd-7-nhusb sys/dev/usb/u3g.c: netbsd-7-nhusb sys/dev/usb/uark.c: netbsd-7-nhusb sys/dev/usb/uatp.c: netbsd-7-nhusb sys/dev/usb/uaudio.c: netbsd-7-nhusb sys/dev/usb/uberry.c: netbsd-7-nhusb sys/dev/usb/ubsa.c: netbsd-7-nhusb sys/dev/usb/ubsa_common.c: netbsd-7-nhusb sys/dev/usb/ubsavar.h: netbsd-7-nhusb sys/dev/usb/ubt.c: netbsd-7-nhusb sys/dev/usb/uchcom.c: netbsd-7-nhusb sys/dev/usb/ucom.c: netbsd-7-nhusb sys/dev/usb/ucomvar.h: netbsd-7-nhusb sys/dev/usb/ucycom.c: netbsd-7-nhusb sys/dev/usb/udl.c: netbsd-7-nhusb sys/dev/usb/udl.h: netbsd-7-nhusb sys/dev/usb/udsbr.c: netbsd-7-nhusb sys/dev/usb/udsir.c: netbsd-7-nhusb sys/dev/usb/uep.c: netbsd-7-nhusb sys/dev/usb/uftdi.c: netbsd-7-nhusb sys/dev/usb/uftdireg.h: netbsd-7-nhusb sys/dev/usb/ugen.c: netbsd-7-nhusb sys/dev/usb/ugensa.c: netbsd-7-nhusb sys/dev/usb/uhci.c: netbsd-7-nhusb sys/dev/usb/uhcireg.h: netbsd-7-nhusb sys/dev/usb/uhcivar.h: netbsd-7-nhusb sys/dev/usb/uhid.c: netbsd-7-nhusb sys/dev/usb/uhidev.c: netbsd-7-nhusb sys/dev/usb/uhidev.h: netbsd-7-nhusb sys/dev/usb/uhmodem.c: netbsd-7-nhusb sys/dev/usb/uhso.c: netbsd-7-nhusb sys/dev/usb/uhub.c: netbsd-7-nhusb sys/dev/usb/uipad.c: netbsd-7-nhusb sys/dev/usb/uipaq.c: netbsd-7-nhusb sys/dev/usb/uirda.c: netbsd-7-nhusb sys/dev/usb/uirdavar.h: netbsd-7-nhusb sys/dev/usb/ukbd.c: netbsd-7-nhusb sys/dev/usb/ukbdmap.c: netbsd-7-nhusb sys/dev/usb/ukyopon.c: netbsd-7-nhusb sys/dev/usb/ukyopon.h: netbsd-7-nhusb sys/dev/usb/ulpt.c: netbsd-7-nhusb sys/dev/usb/umass.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.h: netbsd-7-nhusb sys/dev/usb/umass_quirks.c: netbsd-7-nhusb sys/dev/usb/umass_quirks.h: netbsd-7-nhusb sys/dev/usb/umass_scsipi.c: netbsd-7-nhusb sys/dev/usb/umass_scsipi.h: netbsd-7-nhusb sys/dev/usb/umassvar.h: netbsd-7-nhusb sys/dev/usb/umcs.c: netbsd-7-nhusb sys/dev/usb/umct.c: netbsd-7-nhusb sys/dev/usb/umidi.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.h: netbsd-7-nhusb sys/dev/usb/umodem.c: netbsd-7-nhusb sys/dev/usb/umodem_common.c: netbsd-7-nhusb sys/dev/usb/umodemvar.h: netbsd-7-nhusb sys/dev/usb/ums.c: netbsd-7-nhusb sys/dev/usb/uplcom.c: netbsd-7-nhusb sys/dev/usb/urio.c: netbsd-7-nhusb sys/dev/usb/urio.h: netbsd-7-nhusb sys/dev/usb/usb.c: netbsd-7-nhusb sys/dev/usb/usb.h: netbsd-7-nhusb sys/dev/usb/usb_mem.c: netbsd-7-nhusb sys/dev/usb/usb_mem.h: netbsd-7-nhusb sys/dev/usb/usb_quirks.c: netbsd-7-nhusb sys/dev/usb/usb_quirks.h: netbsd-7-nhusb sys/dev/usb/usb_subr.c: netbsd-7-nhusb sys/dev/usb/usbdevices.config: netbsd-7-nhusb sys/dev/usb/usbdevs: netbsd-7-nhusb sys/dev/usb/usbdevs.h: netbsd-7-nhusb sys/dev/usb/usbdevs_data.h: netbsd-7-nhusb sys/dev/usb/usbdi.c: netbsd-7-nhusb sys/dev/usb/usbdi.h: netbsd-7-nhusb sys/dev/usb/usbdi_util.c: netbsd-7-nhusb sys/dev/usb/usbdi_util.h: netbsd-7-nhusb sys/dev/usb/usbdivar.h: netbsd-7-nhusb sys/dev/usb/usbhid.h: netbsd-7-nhusb sys/dev/usb/usbhist.h: netbsd-7-nhusb sys/dev/usb/usbroothub.c: netbsd-7-nhusb sys/dev/usb/usbroothub.h: netbsd-7-nhusb sys/dev/usb/usbroothub_subr.c: delete sys/dev/usb/usbroothub_subr.h: delete sys/dev/usb/uscanner.c: netbsd-7-nhusb sys/dev/usb/uslsa.c: netbsd-7-nhusb sys/dev/usb/usscanner.c: netbsd-7-nhusb sys/dev/usb/ustir.c: netbsd-7-nhusb sys/dev/usb/uthum.c: netbsd-7-nhusb sys/dev/usb/utoppy.c: netbsd-7-nhusb sys/dev/usb/uts.c: netbsd-7-nhusb sys/dev/usb/uvideo.c: netbsd-7-nhusb sys/dev/usb/uvisor.c: netbsd-7-nhusb sys/dev/usb/uvscom.c: netbsd-7-nhusb sys/dev/usb/uyap.c: netbsd-7-nhusb sys/dev/usb/uyap_firmware.h: netbsd-7-nhusb sys/dev/usb/uyurex.c: netbsd-7-nhusb sys/dev/usb/x1input_rdesc.h: netbsd-7-nhusb sys/dev/usb/xhci.c: netbsd-7-nhusb sys/dev/usb/xhcireg.h: netbsd-7-nhusb sys/dev/usb/xhcivar.h: netbsd-7-nhusb sys/dev/usb/xinput_rdesc.h: netbsd-7-nhusb sys/external/bsd/common/conf/files.linux: netbsd-7-nhusb sys/external/bsd/common/include/linux/err.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/kernel.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/workqueue.h: netbsd-7-nhusb sys/external/bsd/common/linux/linux_work.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/atombios_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/drm/files.drmkms: netbsd-7-nhusb sys/external/bsd/drm2/i915drm/files.i915drmkms: netbsd-7-nhusb sys/external/bsd/drm2/include/linux/err.h: delete sys/external/bsd/drm2/include/linux/workqueue.h: delete sys/external/bsd/drm2/linux/files.drmkms_linux: netbsd-7-nhusb sys/external/bsd/drm2/linux/linux_work.c: delete sys/external/bsd/dwc2/dwc2.c: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2var.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwctwo2netbsd: netbsd-7-nhusb sys/external/bsd/dwc2/conf/files.dwc2: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_coreintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdddma.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdqueue.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hw.h: netbsd-7-nhusb sys/modules/drmkms_linux/Makefile: netbsd-7-nhusb sys/modules/i915drmkms/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libugenhc/ugenhc.c: netbsd-7-nhusb sys/rump/dev/lib/libusb/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libusb/USB.ioconf: netbsd-7-nhusb sys/rump/dev/lib/libusb/usb_at_ugenhc.c: delete sys/rump/dev/lib/libusb/opt/opt_usb.h: delete sys/rump/dev/lib/libusb/opt/opt_usbverbose.h: delete sys/sys/mbuf.h: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.8: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.c: netbsd-7-nhusb Merge netbsd-7-nhusb: - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix - Change the SOFTINT level from NET to SERIAL for the USB softint handler. This gives the callback a chance of running when another softint handler at SOFTINT_NET has blocked holding a lock, e.g. softnet_lock and most of the network stack. - kern/49065 - ifconfig tun0 ... sequence locks up system / lockup: softnet_lock held across usb xfr - kern/50491 - unkillable wait in usbd_transfer while using usmsc0 on raspberry pi 2 - kern/51395 - USB Ethernet makes xhci hang - Various improvements to slhci(4) - Various improvements to dwc2(4)
1.53.2.4	28-Aug-2017	skrll	Sync with HEAD
1.53.2.3	22-Sep-2015	skrll	Sync with HEAD
1.53.2.2	05-Dec-2014	skrll	Use int for return type for [eou]chi_init and motg_init.
1.53.2.1	03-Dec-2014	skrll	The grand renaming of structure members. No functional change.
1.55.8.1	11-May-2017	pgoyette	Sync with HEAD
1.56.8.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.57.18.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.58.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.59.12.1	02-Aug-2025	perseant	Sync with HEAD
1.19	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.18	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.17	09-Apr-2012	plunky	branches: 1.17.40; 1.17.44; Tidy up a little, the way that midi attachment code works - change midi_attach() to omit the 'parent' arg (there are only two callers of this and it is not used) - change midisyn_attach() to midisyn_init(), so not needing a midi_softc, and fix the midi_pcppi driver to set hw_if and hw_hdl directly in its midi_softc before calling midi_attach() - add a device_t to opl_softc structure, change opl drivers to store the device 'self' in opl_softc and fix opl_attach() to use this opl_softc->dev field directly rather than a field in an otherwise unused midi_softc - remove unnecessary midi_softc from opl and cms drivers (child device provides that) reviewed by mrg
1.16	07-Dec-2011	jakllsch	Initialize the opl softc lock pointer to the parent's interrupt lock. Briefly tested only on wss@isapnp.
1.15	28-Apr-2008	martin	branches: 1.15.34; 1.15.38; Remove clause 3 and 4 from TNF licenses
1.14	15-Mar-2008	cube	branches: 1.14.2; 1.14.4; Split device_t and softc for opl(4). That should have been done along with midi(4), because opl_softc embed a midi_softc. Fixes PR#38242 by Onno van der Linden.
1.13	19-Oct-2007	ad	branches: 1.13.12; 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	09-Jun-2006	christos	branches: 1.10.6; 1.10.8; stack police: Introduce an opl_match function that uses malloc instead of allocating ~2K of softc on the stack.
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.8; 1.9.14; merge ktrace-lwp.
1.8	28-Jun-2005	thorpej	branches: 1.8.2; Use ANSI function decls and static.
1.7	04-Feb-2005	perry	de-__P
1.6	04-Dec-2003	keihan	branches: 1.6.8; 1.6.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	02-Oct-2002	thorpej	branches: 1.5.6; Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	13-Nov-2001	lukem	add RCSID
1.1	04-Sep-2001	itohy	branches: 1.1.4; 1.1.6; OPL3 driver for C-Media CMI8x38
1.1.6.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.6.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.6.1	04-Sep-2001	nathanw	file opl_cmpci.c was added on branch nathanw_sa on 2001-09-21 22:35:58 +0000
1.1.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.4.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.1.4.1	04-Sep-2001	thorpej	file opl_cmpci.c was added on branch kqueue on 2001-09-13 01:15:56 +0000
1.5.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.5.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.6.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	12-Feb-2005	yamt	sync with head.
1.6.8.1	29-Apr-2005	kent	sync with -current
1.8.2.3	17-Mar-2008	yamt	sync with head.
1.8.2.2	27-Oct-2007	yamt	sync with head.
1.8.2.1	21-Jun-2006	yamt	sync with head.
1.9.14.1	19-Jun-2006	chap	Sync with head.
1.9.8.1	26-Jun-2006	yamt	sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.8.2	10-Dec-2006	yamt	sync with head.
1.10.8.1	22-Oct-2006	yamt	sync with head
1.10.6.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.2	23-Mar-2008	matt	sync with HEAD
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.12.1	24-Mar-2008	keiichi	sync with head.
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.38.2	29-Apr-2012	mrg	sync to latest -current.
1.15.38.1	18-Feb-2012	mrg	merge to -current.
1.15.34.1	17-Apr-2012	yamt	sync with head
1.17.44.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.17.40.1	10-Jun-2019	christos	Sync with HEAD
1.20	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.19	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.18	09-Apr-2012	plunky	branches: 1.18.40; 1.18.44; Tidy up a little, the way that midi attachment code works - change midi_attach() to omit the 'parent' arg (there are only two callers of this and it is not used) - change midisyn_attach() to midisyn_init(), so not needing a midi_softc, and fix the midi_pcppi driver to set hw_if and hw_hdl directly in its midi_softc before calling midi_attach() - add a device_t to opl_softc structure, change opl drivers to store the device 'self' in opl_softc and fix opl_attach() to use this opl_softc->dev field directly rather than a field in an otherwise unused midi_softc - remove unnecessary midi_softc from opl and cms drivers (child device provides that) reviewed by mrg
1.17	23-Nov-2011	jmcneill	branches: 1.17.2; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.16	28-Apr-2008	martin	branches: 1.16.12; 1.16.34; 1.16.36; Remove clause 3 and 4 from TNF licenses
1.15	15-Mar-2008	cube	branches: 1.15.2; 1.15.4; Split device_t and softc for opl(4). That should have been done along with midi(4), because opl_softc embed a midi_softc. Fixes PR#38242 by Onno van der Linden.
1.14	19-Oct-2007	ad	branches: 1.14.12; 1.14.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.13	16-Nov-2006	christos	branches: 1.13.8; 1.13.22; 1.13.24; 1.13.28; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	11-Dec-2005	christos	branches: 1.11.20; 1.11.22; merge ktrace-lwp.
1.10	28-Jun-2005	thorpej	branches: 1.10.2; Use ANSI function decls and static.
1.9	04-Feb-2005	perry	de-__P
1.8	04-Dec-2003	keihan	branches: 1.8.8; 1.8.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.7	02-Oct-2002	thorpej	branches: 1.7.6; Add trailing ; to CFATTACH_DECL.
1.6	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.5	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.4	13-Nov-2001	lukem	add RCSID
1.3	01-Sep-1999	kleink	branches: 1.3.14; 1.3.16; No longer requires mpuvar.h.
1.2	01-Aug-1999	augustss	branches: 1.2.2; Move the MPU driver isa/ to ic/. (I didn't move it in the repository because the revision history is very short and uninteresting. :)
1.1	12-Jul-1999	kleink	Add a driver for the ESS Technology Solo-1 PCI AudioDrive line of chips.
1.2.2.2	02-Aug-1999	thorpej	Update from trunk.
1.2.2.1	01-Aug-1999	thorpej	file opl_eso.c was added on branch chs-ubc2 on 1999-08-02 22:03:59 +0000
1.3.16.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.16.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.14.2	18-Oct-2002	nathanw	Catch up to -current.
1.3.14.1	14-Nov-2001	nathanw	Catch up to -current.
1.7.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.7.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.6.1	03-Aug-2004	skrll	Sync with HEAD
1.8.10.1	12-Feb-2005	yamt	sync with head.
1.8.8.1	29-Apr-2005	kent	sync with -current
1.10.2.3	17-Mar-2008	yamt	sync with head.
1.10.2.2	27-Oct-2007	yamt	sync with head.
1.10.2.1	30-Dec-2006	yamt	sync with head.
1.11.22.2	10-Dec-2006	yamt	sync with head.
1.11.22.1	22-Oct-2006	yamt	sync with head
1.11.20.1	18-Nov-2006	ad	Sync with head.
1.13.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.13.24.2	23-Mar-2008	matt	sync with HEAD
1.13.24.1	06-Nov-2007	matt	sync with HEAD
1.13.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.8.1	23-Oct-2007	ad	Sync with head.
1.14.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.14.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.14.12.1	24-Mar-2008	keiichi	sync with head.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.36.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.16.34.1	17-Apr-2012	yamt	sync with head
1.16.12.1	09-Dec-2008	ad	Checkpoint work on MIDI.
1.17.2.1	29-Apr-2012	mrg	sync to latest -current.
1.18.44.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.18.40.1	10-Jun-2019	christos	Sync with HEAD
1.19	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.18	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.17	09-Apr-2012	plunky	branches: 1.17.40; 1.17.44; Tidy up a little, the way that midi attachment code works - change midi_attach() to omit the 'parent' arg (there are only two callers of this and it is not used) - change midisyn_attach() to midisyn_init(), so not needing a midi_softc, and fix the midi_pcppi driver to set hw_if and hw_hdl directly in its midi_softc before calling midi_attach() - add a device_t to opl_softc structure, change opl drivers to store the device 'self' in opl_softc and fix opl_attach() to use this opl_softc->dev field directly rather than a field in an otherwise unused midi_softc - remove unnecessary midi_softc from opl and cms drivers (child device provides that) reviewed by mrg
1.16	07-Dec-2011	jakllsch	Initialize the opl softc lock pointer to the parent's interrupt lock. Briefly tested only on wss@isapnp.
1.15	28-Apr-2008	martin	branches: 1.15.34; 1.15.38; Remove clause 3 and 4 from TNF licenses
1.14	15-Mar-2008	cube	branches: 1.14.2; 1.14.4; Split device_t and softc for opl(4). That should have been done along with midi(4), because opl_softc embed a midi_softc. Fixes PR#38242 by Onno van der Linden.
1.13	19-Oct-2007	ad	branches: 1.13.12; 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	04-Feb-2005	perry	de-__P
1.7	04-Dec-2003	keihan	branches: 1.7.8; 1.7.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.6	02-Oct-2002	thorpej	branches: 1.6.6; Add trailing ; to CFATTACH_DECL.
1.5	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.4	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.3	13-Nov-2001	lukem	add RCSID
1.2	15-May-2000	thorpej	branches: 1.2.6; 1.2.8; 1.2.10; ac97.h -> ac97var.h
1.1	01-Nov-1999	augustss	branches: 1.1.4; 1.1.6; Add FM801 audio device driver. From Witold J. Wnuk ww181302@zodiac.mimuw.edu.pl Closes PR kern/8729.
1.1.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.4.2	15-Nov-1999	fvdl	Sync with -current
1.1.4.1	01-Nov-1999	fvdl	file opl_fms.c was added on branch fvdl-softdep on 1999-11-15 00:41:05 +0000
1.2.10.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.10.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.8.2	18-Oct-2002	nathanw	Catch up to -current.
1.2.8.1	14-Nov-2001	nathanw	Catch up to -current.
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	15-May-2000	bouyer	file opl_fms.c was added on branch thorpej_scsipi on 2000-11-20 11:42:29 +0000
1.6.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.6.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.6.1	03-Aug-2004	skrll	Sync with HEAD
1.7.10.1	12-Feb-2005	yamt	sync with head.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.9.2.3	17-Mar-2008	yamt	sync with head.
1.9.2.2	27-Oct-2007	yamt	sync with head.
1.9.2.1	30-Dec-2006	yamt	sync with head.
1.10.22.2	10-Dec-2006	yamt	sync with head.
1.10.22.1	22-Oct-2006	yamt	sync with head
1.10.20.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.2	23-Mar-2008	matt	sync with HEAD
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.12.1	24-Mar-2008	keiichi	sync with head.
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.38.2	29-Apr-2012	mrg	sync to latest -current.
1.15.38.1	18-Feb-2012	mrg	merge to -current.
1.15.34.1	17-Apr-2012	yamt	sync with head
1.17.44.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.17.40.1	10-Jun-2019	christos	Sync with HEAD
1.18	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.17	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.16	09-Apr-2012	plunky	branches: 1.16.40; 1.16.44; Tidy up a little, the way that midi attachment code works - change midi_attach() to omit the 'parent' arg (there are only two callers of this and it is not used) - change midisyn_attach() to midisyn_init(), so not needing a midi_softc, and fix the midi_pcppi driver to set hw_if and hw_hdl directly in its midi_softc before calling midi_attach() - add a device_t to opl_softc structure, change opl drivers to store the device 'self' in opl_softc and fix opl_attach() to use this opl_softc->dev field directly rather than a field in an otherwise unused midi_softc - remove unnecessary midi_softc from opl and cms drivers (child device provides that) reviewed by mrg
1.15	07-Dec-2011	jakllsch	Initialize the opl softc lock pointer to the parent's interrupt lock. Briefly tested only on wss@isapnp.
1.14	28-Apr-2008	martin	branches: 1.14.34; 1.14.38; Remove clause 3 and 4 from TNF licenses
1.13	15-Mar-2008	cube	branches: 1.13.2; 1.13.4; Split device_t and softc for opl(4). That should have been done along with midi(4), because opl_softc embed a midi_softc. Fixes PR#38242 by Onno van der Linden.
1.12	19-Oct-2007	ad	branches: 1.12.12; 1.12.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	16-Nov-2006	christos	branches: 1.11.8; 1.11.22; 1.11.24; 1.11.28; __unused removal on arguments; approved by core.
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	11-Dec-2005	christos	branches: 1.9.20; 1.9.22; merge ktrace-lwp.
1.8	28-Jun-2005	thorpej	branches: 1.8.2; Use ANSI function decls and static.
1.7	04-Feb-2005	perry	de-__P
1.6	04-Dec-2003	keihan	branches: 1.6.8; 1.6.10; netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	02-Oct-2002	thorpej	branches: 1.5.6; Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	13-Nov-2001	lukem	add RCSID
1.1	10-Dec-1998	augustss	branches: 1.1.20; 1.1.22; Add a driver for the S3 SonicVibes chip. From OpenBSD and Constantine Paul Sapuntzakis (csapuntz@cvs.openbsd.org) with some changes by me. XXX The driver still needs work.
1.1.22.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.22.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.20.2	18-Oct-2002	nathanw	Catch up to -current.
1.1.20.1	14-Nov-2001	nathanw	Catch up to -current.
1.5.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.6.4	04-Feb-2005	skrll	Sync with HEAD.
1.5.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.6.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	12-Feb-2005	yamt	sync with head.
1.6.8.1	29-Apr-2005	kent	sync with -current
1.8.2.3	17-Mar-2008	yamt	sync with head.
1.8.2.2	27-Oct-2007	yamt	sync with head.
1.8.2.1	30-Dec-2006	yamt	sync with head.
1.9.22.2	10-Dec-2006	yamt	sync with head.
1.9.22.1	22-Oct-2006	yamt	sync with head
1.9.20.1	18-Nov-2006	ad	Sync with head.
1.11.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.24.2	23-Mar-2008	matt	sync with HEAD
1.11.24.1	06-Nov-2007	matt	sync with HEAD
1.11.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.8.1	23-Oct-2007	ad	Sync with head.
1.12.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.12.1	24-Mar-2008	keiichi	sync with head.
1.13.4.1	16-May-2008	yamt	sync with head.
1.13.2.1	18-May-2008	yamt	sync with head.
1.14.38.2	29-Apr-2012	mrg	sync to latest -current.
1.14.38.1	18-Feb-2012	mrg	merge to -current.
1.14.34.1	17-Apr-2012	yamt	sync with head
1.16.44.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.16.40.1	10-Jun-2019	christos	Sync with HEAD
1.19	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.18	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.17	09-Apr-2012	plunky	branches: 1.17.40; 1.17.44; Tidy up a little, the way that midi attachment code works - change midi_attach() to omit the 'parent' arg (there are only two callers of this and it is not used) - change midisyn_attach() to midisyn_init(), so not needing a midi_softc, and fix the midi_pcppi driver to set hw_if and hw_hdl directly in its midi_softc before calling midi_attach() - add a device_t to opl_softc structure, change opl drivers to store the device 'self' in opl_softc and fix opl_attach() to use this opl_softc->dev field directly rather than a field in an otherwise unused midi_softc - remove unnecessary midi_softc from opl and cms drivers (child device provides that) reviewed by mrg
1.16	07-Dec-2011	jakllsch	Initialize the opl softc lock pointer to the parent's interrupt lock. Briefly tested only on wss@isapnp.
1.15	28-Apr-2008	martin	branches: 1.15.34; 1.15.38; Remove clause 3 and 4 from TNF licenses
1.14	15-Mar-2008	cube	branches: 1.14.2; 1.14.4; Split device_t and softc for opl(4). That should have been done along with midi(4), because opl_softc embed a midi_softc. Fixes PR#38242 by Onno van der Linden.
1.13	19-Oct-2007	ad	branches: 1.13.12; 1.13.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.12	16-Nov-2006	christos	branches: 1.12.8; 1.12.22; 1.12.24; 1.12.28; __unused removal on arguments; approved by core.
1.11	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.10	09-Jun-2006	christos	branches: 1.10.6; 1.10.8; stack police: Introduce an opl_match function that uses malloc instead of allocating ~2K of softc on the stack.
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.8; 1.9.14; merge ktrace-lwp.
1.8	28-Jun-2005	thorpej	branches: 1.8.2; Use ANSI function decls and static.
1.7	08-Nov-2004	kent	ANSIfy and style fixes
1.6	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.5	02-Oct-2002	thorpej	branches: 1.5.6; Add trailing ; to CFATTACH_DECL.
1.4	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.3	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.2	13-Nov-2001	lukem	add RCSID
1.1	30-Mar-2001	minoura	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.1.8.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.6.2	01-May-2001	he	Pull up revision 1.1 (new, requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.1.6.1	30-Mar-2001	he	file opl_yds.c was added on branch netbsd-1-5 on 2001-05-01 11:45:26 +0000
1.1.4.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.1	30-Mar-2001	bouyer	file opl_yds.c was added on branch thorpej_scsipi on 2001-04-21 17:49:16 +0000
1.1.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.1.2.1	30-Mar-2001	nathanw	file opl_yds.c was added on branch nathanw_sa on 2001-04-09 01:57:04 +0000
1.5.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.6.4	14-Nov-2004	skrll	Sync with HEAD.
1.5.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.6.1	03-Aug-2004	skrll	Sync with HEAD
1.8.2.3	17-Mar-2008	yamt	sync with head.
1.8.2.2	27-Oct-2007	yamt	sync with head.
1.8.2.1	21-Jun-2006	yamt	sync with head.
1.9.14.1	19-Jun-2006	chap	Sync with head.
1.9.8.1	26-Jun-2006	yamt	sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.8.2	10-Dec-2006	yamt	sync with head.
1.10.8.1	22-Oct-2006	yamt	sync with head
1.10.6.1	18-Nov-2006	ad	Sync with head.
1.12.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.12.24.2	23-Mar-2008	matt	sync with HEAD
1.12.24.1	06-Nov-2007	matt	sync with HEAD
1.12.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.12.8.1	23-Oct-2007	ad	Sync with head.
1.13.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.13.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.13.12.1	24-Mar-2008	keiichi	sync with head.
1.14.4.1	16-May-2008	yamt	sync with head.
1.14.2.1	18-May-2008	yamt	sync with head.
1.15.38.2	29-Apr-2012	mrg	sync to latest -current.
1.15.38.1	18-Feb-2012	mrg	merge to -current.
1.15.34.1	17-Apr-2012	yamt	sync with head
1.17.44.1	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.17.40.1	10-Jun-2019	christos	Sync with HEAD
1.25	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.24	31-Jul-2012	bouyer	branches: 1.24.2; 1.24.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.23	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.22	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.21	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.20	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.19	04-Apr-2011	dyoung	branches: 1.19.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.18	05-Nov-2010	jakllsch	branches: 1.18.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.17	28-Apr-2008	martin	branches: 1.17.20; 1.17.22; Remove clause 3 and 4 from TNF licenses
1.16	18-Mar-2008	cube	branches: 1.16.2; 1.16.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.15	09-Feb-2007	ad	branches: 1.15.20; 1.15.36; 1.15.40; Merge newlock2 to head.
1.14	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.13	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.12	11-Dec-2005	christos	branches: 1.12.20; 1.12.22; merge ktrace-lwp.
1.11	24-May-2005	lukem	branches: 1.11.2; add __KERNEL_RCSID
1.10	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.9	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.8	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.7	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.6	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.5	03-Jan-2004	thorpej	branches: 1.5.4; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.4	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.3	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.5.4.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.5.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.5.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.5.4.2	03-Aug-2004	skrll	Sync with HEAD
1.5.4.1	03-Jan-2004	skrll	file optiide.c was added on branch ktrace-lwp on 2004-08-03 10:49:10 +0000
1.11.2.3	24-Mar-2008	yamt	sync with head.
1.11.2.2	26-Feb-2007	yamt	sync with head.
1.11.2.1	30-Dec-2006	yamt	sync with head.
1.12.22.2	10-Dec-2006	yamt	sync with head.
1.12.22.1	22-Oct-2006	yamt	sync with head
1.12.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.12.20.1	18-Nov-2006	ad	Sync with head.
1.15.40.2	02-Jun-2008	mjf	Sync with HEAD.
1.15.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.36.1	24-Mar-2008	keiichi	sync with head.
1.15.20.1	23-Mar-2008	matt	sync with HEAD
1.16.4.1	16-May-2008	yamt	sync with head.
1.16.2.1	18-May-2008	yamt	sync with head.
1.17.22.2	21-Apr-2011	rmind	sync with head
1.17.22.1	05-Mar-2011	rmind	sync with head
1.17.20.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.18.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.19.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.19.4.1	30-Oct-2012	yamt	sync with head
1.24.4.1	18-May-2014	rmind	sync with head
1.24.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.24.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.217	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.216	24-Apr-2021	thorpej	branches: 1.216.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.215	10-Nov-2019	chs	branches: 1.215.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.214	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.213	01-Mar-2019	msaitoh	No functional change: - u_int* -> uint* - KNF.
1.212	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.211	10-May-2017	msaitoh	branches: 1.211.8; 1.211.10; Use pci_intr_establish_xname().
1.210	07-Jul-2016	msaitoh	branches: 1.210.8; KNF. Remove extra spaces. No functional change.
1.209	17-Feb-2016	christos	PR/50820: David Binderman: Fail when !(edge\|pulse) instead of !edge && edge
1.208	26-Mar-2015	nakayama	Add missing new lines.
1.207	16-Nov-2014	ozaki-r	branches: 1.207.2; Replace callout_stop with callout_halt In order to call callout_destroy for a callout safely, we have to ensure the function of the callout is not running and pending. To do so, we should use callout_halt, not callout_stop. Discussed with martin@ and riastradh@.
1.206	29-Mar-2014	christos	branches: 1.206.4; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.205	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.204	30-Jan-2012	drochner	branches: 1.204.2; 1.204.6; 1.204.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.203	01-Aug-2011	drochner	branches: 1.203.2; 1.203.6; remove some bloat: -cardbus doesn't use multiple interrupt lines like PCI, and it doesn't use machanisms like interrupt line register and swizzling -- no need to carry around dummy information, this is all dealt with by the bridge (I'm asking myself how "rbus_ppb" can work -- a bridge attached to cardbus just can't work like a normal PCI bridge as far as interrupts are concerned. I thing that should be a hardware specific driver because behavior is not covered by a standard.) -cardbus always uses 3.3V -- no need for a variable to keep track of the voltage
1.202	26-Jul-2011	dyoung	Don't set the iobase and iosize members of pcmciabus_attach_args because they're not used in any meaningful way.
1.201	04-Jan-2011	jruoho	Stop using aprint_debug(9) in an interrupt handler.
1.200	27-Dec-2010	phx	Add TI PCI1510. Enable TI12XX PCI interrupt in MFUNC when MFUNC pin0 and pin1 are zero. Ignore the other pins, which may be non-zero (e.g. IRQSER in pin3).
1.199	20-Apr-2010	dyoung	Change sc_pil_intr_enable from an int to a bool. No functional change intended.
1.198	19-Apr-2010	dyoung	Add default implementations for bus_space_is_equal(9), bus_space_tag_create(9), and bus_space_tag_destroy(9). Use bus_space_is_equal(9) throughout the kernel to compare bus_space_tag_t's. Tested on i386 and on sparc64.
1.197	15-Mar-2010	dyoung	rbus is always #defined as 1, so unifdef -D rbus=1.
1.196	25-Feb-2010	dyoung	branches: 1.196.2; Cardbus_free_tag() and cardbus_free_tag() are no-ops, so get rid of them to simplify the code a bit.
1.195	24-Feb-2010	dyoung	Start to tuck Cardbus under the PCI abstraction. Step #1, textual substitution: for all practical purposes, pcitag_t and cardbustag_t are interchangeable, so just use pcitag_t. Ditto pcireg_t and cardbusreg_t. Poison new uses of cardbusreg_t and cardbustag_t by deleting the type definitions.
1.194	24-Feb-2010	dyoung	A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.193	08-Jan-2010	dyoung	branches: 1.193.2; Expand PMF_FN_* macros.
1.192	15-Dec-2009	snj	Move to 2-clause license. Approved by HAYAKAWA Koichi (copyright holder).
1.191	07-Aug-2009	blymn	Destroy mutex and condvar initialised in pccbbattach().
1.190	07-Aug-2009	blymn	Clean up extraneous white space.
1.189	23-Jul-2009	dyoung	Use kpause(9), cv_timedwait(9), and cv_signal(9) instead of tsleep(9) and wakeup(9). Use mstohz(9). XXX Protection against spurious wakeups is still needed, but this patch XXX makes the code no worse than before in this regard.
1.188	21-May-2009	dyoung	Let us detach pcmcia0 and cardbus0 at cardslot0, and cardslot0 at cardbus0.
1.187	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.186	03-May-2009	dyoung	In pccbb_pci_callback(), set CBB_SPECMAPPED in sc_flags if we've acquired bus space from bus_space_alloc(9) or from rbus_space_alloc(), so that we use the appropriate routine to release the bus space in pccbbdetach(). This stops panics during cbb(4) detachment (esp. at shutdown) on some systems. Thanks to Kurt Schreiner for testing the fix, Kurt, Patrick Welche, Pouya D. Tafti, and Matt Green for reporting.
1.185	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.184	05-Mar-2009	msaitoh	Fix typo(s/interupt/interrupt/)
1.183	13-Feb-2009	bouyer	Fix bus_addr_t/bus_size_t confusion
1.182	06-Feb-2009	dyoung	branches: 1.182.2; When the power status changes, always note the change and wake the socket events thread. Sometimes we would not note the power-status change if the socket also indicated a change in card status.
1.181	11-Jan-2009	christos	merge christos-time_t
1.180	25-Oct-2008	christos	branches: 1.180.2; 1.180.4; 1.180.8; 1.180.12; another TI bridge.
1.179	06-Aug-2008	dyoung	Identify the O2 Micro OZ711E0 and treat it the same as all of the other O2 Micro bridges, which seem to spuriously report bus errors if parity-error detection is enabled.
1.178	03-Jul-2008	drochner	branches: 1.178.2; On cardbus configuration reads, check for a master abort in the cbb and return all-ones explicitely. Since pccbb.c rev. 1.151 the cbb is set to translate cardbus master aborts which makes that the PCI configuration read on the primary bus returns random junk. This can cause the probing code to get confused. The issue was reported by KIYOHARA Takashi and Jonathan A. Kollasch, the patch was tested by KIYOHARA Takashi.
1.177	26-Jun-2008	drochner	delete more dead code, remove use of pcic_handle except of some local uses in pcmcia ressource management which is copied from ISA pcic, so use our softc as handle exported to generic pcmcia code and kill the ph_parent backpointer - this makes the code simpler and smaller
1.176	26-Jun-2008	drochner	remove some more dead code, and use the softc as handle in more internal functions rather than pcic_handle
1.175	26-Jun-2008	drochner	remove references to nonexisting members of pcic_handle in commented out and debug code, from Patrick Welche (we could as well drop pcic_handle completely and use our softc as the opaque handle for the pcmcia code - this would simplify the code even more)
1.174	26-Jun-2008	drochner	constify the pcmcia/cardbus method tables
1.173	25-Jun-2008	drochner	Avoid the use of i82365var.h - we don't use any functions from there. Extract some data structures for ressource management into our private header instead. This allows to use a typed pointer instead of a generic one which saves a lot of typecasts. Also remove something marked as "dirty hack" which I admittedly don't understand, but it doesn't look useful...
1.172	25-Jun-2008	drochner	split device/softc for pccbb, the cardbus half of cardslot and cardbus (not yet pcmcia, there is a lot of md code involved)
1.171	24-Jun-2008	drochner	clean up the cardbus interrupt stuff: There were cardbus_intr_line_t and cardbus_intr_handle_t used intermixed for the same variable, and that variable is pretty much useless because cardbus doesn't follow the PCI interrupt swizzling etc scheme. Useless interrupt numbers were printed on cardbus device attach. So as a first step to sanity, kill cardbus_intr_handle_t and poison cardbus_intr_line_t to discourage printing it as a %d. Use cardbus_intr_line_t consistently throughout the code. Remove the "interrupting at foo" messages because the information is misleading. We could come up with a better interrupt vector information, but because cardbus interrupts are mediated by pccbb it would still be misleading.
1.170	24-Jun-2008	drochner	clean up some useless code and variables in cardbus attachment
1.169	27-May-2008	dyoung	branches: 1.169.2; Distinguish O2 Micro bridges from generic bridges. PCI Parity Error reporting seems to be broken on O2 Micro bridges, so do not activate it. Should fix kern/38698, reported by Andreas Gustafsson.
1.168	22-May-2008	dyoung	Use __arraycount().
1.167	04-Mar-2008	dyoung	branches: 1.167.2; 1.167.4; 1.167.6; Check device_has_power() instead of device_is_active() in pccbbintr(), since the driver must acknowledge some interrupts while it leaves the suspended state. This protects against an interrupt storm that will hang the system while I resume rtw0 at cardbus1.
1.166	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.165	02-Feb-2008	dyoung	branches: 1.165.2; 1.165.6; Fix a two races between the initialization code and cbb(4) interrupts: establish the interrupt handler before enabling interrupts, so that we will not get an interrupt storm that keeps the system from making any progress. (We may have only been susceptible to the storm if cbb(4) shared an interrupt with, say, uhci(4)). Also, finish setting up the pcic_handle before enabling interrupts, so that the Pcic_read() call in pccbbintr() will not jump to address 0! Now, pccbb_intrinit() encapsulates all of the interrupt initialization. Explicitly enable card-status changed (CSTSCHG) events. Paranoia: ensure that writes to the socket event register (interrupt acknowledgements) are flushed to the h/w by reading the register (gross). We really need something like bus_space_sync(BUS_SPACE_BARRIER_WRITE_BEFORE_INTERRUPT). More paranoia: in pccbbintr(), if the device is inactive, get out.
1.164	01-Feb-2008	dyoung	Use aprint_*_dev() and device_xname() instead of accessing dv_xname directly.
1.163	15-Jan-2008	christos	Remove ricoh hack; FreeBSD does not seem to do it, and it was a remnant from the branch. Things should work properly now.
1.162	14-Jan-2008	dyoung	Use device_t, device_private().
1.161	14-Jan-2008	dyoung	In pccbbdetach(), panic if interrupt handlers are still registered after detaching children. Use PCI_CB_MEMBASE0, PCI_CB_MEMLIMIT0, PCI_CB_IOBASE0, PCI_CB_IOLIMIT0 instead of magic numbers.
1.160	02-Jan-2008	dyoung	Introduce a struct cardbus_chipset_tag, and make cardbus_chipset_tag_t a pointer to that struct, so that we cannot assign pointers of arbitrary type to cardbus_chipset_tag_t. Tweak cbb(4) to accomodate this change. Make Cardbus_conf_read() and Carbus_conf_write() pass the right arguments to cardbus_functions->cardbus_conf_{read,write}() for a change. Let's hope this stops the crash in cardbus_function_enable() that macallan@ reported to me.
1.159	20-Dec-2007	dyoung	Use LIST_FOREACH().
1.158	16-Dec-2007	dyoung	Improve pci, cbb, cardslot, cardbus, and pcmcia to support detachment of this entire device tree: pci0 at mainbus0 elansc0 at pci0 gpio0 at elansc0 cbb0 at pci0 cardslot0 at cbb0 cardbus0 at cardslot0 pcmcia0 at cardslot0 cbb1 at pci0 cardslot1 at cbb1 cardbus1 at cardslot1 rtw0 at cardbus1 pcmcia1 at cardslot1 sip0 at pci0 nsphyter0 at sip0 sip1 at pci0 nsphyter1 at sip1 Whew!
1.157	11-Dec-2007	martin	As a stopgap fix: disable a pci_conf_ access at offset 0x82 on archs that would mind ;-) IMHO this access violates the PCI spec, so this should be solved differently.
1.156	09-Dec-2007	jmcneill	branches: 1.156.2; Merge jmcneill-pm branch.
1.155	01-Dec-2007	jmcneill	branches: 1.155.2; 1.155.4; aprintify
1.154	24-Nov-2007	dyoung	Use TAILQ_FOREACH() to search window lists. Do not insert/remove windows inside the loop.
1.153	24-Nov-2007	dyoung	Use the PCI #define for the secondary bridge reset instead of using the duplicate Cardbus #define. Write a note to self: pccbb interrupts need more work. No functional change intended.
1.152	21-Nov-2007	dyoung	Sometimes a change of CSTSCHG# accompanies the first interrupt from an Atheros WLAN. That generates a CB_SOCKET_EVENT_CSTS event on the bridge. The event isn't interesting to pccbb(4), so we used to ignore the interrupt. Now, let the child devices try to handle the interrupt, instead. The Atheros NIC produces interrupts more reliably, now: used to be that it would only interrupt if the driver avoided powering down the NIC's cardslot, and then the NIC would only work after it was reset a second time.
1.151	16-Nov-2007	dyoung	Improve Cardbus performance, error handling, and error reporting. Let the TI1420 PCI-Cardbus bridge do burst reads from the primary (PCI) bus. This ought to improve Tx performance on Cardbus NICs. This optimization may apply to other TI bridges, but I only have a datasheet for the TI1420. :-/ Activate PCI Parity Error and System Error reporting on PCI-Cardbus bridges. To avoid data destruction, set the Master Abort mode to 1. Stops the bridge from silently discarding writes from the secondary bus to the primary bus (Cardbus writes to PCI). Also, stops the bridge from fulfilling a read by a bus master on the secondary bus that failed on the primary bus with 0xffffffff (Cardbus reads from PCI). Now the bus will indicate an error condition (SERR) instead of silently destroying/corrupting data. Forward system error indications from the secondary to the primary bus. Detect parity errors on the secondary. Set a Cardbus card's Latency Timer to something reasonable, according to the Cardbus card's Minimum Grant and the bandwidth available on the PCI bus. Restore the Latency Timer when re-enabling a card (e.g., after power reactivation).
1.150	25-Oct-2007	joerg	branches: 1.150.2; Remove a completely bogus check if cbb has a valid PCI interrupt routed. If you have a system where it doesn't get one and you still want pcic, you can just disable cbb. This check is invalid as soon as the PCI interrupt register is not used and broke on my Dell Latitude for that reason.
1.149	22-Oct-2007	joerg	Correctly initialise sc_insert_ch and switch to callout_setfunc/callout_schedule.
1.148	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.147	11-Aug-2007	dyoung	branches: 1.147.2; 1.147.6; Read the PCI latency timer from the right register, using the right macro.
1.146	11-Aug-2007	dyoung	Make pccbb(4) use more PCI configuration register definitions, especially the Bridge Control Register definition, instead of rolling its own. As we read/modify/write configuration registers, use a variable that is named for the register we're r/m/w'ing, instead of using the variable 'reg' over and over. This helps both a human reader verify that we're not reading register X, modifying it, and writing back to register Y (oops); the compiler can help a little by warning that a variable is used before it is initialized.
1.145	16-Jul-2007	christos	branches: 1.145.2; 1.145.6; PR/36652: Dave Barnes: cbb0: Bad Vcc status once set will not clear -ThinkPad 570 / PC: Don't write to SOCKET_STAT, which is read-only, instead write to SOCKET_FORCE.
1.144	04-Feb-2007	dyoung	branches: 1.144.6; Fix a bug in Cardbus power activation. Most Cardbus bridges supported by pccbb(4) fire a power-cycle interrupt when the power state of a cardslot changes from 'off' to 'on'. TI bridges fire a power-cycle interrupt on both on->off and off->on changes. When pccbb_power() powered-down a cardslot, it did not wait around for the power-cycle interrupt. When pccbb_power() powered-up a cardslot, it did wait for the interrupt. If a pccbb_power(UP) followed a pccbb_power(DOWN) very closely, pccbb_power() used to interpret the power-cycle interrupt for the up->down transition as "power-up complete," read the power-state bit and, finding that power had NOT been activated, complain, "cbb0: power on failed?" Then pccbb_power() exited before power-activation was complete, falsely indicating that the power-activation was complete. After that, a driver attach/enable routine would blithely configure a card that was not fully powered-up. An operator who ran a command such as 'ifconfig rtw0 down up' or 'ifconfig ath0 down up' would read 'cbb0: power on failed?' in the system log, and their NIC would misbehave. This excerpt from a comment in the source should suffice to explain how I fixed the bug, /* * Wait as long as 200ms for a power-cycle interrupt. If * interrupts are enabled, but the socket has already * changed to the desired status, keep waiting for the * interrupt. "Consuming" the interrupt in this way keeps * the interrupt from prematurely waking some subsequent * pccbb_power call. And this explains why this patch will work for Ricoh bridges that do not fire an interrupt on the on->off transition: * XXX Not every bridge interrupts on the ->OFF transition. * XXX That's ok, we will time-out after 200ms. * * XXX The power cycle event will never happen when attaching * XXX a 16-bit card. That's ok, we will time-out after * XXX 200ms. */ M. Warner Losh and Charles M. Hannum provided valuable input on this patch.
1.143	04-Feb-2007	dyoung	Cosmetic: KNF indentation, curly braces, and argument declarations.
1.142	04-Feb-2007	dyoung	Convert the rather long and backslash-ridden DELAY_MS macro to a much shorter static subroutine, delay_ms().
1.141	04-Feb-2007	dyoung	Use the right subroutine name for the debug message.
1.140	26-Jan-2007	dyoung	Fix grammar in comment. From Patrick Welche.
1.139	26-Jan-2007	dyoung	Name magic numbers PCI_PMCSR_PME_EN and PCI_PMCSR.
1.138	21-Dec-2006	yamt	merge yamt-splraiseipl branch. - finish implementing splraiseipl (and makeiplcookie). http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html - complete workqueue(9) and fix its ipl problem, which is reported to cause audio skipping. - fix netbt (at least compilation problems) for some ports. - fix PR/33218.
1.137	16-Nov-2006	christos	branches: 1.137.2; 1.137.4; __unused removal on arguments; approved by core.
1.136	24-Oct-2006	itohy	3c575CT Ethernet card requires a little more time during reset and after reset, or the card won't be detected on some machines. Change the time from 40ms to 50ms. XXX The card doesn't work correctly yet.
1.135	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.134	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.133	08-Jul-2006	christos	branches: 1.133.4; 1.133.6; - Tidy up some debugging/diagnostic messages. - Add a delay for the topic95b after we power on. It makes things work without CBB_DEBUG on!
1.132	08-Jul-2006	christos	- all the polling code is stale and broken. Mention it. - the topic95B bridge does not set CB_SOCKET_STAT_PWRCYCLE - put back the delay code that was removed again. This makes my libretto 100CT work again.
1.131	04-Jul-2006	christos	fix typo
1.130	04-Jul-2006	christos	Apply fvdl's acpi pci interrupt configuration code. - MPACPI is no more. - MPACPI_SCANPCI -> ACPI_SCANPCI
1.129	17-Jun-2006	jmcneill	For Ricoh 5c47x class controllers, restore the "misc control" PCI config register on resume.
1.128	05-Apr-2006	dyoung	branches: 1.128.2; 1.128.4; Fix typo in comment, s/will be never/will never/.
1.127	18-Dec-2005	sekiya	branches: 1.127.4; 1.127.6; 1.127.8; 1.127.10; 1.127.12; Use aprint_debug() instead of printf() for obvious debug statement.
1.126	11-Dec-2005	christos	merge ktrace-lwp.
1.125	09-Sep-2005	drochner	remove the useless (always 0) device number from cardbus driver state, autoconf attach structures and configuration cycle functions (just leave the kernel configuration attribute for now to avoid breaking config(1) files)
1.124	23-Jun-2005	wiz	branches: 1.124.2; s/PCIBIOS_FIXUP/PCI_FIXUP/ in printf, from Patrick Welche in PR 30583.
1.123	22-Jun-2005	sekiya	Oops, a unilateral DEBUG define crept in with my last commit.
1.122	20-Jun-2005	sekiya	Don't abort the attach if we are routing interrupts through the ioapic, there
1.121	01-Jun-2005	sekiya	Fix compilation problem when CBB_DEBUG is defined.
1.120	01-Jun-2005	sekiya	"resister" -> "register"
1.119	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.118	23-Mar-2005	christos	PR/29600: YAMAMOTO Takashi: cbb hangs on boot. From the PR: It seems that we never get a power interrupt for 16 bit cards. The tsleep has been changed to wait for 200ms maximum, before giving up instead of waiting forever. XXX: Maybe that is too small for some laptops?
1.117	27-Feb-2005	perry	branches: 1.117.2; nuke trailing whitespace
1.116	04-Feb-2005	perry	de-__P
1.115	27-Jan-2005	jmcneill	Fix typo.
1.114	27-Jan-2005	jmcneill	Don't restore PCI configuration state if the chip isn't in D0, from Kentaro A. Kurahone.
1.113	27-Jan-2005	jmcneill	Capture PCI configuration state on suspend and restore it on resume, from Kentaro A. Kurahone.
1.112	16-Jan-2005	mycroft	branches: 1.112.2; 1.112.4; Just to be absolutely correct, wait an additional 1ms after the "power cycle" event, to cover the RESET Width (Tw (RESET)).
1.111	16-Jan-2005	mycroft	Christos's crappy Libretto 100CT has an extremely long ramp for socket Vcc (violating the PC Card spec), so... use the "power cycle" socket event to determine when we've reached Vcc before proceeding, rather than using a fixed amount of time. This has the double advantage that it makes the card attach time even shorter on sane systems -- the minimum is now ~38ms on my i8500 rather than 222ms. Probably a similar change should be made to pcic, but it was hard enough figuring out whether it would work with pccbb. The chip specs suck. For now, I'm leaving in a couple of additional printf()s in the hope that I will get some interesting data from them.
1.110	16-Aug-2004	mrg	fix a wrong variable name in a debug message.
1.109	16-Aug-2004	mycroft	Simplify some manipulation of PCIC_INTR. Further attempt to mitigate "bad Vcc" errors.
1.108	16-Aug-2004	mycroft	Some tweaks, per the PC Card spec: * Assert RESET before powering off a socket. * Turn on the output enable bit earlier so the interface actually drives CEn and RESET. * Tighten up the power-on timing a bit. * Mention the specific timing values named in the spec. For pccbb, be careful to always power off before zeroing PWRCTL.
1.107	16-Aug-2004	mycroft	Do not attempt to fiddle with the power-control bits in PWRCTL; just read them back after touching the PCI registers. This shouldn't be necessary, but somehow the controller detects the need for VPP2=12V and automatically applies it, and gives us a "bad Vcc" error if we turn it off accidentally.
1.106	15-Aug-2004	mycroft	And the reason we had to call pccbb_power() twice in the enable routine is... we were writing to PCIC_PWRCTL in between and causing the socket to power off. D'oh.
1.105	12-Aug-2004	mycroft	I can't actually use the FORCE register this way. Just try turning the socket power off.
1.104	12-Aug-2004	mycroft	A bunch of random cleanup: * Like the i82365 code, add a "delay" function that uses tsleep() to wait, and use this in the socket enable/disable paths. This gets rid of the annoying system pauses during card insertion and removal. (There are still some issues related to this in various drivers -- notably big delay()s in wi and xi.) * Move the power-change delay out of pccbb_power() and into the PCMCIA backend code -- specifically, once in the disable path and once in the enable path. We were being pretty schizo about this before. Make these use tsleep(). (Note: This should be safe because card insertion/removal is handled by a kernel process, not in an interrupt handler. It works for me with DIAGNOSTIC.) * If we get a "bad Vcc" error, attempt to force the socket to power off, and return an error. If we don't do this, we will get "bad Vcc" errors forever and never be able to use another card without rebooting, which is dumb. XXX I haven't been able to test this very well, because it doesn't fail for me in the first place. :-) * Clean up the socket mappings earlier in the enable path. * Try to be consistent about clearing PWRCTL (which contains OE) before turning off power.
1.103	11-Aug-2004	mycroft	Oops; fix a compile error.
1.102	11-Aug-2004	mycroft	Be more liberal in torching the old state in the disable and settype routines.
1.101	11-Aug-2004	mycroft	Rather than having a call up from the low-level driver to get the card type, instead have a call down from the PCMCIA mid-layer to set it. Use this from pcmcia_function_enable(). (Currently the policy is the same, but this would allow for more flexibility in deciding which mode to use.) Now it is safe to hold the socket enabled during attach, so do that. Only one enable/disable cycle to attach a card now!
1.100	06-Aug-2004	mycroft	Rework the power sequencing so that I stop getting unrecoverable "bad Vcc" errors with PCMCIA cards. I've been using this change for almost a year, but accidentally blew it away recently, and boy did that suck.
1.99	28-Jul-2004	he	Add entries for TI CardBus bridges 1520 and 4520Yenta, as found in IBM ThinkPad 40p and 41p respectively, so that the CardBus probes and works on those machines. Fixes PR#26380.
1.98	22-Jul-2004	mycroft	If the bus number has not been initialized, do not attach a cardslot, and output a diagnostic: pccbbN: secondary bus number uninitialized; try PCIBIOS_BUS_FIXUP
1.97	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.96	28-Mar-2004	nakayama	branches: 1.96.2; Add some initialization code for TI bridges. Mostly from FreeBSD, and tested on i386 w/ PCI1250 and sparc64 w/ PCI1410.
1.95	19-Dec-2003	christos	More unsigned long casts to make this work on ppc. Fix formatting of debug statements to 0x when there are hex numbers printed.
1.94	19-Dec-2003	christos	PR/23802: KIYOHARA Takashi: Bad printf format with CBB_DEBUG. While I am there, make all printfs consistent 0x%lx and (unsigned long) cast.
1.93	10-Dec-2003	briggs	Power up the card without OE enabled to avoid letting SanDisks enter True IDE mode. Hinted at by Charles Hannum a while back. This lets my PCI-Cardbus adapter read SanDisk CF cards much more reliably. Also bring in a few defines from FreeBSD for some more TI registers and such.
1.92	25-Oct-2003	christos	Fix uninitialized variable warnings
1.91	23-Oct-2003	briggs	Try to deal better with card removal during the card enabling process.
1.90	19-Jun-2003	msaitoh	branches: 1.90.2; regard as the card is removed if one of CCD1# and CCD2# is deasserted. This fixes the porblem that the cbb misses a remove event.
1.89	16-Jun-2003	thorpej	Rename IPL_IMP -> IPL_VM.
1.88	22-Mar-2003	nakayama	PCI-CardBus bridge support for sparc64: - add pccbb_attach_hook in pccbbattach for MD initializations. - omit arithmetics to bus_space_handle_t. - remove use of IST_LEVEL; not defined on sparc64 and unused.
1.87	20-Jan-2003	simonb	Remove unreachable "break"s after "return"s.
1.86	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.85	23-Dec-2002	tsutsui	Replace magic numbers for power management control with PCI_PMCSR* macros. XXX Should we use pci_get_powerstate() and pci_set_powerstate() in pci.c?
1.84	09-Nov-2002	martin	Add TI PCI4410 PCI-CardBus Bridge w/ OHCI IEEE 1394, from FUKAUMI Naoki in PR 18989.
1.83	19-Oct-2002	atatat	Don't print messages about pccbb_intr_disestablish() unless debugging is actually enabled.
1.82	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.81	01-Oct-2002	onoe	make it compilable with CBB_DEBUG.
1.80	01-Oct-2002	haya	Use LIST in sys/queue. Add some comments.
1.79	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.78	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.77	31-May-2002	mycroft	Patch the CLKRUN enable bit on ToPIC chips.
1.76	31-May-2002	haya	Add comment.
1.75	05-Mar-2002	haya	branches: 1.75.2; 1.75.6; Enforce 10-bit width for I/O addresses of 16-bit PC Cards when I/O start address is assigned by a driver.
1.74	10-Jan-2002	haya	Bug fix: correct chattering suppressor. Card insertion will be detected properly by this fix.
1.73	24-Dec-2001	christos	change if defined foo to ifdef foo and change a c++ comment into a c comment.
1.72	15-Nov-2001	lukem	don't need <sys/types.h> when including <sys/param.h>
1.71	13-Nov-2001	lukem	add RCSID
1.70	02-Nov-2001	haya	Add PCI power management control. If the power state of a bridge is not D0 at device attaching or resuming, set the bridge D0 mode.
1.69	17-Oct-2001	haya	branches: 1.69.2; Move some chip initialise code into pccbb_chipinit() function.
1.68	30-Sep-2001	yamt	force 0x1000 align for memory base before calling rbus_space_alloc(). resource management may be inconsistent without it.
1.67	30-Aug-2001	haya	branches: 1.67.2; Add some register definitions for TI PCI12XX and Toshiba ToPIC. Add ToPIC97 and ToPIC100 chip initialisation routine. Add more TI PCI12XX chip initialisation.
1.66	24-Aug-2001	haya	Buf fix: use PWAIT priority and do not set PCATCH.
1.65	06-Jul-2001	mcr	branches: 1.65.2; split pccbb_intr_route() into seperate function.
1.64	19-May-2001	soren	Add TI 1410 to yc_chipsets[].
1.63	30-Apr-2001	jmc	Cast args correctly to match format strings.
1.62	12-Apr-2001	thorpej	splimp -> splvm (use IPL_IMP needs to be revisited).
1.61	22-Feb-2001	enami	branches: 1.61.2; - Initialize auto variable (win[{0,1}].win_flags). - If PCCBB_MEM_CACHABLE isn't set, clear CB_BCR_PREFETCH_MEMWIN[01].
1.60	21-Feb-2001	jdolecek	make some more constant arrays 'const'
1.59	21-Feb-2001	minoura	Correct the 1st argument for pci_conf_{read,write}. On non-i386 platforms the previous revision could not be compiled. Pointed out by mjl.
1.58	20-Feb-2001	minoura	Re-initialize more registers after resuming from suspend. Some APM BIOSes do not restore them. Reviewd by haya.
1.57	09-Feb-2001	haya	BUG FIX: calculate address line mask in pccbb_pcmcia_io_alloc() when start address is not set.
1.56	30-Jan-2001	itohy	Fix problem where tsleep() is used not on the cardbus thread. This change makes PCMCIA cards work again on my machine. Unfortunately, some of the DELAY_MS() are called when the current process is not the cardbus thread, that is, on interrupt contexts. I confirmed these are actually used in interrupt context and replaced them with delay()s. However, I haven't confirmed other DELAY_MS()s are always on the cardbus thread. This shall be a temporary fix. The driver needs being rewritten not to use such delay()s, of course.
1.55	24-Jan-2001	haya	IMPROVEMENT: use tsleep() instead of delay(). Kernel do not stop during card insertion, deletion and power change.
1.54	22-Jan-2001	augustss	Add a comment.
1.53	17-Jan-2001	haya	Fix comment.
1.52	30-Dec-2000	augustss	Be less talkative in attach.
1.51	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.50	08-Dec-2000	mycroft	Turn off the damned TI `VCC protection' bit.
1.49	08-Dec-2000	haya	WORKAROUND: CardBus bridge avoids turning off the power of channel in shutdown hook when the bridge chip is TI113X. Because PowerBook 2400 hangs here.
1.48	28-Nov-2000	haya	BUG FIX: change enable sequence for 16-bit pcmcia cards. This change provides better (not perfect) reset sequence. The most significant change is asserting output enable reg before power up. Stop routing interrupt during reset.
1.47	27-Nov-2000	haya	BUG FIX: pccbb turns off PC Cards' power on the bus in shutdown hook. Some machines cannot reboot without this fix.
1.46	25-Oct-2000	haya	Bugfix: reset 16-bit pcmcia in chip initialisation sequence.
1.45	28-Aug-2000	haya	Correct PCI interrput setup routine for Texas Instruments PCI113X. Now PCI113X is able to handle hot insertion and deletion properly.
1.44	09-Jul-2000	jhawk	Clarify that the RX5C47X requires writes to the PCI_LEGACY to disable and not PCI_BCR_INTR; this is what rev 1.9 did under the label of "Maintenance."
1.43	09-Jul-2000	jhawk	Move legacy pcic-mode disabling code from pccbb_chipinit() to pccbbattach(). This is necessary as pccbb_chipinit() is deferred, and may not run until after the pcic is already attached. Now pcics are properly disabled on Sony VAIO, f'rinstance.
1.42	16-Jun-2000	cgd	branches: 1.42.2; beat back the undead: __BROKEN_INDIRECT_CONFIG had risen, and was terrorizing innocent hackers...
1.41	08-Jun-2000	haya	Change IPL when bridge's interrupt handler calls child device's interrupt handler (This solution is not very good).
1.40	07-Jun-2000	haya	Add CB_BCR_RESET_ENABLE in bridge control resister's bit definition.
1.39	08-May-2000	kleink	branches: 1.39.2; Fix a inverted logic botch introduced in rev. 1.35, typically causing cb_chipset() to return the wrong chipset type.
1.38	06-Apr-2000	haya	Add CardBus interrupt enable flag. When this flag is asserted, a CardBus interrupt is routed to the child devices and vice versa. This flag is negated when a pccbb receives a suspend or stanby event.
1.37	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.36	22-Mar-2000	haya	1. Do not apply 5 V on a CardBus card. 2. Do not avoid ISA hole when a device requires more than 256-byte io space. 3. Remove sending CardBus bridge's function number to cardbus stack. This number is not needed for anyone.
1.35	14-Mar-2000	enami	Simplify the code a bit.
1.34	14-Mar-2000	enami	Don't think that a card is inserted if there is a card already in the slot. Note that powerdown hook calls pccbbintr so that it detect card insertion or removal during suspension.
1.33	14-Mar-2000	enami	Cosmetic change.
1.32	14-Mar-2000	enami	Fix typos in comments.
1.31	12-Mar-2000	mycroft	A few things: * Initialize PCIC_INTR and PCIC_CSC_INTR on all chips, not just TI. * Leave card detect interrupts enabled in PCIC_CSC_INTR. (This requires reading PCIC_CSC in the interrupt handler on some chips, so do that too.) * Leave PCIC_INTR_ENABLE set, as some chips require this to post card detect interrupts while in PCMCIA mode. * Leave PCIC_INTR_RESET (which is inverted) set at all times except when powering up a socket, as some chips will use it to drive RESET# low even while a CardBus card is in the slot. This makes PCMCIA and CardBus cards works with some TI chips. Cross-checked with a Toshiba ToPIC 95B.
1.30	12-Mar-2000	mycroft	Do not fiddle with the serial/parallel IRQ mode selection on TI chips. Without extensive probing, we cannot possibly know that we're doing the right thing. Also some stylistic changes.
1.29	12-Mar-2000	mycroft	Revert the previous change. It does not match the commit message and was incomplete (and probably quite broken).
1.28	10-Mar-2000	haya	Only apply power on 3.3 V CardBus card (there are no 5 V CardBus card.)
1.27	01-Mar-2000	thorpej	Simplify the winlist code.
1.26	23-Feb-2000	haya	Pccbb maintains the interrupt handlers for 16-bit PC Cards by itself. Add some comments.
1.25	22-Feb-2000	enami	Establish a powerhook and make sure the card detect interrupt is enabled after resume from suspend, as i82365.c does. This is a workaround for newer Sony VAIO notebook. Patch is submitted by Atsushi Onoe (onoe@sm.sony.co.jp) in PR#9463.
1.24	21-Feb-2000	thorpej	Sync PCMCIA memory mapping with a change from rev 1.47 of i82365.c: always map attr memory 8 bit, since its defined that way. (chopps)
1.23	06-Feb-2000	cgd	clean up some printfs
1.22	05-Feb-2000	chopps	run this through indent with some comment cleanup -- more knfing needed.
1.21	26-Jan-2000	haya	Pccbb manages interrupt handlers by itself for its child device.
1.20	25-Jan-2000	joda	get the device name from pcidevs instead of having a list here
1.19	25-Jan-2000	haya	Do pccbb_chipinit() after YENTA register space allocation. Add TI PCI 1451 PCI-CardBus entry.
1.18	25-Jan-2000	chopps	Don't enable the interrupt until after we cycle the power, disable interrupt at disable, and make this code look more like the std pcic code. From Mycroft and I.
1.17	24-Jan-2000	joda	The TI 2030 is a PCI-PCI bridge and doesn't belong here
1.16	23-Jan-2000	mycroft	Remove code that should not have been added.
1.15	23-Jan-2000	augustss	Add parens so a ?: expression binds correctly.
1.14	17-Jan-2000	joda	add shutdown hook
1.13	13-Jan-2000	joda	remove stale comment
1.12	13-Jan-2000	joda	(pccbb_pcmcia_do_mem_map): handle 8-bit memory
1.11	13-Jan-2000	joda	make this slightly more inline with other code; and while at it, fix some oddities in pccbb_winset
1.10	12-Jan-2000	haya	Add an entry for Toshiba ToPIC100 CardBus-PCI bridge. Maintainance for 3Com multi-function, 16-bit pcmcia card.
1.9	20-Dec-1999	haya	branches: 1.9.2; Maintanance for Ricoh RL5C475. VAIO 505TX uses this chip.
1.8	15-Nov-1999	joda	(pccbb_pci_callback): no point in reading the bus number register twice
1.7	01-Nov-1999	haya	Maintainance the name of Ricoh PCI-CardBus bridges.
1.6	30-Oct-1999	augustss	Make this file compile again by using the same names as in pcidevs.
1.5	27-Oct-1999	haya	Do not stop no matter when pci_mapreg_map fails to map bridge socket registers. When it fails, reset Socket Base Address register and map registers using rbus.
1.4	19-Oct-1999	haya	branches: 1.4.2; Disable legacy io port before the bridge is probed as a pcic on isa. Comments ware changed to comply the KNF. Add PCCBB_PCMCIA_16BITONLY to support TI PCI103X series.
1.3	15-Oct-1999	augustss	branches: 1.3.2; Comment out some debug settings. Change attach printing so it looks more standard.
1.2	15-Oct-1999	haya	add NetBSD ID tag.
1.1	15-Oct-1999	haya	This is the first check-in of CardBus driver. CardBus driver contains CardBus bus stub, YENTA PCI-CardBus bridge (cbb), 3Com 3C575TX driver (ex) and Intel fxp driver. TODO: o Conform to the KNF more strictly. o Be unified with pcmcia code as much as possible. o Add more drivers for CardBus card, such as APA-1480 or USB card. The affected files are listed below. sys/arch/i386/conf/files.i386 sys/arch/macppc/conf/files.macppc sys/conf/files sys/dev/ic/elinkxl.c sys/dev/ic/elinkxlvar.h sys/dev/ic/i82365.c sys/dev/ic/i82365var.h sys/dev/isa/i82365_isasubr.c sys/dev/pci/files.pci sys/dev/pcmcia/pcmcia.c sys/dev/pcmcia/pcmciachip.h The added files are listed below. sys/arch/i386/conf/CARDBUS sys/arch/i386/include/rbus_machdep.h sys/arch/i386/i386/rbus_machdep.c sys/arch/macppc/include/rbus_machdep.h sys/arch/macppc/macppc/rbus_machdep.c sys/dev/cardbus/if_ex_cardbus.c sys/dev/cardbus/Makefile.cardbusdevs sys/dev/cardbus/cardbus.c sys/dev/cardbus/cardbus_map.c sys/dev/cardbus/cardbusdevs sys/dev/cardbus/cardbusdevs.h sys/dev/cardbus/cardbusdevs_data.h sys/dev/cardbus/cardbusvar.h sys/dev/cardbus/cardslot.c sys/dev/cardbus/cardslotvar.h sys/dev/cardbus/devlist2h.awk sys/dev/cardbus/files.cardbus sys/dev/cardbus/if_fxp_cardbus.c sys/dev/cardbus/pccardcis.h sys/dev/cardbus/rbus.c sys/dev/cardbus/rbus.h sys/dev/pci/pccbb.c sys/dev/pci/pccbbreg.h sys/dev/pci/pccbbvar.h
1.3.2.10	21-Apr-2001	bouyer	Sync with HEAD
1.3.2.9	12-Mar-2001	bouyer	Sync with HEAD.
1.3.2.8	11-Feb-2001	bouyer	Sync with HEAD.
1.3.2.7	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.3.2.6	05-Jan-2001	bouyer	Sync with HEAD
1.3.2.5	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.3.2.4	08-Dec-2000	bouyer	Sync with HEAD.
1.3.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.4.2.2	15-Nov-1999	fvdl	Sync with -current
1.4.2.1	19-Oct-1999	fvdl	file pccbb.c was added on branch fvdl-softdep on 1999-11-15 00:41:06 +0000
1.9.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.39.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.42.2.7	09-Feb-2002	he	Pull up revision 1.74 (requested by haya): Correct chattering suppressor. This makes card insertion detection work properly.
1.42.2.6	22-Jan-2002	he	Pull up revision 1.57 (requested by haya): Calculate address mask appropriately for 16-bit PC-cards.
1.42.2.5	24-Dec-2001	he	Pull up revision 1.70 (partial, via patch, requested by haya): Add PCI power management control to the PCI-CardBus bridge. Fixes PR#14149.
1.42.2.4	01-May-2001	he	Pull up revisions 1.46,1.48 (requested by fvdl): Deal correctly with 16-bit pcmcia cards. Fixes PR#12736 and PR#12737.
1.42.2.3	15-Dec-2000	he	Pull up revisions 1.47,1.49 (requested by haya): Turn off power on shutdown, except on ToPIC (TI 113X) bridges on shutdown to avoid hangs on some systems.
1.42.2.2	08-Sep-2000	haya	Pull up rev 1.45: bugfix for Texas Instruments PCI113X. This pullup fixes that the bridge does not propagate any interrupt signals when a card is inserted or removed. This pullup is approved by <thorpej@netbsd.org>.
1.42.2.1	12-Jul-2000	jhawk	Pullup rev 1.43-1.44, approved by thorpej: 1.43: Move legacy pcic-mode disabling code from pccbb_chipinit() to pccbbattach(). This is necessary as pccbb_chipinit() is deferred, and may not run until after the pcic is already attached. Now pcics are properly disabled on Sony VAIO, f'rinstance. rev 1.44: Clarify that the RX5C47X requires writes to the PCI_LEGACY to disable and not PCI_BCR_INTR; this is what rev 1.9 did under the label of "Maintenance."
1.61.2.15	03-Jan-2003	thorpej	Sync with HEAD.
1.61.2.14	29-Dec-2002	thorpej	Sync with HEAD.
1.61.2.13	11-Nov-2002	nathanw	Catch up to -current
1.61.2.12	18-Oct-2002	nathanw	Catch up to -current.
1.61.2.11	20-Jun-2002	nathanw	Catch up to -current.
1.61.2.10	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.61.2.9	28-Feb-2002	nathanw	Catch up to -current.
1.61.2.8	11-Jan-2002	nathanw	More catchup.
1.61.2.7	08-Jan-2002	nathanw	Catch up to -current.
1.61.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.61.2.5	22-Oct-2001	nathanw	Catch up to -current.
1.61.2.4	08-Oct-2001	nathanw	Catch up to -current.
1.61.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.61.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.61.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.65.2.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.65.2.6	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.65.2.5	16-Mar-2002	jdolecek	Catch up with -current.
1.65.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.65.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.65.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.65.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.67.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.69.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.75.6.1	20-Jun-2002	gehenna	catch up with -current.
1.75.2.1	17-Mar-2002	thorpej	IPL_IMP -> IPL_VM
1.90.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.90.2.7	01-Apr-2005	skrll	Sync with HEAD.
1.90.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.90.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.90.2.4	17-Jan-2005	skrll	Sync with HEAD.
1.90.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.90.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.90.2.1	03-Aug-2004	skrll	Sync with HEAD
1.96.2.2	28-Jul-2004	jmc	Pullup rev 1.99 (requested by he in ticket #724) Add entries for TI CardBus bridges 1520 and 4520Yenta, as found in IBM ThinkPad 40p and 41p respectively, so that the CardBus probes and works on those machines. PR#26380.
1.96.2.1	23-Jul-2004	he	Pull up revision 1.98 (requested by mycroft in ticket #692): If the bus number has not been initialized, do not attach a cardslot, and instead output a diagnostic pointing the administrator towards the PCIBIOS_BUS_FIXUP option.
1.112.4.3	26-Mar-2005	yamt	sync with head.
1.112.4.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.112.4.1	12-Feb-2005	yamt	sync with head.
1.112.2.2	29-Apr-2005	kent	sync with -current
1.112.2.1	16-Jan-2005	kent	file pccbb.c was added on branch kent-audio2 on 2005-04-29 11:29:07 +0000
1.117.2.2	21-Jan-2006	snj	Pull up following revision(s) (requested by uwe in ticket #1120): sys/dev/pci/pccbb.c: revision 1.127 Use aprint_debug() instead of printf() for obvious debug statement.
1.117.2.1	27-Mar-2005	tron	Pull up revision 1.118 (requested by christos in ticket #64): PR/29600: YAMAMOTO Takashi: cbb hangs on boot. From the PR: It seems that we never get a power interrupt for 16 bit cards. The tsleep has been changed to wait for 200ms maximum, before giving up instead of waiting forever. XXX: Maybe that is too small for some laptops?
1.124.2.9	17-Mar-2008	yamt	sync with head.
1.124.2.8	04-Feb-2008	yamt	sync with head.
1.124.2.7	21-Jan-2008	yamt	sync with head
1.124.2.6	07-Dec-2007	yamt	sync with head
1.124.2.5	27-Oct-2007	yamt	sync with head.
1.124.2.4	03-Sep-2007	yamt	sync with head.
1.124.2.3	26-Feb-2007	yamt	sync with head.
1.124.2.2	30-Dec-2006	yamt	sync with head.
1.124.2.1	21-Jun-2006	yamt	sync with head.
1.127.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.127.10.1	19-Apr-2006	elad	sync with head.
1.127.8.3	11-Aug-2006	yamt	sync with head
1.127.8.2	26-Jun-2006	yamt	sync with head.
1.127.8.1	11-Apr-2006	yamt	sync with head
1.127.6.1	22-Apr-2006	simonb	Sync with head.
1.127.4.1	09-Sep-2006	rpaulo	sync with head
1.128.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.128.2.1	19-Jun-2006	chap	Sync with head.
1.133.6.3	10-Dec-2006	yamt	sync with head.
1.133.6.2	02-Nov-2006	yamt	use splraiseipl.
1.133.6.1	22-Oct-2006	yamt	sync with head
1.133.4.4	09-Feb-2007	ad	Sync with HEAD.
1.133.4.3	01-Feb-2007	ad	Sync with head.
1.133.4.2	12-Jan-2007	ad	Sync with head.
1.133.4.1	18-Nov-2006	ad	Sync with head.
1.137.4.2	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.137.4.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.137.2.3	26-Oct-2007	xtraeme	Pull up following revision(s) (requested by joerg in ticket #949): sys/dev/pci/pccbb.c: revision 1.150 Remove a completely bogus check if cbb has a valid PCI interrupt routed. If you have a system where it doesn't get one and you still want pcic, you can just disable cbb. This check is invalid as soon as the PCI interrupt register is not used and broke on my Dell Latitude for that reason.
1.137.2.2	26-Oct-2007	xtraeme	Pull up following revision(s) (requested by joerg in ticket #942): sys/dev/pci/pccbb.c: revision 1.149 Correctly initialise sc_insert_ch and switch to callout_setfunc/callout_schedule.
1.137.2.1	30-Jul-2007	liamjfoy	Pull up following revision(s) (requested by dyoung in ticket #739): sys/dev/pci/pccbb.c: revision 1.140 sys/dev/pci/pccbb.c: revision 1.141 sys/dev/pci/pccbb.c: revision 1.142 sys/dev/pci/pccbb.c: revision 1.143 sys/dev/pci/pccbb.c: revision 1.144 sys/dev/pci/pccbbvar.h: revision 1.28 sys/dev/pci/pcireg.h: revision 1.53 sys/dev/pci/pccbb.c: revision 1.139 Name magic numbers PCI_PMCSR_PME_EN and PCI_PMCSR. Fix grammar in comment. From Patrick Welche. Use the right subroutine name for the debug message. Convert the rather long and backslash-ridden DELAY_MS macro to a much shorter static subroutine, delay_ms(). Cosmetic: KNF indentation, curly braces, and argument declarations. sc_pwrcycle is shared between the interrupt handler and Cardbus event thread, so make it volatile. Fix a bug in Cardbus power activation. Most Cardbus bridges supported by pccbb(4) fire a power-cycle interrupt when the power state of a cardslot changes from 'off' to 'on'. TI bridges fire a power-cycle interrupt on both on->off and off->on changes. When pccbb_power() powered-down a cardslot, it did not wait around for the power-cycle interrupt. When pccbb_power() powered-up a cardslot, it did wait for the interrupt. If a pccbb_power(UP) followed a pccbb_power(DOWN) very closely, pccbb_power() used to interpret the power-cycle interrupt for the up->down transition as "power-up complete," read the power-state bit and, finding that power had NOT been activated, complain, "cbb0: power on failed?" Then pccbb_power() exited before power-activation was complete, falsely indicating that the power-activation was complete. After that, a driver attach/enable routine would blithely configure a card that was not fully powered-up. An operator who ran a command such as 'ifconfig rtw0 down up' or 'ifconfig ath0 down up' would read 'cbb0: power on failed?' in the system log, and their NIC would misbehave. This excerpt from a comment in the source should suffice to explain how I fixed the bug, /* * Wait as long as 200ms for a power-cycle interrupt. If * interrupts are enabled, but the socket has already * changed to the desired status, keep waiting for the * interrupt. "Consuming" the interrupt in this way keeps * the interrupt from prematurely waking some subsequent * pccbb_power call. And this explains why this patch will work for Ricoh bridges that do not fire an interrupt on the on->off transition: * XXX Not every bridge interrupts on the ->OFF transition. * XXX That's ok, we will time-out after 200ms. * * XXX The power cycle event will never happen when attaching * XXX a 16-bit card. That's ok, we will time-out after * XXX 200ms. */ M. Warner Losh and Charles M. Hannum provided valuable input on this patch.
1.144.6.2	23-Oct-2007	ad	Sync with head.
1.144.6.1	20-Aug-2007	ad	Sync with HEAD.
1.145.6.8	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.145.6.7	01-Dec-2007	jmcneill	Sync with HEAD.
1.145.6.6	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.145.6.5	21-Nov-2007	joerg	Sync with HEAD.
1.145.6.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.145.6.3	28-Oct-2007	joerg	Sync with HEAD.
1.145.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.145.6.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.145.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.147.6.4	21-Nov-2007	bouyer	Sync with HEAD
1.147.6.3	18-Nov-2007	bouyer	Sync with HEAD
1.147.6.2	13-Nov-2007	bouyer	Sync with HEAD
1.147.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.147.2.3	23-Mar-2008	matt	sync with HEAD
1.147.2.2	09-Jan-2008	matt	sync with HEAD
1.147.2.1	06-Nov-2007	matt	sync with HEAD
1.150.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.150.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.150.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.150.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.155.4.1	11-Dec-2007	yamt	sync with head.
1.155.2.1	26-Dec-2007	ad	Sync with head.
1.156.2.4	19-Jan-2008	bouyer	Sync with HEAD
1.156.2.3	08-Jan-2008	bouyer	Sync with HEAD
1.156.2.2	02-Jan-2008	bouyer	Sync with HEAD
1.156.2.1	13-Dec-2007	bouyer	Sync with HEAD
1.165.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.165.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.165.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.165.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.165.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.165.2.1	24-Mar-2008	keiichi	sync with head.
1.167.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.167.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.167.4.6	11-Aug-2010	yamt	sync with head.
1.167.4.5	11-Mar-2010	yamt	sync with head
1.167.4.4	19-Aug-2009	yamt	sync with head.
1.167.4.3	20-Jun-2009	yamt	sync with head
1.167.4.2	16-May-2009	yamt	sync with head
1.167.4.1	04-May-2009	yamt	sync with head.
1.167.2.1	04-Jun-2008	yamt	sync with head
1.169.2.2	03-Jul-2008	simonb	Sync with head.
1.169.2.1	27-Jun-2008	simonb	Sync with head.
1.178.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.178.2.1	19-Oct-2008	haad	Sync with HEAD.
1.180.12.1	21-Apr-2010	matt	sync to netbsd-5
1.180.8.2	04-Jan-2009	christos	fix time_t format.
1.180.8.1	25-Oct-2008	christos	file pccbb.c was added on branch christos-time_t on 2009-01-04 02:16:41 +0000
1.180.4.1	30-Sep-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #1040): sys/dev/isa/if_lc_isa.c: revision 1.30 sys/dev/pci/pccbb.c: revision 1.183 Fix bus_addr_t/bus_size_t confusion
1.180.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.180.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.180.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.182.2.2	23-Jul-2009	jym	Sync with HEAD.
1.182.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.193.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.196.2.2	05-Mar-2011	rmind	sync with head
1.196.2.1	30-May-2010	rmind	sync with head
1.203.6.1	18-Feb-2012	mrg	merge to -current.
1.203.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.203.2.1	17-Apr-2012	yamt	sync with head
1.204.10.1	18-May-2014	rmind	sync with head
1.204.6.2	03-Dec-2017	jdolecek	update from HEAD
1.204.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.204.2.1	07-Dec-2014	martin	Pull up following revision(s) (requested by ozaki-r in ticket #1201): sys/kern/kern_ktrace.c: revision 1.166 sys/dev/isa/aps.c: revision 1.16 sys/dev/sysmon/sysmonvar.h: revision 1.45 sys/dev/ir/irframe_tty.c: revision 1.60 sys/dev/sysmon/sysmon_envsys_events.c: revision 1.111-1.112 (patch) sys/dev/pci/pccbb.c: revision 1.207 sys/dev/wscons/wskbd.c: revision 1.135 sys/dev/usb/ohci.c: revision 1.254 sys/net/if_ecosubr.c: revision 1.41 sys/dev/pcmcia/btbc.c: revision 1.17 sys/arch/x86/x86/via_padlock.c: revision 1.23 sys/dev/sdmmc/sdmmc.c: revision 1.23 (patch) sys/dev/bluetooth/btkbd.c: revision 1.17 sys/dev/bluetooth/bcsp.c: revision 1.25 sys/arch/x86/pci/fwhrng.c: revision 1.8 sys/dev/ic/nslm7x.c: revision 1.61 share/man/man9/callout.9: revision 1.28 (patch) Replace callout_stop with callout_halt and ensure the callout is not running before destroying it.
1.206.4.1	01-Dec-2014	martin	Pull up following revision(s) (requested by ozaki-r in ticket #279): sys/kern/kern_ktrace.c: revision 1.166 sys/dev/isa/aps.c: revision 1.16 sys/dev/sysmon/sysmonvar.h: revision 1.45 sys/dev/ir/irframe_tty.c: revision 1.60 sys/dev/sysmon/sysmon_envsys_events.c: revision 1.111 sys/dev/sysmon/sysmon_envsys_events.c: revision 1.112 sys/dev/pci/pccbb.c: revision 1.207 sys/dev/wscons/wskbd.c: revision 1.135 sys/dev/usb/ohci.c: revision 1.254 sys/net/if_ecosubr.c: revision 1.41 sys/dev/pcmcia/btbc.c: revision 1.17 sys/arch/x86/x86/via_padlock.c: revision 1.23 sys/dev/sdmmc/sdmmc.c: revision 1.23 sys/dev/bluetooth/btkbd.c: revision 1.17 sys/dev/bluetooth/bcsp.c: revision 1.25 sys/arch/x86/pci/fwhrng.c: revision 1.8 sys/dev/ic/nslm7x.c: revision 1.61 share/man/man9/callout.9: revision 1.28 Replace callout_stop with callout_halt In order to call callout_destroy for a callout safely, we have to ensure the function of the callout is not running and pending. To do so, we should use callout_halt, not callout_stop. Discussed with martin@ and riastradh@. Make it clear that we should use not callout_stop but callout_halt before callout_destroy Replace callout_stop with callout_halt In order to call callout_destroy for a callout safely, we have to ensure the function of the callout is not running and pending. To do so, we should use callout_halt, not callout_stop. In this case, we need to pass an interlock to callout_halt to wait for the callout complete. Reviewed by riastradh@. Kill sme_callout_mtx and use sme_mtx instead We can use sme_mtx for the callout as well. Actually we should do so because sme_events_list and some other data that are touched in the callout should be protected by sme_mtx, not sme_callout_mtx. Discussed with riastradh@ in http://mail-index.netbsd.org/tech-kern/2014/11/11/msg017956.html Replace callout_stop with callout_halt In order to call callout_destroy for a callout safely, we have to ensure the function of the callout is not running and pending. To do so, we should use callout_halt, not callout_stop. In this case, we need to pass an interlock to callout_halt to wait for the callout complete. And also we make sure that SME_CALLOUT_INITIALIZED is unset before calling callout_halt to prevent the callout from calling callout_schedule. This is the same as what we did in sys/netinet6/mld6.c@1.61. Reviewed by riastradh@.
1.207.2.4	28-Aug-2017	skrll	Sync with HEAD
1.207.2.3	09-Jul-2016	skrll	Sync with HEAD
1.207.2.2	19-Mar-2016	skrll	Sync with HEAD
1.207.2.1	06-Apr-2015	skrll	Sync with HEAD
1.210.8.1	11-May-2017	pgoyette	Sync with HEAD
1.211.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.211.10.1	10-Jun-2019	christos	Sync with HEAD
1.211.8.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.215.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.216.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.17	06-May-2023	andvar	few more typos in the word "register", mainly s/resister/register/. In one comment I assume that it was meant to be pull-up resistor.
1.16	15-Oct-2019	msaitoh	Use unsigned to avoid undefined behavior. Found by kUBSan.
1.15	15-Dec-2009	snj	branches: 1.15.64; 1.15.70; Move to 2-clause license. Approved by HAYAKAWA Koichi (copyright holder).
1.14	03-Jul-2008	drochner	On cardbus configuration reads, check for a master abort in the cbb and return all-ones explicitely. Since pccbb.c rev. 1.151 the cbb is set to translate cardbus master aborts which makes that the PCI configuration read on the primary bus returns random junk. This can cause the probing code to get confused. The issue was reported by KIYOHARA Takashi and Jonathan A. Kollasch, the patch was tested by KIYOHARA Takashi.
1.13	16-Nov-2007	dyoung	branches: 1.13.14; 1.13.18; 1.13.20; 1.13.22; Improve Cardbus performance, error handling, and error reporting. Let the TI1420 PCI-Cardbus bridge do burst reads from the primary (PCI) bus. This ought to improve Tx performance on Cardbus NICs. This optimization may apply to other TI bridges, but I only have a datasheet for the TI1420. :-/ Activate PCI Parity Error and System Error reporting on PCI-Cardbus bridges. To avoid data destruction, set the Master Abort mode to 1. Stops the bridge from silently discarding writes from the secondary bus to the primary bus (Cardbus writes to PCI). Also, stops the bridge from fulfilling a read by a bus master on the secondary bus that failed on the primary bus with 0xffffffff (Cardbus reads from PCI). Now the bus will indicate an error condition (SERR) instead of silently destroying/corrupting data. Forward system error indications from the secondary to the primary bus. Detect parity errors on the secondary. Set a Cardbus card's Latency Timer to something reasonable, according to the Cardbus card's Minimum Grant and the bandwidth available on the PCI bus. Restore the Latency Timer when re-enabling a card (e.g., after power reactivation).
1.12	11-Aug-2007	dyoung	branches: 1.12.2; 1.12.6; 1.12.8; Make pccbb(4) use more PCI configuration register definitions, especially the Bridge Control Register definition, instead of rolling its own. As we read/modify/write configuration registers, use a variable that is named for the register we're r/m/w'ing, instead of using the variable 'reg' over and over. This helps both a human reader verify that we're not reading register X, modifying it, and writing back to register Y (oops); the compiler can help a little by warning that a variable is used before it is initialized.
1.11	17-Jun-2006	jmcneill	branches: 1.11.14; 1.11.24; 1.11.28; For Ricoh 5c47x class controllers, restore the "misc control" PCI config register on resume.
1.10	25-Apr-2006	dyoung	branches: 1.10.2; 1.10.4; Fix a typo in register definitions: looks like somebody copied the CB_SOCKET_STAT_3VSOCK definition to CB_SOCKET_STAT_XVSOCK and CB_SOCKET_STAT_YVSOCK definitions without updating the numbers. It doesn't look like CB_SOCKET_STAT_XVSOCK and CB_SOCKET_STAT_YVSOCK are used anywhere, so this will not make any functional difference.
1.9	11-Dec-2005	christos	branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; 1.9.12; merge ktrace-lwp.
1.8	10-Dec-2003	briggs	branches: 1.8.16; Power up the card without OE enabled to avoid letting SanDisks enter True IDE mode. Hinted at by Charles Hannum a while back. This lets my PCI-Cardbus adapter read SanDisk CF cards much more reliably. Also bring in a few defines from FreeBSD for some more TI registers and such.
1.7	30-Aug-2001	haya	branches: 1.7.20; Add some register definitions for TI PCI12XX and Toshiba ToPIC. Add ToPIC97 and ToPIC100 chip initialisation routine. Add more TI PCI12XX chip initialisation.
1.6	08-Dec-2000	mycroft	branches: 1.6.2; 1.6.4; Turn off the damned TI `VCC protection' bit.
1.5	07-Jun-2000	haya	Add CB_BCR_RESET_ENABLE in bridge control resister's bit definition.
1.4	13-Jan-2000	joda	branches: 1.4.2; add %b strings for sockevent and sockstate registers
1.3	01-Nov-1999	haya	branches: 1.3.4; Maintainance the name of Ricoh PCI-CardBus bridges.
1.2	15-Oct-1999	haya	branches: 1.2.2; 1.2.4; add NetBSD ID tag.
1.1	15-Oct-1999	haya	This is the first check-in of CardBus driver. CardBus driver contains CardBus bus stub, YENTA PCI-CardBus bridge (cbb), 3Com 3C575TX driver (ex) and Intel fxp driver. TODO: o Conform to the KNF more strictly. o Be unified with pcmcia code as much as possible. o Add more drivers for CardBus card, such as APA-1480 or USB card. The affected files are listed below. sys/arch/i386/conf/files.i386 sys/arch/macppc/conf/files.macppc sys/conf/files sys/dev/ic/elinkxl.c sys/dev/ic/elinkxlvar.h sys/dev/ic/i82365.c sys/dev/ic/i82365var.h sys/dev/isa/i82365_isasubr.c sys/dev/pci/files.pci sys/dev/pcmcia/pcmcia.c sys/dev/pcmcia/pcmciachip.h The added files are listed below. sys/arch/i386/conf/CARDBUS sys/arch/i386/include/rbus_machdep.h sys/arch/i386/i386/rbus_machdep.c sys/arch/macppc/include/rbus_machdep.h sys/arch/macppc/macppc/rbus_machdep.c sys/dev/cardbus/if_ex_cardbus.c sys/dev/cardbus/Makefile.cardbusdevs sys/dev/cardbus/cardbus.c sys/dev/cardbus/cardbus_map.c sys/dev/cardbus/cardbusdevs sys/dev/cardbus/cardbusdevs.h sys/dev/cardbus/cardbusdevs_data.h sys/dev/cardbus/cardbusvar.h sys/dev/cardbus/cardslot.c sys/dev/cardbus/cardslotvar.h sys/dev/cardbus/devlist2h.awk sys/dev/cardbus/files.cardbus sys/dev/cardbus/if_fxp_cardbus.c sys/dev/cardbus/pccardcis.h sys/dev/cardbus/rbus.c sys/dev/cardbus/rbus.h sys/dev/pci/pccbb.c sys/dev/pci/pccbbreg.h sys/dev/pci/pccbbvar.h
1.2.4.1	15-Nov-1999	fvdl	Sync with -current
1.2.2.2	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.4.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.4.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.6.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.6.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.7.20.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.20.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.20.1	03-Aug-2004	skrll	Sync with HEAD
1.8.16.3	07-Dec-2007	yamt	sync with head
1.8.16.2	03-Sep-2007	yamt	sync with head.
1.8.16.1	21-Jun-2006	yamt	sync with head.
1.9.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.9.10.1	11-May-2006	elad	sync with head
1.9.8.2	26-Jun-2006	yamt	sync with head.
1.9.8.1	24-May-2006	yamt	sync with head.
1.9.6.1	01-Jun-2006	kardel	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.10.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.10.2.1	19-Jun-2006	chap	Sync with head.
1.11.28.2	21-Nov-2007	joerg	Sync with HEAD.
1.11.28.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.11.24.1	15-Aug-2007	skrll	Sync with HEAD.
1.11.14.1	20-Aug-2007	ad	Sync with HEAD.
1.12.8.1	19-Nov-2007	mjf	Sync with HEAD.
1.12.6.1	18-Nov-2007	bouyer	Sync with HEAD
1.12.2.1	09-Jan-2008	matt	sync with HEAD
1.13.22.1	03-Jul-2008	simonb	Sync with head.
1.13.20.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.13.18.2	11-Mar-2010	yamt	sync with head
1.13.18.1	04-May-2009	yamt	sync with head.
1.13.14.1	28-Sep-2008	mjf	Sync with HEAD.
1.15.70.1	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #671): sys/dev/pci/if_bce.c: revision 1.53 sys/dev/pci/pccbbreg.h: revision 1.16 sys/dev/ic/rt2860.c: revision 1.34 sys/dev/pci/if_alc.c: revision 1.45 sys/dev/pci/if_mcx.c: revision 1.5 sys/dev/pci/if_pcn.c: revision 1.72 sys/dev/pci/if_ale.c: revision 1.37 sys/dev/pci/if_age.c: revision 1.65 sys/dev/ieee1394/fwohcireg.h: revision 1.20 sys/dev/ieee1394/fwohci.c: revision 1.143 sys/dev/ieee1394/firewire.c: revision 1.49 sys/dev/ic/am79900reg.h: revision 1.10 Use unsigned to avoid undefined behavior. Found by kUBSan.
1.15.64.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.41	20-Apr-2010	dyoung	Change sc_pil_intr_enable from an int to a bool. No functional change intended.
1.40	15-Mar-2010	dyoung	rbus is always #defined as 1, so unifdef -D rbus=1.
1.39	15-Dec-2009	snj	branches: 1.39.2; 1.39.4; Move to 2-clause license. Approved by HAYAKAWA Koichi (copyright holder).
1.38	23-Jul-2009	dyoung	Use kpause(9), cv_timedwait(9), and cv_signal(9) instead of tsleep(9) and wakeup(9). Use mstohz(9). XXX Protection against spurious wakeups is still needed, but this patch XXX makes the code no worse than before in this regard.
1.37	26-Jun-2008	drochner	delete more dead code, remove use of pcic_handle except of some local uses in pcmcia ressource management which is copied from ISA pcic, so use our softc as handle exported to generic pcmcia code and kill the ph_parent backpointer - this makes the code simpler and smaller
1.36	25-Jun-2008	drochner	Avoid the use of i82365var.h - we don't use any functions from there. Extract some data structures for ressource management into our private header instead. This allows to use a typed pointer instead of a generic one which saves a lot of typecasts. Also remove something marked as "dirty hack" which I admittedly don't understand, but it doesn't look useful...
1.35	25-Jun-2008	drochner	split device/softc for pccbb, the cardbus half of cardslot and cardbus (not yet pcmcia, there is a lot of md code involved)
1.34	27-May-2008	dyoung	branches: 1.34.2; Distinguish O2 Micro bridges from generic bridges. PCI Parity Error reporting seems to be broken on O2 Micro bridges, so do not activate it. Should fix kern/38698, reported by Andreas Gustafsson.
1.33	15-Jan-2008	christos	branches: 1.33.6; 1.33.8; 1.33.10; 1.33.12; Remove ricoh hack; FreeBSD does not seem to do it, and it was a remnant from the branch. Things should work properly now.
1.32	14-Jan-2008	dyoung	Use device_t, device_private().
1.31	16-Dec-2007	dyoung	Improve pci, cbb, cardslot, cardbus, and pcmcia to support detachment of this entire device tree: pci0 at mainbus0 elansc0 at pci0 gpio0 at elansc0 cbb0 at pci0 cardslot0 at cbb0 cardbus0 at cardslot0 pcmcia0 at cardslot0 cbb1 at pci0 cardslot1 at cbb1 cardbus1 at cardslot1 rtw0 at cardbus1 pcmcia1 at cardslot1 sip0 at pci0 nsphyter0 at sip0 sip1 at pci0 nsphyter1 at sip1 Whew!
1.30	09-Dec-2007	jmcneill	branches: 1.30.2; Merge jmcneill-pm branch.
1.29	10-Aug-2007	dyoung	branches: 1.29.2; 1.29.8; 1.29.10; 1.29.12; Remove unused definitions CB_CHIPS_LAST and cb_chipset_name. Add a type code for the Texas Instruments PCI1420 PCI-CardBus bridge, CB_TI1420, which I will use in the near future.
1.28	04-Feb-2007	dyoung	branches: 1.28.6; 1.28.14; 1.28.18; sc_pwrcycle is shared between the interrupt handler and Cardbus event thread, so make it volatile.
1.27	21-Dec-2006	yamt	merge yamt-splraiseipl branch. - finish implementing splraiseipl (and makeiplcookie). http://mail-index.NetBSD.org/tech-kern/2006/07/01/0000.html - complete workqueue(9) and fix its ipl problem, which is reported to cause audio skipping. - fix netbt (at least compilation problems) for some ports. - fix PR/33218.
1.26	17-Jun-2006	jmcneill	branches: 1.26.4; 1.26.6; 1.26.8; 1.26.10; For Ricoh 5c47x class controllers, restore the "misc control" PCI config register on resume.
1.25	11-Dec-2005	christos	branches: 1.25.4; 1.25.8; 1.25.14; 1.25.16; merge ktrace-lwp.
1.24	04-Feb-2005	perry	branches: 1.24.6; de-__P
1.23	27-Jan-2005	jmcneill	Capture PCI configuration state on suspend and restore it on resume, from Kentaro A. Kurahone.
1.22	16-Jan-2005	mycroft	branches: 1.22.2; 1.22.4; Christos's crappy Libretto 100CT has an extremely long ramp for socket Vcc (violating the PC Card spec), so... use the "power cycle" socket event to determine when we've reached Vcc before proceeding, rather than using a fixed amount of time. This has the double advantage that it makes the card attach time even shorter on sane systems -- the minimum is now ~38ms on my i8500 rather than 222ms. Probably a similar change should be made to pcic, but it was hard enough figuring out whether it would work with pccbb. The chip specs suck. For now, I'm leaving in a couple of additional printf()s in the hope that I will get some interesting data from them.
1.21	22-Jul-2004	mycroft	If the bus number has not been initialized, do not attach a cardslot, and output a diagnostic: pccbbN: secondary bus number uninitialized; try PCIBIOS_BUS_FIXUP
1.20	10-Dec-2003	briggs	branches: 1.20.2; Power up the card without OE enabled to avoid letting SanDisks enter True IDE mode. Hinted at by Charles Hannum a while back. This lets my PCI-Cardbus adapter read SanDisk CF cards much more reliably. Also bring in a few defines from FreeBSD for some more TI registers and such.
1.19	22-Mar-2003	nakayama	branches: 1.19.2; PCI-CardBus bridge support for sparc64: - add pccbb_attach_hook in pccbbattach for MD initializations. - omit arithmetics to bus_space_handle_t. - remove use of IST_LEVEL; not defined on sparc64 and unused.
1.18	01-Oct-2002	haya	Use LIST in sys/queue. Add some comments.
1.17	02-Nov-2001	haya	Add PCI power management control. If the power state of a bridge is not D0 at device attaching or resuming, set the bridge D0 mode.
1.16	06-Jul-2001	mcr	branches: 1.16.2; 1.16.6; split pccbb_intr_route() into seperate function.
1.15	20-Feb-2001	minoura	branches: 1.15.2; Re-initialize more registers after resuming from suspend. Some APM BIOSes do not restore them. Reviewd by haya.
1.14	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.13	08-Jun-2000	haya	Change IPL when bridge's interrupt handler calls child device's interrupt handler (This solution is not very good).
1.12	23-Mar-2000	thorpej	branches: 1.12.2; New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.11	22-Mar-2000	haya	1. Do not apply 5 V on a CardBus card. 2. Do not avoid ISA hole when a device requires more than 256-byte io space. 3. Remove sending CardBus bridge's function number to cardbus stack. This number is not needed for anyone.
1.10	15-Mar-2000	haya	Remove ancient code.
1.9	12-Mar-2000	mycroft	Revert the previous change.
1.8	12-Mar-2000	thorpej	Add missing sc_pil_intr_enable member to pccbb_softc. XXX Double check this, please. The code appears to only use it as a boolean, but...
1.7	01-Mar-2000	thorpej	Simplify the winlist code.
1.6	05-Feb-2000	chopps	indent this file
1.5	26-Jan-2000	haya	Pccbb manages interrupt handlers by itself for its child device.
1.4	01-Nov-1999	haya	branches: 1.4.4; Maintainance the name of Ricoh PCI-CardBus bridges.
1.3	19-Oct-1999	haya	branches: 1.3.2; Disable legacy io port before the bridge is probed as a pcic on isa. Comments ware changed to comply the KNF. Add PCCBB_PCMCIA_16BITONLY to support TI PCI103X series.
1.2	15-Oct-1999	haya	branches: 1.2.2; add NetBSD ID tag.
1.1	15-Oct-1999	haya	This is the first check-in of CardBus driver. CardBus driver contains CardBus bus stub, YENTA PCI-CardBus bridge (cbb), 3Com 3C575TX driver (ex) and Intel fxp driver. TODO: o Conform to the KNF more strictly. o Be unified with pcmcia code as much as possible. o Add more drivers for CardBus card, such as APA-1480 or USB card. The affected files are listed below. sys/arch/i386/conf/files.i386 sys/arch/macppc/conf/files.macppc sys/conf/files sys/dev/ic/elinkxl.c sys/dev/ic/elinkxlvar.h sys/dev/ic/i82365.c sys/dev/ic/i82365var.h sys/dev/isa/i82365_isasubr.c sys/dev/pci/files.pci sys/dev/pcmcia/pcmcia.c sys/dev/pcmcia/pcmciachip.h The added files are listed below. sys/arch/i386/conf/CARDBUS sys/arch/i386/include/rbus_machdep.h sys/arch/i386/i386/rbus_machdep.c sys/arch/macppc/include/rbus_machdep.h sys/arch/macppc/macppc/rbus_machdep.c sys/dev/cardbus/if_ex_cardbus.c sys/dev/cardbus/Makefile.cardbusdevs sys/dev/cardbus/cardbus.c sys/dev/cardbus/cardbus_map.c sys/dev/cardbus/cardbusdevs sys/dev/cardbus/cardbusdevs.h sys/dev/cardbus/cardbusdevs_data.h sys/dev/cardbus/cardbusvar.h sys/dev/cardbus/cardslot.c sys/dev/cardbus/cardslotvar.h sys/dev/cardbus/devlist2h.awk sys/dev/cardbus/files.cardbus sys/dev/cardbus/if_fxp_cardbus.c sys/dev/cardbus/pccardcis.h sys/dev/cardbus/rbus.c sys/dev/cardbus/rbus.h sys/dev/pci/pccbb.c sys/dev/pci/pccbbreg.h sys/dev/pci/pccbbvar.h
1.2.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.3.2.2	15-Nov-1999	fvdl	Sync with -current
1.3.2.1	19-Oct-1999	fvdl	file pccbbvar.h was added on branch fvdl-softdep on 1999-11-15 00:41:06 +0000
1.4.4.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.12.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.15.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.15.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.15.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.16.6.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.16.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.16.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.2.3	04-Feb-2005	skrll	Sync with HEAD.
1.19.2.2	17-Jan-2005	skrll	Sync with HEAD.
1.19.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.2.1	23-Jul-2004	he	Pull up revision 1.21 (requested by mycroft in ticket #692): If the bus number has not been initialized, do not attach a cardslot, and instead output a diagnostic pointing the administrator towards the PCIBIOS_BUS_FIXUP option.
1.22.4.1	12-Feb-2005	yamt	sync with head.
1.22.2.2	29-Apr-2005	kent	sync with -current
1.22.2.1	16-Jan-2005	kent	file pccbbvar.h was added on branch kent-audio2 on 2005-04-29 11:29:07 +0000
1.24.6.5	21-Jan-2008	yamt	sync with head
1.24.6.4	03-Sep-2007	yamt	sync with head.
1.24.6.3	26-Feb-2007	yamt	sync with head.
1.24.6.2	30-Dec-2006	yamt	sync with head.
1.24.6.1	21-Jun-2006	yamt	sync with head.
1.25.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.25.14.1	19-Jun-2006	chap	Sync with head.
1.25.8.1	26-Jun-2006	yamt	sync with head.
1.25.4.1	09-Sep-2006	rpaulo	sync with head
1.26.10.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.26.8.1	30-Jul-2007	liamjfoy	Pull up following revision(s) (requested by dyoung in ticket #739): sys/dev/pci/pccbb.c: revision 1.140 sys/dev/pci/pccbb.c: revision 1.141 sys/dev/pci/pccbb.c: revision 1.142 sys/dev/pci/pccbb.c: revision 1.143 sys/dev/pci/pccbb.c: revision 1.144 sys/dev/pci/pccbbvar.h: revision 1.28 sys/dev/pci/pcireg.h: revision 1.53 sys/dev/pci/pccbb.c: revision 1.139 Name magic numbers PCI_PMCSR_PME_EN and PCI_PMCSR. Fix grammar in comment. From Patrick Welche. Use the right subroutine name for the debug message. Convert the rather long and backslash-ridden DELAY_MS macro to a much shorter static subroutine, delay_ms(). Cosmetic: KNF indentation, curly braces, and argument declarations. sc_pwrcycle is shared between the interrupt handler and Cardbus event thread, so make it volatile. Fix a bug in Cardbus power activation. Most Cardbus bridges supported by pccbb(4) fire a power-cycle interrupt when the power state of a cardslot changes from 'off' to 'on'. TI bridges fire a power-cycle interrupt on both on->off and off->on changes. When pccbb_power() powered-down a cardslot, it did not wait around for the power-cycle interrupt. When pccbb_power() powered-up a cardslot, it did wait for the interrupt. If a pccbb_power(UP) followed a pccbb_power(DOWN) very closely, pccbb_power() used to interpret the power-cycle interrupt for the up->down transition as "power-up complete," read the power-state bit and, finding that power had NOT been activated, complain, "cbb0: power on failed?" Then pccbb_power() exited before power-activation was complete, falsely indicating that the power-activation was complete. After that, a driver attach/enable routine would blithely configure a card that was not fully powered-up. An operator who ran a command such as 'ifconfig rtw0 down up' or 'ifconfig ath0 down up' would read 'cbb0: power on failed?' in the system log, and their NIC would misbehave. This excerpt from a comment in the source should suffice to explain how I fixed the bug, /* * Wait as long as 200ms for a power-cycle interrupt. If * interrupts are enabled, but the socket has already * changed to the desired status, keep waiting for the * interrupt. "Consuming" the interrupt in this way keeps * the interrupt from prematurely waking some subsequent * pccbb_power call. And this explains why this patch will work for Ricoh bridges that do not fire an interrupt on the on->off transition: * XXX Not every bridge interrupts on the ->OFF transition. * XXX That's ok, we will time-out after 200ms. * * XXX The power cycle event will never happen when attaching * XXX a 16-bit card. That's ok, we will time-out after * XXX 200ms. */ M. Warner Losh and Charles M. Hannum provided valuable input on this patch.
1.26.6.1	02-Nov-2006	yamt	use splraiseipl.
1.26.4.2	09-Feb-2007	ad	Sync with HEAD.
1.26.4.1	12-Jan-2007	ad	Sync with head.
1.28.18.2	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.28.18.1	16-Aug-2007	jmcneill	Sync with HEAD.
1.28.14.1	15-Aug-2007	skrll	Sync with HEAD.
1.28.6.1	20-Aug-2007	ad	Sync with HEAD.
1.29.12.1	11-Dec-2007	yamt	sync with head.
1.29.10.1	26-Dec-2007	ad	Sync with head.
1.29.8.1	18-Feb-2008	mjf	Sync with HEAD.
1.29.2.2	23-Mar-2008	matt	sync with HEAD
1.29.2.1	09-Jan-2008	matt	sync with HEAD
1.30.2.2	19-Jan-2008	bouyer	Sync with HEAD
1.30.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.33.12.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.33.12.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.33.10.4	11-Aug-2010	yamt	sync with head.
1.33.10.3	11-Mar-2010	yamt	sync with head
1.33.10.2	19-Aug-2009	yamt	sync with head.
1.33.10.1	04-May-2009	yamt	sync with head.
1.33.8.1	04-Jun-2008	yamt	sync with head
1.33.6.2	29-Jun-2008	mjf	Sync with HEAD.
1.33.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.34.2.1	27-Jun-2008	simonb	Sync with head.
1.39.4.1	30-May-2010	rmind	sync with head
1.39.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.168	23-Jun-2024	riastradh	pci: Pass cookie through pci_find_device, pci_enumerate_bus, take 2. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This will allow pci_find_device callers to pass a cookie through to the match function so they can keep state or pass in extra parameters like b/d/f numbers, which will allow us to nix some horrible kludges in the Linux PCI API emulation for drm (and, perhaps, Intel wifi). This change drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway. Take 2: Make sure to handle NULL match function.
1.167	20-May-2024	riastradh	Revert "pci: Pass cookie through pci_find_device, pci_enumerate_bus." Evidently something is wrong with this, to be diagnosed and redone once the builds and tests are in better shape.
1.166	20-May-2024	riastradh	pci: Pass cookie through pci_find_device, pci_enumerate_bus. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway.
1.165	24-Aug-2022	riastradh	branches: 1.165.4; pmf(9): *_child_register never fails. Make it return void. No kernel bump because this isn't documented or used in any modules, only in dev/pci/pci.c and dev/cardbus/cardbus.c which are as far as I know always statically linked into the kernel. The next change, however, will require a revbump -- to make pmf_device_register return void so we can prune vast swaths of dead error branches.
1.164	21-Jan-2022	thorpej	Replace devhandle_invalidate(), which invalidates a devhandle, with devhandle_invalid(), which returns an invalid devhandle.
1.163	10-Oct-2021	msaitoh	Use PCI-SIG official acronyms: - RP stands for Root Port. - RC stands for Root Complex. - RCIEP stands for Root Complex Integrated End Point.
1.162	15-Sep-2021	thorpej	Adjust the device_call() calling convention so as to provide type checking of the arguments passed to the call, using auto-generated argument structures and binding macros.
1.161	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.160	12-May-2021	thorpej	branches: 1.160.4; - Define a device call for PCI bus instances to fetch a direct child's device handle given the device's device/function #s (extracted from a pcitag_t). Use it to associate the handle with the child device at config_found() time. - Implement this device call for ACPI and OpenFirmware. - Enable the OpenFirmware variant for evbarm FDT, macppc, ofppc, sparc64. - Obsolete acpi_device_register(); it is no longer needed. - Obsolete setting the OpenFirmware handle in PCI devices in the sparc64 device_register(); it is no longer needed.
1.159	24-Apr-2021	thorpej	branches: 1.159.2; 1.159.4; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.158	29-Jan-2021	skrll	branches: 1.158.2; Trailing whitespace
1.157	02-Feb-2020	jmcneill	branches: 1.157.6; pci_probe_device: make sure PCI_ID_REG returns valid data before touching any other registers
1.156	15-Oct-2019	jmcneill	branches: 1.156.2; PCIe downstream ports only have a single child device, so limit probing to dev 0.
1.155	14-Oct-2019	jmcneill	The PCI Local Bus Specification says that we should probe devices by reading PCI_ID_REG. pci_enumerate_bus was incorrectly reading PCI_BHLC_REG first, which surprisingly has worked for the past 16 years. Unfortunately, this undefined behavior does the wrong thing on Amazon's Annapurna Labs PCIe host controller.
1.154	15-Dec-2018	msaitoh	branches: 1.154.4; Clearing PCI_PMCSR_PME_STS(W1C) bit is required to stop asserting PME#. This change would prevent unexpected rebooting from shutdown -p or unexpected resuming from suspend.
1.153	01-Dec-2018	msaitoh	Save control registers in PCI-X, PCIe, MSI and MSI-X capability area when suspend and restore them when resume. For PCIe cababilities register, it's required to check the existence of each register to not to write the next area. This chagnge fixes a stability of suspend/resume.
1.152	05-Apr-2017	msaitoh	branches: 1.152.6; 1.152.12; 1.152.14; IF PCIe's Alternative Routing-ID Interpretation (ARI) is enabled in a bridge, all area under the bus should be checked in pci_enumerate_bus() to enumerate all existing devices. The similar code also be required for "pcictl pciN list". (not yet)
1.151	23-Jan-2016	macallan	branches: 1.151.2; 1.151.4; add another special case so SM50x's framebuffer gets mapped with prefetching / write combining enabled. This needs to be generalized somewhere...
1.150	02-Nov-2015	knakahara	Add verbose messages when the kernel disables MSI/MSI-X.
1.149	02-Oct-2015	msaitoh	PCI Extended Configuration stuff written by nonaka@: - Add PCI Extended Configuration Space support into x86. - Check register offset of pci_conf_read() in MD part. It returns (pcireg_t)-1 if it isn't accessible. - Decode Extended Capability in PCI Extended Configuration Space. Currently the following extended capabilities are decoded: - Advanced Error Reporting - Virtual Channel - Device Serial Number - Power Budgeting - Root Complex Link Declaration - Root Complex Event Collector Association - Access Control Services - Alternative Routing-ID Interpretation - Address Translation Services - Single Root IO Virtualization - Page Request - TPH Requester - Latency Tolerance Reporting - Secondary PCI Express - Process Address Space ID - LN Requester - L1 PM Substates The following extended capabilities are not decoded yet: - Root Complex Internal Link Control - Multi-Function Virtual Channel - RCRB Header - Vendor Unique - Configuration Access Correction - Multiple Root IO Virtualization - Multicast - Resizable BAR - Dynamic Power Allocation - Protocol Multiplexing - Downstream Port Containment - Precision Time Management - M-PCIe - Function Reading Status Queueing - Readiness Time Reporting - Designated Vendor-Specific
1.148	24-Aug-2015	pooka	add some _KERNEL_OPT as the finishing touch
1.147	13-Aug-2015	msaitoh	- Don't take pci_attach_args as an argument in pci_msi[x]_count(). - Move prototypes of pci_msi[x]_count() from x86/x86/pci_machdep_common to sys/dev/pci/pcivar.h. - Move pci_msi[x]_count() from x86/pci/pci_msi_machdep.c to sys/dev/pci/pci.c
1.146	27-Apr-2015	knakahara	add x86 MD MSI/MSI-X support code.
1.145	05-Sep-2014	matt	branches: 1.145.2; Don't use class or typename as a variable name.
1.144	15-Sep-2013	martin	Comment out a currently unused variable, but leave it around for debug/ documentation purposes.
1.143	20-Oct-2012	matt	branches: 1.143.2; Add a __PCI_DEV_FUNCORDER hook.
1.142	29-Aug-2011	jmcneill	branches: 1.142.2; 1.142.12; have pci register itself as a module, and make PCI driver modules depend on it
1.141	24-Aug-2011	dyoung	Add to pci_bus_devorder() an argument that tells the number of slots available in the devs array. Change the type of the devs array from char to uint8_t. Treat the return value of pci_bus_devorder() as the number of slots that it filled. Don't use the __PCI_BUS_DEVORDER #definition to configure the kernel but let the linker do it. Make pci_bus_devorder() available on all architectures by adding a default implementation that will DTRT on all architectures but hpcmips, the only architecture to #define __PCI_BUS_DEVORDER. On hpcmips, adapt the implementation to the new calling convention. XXX I can compile an hpcmips GENERIC kernel, but I don't have a XXX hpcmips box to test it on.
1.140	17-May-2011	dyoung	PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.139	10-May-2011	dyoung	Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.138	05-Apr-2011	dyoung	Lightly KNF. Changes to the generated assembly language appeared to correspond to line numbers in KASSERT() statements.
1.137	05-Apr-2011	macallan	fix a typo
1.136	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.135	02-Mar-2011	matt	When probing the PCI bars, make sure to only probe BARS (PPBs and PCB has fewer bars than type 0 devices).
1.134	27-Feb-2011	jmcneill	follow the lead of genfb_pci and call pci_mapreg_probe before pci_mapreg_info
1.133	24-Feb-2011	macallan	add a hack to allow mapping the video memory portions of a mach64's aperture with BUS_SPACE_MAP_PREFETCHABLE This probably needs to go in its own function to avoid more clutter when adding more hacks. tested on macppc
1.132	10-Feb-2011	jmcneill	pcimmap: if the requested page is marked prefetchable in a child device's BAR, pass the BUS_SPACE_MAP_PREFETCHABLE flag down to bus_space_mmap
1.131	01-Feb-2011	dyoung	During shutdown, restore the each PCI device's bus-mastering enable to the state that we found it in before attaching a driver to it.
1.130	13-Nov-2010	uebayasi	branches: 1.130.2; 1.130.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.129	06-Jun-2010	pgoyette	Update pciverbose module to use module_autoload() rather than module_load(). Load the module right before each attempt to use its features, and let the module subsystem handle unloading.
1.128	24-May-2010	pgoyette	Extract the vendor/product tables and related access routines into a separate kernel module. Update pci bus attach routine to load the module (if available) when we're about to start scanning the bus, and unload the module after the scan is finished. On architectures which support loading of modules by the boot loader, the 'pciverbose' module can be loaded and executed without needing to rebuild the kernel. On all architectures, using 'options PCIVERBOSE' in the kernel configuration file will create a 'builtin' module which is functionally equivalent to previous behavior. XXX Although not nearly as large as the vendor and product tables, XXX the PCI class and subclass tables might also be offloaded into XXX the module at a future time. XXX Cardbus (and possibly other) drivers should also be modified to XXX load the module before scanning/attaching devices.
1.127	24-Feb-2010	dyoung	branches: 1.127.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.126	23-Feb-2010	dyoung	Remove unused functions pci_disable_retry() and cardbus_disable_retry().
1.125	08-Jan-2010	dyoung	branches: 1.125.2; Expand PMF_FN_* macros.
1.124	19-Aug-2009	dyoung	Since the intention in capturing a device's PCI Configuration Space in pci_probe_device() is for pcidevdetached() to restore the device's original PCI Configuration Space after it is detached, let's capture a device's PCI configuration before attaching a driver to it instead of after. Capturing the configuration after attaching is too late, because a driver such as ath(4) may put a device to sleep promptly after attaching it, and putting a PCI device to sleep often entails changing PCI configuration from defaults.
1.123	04-Jul-2009	cegger	remove useless parenthesis
1.122	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.121	16-Dec-2008	christos	branches: 1.121.2; replace bitmask_snprintf(9) with snprintb(3)
1.120	16-Nov-2008	bouyer	Do not depend on physmem for pci_dma64_available(). On some systems (e.g. Xen, or any system with non-contigous physical memory), physmem may not represent the hightest physical address available. Discussed on tech-kern@ some weeks ago.
1.119	19-Sep-2008	joerg	branches: 1.119.2; 1.119.4; Complain about devices with misaligned PCI capability pointers instead of panicing. Christoph Schug reported that such a device is made by Nvidia.
1.118	12-Jun-2008	cegger	branches: 1.118.2; use device_lookup_private to get softc
1.117	05-May-2008	dyoung	branches: 1.117.2; 1.117.4; Restore PCI devices to a state that is closer to the state that NetBSD finds them in: save and restore power management state (D0..D3) and PCI Configuration Registers 0x0 through 0x40 during device attachment and detachment, respectively. Among other things, this will fix sip(4) detachment and re-attachment.
1.116	09-Apr-2008	dyoung	branches: 1.116.2; 1.116.4; Export some PCI autoconf routines for use by elansc(4), for example.
1.115	23-Mar-2008	cube	Split device_t and softc for pci(4).
1.114	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.113	28-Feb-2008	drochner	Extend the pmf suspend/resume hooks by a shutdown method, so drivers can register a shutdown handler explicitely. Install a pci bus shutdown handler which disables bus master accesses for all childs, so the drivers don't need to care. This will hopefully be sufficient to replace the shutdownhooks (together with the powerhooks). (It has been suggested to use some general event notification framework for shutdown handlers, but there might be cases where shutdown handlers must be run in an order following the device hierarchy, which wouldn't be easy with event handlers not tied to drivers.) approved by David Young
1.112	23-Feb-2008	dyoung	Simplify the switch-statement in pci_set_powerstate_int(), and fix a couple of typos so that it will work as I think that it was intended to: delay for 10000us or 200us on transitions to/from D3 or D2 state, respectively. Discussed this with jmcneill@.
1.111	21-Feb-2008	drochner	-fix the delay logics in "set powerstate" for the cases that D2 or D3 are switched to (was harmless because we don't do D2 yet and also don't (hmm - shouldn't) access devices in D3 (which would only make sense if we'd support D3hot) -zero the io/mem/master enable bits before entering D3 (The special handling of PCI_CLASS_DISPLAY devices is questionable here -- we can't care about the console if we are seriously follow the spec, and upstream bridges aren't considered anyway.) -add exact references to the PCI PM spec
1.110	28-Jan-2008	jmcneill	branches: 1.110.2; 1.110.6; PCI Power Management 1.2 says that system software should wait 10ms after transitioning a device to or from D3hot, 200us to or from D2, and no wait for D1 transitions. Update pci_set_powerstate_int to match the spec, rather than naively waiting 1ms after every state transition.
1.109	03-Jan-2008	dyoung	Disable pci_disable_retry() and cardbus_disable_retry(), with a long comment on pci_disable_retry() (cross-referenced at cardbus_disable_retry()) that explains why.
1.108	16-Dec-2007	dyoung	Improve pci, cbb, cardslot, cardbus, and pcmcia to support detachment of this entire device tree: pci0 at mainbus0 elansc0 at pci0 gpio0 at elansc0 cbb0 at pci0 cardslot0 at cbb0 cardbus0 at cardslot0 pcmcia0 at cardslot0 cbb1 at pci0 cardslot1 at cbb1 cardbus1 at cardslot1 rtw0 at cardbus1 pcmcia1 at cardslot1 sip0 at pci0 nsphyter0 at sip0 sip1 at pci0 nsphyter1 at sip1 Whew!
1.107	09-Dec-2007	jmcneill	branches: 1.107.2; Merge jmcneill-pm branch.
1.106	01-Dec-2007	ad	branches: 1.106.2; 1.106.4; Empty vessels make the most noise
1.105	12-Nov-2007	joerg	Merge pci_disable_retry function from jmcneill-pm as it is found in various drivers.
1.104	16-Oct-2007	joerg	branches: 1.104.2; 1.104.4; Merge 1.103.22.4 from jmcneill-pm: Decode capabilities for PCI-PCI bridges as well.
1.103	16-Nov-2006	christos	branches: 1.103.8; 1.103.22; 1.103.24; 1.103.26; __unused removal on arguments; approved by core.
1.102	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.101	02-Oct-2006	jmcneill	Print error on powerhook_establish failure, not success. PR# 34698.
1.100	25-Sep-2006	jmcneill	Create a PCI bus powerhook to handle setting PCI D# power state transitions on suspend/standby/resume.
1.99	17-Jun-2006	christos	branches: 1.99.4; 1.99.6; re-factor the pci powestate api. reviewed by gimpy
1.98	29-Mar-2006	thorpej	branches: 1.98.2; 1.98.4; Replace device_locators() with device_locator(), and use it.
1.97	11-Dec-2005	christos	branches: 1.97.4; 1.97.6; 1.97.8; 1.97.10; 1.97.12; merge ktrace-lwp.
1.96	26-Aug-2005	drochner	use XXXCF_NLOCS constants instead of magic numbers
1.95	25-Aug-2005	drochner	kill a number of autoconf submatch functions which follow the standard scheme: if (<configured> != <wildcard> && <configured> != <real>) then fail else ask device match function This is handled by config_stdsubmatch() now.
1.94	25-Aug-2005	drochner	replace the "locdesc_t" structure carrying the number of locators explicitely by a plain integer array the length in now known to all relevant parties, so this avoids duplication of information, and we can allocate that thing in drivers without hacks
1.93	28-Jun-2005	thorpej	branches: 1.93.2; Use ANSI function decls and static.
1.92	27-Feb-2005	perry	nuke trailing whitespace
1.91	04-Feb-2005	perry	de-__P
1.90	26-Jan-2005	jmcneill	Add support functions for capturing and restoring PCI configuration registers for power management code.
1.89	13-Sep-2004	drochner	branches: 1.89.4; 1.89.6; remove macros which just hide cfdata internals (and are used at one place only)
1.88	30-Aug-2004	drochner	Phase out the use of a string as first "attach args" member to control which bustype should be attached with a specific call to config_found() (from a "mainbus" or a bus bridge). Do it for isa/eisa/mca and pci/agp for now. These buses all attach to an mi interface attribute "isabus", "eisabus" etc., and the autoconf framework now allows to specify an interface attribute on config_found() and config_search(), which limits the search of matching config data to these which attach to that specific attribute. So we basically have to call config_found_ia(..., "foobus", ...) where such a bus is attached. As a consequence, where a "mainbus" or alike also attaches other devices (eg CPUs) which do not attach to a specific attribute yet, we need at least pass an attribute name (different from "foobus") so that the foo bus is not found at these places. This made some minor changes necessary which are not obviously related to the mentioned buses.
1.87	17-Aug-2004	drochner	make PCI devices attachable/detachable (as far as the particular drivers allow), and allow to rescan a bus selectively (ie only the device/ function I'm looking at)
1.86	29-Jul-2004	drochner	make the "generic" PCI bus enumeration code the standard case which gets used if nothing else is defined in MD headers, introduce a "PCI_MACHDEP_ENUMERATE_BUS" CPP definition which can be used by MD headers (just 1 port atm) to plug in special code
1.85	17-May-2004	kochi	Fix typo (func8 -> func7)
1.84	08-May-2004	christos	GC pci_{g,s}et_powerstate into pci_powerstate(). Idea from mycroft and gimpy. Nothing uses them yet.
1.83	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.82	18-Aug-2003	itojun	support new quirk bits on debug output
1.81	15-Aug-2003	itojun	- check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.80	15-Jun-2003	fvdl	branches: 1.80.2; Handle 64bit DMA addresses on PCI for platforms that can (currently only enabled on amd64). Add a dmat64 field to various PCI attach structures, and pass it down where needed. Implement a simple new function called pci_dma64_available(pa) to test if 64bit DMA addresses may be used. This returns 1 iff _PCI_HAVE_DMA64 is defined in <machine/pci_machdep.h>, and there is more than 4G of memory.
1.79	03-May-2003	thorpej	Fix inverted logic in pci_vpd_write(). From Jachym Holecek, PR kern/21440.
1.78	29-Apr-2003	thorpej	Use aprint*().
1.77	25-Mar-2003	thorpej	Add PCI VPD access routines. From psi.cz!freza, PR kern/20889.
1.76	19-Mar-2003	christos	It is not appropriate for pcivar.h to include "locators.h" The two files that need it, should include it themselves.
1.75	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.74	23-Oct-2002	perry	fix lint warnings about things being stringified under cpp -traditional
1.73	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.72	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.71	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.70	27-Sep-2002	thorpej	Introduce a new routine, config_match(), which invokes the cfattach->ca_match function in behalf of the caller. Use it rather than invoking cfattach->ca_match directly.
1.69	27-Sep-2002	thorpej	Rather than referencing the cfdriver directly in the cfdata entries, instead use a string naming the driver. The cfdriver is then looked up in a list which is built at run-time.
1.68	05-Jul-2002	thorpej	Return 0, not NULL, when we mean 0.
1.67	20-Jun-2002	itojun	minor KNF
1.66	18-Jun-2002	tshiozak	add support for the per-device power management capability. int pci_set_powerstate(pci_chipset_tag_t pc, pcitag_t tag, int newstate) set power state of the device to newstate. int pci_get_powerstate(pci_chipset_tag_t pc, pcitag_t tag) get current power state of the device. In the future, these functions will be used for ACPI support.
1.65	18-May-2002	sommerfeld	branches: 1.65.2; Add "pa_rawintrpin" containing unswizzled interrupt pin to pci_attach_args.
1.64	18-May-2002	sommerfeld	In pci_probe_device(), correct test for whether the device is behind a pci-pci bridge (and thus needs its interrupts "swizzled"). Formerly, pci_probe_device() assumed pci busses other than bus zero were bridged; however, much recent x86 hardware supports multiple primary pci busses differently -- there is a system-wide bus numbering scheme. Now, we instead look at the (newly introduced) sc_bridgetag value in the parent bus to figure out if there's a ppb or equivalent in the way. This fixed at least one case where the i386 MP branch gets interrupt mapping wrong.
1.63	16-May-2002	thorpej	branches: 1.63.2; Make sure pa.pa_bus is initialized. (Thanks, uwe!)
1.62	16-May-2002	thorpej	* Add "pcitag_t pba_bridgetag" to pci_attach_args. This is set to NULL for root PCI busses. For busses behind a bridge, it points to a persistent copy of the bridge's pcitag_t. This can be very useful for machine-dependent PCI bus enumeration code. Implement a machine-dependent pci_enumerate_bus() for sparc64 which uses OFW device nodes to enumerate the bus. When a PCI bus that is behind a bridge is attached, pci_attach_hook() allocates a new PCI chipset tag for the new bus and sets it's "curnode" to the OFW node of the bridge. This is used as a starting point when enumerating that bus. Root busses get the OFW node of the host bridge (psycho). * Garbage-collect "ofpci" and "ofppb" from the sparc64 port.
1.61	15-May-2002	thorpej	Let machine-dependent code specify how to enumerate the bus. Currently, everyone uses pci_enumerate_bus_generic().
1.60	15-May-2002	thorpej	Put back __PCI_BUS_DEVORDER for now.
1.59	15-May-2002	thorpej	Split the code that enumerates the PCI bus and that actually probes for a device into two functions: * pci_probe_device() actually probes/attaches the device specified by the provide pcitag_t. * pci_enumerate_bus() enumerates the bus, and calls pci_probe_device() for each device on the bus. A pci_enumerate_bus_generic() is provided which implements the old method of doing this: If something found at dev0/func0, determine number of functions and probe each one. Machine-dependent code will be able to specify the bus enumeration routine in the future.
1.58	13-Nov-2001	lukem	add RCSID
1.57	17-Oct-2001	thorpej	Fix the semantics of __PCI_DEV_FUNCORDER so that they're actually useful: If __PCI_DEV_FUNCORDER is defined, don't do the song-and-dance to check if a device is multi-function; machdep code is going to tell us exactly which functions to probe. Note this required changing how pci_func_devorder() works in the sparc64 PCI machdep code; now the "curnode" is assumed to point to the bus, rather than some function (typically 0) on the device, just as pci_bus_devorder() makes that assumption. All this should allow the PCI code to actually locate the second HME device on a Sun Netra t1, which is at 3,1 -- previously, the PCI code would have missed it because there is no device at 3,0. (Sun deserves a brick to the head for this one -- this seems clearly out of line with the PCI spec.)
1.56	13-Sep-2001	thorpej	Add an ioctl interface to the PCI bus. Add ioctls to read/write PCI configuration space registers, and to fetch bus info.
1.55	10-Sep-2001	fvdl	Add a pci_find_device function, to find a PCI device using a match function on all (probed) PCI buses: int pci_find_device(struct pci_attach_args pa, int (match)(struct pci_attach_args *)); The pci_attach_args structure pointed to by pa is filled in if the device is found, and 1 is returned. Otherwise 0 is returned. This function is, unfortunately, needed by the i810 agp code. It's also of use for LKMs. Also frob pci_probe_bus to take 2 extra args when used by pci_find_device.
1.54	11-Jul-2001	mrg	branches: 1.54.2; don't try to print pcitag_t values; these are opaque values that should not be interpreted by MI code.
1.53	22-May-2001	thorpej	branches: 1.53.2; If the cache line size in the BHLC register is not configured, then clear the MRL/MRM/MWI command-ok flags when attaching a device.
1.52	04-May-2001	bouyer	Add pci bus number to pci_attach_args, as suggested by Jason on tech-kern (needed for pciide)
1.51	02-Mar-2001	mrg	branches: 1.51.2; allow pci_bus_maxdevs() to be replaced with pci_bus_devorder() and pci_dev_funcorder() that have the following signatures: int pci_bus_devorder(pci_chipset_tag_t pc, int bus, char list[32]); int pci_dev_funcorder(pci_chipset_tag_t pc, int bus, int device, char list[8]); they control the order of PCI bus probe at the device and function level, by filling in a value from 0 to 31 for pci_bus_devorder() or 0 to 7 for pci_dev_funcorder, with a value of -1 to signify no more entries. when device properties arrive, these will be replaced with some facility based on properties (design/implementation unknown currently.)
1.50	12-Feb-2001	mrg	backout the parts of the previous change (7 months ago) that were not part of the sys/vm removal, but some (fortunately disabled) work-in-progress.
1.49	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.48	10-May-2000	kleink	branches: 1.48.4; Add a comment describing the previous change.
1.47	08-May-2000	kleink	Make pci_get_capability() work correctly on both header type 0 and type 2 devices; both have different Capability List Pointer registers.
1.46	30-Mar-2000	augustss	Remove register declarations.
1.45	22-Mar-2000	cgd	if the PCI_CONFIG_DUMP option is defined, set the patchable pci_config_dump variable to 1. otherwise, it gets set to 0. If that's set, when configuring PCI devices spew the very verbose configuration space header dump that was previously "#if 0"'d. If you use this option, or patch the variable to 1, it's strongly encouraged that you have used the MSGBUFSIZE option to enlarge the kernel message buffer.
1.44	04-Nov-1999	thorpej	Make sure devices are told they can use rd/line, rd/mult, and wr/inval.
1.43	04-Nov-1999	thorpej	Print out if the read {line,multiple} and write/invalidate commands are okay to use.
1.42	06-May-1999	thorpej	branches: 1.42.2; 1.42.4; 1.42.8; Back out previous. Thanks to cgd for pointing out another way to do this.
1.41	06-May-1999	thorpej	Add a `bus' member to the pci_attach_args. This is not normally used, but some child drivers might need to know this information.
1.40	07-Nov-1998	drochner	branches: 1.40.6; 1.40.8; add support for "extended capabilities" (new in PCI spec 2.2)
1.39	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.38	09-Jun-1998	thorpej	Remove the PCI-ISA bridge callback mechanism; it's no longer needed.
1.37	31-May-1998	cgd	use the new PCI 'quirks' mechanism to help determine the number of functions on a given device. Also, clean up the #if 0'd major-debugging-spew code so that it's all one piece, so that it's a bit prettier, and so that it prints out quirk information.
1.36	18-May-1998	cgd	add an #if 0'd chunk which will pci_conf_print() every device (spews lots of data, e.g. ~18k on a PCI system with few add-in devices; use with MSGBUFSIZE=...). Useful to have here so that people who want as much data about the PCI configuration in a machine can get it without having to craft their own code. Also, clean up a few of the other #if 0'd printfs.
1.35	05-May-1998	mjacob	more info (in heavy debugging cases) to print
1.34	17-Apr-1998	drochner	Keep all relevant state information in a "pci_softc" so that it is accessible at runtime (for LKMs in particular). Remove BROKEN_INDIRECT_CONFIG.
1.33	28-Mar-1998	cgd	check vendor against PCI_VENDOR_INVALID instead of 0xffff. Suggested by Soren S. Jorvang.
1.32	20-Mar-1998	cgd	check only the Vendor ID (rather than both the Vendor ID and the Product ID) when determining if the Vendor ID is invalid. The spec says that Vendor ID of 0xffff is invalid, so, it doesn't _matter_ what the product ID is in that case. Treat Vendor ID 0 as invalid because we always have.
1.31	31-Jan-1998	thorpej	branches: 1.31.2; When a PCI bus is attached, print out which bus spaces were enabled by the parent.
1.30	12-Jan-1998	thorpej	branches: 1.30.2; Update for config changes.
1.29	30-Aug-1997	mycroft	branches: 1.29.2; Pass down bus_dma_tag_t's as appropriate (per Jason's bus_dma code).
1.28	30-Aug-1997	mycroft	Remove pci_*_find() from here.
1.27	10-Apr-1997	cgd	pass memory- and i/o-enabled flags down via the PCI bus and device attach arguments, so that a device can tell if its memory and I/O spaces are enabled. The flags are cleared, depending on the contents of devices CSR registers, in the machine-independent PCI bus code.
1.26	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.25	23-Nov-1996	cgd	Provide a routine so that ISA/EISA bridges can set up a callback so that their child busses can be attached after the PCI bus autoconfiguration for their parent bus is done. This works because: (1) there can be at most one ISA/EISA bridge per PCI bus, and (2) any ISA/EISA bridges must be attached to primary PCI busses (i.e. bus zero). That boils down to: there can only be one of these outstanding at a time, it is cleared when configuring PCI bus 0 before any subdevices have been found, and it is run after all subdevices of PCI bus 0 have been found. This (or something like it) is needed because there are some (legacy) PCI devices which can show up as ISA/EISA devices as well (the prime example of which are VGA controllers). If you attach ISA from a PCI-ISA/EISA bridge, and the bridge is seen before the video board is, the board can show up as an ISA device, and that can (bogusly) complicate the PCI device's attach code, or make the PCI device not be properly attached at all. This could be done with machine-dependent code, but as more ports add support for PCI (and PCI-ISA/EISA bridges) more will need it. The i386 port could (perhaps should) be converted to use it as well.
1.24	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.23	13-Oct-1996	christos	backout kprintf changes
1.22	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.21	27-Aug-1996	cgd	change cfprint_t type definition to take a const char , rather than a char , because that's what was really intended, and because if the print function modifies the string, various things could become unhappy (so the string should _not_ be modified).
1.20	26-Jul-1996	mycroft	Add PCI_MAPREG_{IO,MEM}_SIZE(), and use them.
1.19	03-May-1996	christos	branches: 1.19.4; remove unused variables
1.18	27-Mar-1996	cgd	modify these to provide a new, better-specified PCI interface (soon to be documented on mailing lists; eventually in section 9 manual pages), most importantly: (1) support interrupt pin swizzling on non-i386 systems with PCI-PCI bridges (per PPB spec; done, but meaningless, on i386). (2) provide pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register. (3) provide pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (4) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.17	27-Mar-1996	cgd	add definitions for the BIST/Header Type/Latency Timer/Cache Line Size configuration space register, and use it to determine whether or not a given PCI device uses multiple functions.
1.16	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.15	14-Mar-1996	cgd	These devices don't actually need softc's that are supersets of 'struct device.' No point in adding that complexity + space if they're not needed, so axe them.
1.14	14-Mar-1996	cgd	(1) provide #defines for cf_loc[] entries for devices that attach to pcibus and pci. (2) remove the #ifdef i386 from pci.c, and provide a machine-dependent hook (pci_md_attach_hook()) to do any machine-dependent attachment gunk, e.g. on the i386 printing out the configuration mode (if bus 0) (3) don't pass max device number for a given bus in, use PCI_MAX_DEVICE_NUMBER, which can be defined on a per-machine basis. (defaults to 32. on i386, it's 32 if pci conf mode == 1, 16 if 2.)
1.13	08-Mar-1996	cgd	remember the bus_chipset_tag_t given by our parent, and pass it on to children.
1.12	04-Mar-1996	cgd	the i386 port no longer attaches isa, eisa, and pci at root.
1.11	02-Mar-1996	cgd	fix uninitialized variable; it shouldn't have been a variable at all. (when probing to see if a device is there, always probe function 0.)
1.10	28-Feb-1996	cgd	make PCI bus match/attach and sub-device attachment machine-independent.
1.9	17-Jun-1995	cgd	oops; forgot this in last commit
1.8	23-May-1995	cgd	split single-subdevice lookup & attachment into a subroutine (pci_attach_subdev()). remove pciattach() function and the pcicd cfdriver struct, the former because thre are a lot of attachment actions which really are machine-dependent (perhaps even "most"), and the latter because now that both pcimatch() and pciattach() are machine-dependent it's bad style to declare them here and it gains nothing.
1.7	27-Jan-1995	cgd	include files from the correct places.
1.6	04-Nov-1994	mycroft	Make a wrapper match function to check the bus and device numbers, rather than insisting that every driver do it.
1.5	03-Nov-1994	mycroft	Rename pciprobe() to pcimatch(), and move it to pci_machdep.c.
1.4	03-Nov-1994	mycroft	Always use direct configuration.
1.3	27-Oct-1994	cgd	new RCS ID format.
1.2	10-Aug-1994	mycroft	Update some comments.
1.1	09-Aug-1994	mycroft	branches: 1.1.2; Add PCI autoconfiguration support.
1.1.2.1	10-Aug-1994	mycroft	update from trunk
1.19.4.1	10-Dec-1996	mycroft	From trunk: Deal with devices that do not allow allow of the I/O address bits to be set.
1.29.2.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.29.2.1	30-Aug-1997	thorpej	file pci.c was added on branch marc-pcmcia on 1997-09-01 20:29:42 +0000
1.30.2.2	12-Jan-1998	thorpej	Update for config changes.
1.30.2.1	12-Jan-1998	thorpej	file pci.c was added on branch thorpej-bus-dma on 1998-01-12 09:40:12 +0000
1.31.2.2	31-Jan-1998	thorpej	When a PCI bus is attached, print out which bus spaces were enabled by the parent.
1.31.2.1	31-Jan-1998	thorpej	file pci.c was added on branch bouyer-scsipi on 1998-01-31 00:37:40 +0000
1.40.8.1	21-Jun-1999	thorpej	Sync w/ -current.
1.40.6.1	27-Jun-2000	he	Pull up revision 1.45 (via patch, requested by cgd): If the 'PCI_CONFIG_DUMP' option is defined, print a verbose dump of all PCI devices' configuration space.
1.42.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.42.4.1	15-Nov-1999	fvdl	Sync with -current
1.42.2.2	12-Mar-2001	bouyer	Sync with HEAD.
1.42.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.48.4.1	15-May-2001	he	Pull up revision 1.52 (requested by bouyer): Add a pa_bus member to pci_attach_args, so that the PCI bus number can be used in pci device drivers.
1.51.2.10	03-Jan-2003	thorpej	Sync with HEAD.
1.51.2.9	11-Nov-2002	nathanw	Catch up to -current
1.51.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.51.2.7	01-Aug-2002	nathanw	Catch up to -current.
1.51.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.51.2.5	14-Nov-2001	nathanw	Catch up to -current.
1.51.2.4	22-Oct-2001	nathanw	Catch up to -current.
1.51.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.51.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.51.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.53.2.6	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.53.2.5	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.53.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.53.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.53.2.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.53.2.1	03-Aug-2001	lukem	update to -current
1.54.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.63.2.3	15-Jul-2002	gehenna	catch up with -current.
1.63.2.2	30-May-2002	gehenna	Catch up with -current.
1.63.2.1	16-May-2002	gehenna	file pci.c was added on branch gehenna-devsw on 2002-05-30 14:46:11 +0000
1.65.2.1	15-Aug-2003	tron	Pull up revision 1.81 (requested by itojun in ticket #1412): - check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.80.2.8	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.80.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.80.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.80.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.80.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.80.2.3	03-Sep-2004	skrll	Sync with HEAD
1.80.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.80.2.1	03-Aug-2004	skrll	Sync with HEAD
1.89.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.89.6.1	12-Feb-2005	yamt	sync with head.
1.89.4.1	29-Apr-2005	kent	sync with -current
1.93.2.10	24-Mar-2008	yamt	sync with head.
1.93.2.9	17-Mar-2008	yamt	sync with head.
1.93.2.8	27-Feb-2008	yamt	sync with head.
1.93.2.7	04-Feb-2008	yamt	sync with head.
1.93.2.6	21-Jan-2008	yamt	sync with head
1.93.2.5	07-Dec-2007	yamt	sync with head
1.93.2.4	15-Nov-2007	yamt	sync with head.
1.93.2.3	27-Oct-2007	yamt	sync with head.
1.93.2.2	30-Dec-2006	yamt	sync with head.
1.93.2.1	21-Jun-2006	yamt	sync with head.
1.97.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.97.10.1	19-Apr-2006	elad	sync with head.
1.97.8.2	26-Jun-2006	yamt	sync with head.
1.97.8.1	01-Apr-2006	yamt	sync with head.
1.97.6.1	22-Apr-2006	simonb	Sync with head.
1.97.4.1	09-Sep-2006	rpaulo	sync with head
1.98.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.98.2.1	19-Jun-2006	chap	Sync with head.
1.99.6.2	10-Dec-2006	yamt	sync with head.
1.99.6.1	22-Oct-2006	yamt	sync with head
1.99.4.1	18-Nov-2006	ad	Sync with head.
1.103.26.1	18-Oct-2007	yamt	sync with head.
1.103.24.3	23-Mar-2008	matt	sync with HEAD
1.103.24.2	09-Jan-2008	matt	sync with HEAD
1.103.24.1	06-Nov-2007	matt	sync with HEAD
1.103.22.13	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.103.22.12	04-Dec-2007	jmcneill	Don't neglect to install a power handler in the event of no spaces being available.
1.103.22.11	03-Dec-2007	joerg	Sync with HEAD.
1.103.22.10	01-Dec-2007	jmcneill	It seems that a hack to force PCI bridges to enable i/o and mem space snuck in with the previous commit. Revert that piece.
1.103.22.9	01-Dec-2007	jmcneill	Don't actually power off vga in the PCI pnp bus handler, as it may be the console device and it's nice to be able to figure out why we have crashed if we fail to sleep.
1.103.22.8	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.103.22.7	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.103.22.6	11-Sep-2007	jmcneill	Missing break after PNP_REQUEST_SET_STATE.
1.103.22.5	23-Aug-2007	joerg	Introduce pci_net_generic_power, which should be enough for most network drivers and be a good foundation for C&P for the rest. For iwi(4), don't reset the PCI retry register again, pci_conf_restore should take care of that already. For bge(4), add a NetBSD style if_stop.
1.103.22.4	21-Aug-2007	joerg	Decode capabilities for PCI-PCI bridges as well.
1.103.22.3	06-Aug-2007	jmcneill	Fix compilation with DIAGNOSTIC.
1.103.22.2	05-Aug-2007	jmcneill	In pci_conf_restore, don't bother updating a register if the current value matches the new value.
1.103.22.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.103.8.1	23-Oct-2007	ad	Sync with head.
1.104.4.4	18-Feb-2008	mjf	Sync with HEAD.
1.104.4.3	27-Dec-2007	mjf	Sync with HEAD.
1.104.4.2	08-Dec-2007	mjf	Sync with HEAD.
1.104.4.1	19-Nov-2007	mjf	Sync with HEAD.
1.104.2.1	13-Nov-2007	bouyer	Sync with HEAD
1.106.4.1	11-Dec-2007	yamt	sync with head.
1.106.2.1	26-Dec-2007	ad	Sync with head.
1.107.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.107.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.110.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.110.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.110.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.110.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.110.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.110.2.1	24-Mar-2008	keiichi	sync with head.
1.116.4.6	11-Aug-2010	yamt	sync with head.
1.116.4.5	11-Mar-2010	yamt	sync with head
1.116.4.4	16-Sep-2009	yamt	sync with head
1.116.4.3	18-Jul-2009	yamt	sync with head.
1.116.4.2	04-May-2009	yamt	sync with head.
1.116.4.1	16-May-2008	yamt	sync with head.
1.116.2.2	17-Jun-2008	yamt	sync with head.
1.116.2.1	18-May-2008	yamt	sync with head.
1.117.4.1	18-Jun-2008	simonb	Sync with head.
1.117.2.2	10-Oct-2008	skrll	Sync with HEAD.
1.117.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.118.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.118.2.1	19-Oct-2008	haad	Sync with HEAD.
1.119.4.2	16-Feb-2011	bouyer	Pull up following revision(s) (requested by dyoung in ticket #1540): sys/dev/pci/pci.c: revision 1.131 During shutdown, restore the each PCI device's bus-mastering enable to the state that we found it in before attaching a driver to it.
1.119.4.1	20-Nov-2008	snj	branches: 1.119.4.1.2; 1.119.4.1.4; 1.119.4.1.6; Pull up following revision(s) (requested by bouyer in ticket #66): sys/dev/pci/pci.c: revision 1.120 Do not depend on physmem for pci_dma64_available(). On some systems (e.g. Xen, or any system with non-contigous physical memory), physmem may not represent the hightest physical address available. Discussed on tech-kern@ some weeks ago.
1.119.4.1.6.1	16-Feb-2011	bouyer	Pull up following revision(s) (requested by dyoung in ticket #1540): sys/dev/pci/pci.c: revision 1.131 During shutdown, restore the each PCI device's bus-mastering enable to the state that we found it in before attaching a driver to it.
1.119.4.1.4.1	24-Dec-2011	matt	Add support for __PCI_DEV_FUNCORDER
1.119.4.1.2.1	16-Feb-2011	bouyer	Pull up following revision(s) (requested by dyoung in ticket #1540): sys/dev/pci/pci.c: revision 1.131 During shutdown, restore the each PCI device's bus-mastering enable to the state that we found it in before attaching a driver to it.
1.119.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.119.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.121.2.2	23-Jul-2009	jym	Sync with HEAD.
1.121.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.125.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.125.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.127.2.5	31-May-2011	rmind	sync with head
1.127.2.4	21-Apr-2011	rmind	sync with head
1.127.2.3	05-Mar-2011	rmind	sync with head
1.127.2.2	03-Jul-2010	rmind	sync with head
1.127.2.1	30-May-2010	rmind	sync with head
1.130.4.3	05-Mar-2011	bouyer	Sync with HEAD
1.130.4.2	17-Feb-2011	bouyer	Sync with HEAD
1.130.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.130.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.142.12.4	03-Dec-2017	jdolecek	update from HEAD
1.142.12.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.142.12.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.142.12.1	12-Sep-2012	tls	Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for physio (I/O to raw devices); needs more doing to get it going with the filesystems, but it shouldn't damage data. All work's been done on amd64 so far. Not hard to add support to other ports. If others want to pitch in, one very helpful thing would be to sort out when and how IDE disks can do 128K or larger transfers, and adjust the various PCI IDE (or at least ahcisata) drivers and wd.c accordingly -- it would make testing much easier. Another very helpful thing would be to implement a smart minphys() for RAIDframe along the lines detailed in the MAXPHYS-NOTES file.
1.142.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.142.2.1	30-Oct-2012	yamt	sync with head
1.143.2.1	18-May-2014	rmind	sync with head
1.145.2.5	28-Aug-2017	skrll	Sync with HEAD
1.145.2.4	19-Mar-2016	skrll	Sync with HEAD
1.145.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.145.2.2	22-Sep-2015	skrll	Sync with HEAD
1.145.2.1	06-Jun-2015	skrll	Sync with HEAD
1.151.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.151.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.152.14.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.152.14.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.152.14.1	10-Jun-2019	christos	Sync with HEAD
1.152.12.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.152.6.2	31-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1178): sys/dev/pci/pci.c: revision 1.154 Clearing PCI_PMCSR_PME_STS(W1C) bit is required to stop asserting PME#. This change would prevent unexpected rebooting from shutdown -p or unexpected resuming from suspend.
1.152.6.1	07-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1128): sys/dev/pci/pcivar.h: revision 1.113 sys/dev/pci/pci.c: revision 1.153 Save control registers in PCI-X, PCIe, MSI and MSI-X capability area when suspend and restore them when resume. For PCIe capabilities register, it's required to check the existence of each register to not to write the next area. This change fixes a stability of suspend/resume.
1.154.4.3	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1384: sys/dev/pci/pcireg.h 1.152-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.222, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158, 1.163 sys/dev/pci/ppb.c 1.74 - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.154.4.2	06-Nov-2019	martin	Pull up following revision(s) (requested by mrg in ticket #401): sys/dev/pci/pci.c: revision 1.156 PCIe downstream ports only have a single child device, so limit probing to dev 0.
1.154.4.1	15-Oct-2019	martin	Pull up following revision(s) (requested by jmcneill in ticket #331): sys/dev/pci/pci.c: revision 1.155 The PCI Local Bus Specification says that we should probe devices by reading PCI_ID_REG. pci_enumerate_bus was incorrectly reading PCI_BHLC_REG first, which surprisingly has worked for the past 16 years. Unfortunately, this undefined behavior does the wrong thing on Amazon's Annapurna Labs PCIe host controller.
1.156.2.1	29-Feb-2020	ad	Sync with head.
1.157.6.1	03-Apr-2021	thorpej	Sync with HEAD.
1.158.2.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.158.2.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.159.4.1	31-May-2021	cjep	sync with head
1.159.2.1	13-May-2021	thorpej	Sync with HEAD.
1.160.4.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.165.4.1	04-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #928): sys/external/bsd/drm2/dist/drm/drm_gem.c: revision 1.25 sys/external/bsd/drm2/dist/drm/radeon/radeon_ci_dpm.c: revision 1.7 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/priv.h: revision 1.4 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_acpi.c: revision 1.4 sys/external/bsd/drm2/dist/drm/i915/display/intel_opregion.h: revision 1.6 sys/external/bsd/drm2/dist/drm/i915/i915_drv.h: revision 1.49 sys/external/bsd/drm2/include/linux/mxm-wmi.h: revision 1.1 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/pci/nouveau_nvkm_subdev_pci_pcie.c: revision 1.4 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_base.c: revision 1.13 sys/external/bsd/common/include/linux/bitops.h: revision 1.17 sys/external/bsd/drm2/nouveau/files.nouveau: revision 1.40 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.30 sys/external/bsd/drm2/dist/drm/radeon/radeon_drv.h: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/pci/nouveau_nvkm_subdev_pci_pcie.c: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/mxm/nouveau_nvkm_subdev_mxm_base.c: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_gart.c: revision 1.12 sys/external/bsd/drm2/dist/drm/radeon/radeon_rv770.c: revision 1.3 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/disp/nouveau_nvkm_engine_disp_sorgm200.c: revision 1.3 sys/external/bsd/common/include/linux/printk.h: revision 1.14 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/instmem/nouveau_nvkm_subdev_instmem_gk20a.c: revision 1.10 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_vi.c: revision 1.4 sys/external/bsd/drm2/include/linux/acpi.h: revision 1.11 sys/external/bsd/drm2/drm/drm_cdevsw.c: revision 1.31 sys/external/bsd/drm2/dist/drm/radeon/radeon_si.c: revision 1.6 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_acpi.c: revision 1.5 sys/external/bsd/drm2/dist/drm/i915/display/intel_acpi.h: revision 1.5 sys/external/bsd/drm2/include/acpi/video.h: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_evergreen.c: revision 1.6 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_acpi.h: revision 1.4 sys/arch/sparc64/include/pci_machdep.h: revision 1.31 sys/arch/sparc64/dev/pci_machdep.c: revision 1.83 sys/external/bsd/drm2/include/linux/kref.h: revision 1.14 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_pci.c: revision 1.12 sys/external/bsd/drm2/linux/linux_dma_buf.c: revision 1.17 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/bios/nouveau_nvkm_subdev_bios_shadowacpi.c: revision 1.4 sys/external/bsd/drm2/drm/drm_module.c: revision 1.32 sys/external/bsd/drm2/dist/drm/i915/i915_gem.h: revision 1.8 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_dmabuf.c: revision 1.7 sys/external/bsd/drm2/include/linux/smp.h: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_si.c: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_device.c: revision 1.20 sys/arch/x86/x86/bus_dma.c: revision 1.91 sys/external/bsd/drm2/radeon/files.radeon: revision 1.40 sys/external/bsd/drm2/include/acpi/acpi_bus.h: revision 1.1 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_drv.h: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_device.c: revision 1.21 sys/external/bsd/common/include/asm/barrier.h: revision 1.20 sys/external/bsd/drm2/include/linux/nbsd-namespace-acpi.h: revision 1.2 sys/external/bsd/common/include/asm/barrier.h: revision 1.21 sys/modules/drmkms/drmkms_pci.h: revision 1.1 sys/external/bsd/drm2/dist/drm/drm_dp_helper.c: revision 1.17 sys/external/bsd/drm2/radeon/radeon_pci.c: revision 1.23 sys/external/bsd/drm2/linux/linux_xa.c: revision 1.4 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.23 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.24 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.25 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.26 sys/dev/pci/pcivar.h: revision 1.120 sys/arch/xen/include/pci_machdep.h: revision 1.24 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.26 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.27 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.27 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.28 sys/external/bsd/drm2/dist/drm/radeon/radeon_cik.c: revision 1.8 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.28 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.29 sys/external/bsd/drm2/include/linux/pci.h: revision 1.57 sys/external/bsd/drm2/include/linux/pci.h: revision 1.58 sys/external/bsd/drm2/dist/drm/radeon/radeon_acpi.c: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_display.c: revision 1.6 sys/external/bsd/drm2/dist/drm/amd/powerplay/hwmgr/amdgpu_hwmgr.c: revision 1.3 sys/external/bsd/drm2/dist/drm/amd/display/dc/core/amdgpu_dc_stream.c: revision 1.3 share/man/man9/bus_dma.9: revision 1.69 sys/external/bsd/drm2/drm/drm_gem_cma_helper.c: revision 1.15 sys/external/bsd/drm2/dist/drm/radeon/radeon_acpi.c: revision 1.6 sys/external/bsd/drm2/dist/drm/radeon/radeon.h: revision 1.12 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_cik.c: revision 1.7 sys/dev/acpi/acpi_mcfg.c: revision 1.29 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_acpi.c: revision 1.6 sys/external/bsd/drm2/dist/drm/radeon/radeon_r600.c: revision 1.7 sys/external/bsd/drm2/dist/drm/radeon/radeon_bios.c: revision 1.13 sys/modules/amdgpu/Makefile: revision 1.9 sys/external/bsd/drm2/dist/drm/radeon/radeon_bios.c: revision 1.14 sys/external/bsd/common/linux/linux_tasklet.c: revision 1.12 sys/external/bsd/drm2/dist/drm/nouveau/include/nvkm/core/device.h: revision 1.10 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.23 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu.h: revision 1.9 sys/external/bsd/drm2/include/linux/interval_tree.h: revision 1.14 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.26 sys/external/bsd/drm2/dist/drm/amd/powerplay/hwmgr/amdgpu_smu7_hwmgr.c: revision 1.5 sys/dev/pci/pci.c: revision 1.168 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.27 sys/external/bsd/drm2/dist/drm/radeon/radeon_si_dpm.c: revision 1.9 sys/external/bsd/drm2/pci/files.drmkms_pci: revision 1.18 sys/external/bsd/drm2/linux/linux_sync_file.c: revision 1.3 sys/external/bsd/drm2/amdgpu/files.amdgpu: revision 1.31 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_tegra.c: revision 1.4 sys/external/bsd/drm2/dist/drm/drm_gem.c: revision 1.24 sys/arch/xen/xen/xpci_xenbus.c: revision 1.29 drm: Eliminate __HAVE_ATOMIC_AS_MEMBAR conditionals. Discussed on tech-kern: https://mail-index.netbsd.org/tech-kern/2023/02/23/msg028729.html linux asm/barrier.h: Fix !MULTIPROCESSOR build. remove "nouveau" from a comment. noted by jmcneill. drm: KASSERT(A && B) -> KASSERT(A); KASSERT(B) comment a function that has a clear overbounds read but it isn't used. found by GCC 12. nix the NetBSD specific GEM_BUG_ON(). avoids GCC 12 warnings, and matches upstream closer. avoid uninitialised variable usage in drm_gem_cma_create_internal(). in the case nothing has returned 'error', 'nsegs' and the dma info are (potentially) uninitialised, so consider this an error. found by GCC 12. avoid a GCC 12 warning. there's a 1-element long array and a loop conditional that tries to see if indexes for it are not identical. as these indexes will always both be 0, the only valid index, the condition is always false. GCC 12 triggers a strange warning on this code that can never run (see below), so simply assert the array size is 1 and comment the rest. amdgpu_dc_stream.c:470:55: error: array subscript [0, 0] is outside array bounds of 'struct dc_writeback_info[1]' [-Werror=array-bounds] 470 \| stream->writeback_info[j] = stream->writeback_info[i]; convert a KASSERT() into an if () panic() sequence to appease GCC 12. OK riastradh@. drm: Fix conditionals around drmkms_pci and agp. Kernel should build now with all pci drm drivers stripped out but DRM_LEGACY still enabled. (Might not be very useful, but it'll build. Maybe we should also have DRM_LEGACY_PCI so those drivers can be modloaded later.) drmkms: Fix module build. avoid an unlikely array bounds issue picked up by GCC 12. nvkm_pcie_speed() can return -1, which is then used as an array index, so make this default return PCIe 1.0 speeds. drm: enable almost all PCIe functionality linux_pci.c revisions 1.24 and 1.25 implemented most of the remaining missing PCIe backends, but only enabled them for some amdgpu portions. this enables all code marked with "XXX amdgpu pcie", "XXX radeon pcie", and "XXX pcie speed". for most of it, simply removing #ifndefs __NetBSD__ to enable compliation was required, once the new "bus->max_bus_speed" member was added to struct pci_bus. add an "always fails" backend for pci_enable_atomic_ops_to_root() which seems to only be necessary for virtual GPU functionality (and could be implemented if needed.) tested on radeon 5450, 7750, R7 240 [radeon], and RX 550 [amdgpu], and nvidia 750 and 1030 [nouveau]. this still does not quite work on nvidia cards. there are two problems that remain: - the call to set the link speed is skipped because the speed is set to the default value of "-1". nvkm_pcie_set_link() will actually determine the right value for this and for some cards, calling this function if the current speed is -1 helps set the link speed. it may be that on linux other paths we don't have enabled properly would set this (there's one via debugfs, and a jetson specific one, though perhaps setting either AC or DC speed values as boot options (after hooking up these for netbsd) would currently work. - worse, cards newer than kepler - geforce 900, 1000, and newer, are all lacking the backing support to set pcie link speed. the GT 1030 card i have been testing with remains at pcie 1.0. radeon: fix and enable ACPI methods for getting ROM BIOS The hacky way of getting the BIOS mapped only works on x86. ACPI should be preferred if available. Makes BIOS reading though VFCT work on aarch64 with EDK2. (But only if EDK2 has POSTed the GPU.) XXX amdgpu should get the same treatment. drm: put_cpu() should enable preemption, not disable it again drm(4): make pr_debug equivalent to aprint_debug significantly reduces the default spam from amdgpu(4). drm: Set CONFIG_ACPI in linux/acpi.h and make it build. Leave a little ACPI-related functionality disabled for now, like getting EDID out of ACPI -- needs a bit more work to make this work, and I don't have hardware to work on that. Should help with failures of the forms: - unable to locate a BIOS ROM - bios: unable to locate usable image on various machines. radeon_acpi.c: ifdef out unused function on NetBSD. Should fix syzkaller build. drm(4): Fix st_rdev in stat. dminor->index already has the 64*type adjustment, as allocated in drm_minor_alloc. PR kern/58180 linux_sync_file: Fix missing init/fini steps. Noted by rjs@. PR kern/58210 ttm: Sync ttm_bo_uvm_fault_idle better with Linux. PR xsrc/58133 ttm: Undo mistake in previous. PR xsrc/58133 linux: Add a few more cases to pci_get_class. Should fix crash on boot with amdgpu now that the ACPI business is enabled. i915: Fix dmabuf mmap object. drm: Fix missing bounds checks in dma buf mmap. drm_gem.c: Fix sense of assertion. This is the opposite of WARN_ON. Noted by rjs@. drm_gem.c: Enable drm_gem_fence_array_add now that we emulate xa. linux_xa: Delete and replace collision in xa_store as intended. Don't free the colliding node that's still in the tree. Noted by rjs@. i915_gem_mman.c: Apply mmap types via pmap flags. This way, userland gets buffers mapped write-combining or uncached as needed. PR xsrc/58307 x86: Teach bus_dmamem_map about BUS_DMA_PREFETCHABLE. PR port-amd64/58308 bus_dma(9): Document BUS_DMA_PREFETCHABLE. Like BUS_DMA_NOCACHE. Doesn't absolve you of the need for bus_dmamap_sync, but if you later pass the vaddr to bus_dmamap_load, the DMA map might notice the mapping is write-combining and use this to make bus_dmamap_sync cheaper. PR kern/58309 nouveau_nvkm_subdev_instmem_gk20a.c: Use BUS_DMA_PREFETCHABLE. Matches Linux's pgprot_writecombine. Unclear where the appropriate bus_dmamap_sync happens, or is supposed to happen -- not using it would be wrong, but asking for a prefetchable mapping may paper over symptoms, at least! ttm: Sync more with Linux. Add the original copyright and attribution since this is now, intentionally, a modified copy of the original and not just roughly the same algorithm. ttm: Respect PGO_ALLPAGES. Not sure this is useful but it reduces XXX's and makes this match udv_fault better so it's easier to understand. ttm: Sync cacheability flag logic with Linux. ttm: Add XXX about readahead fault failures. pci: Pass cookie through pci_find_device, pci_enumerate_bus, take 2. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This will allow pci_find_device callers to pass a cookie through to the match function so they can keep state or pass in extra parameters like b/d/f numbers, which will allow us to nix some horrible kludges in the Linux PCI API emulation for drm (and, perhaps, Intel wifi). This change drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway. Take 2: Make sure to handle NULL match function. linux_pci: Nix pci enumeration kludges. Now that we can pass a cookie through, this stuff will be a little less fragile. i915: Omit needless i915_gem_object_pin/unpin_pages cycle in fault. vm_fault_cpu and vm_fault_gtt, called by i915_gem_fault, already do the pinning and unpinning internally, so there is no need for i915_gem_fault to do it. No functional change intended, except that the transient pin count will be one lower than before during the fault routine (but it will still be positive). i915: Match Linux fault routine return code actions. Omit needless EINTR interception -- this is now handled by i915_error_to_vmf_fault. Earlier revert was over a false alarm -- bisection shows the new warnings arose from linux_pci.c 1.29 here: https://mail-index.netbsd.org/source-changes/2024/06/23/msg151929.html linux_pci: Fix shifto in pci_get_class. It looks like Linux's pci_get_class also matches the interface part of the PCI class register (but not the revision part), and I hadn't noticed that in the previous shim structured differently. With GCC12 kernel ALL/amd64 triggers "'sor' may be used uninitialized". If "sublinks & 3" is zero GCC is right and sor[1] may be returned unitialized. Fix by initializing "sor" to zero to return -1 instead of uninitialized value. Ok: Taylor R Campbell <riastradh@> amdgpu: Map BAR 2, not BAR 5, on pre-bonaire chips. PR kern/58384 amdgpu: Map consecutive pages, not the same one over and over again. PR kern/58385 linux/bitops: Fix overestimate for BITS_TO_LONGS(9) Fortunately, this seems harmless except for allocating excessive buffer memory. Pointed out by nonaka@, OK riastradh@.
1.1	15-Sep-2021	thorpej	Device call interface definition files for the "device" and "pci" subsystems.
1.1	15-Sep-2021	thorpej	Adjust the device_call() calling convention so as to provide type checking of the arguments passed to the call, using auto-generated argument structures and binding macros.
1.4	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.3	11-Jan-2001	martin	Move interrupt mapping before initialization of the card (which might cause interrupts). This works around a problem on sparc64, where interrupts without installed handler seem to cause heavy lossage (and interrupts are already enabled at this stage of autoconfiguration).
1.2	09-Jan-2001	martin	Ifdef some routines only needed for ELSA Quickstep cards when compiling without support for those.
1.1	05-Jan-2001	martin	branches: 1.1.1; Initial revision
1.1.1.1	05-Jan-2001	martin	branches: 1.1.1.1.2; Initial import of ISDN4BSD release 0.96
1.1.1.1.2.4	12-Mar-2001	bouyer	Sync with HEAD.
1.1.1.1.2.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.1.1.1.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.1.1.2.1	05-Jan-2001	bouyer	file pci_isic.c was added on branch thorpej_scsipi on 2001-01-05 17:36:12 +0000
1.2	18-Feb-2001	martin	Filename cleanup police (per discussion with Soren S. Jorvang): remove all (legacy) "i4b_" prefixes outside of sys/netisdn. Prefix all card specific driver support files with the basename of the driver bus attachement file. Renamed here: pci_isic.h -> isic_pci.h pci_isic.c -> isic_pci.c i4b_avm_fritz_pci.c -> isic_pci_avm_fritz_pci.c i4b_elsa_qs1p.c -> isic_pci_elsa_qs1p.c
1.1	05-Jan-2001	martin	branches: 1.1.1; Initial revision
1.1.1.1	05-Jan-2001	martin	branches: 1.1.1.1.2; Initial import of ISDN4BSD release 0.96
1.1.1.1.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.1.1.1.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.1.1.1.2.1	05-Jan-2001	bouyer	file pci_isic.h was added on branch thorpej_scsipi on 2001-01-05 17:36:12 +0000
1.44	29-Dec-2020	skrll	KNF a comment
1.43	29-Dec-2020	skrll	More of the patch that adds support for Enhanced Allocations as seen in the Cavium ThunderX based GIGABYTE MT30-GS2-00 From thorpej@. Thanks! (botched the copying of this from a branch previously)
1.42	29-Dec-2020	skrll	Add support for Enhanced Allocations as seen in the Cavium ThunderX based GIGABYTE MT30-GS2-00 From thorpej@. Thanks!
1.41	28-Dec-2020	skrll	Trailing whitespace
1.40	05-May-2020	bouyer	branches: 1.40.2; disable I/O or mem decode before probing BAR size. Bar size is probed writing 0xffffffff to the BAR and reading back; but while doing this the decoding address is not guaranteed to be valid and could have side effect. Xen PVH enforces disabling decoding before writing to a BAR. Proposed on tech-kern@, got positive comments
1.39	02-Dec-2019	riastradh	Leave it entirely up to the driver whether to map prefetchable. We have only a few drivers that pass BUS_SPACE_MAP_PREFETCHABLE to pci_mapreg_map -- dev/pci/if_hme_pci.c, dev/pci/igma.c, dev/pci/radeonfb.c, dev/pci/wcfb.c -- and they all do it for ROM or framebuffers. Based on a subthread from macallan@ in the earlier discussion: https://mail-index.NetBSD.org/tech-kern/2017/03/23/msg021685.html The thrust is that: - a driver not asking for prefetchable mappings shouldn't get it - some devices are correctly used with prefetchable mappings, but for some reason fail to set the prefetchable bit in the BAR - nobody could identify any classes of device for which (a) the driver asks for prefetchable mappings, but (b) certain matching devices can't actually be used with prefetchable mappings This brings pci_mapreg_map in line with what the documentation says; the documentation never advertised that the prefetchable bit in the BAR could cause the bus_space mapping to be prefetchable.
1.38	02-Dec-2019	riastradh	Use BUS_SPACE_MAP_PREFETCHABLE only if BAR and driver agree on it. - A driver that expects prefetchable memory and knows to issue the needed bus_space_barrier calls can pass BUS_SPACE_MAP_PREFETCHABLE to indicate a desire to map the memory prefetchable if the BAR allows it. (A driver that _really wants_ BUS_SPACE_MAP_PREFETCHABLE even if the BAR claims _not_ to be prefetchable can use pci_mapreg_info and bus_space_map explicitly -- this is not different from what we have today.) - For a driver that _does not_ expect prefetchable memory, the appearance of the prefetchable bit in the BAR shouldn't cause it to use BUS_SPACE_MAP_PREFETCHABLE, because the driver will not issue the needed bus_space_barrier calls to get sensible results. Note: `Prefetchable' here, sometimes called `write-combining', means reads have no side effects, and writes are idempotent, so it is safe to issue reads out of order and safe to combine writes. Mappings with BUS_SPACE_MAP_PREFETCHABLE are often more weakly ordered than normal memory -- e.g., on x86, in WC-type memory regions, loads can be reordered with loads, stores can be reordered with stores, which is not possible with any other type of memory regions. Discussed on tech-kern a while ago: https://mail-index.NetBSD.org/tech-kern/2017/03/22/msg021678.html This is option A, which received the most support. This should help unconfuse drivers that do not expect prefetchable mappings, like Yamaguchi-san tripped over recently: https://mail-index.NetBSD.org/tech-kern/2019/12/02/msg025785.html
1.37	01-Mar-2019	msaitoh	No functional change: - u_int* -> uint* - KNF.
1.36	19-May-2018	jakllsch	branches: 1.36.2; Re-enable PCI window decode enable in PCI Command Register upon successful pci_mapreg_map(). Loosely related to PR #53286.
1.35	19-May-2018	jakllsch	Refine previous change to enable PCI window decoding in Command Register upon mapping; conditionalize on a global variable, that is set to true on x86 machines booting under EFI. For now, initialize the global variable at compile time to false. This is intended to limit potential problems for other NetBSD ports, should this changeset be pulled up to netbsd-8. Related to PR #53286.
1.34	16-May-2018	jakllsch	Enable the appropriate memory or I/O space decode in the PCI Command/Status Register upon mapping a BAR. This should fix PR #53286. It's also possible there are other similar PRs that might be fixed by this.
1.33	17-Mar-2017	msaitoh	branches: 1.33.6; 1.33.12; Modify some variables to make code more understandable.
1.32	26-Dec-2014	msaitoh	branches: 1.32.2; 1.32.4; Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fix ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.31	16-Oct-2014	riastradh	branches: 1.31.2; Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.30	20-Oct-2012	matt	branches: 1.30.12; Only probe the upper BAR of a mem64 BAR if bit 31 is readonly
1.29	17-May-2011	dyoung	branches: 1.29.4; 1.29.10; 1.29.14; PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.28	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.27	27-Feb-2011	jmcneill	revert previous; instead of commenting out the error message how about fixing the code that causes it to be printed in the first place
1.26	27-Feb-2011	jruoho	Remove the following aprint_debug(9): "pci_mem_find: void region". I have a system where this is printed about a hundred times. No functional change.
1.25	11-May-2010	dyoung	branches: 1.25.2; Cosmetic: remove superfluous parentheses from return statements. Don't test truth of pointers, or compare with 0, but compare with NULL, instead. The compiler generated the same assembly language before & after this change.
1.24	22-Jul-2008	bjs	branches: 1.24.4; 1.24.12; 1.24.14; 1.24.16; Add pci_mapreg_submap(): This function is pci_mapreg_map() with two additional arguments, offset and maxsize. This new functionality eases handling certain tasks within the direct rendering manager, though I hope others will also find it useful. pci_mapreg_map() is now merely a wrapper around pci_mapreg_submap(); the latter contains all of the code from the former. ok christos@
1.23	28-Apr-2008	martin	branches: 1.23.2; 1.23.4; 1.23.6; Remove clause 3 and 4 from TNF licenses
1.22	01-Dec-2007	jmcneill	branches: 1.22.14; 1.22.16; 1.22.18; aprintify
1.21	20-Nov-2007	joerg	Merge from jmcneill-pm: XXX allow pci_find_mem to map a 64-bit BAR.
1.20	10-Apr-2007	macallan	branches: 1.20.8; 1.20.10; 1.20.14; 1.20.16; turn the other 'void region' message into aprint_debug() as well
1.19	10-Apr-2007	macallan	demote 'pci_io_find: void region' to aprint_debug()
1.18	16-Nov-2006	christos	branches: 1.18.2; 1.18.4; 1.18.8; 1.18.10; 1.18.14; __unused removal on arguments; approved by core.
1.17	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.16	01-Mar-2006	gdamore	branches: 1.16.14; 1.16.16; Add pci_find_rom() API as discussed on tech-kern.
1.15	27-Feb-2006	gdamore	Add PCI_MAPREG_TYPE_ROM and allow it to be used with pci_mapreg_map(). Fix to configure (but do not enable) BARs for expansion ROMs. Reviewed by briggs@
1.14	11-Dec-2005	christos	branches: 1.14.2; 1.14.4; 1.14.6; merge ktrace-lwp.
1.13	27-Feb-2005	perry	branches: 1.13.4; nuke trailing whitespace
1.12	30-May-2002	drochner	branches: 1.12.6; 1.12.14; 1.12.16; implement a check whether a BAR is present at all at a given configuration space address and use it where the mappings of the VGA card are registered before descenting too deep into "memory" type specific code (pci_mem_find() gets noisy if it doesn't like the register)
1.11	29-May-2002	bouyer	When testing for a void region, use PCI_MAPREG_MEM_SIZE() for 32bit memaddr. If the 32bit mask read is 0, wmask will be 0xffffffff00000000, and PCI_MAPREG_MEM64_SIZE() won't make this 0.
1.10	13-Nov-2001	lukem	branches: 1.10.8; 1.10.10; add RCSID
1.9	29-Nov-2000	thorpej	branches: 1.9.2; 1.9.4; ANSI'ify.
1.8	28-Jun-2000	thorpej	Don't check for a BAR address past PCI_MAPREG_END; some devices have BARs way out in left field.
1.7	10-May-2000	thorpej	branches: 1.7.4; Add support for mapping 64-bit PCI memory space. If the region is mapped in a way that is inaccessible by a 32-bit bus_addr_t, then print a message to that effect and return failure. Original patches by Bill Studenmund, with a few small changes by me.
1.6	25-Jan-2000	drochner	use BUS_SPACE_MAP_PREFETCHABLE instead BUS_SPACE_MAP_CACHEABLE where the PCI BAR bit is referred to
1.5	15-Aug-1998	mycroft	branches: 1.5.12; Assign my copyrights to TNF.
1.4	06-Oct-1997	thorpej	Back out last change. (Partially my fault, for not reviewing it close enough.)
1.3	03-Oct-1997	lonhyn	pci_mapreg_info() has been changed to take arguments like pci_mapreg_map()
1.2	23-Sep-1997	mycroft	If the region size turns out to be 0, return an error. This is what will happen if we try to use a mapping register that's not actually defined.
1.1	30-Aug-1997	mycroft	branches: 1.1.2; * Rewrite pci_io_find() and pci_mem_find() using my original pci_map_io() and pci_map_mem(), with appropriate changes for bus_space. * Add extra arguments for use by pci_mapreg_(), and make the functions static. Rewrite pci_mapreg_info() and pci_mapreg_map() as wrappers around pci_*_find(), using the documented interface.
1.1.2.3	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.1.2.1	30-Aug-1997	thorpej	file pci_map.c was added on branch marc-pcmcia on 1997-09-01 20:29:44 +0000
1.5.12.2	08-Dec-2000	bouyer	Sync with HEAD.
1.5.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.4.1	28-Jun-2000	thorpej	Pull up rev. 1.8: Don't check for a BAR address past PCI_MAPREG_END; some devices have BARs way out in left field.
1.9.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.2.2	20-Jun-2002	nathanw	Catch up to -current.
1.9.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.10.10.1	10-Oct-2003	tron	Pull up revision 1.11 (requested by kleink in ticket #1513): When testing for a void region, use PCI_MAPREG_MEM_SIZE() for 32bit memaddr. If the 32bit mask read is 0, wmask will be 0xffffffff00000000, and PCI_MAPREG_MEM64_SIZE() won't make this 0.
1.10.8.2	20-Jun-2002	gehenna	catch up with -current.
1.10.8.1	30-May-2002	gehenna	Catch up with -current.
1.12.16.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.12.14.1	29-Apr-2005	kent	sync with -current
1.12.6.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.4.3	07-Dec-2007	yamt	sync with head
1.13.4.2	03-Sep-2007	yamt	sync with head.
1.13.4.1	21-Jun-2006	yamt	sync with head.
1.14.6.1	22-Apr-2006	simonb	Sync with head.
1.14.4.1	09-Sep-2006	rpaulo	sync with head
1.14.2.1	01-Mar-2006	yamt	sync with head.
1.16.16.2	10-Dec-2006	yamt	sync with head.
1.16.16.1	22-Oct-2006	yamt	sync with head
1.16.14.1	18-Nov-2006	ad	Sync with head.
1.18.14.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.18.10.1	11-Jul-2007	mjf	Sync with head.
1.18.8.1	10-Apr-2007	ad	Sync with head.
1.18.4.1	15-Apr-2007	yamt	sync with head.
1.18.2.1	17-May-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1142): sys/dev/pci/pci_map.c: revision 1.21 allow pci_find_mem to map a 64-bit BAR.
1.20.16.1	08-Dec-2007	mjf	Sync with HEAD.
1.20.14.1	21-Nov-2007	bouyer	Sync with HEAD
1.20.10.1	09-Jan-2008	matt	sync with HEAD
1.20.8.3	01-Dec-2007	jmcneill	Sync with HEAD.
1.20.8.2	20-Nov-2007	joerg	Change the XXX to actually explain what it is doing.
1.20.8.1	27-Aug-2007	jmcneill	XXX allow pci_find_mem to map a 64-bit BAR, needs revisiting.
1.22.18.3	11-Aug-2010	yamt	sync with head.
1.22.18.2	04-May-2009	yamt	sync with head.
1.22.18.1	16-May-2008	yamt	sync with head.
1.22.16.1	18-May-2008	yamt	sync with head.
1.22.14.2	28-Sep-2008	mjf	Sync with HEAD.
1.22.14.1	02-Jun-2008	mjf	Sync with HEAD.
1.23.6.1	19-Oct-2008	haad	Sync with HEAD.
1.23.4.1	28-Jul-2008	simonb	Sync with head.
1.23.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.24.16.3	31-May-2011	rmind	sync with head
1.24.16.2	21-Apr-2011	rmind	sync with head
1.24.16.1	30-May-2010	rmind	sync with head
1.24.14.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.24.12.1	03-Jan-2012	matt	Only probe the upper BAR of a mem64 if the bit31 of the lower BAR isn't writable.
1.24.4.1	23-Jan-2015	martin	Pull up the following changes, requested by msaitoh in ticket #1942: sys/arch/x86/pci/ichlpcib.c 1.40, 1.45 via patch sys/dev/ic/i82801lpcreg.h 1.12 sys/dev/pci/pci_map.c 1.32 via patch - Fix a bug that ichlpcib(4) maps I/O area incorrectly. It might also fixes ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Fix the calculation of the map size in pci_mapreg_submap(). - Use '\n' at the end of aprint_error_dev() format strings.
1.25.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.29.14.2	03-Dec-2017	jdolecek	update from HEAD
1.29.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.29.10.1	16-Jan-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1229): sys/arch/x86/pci/ichlpcib.c: revision 1.40, 1.45 sys/dev/pci/pcivar.h: revision 1.101 sys/dev/pci/pci_map.c: revision 1.32 sys/dev/ic/i82801lpcreg.h: revision 1.12 Use '\n' at the end of all aprint_error_dev() format strings. -- Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fixes ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.29.4.1	30-Oct-2012	yamt	sync with head
1.30.12.2	08-Jan-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #394): sys/dev/pci/pcivar.h: revision 1.101 sys/dev/pci/pci_map.c: revision 1.32 sys/dev/ic/i82801lpcreg.h: revision 1.12 sys/arch/x86/pci/ichlpcib.c: revision 1.45 Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fix ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.30.12.1	17-Oct-2014	martin	Pull up following revision(s) (requested by riastradh in ticket #144): sys/dev/pci/radeonfb.c: revision 1.85 sys/dev/pci/pcivar.h: revision 1.100 sys/dev/pci/pci_map.c: revision 1.31 sys/external/bsd/drm2/include/linux/pci.h: revision 1.9 Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.31.2.2	28-Aug-2017	skrll	Sync with HEAD
1.31.2.1	06-Apr-2015	skrll	Sync with HEAD
1.32.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.32.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.33.12.1	21-May-2018	pgoyette	Sync with HEAD
1.33.6.1	07-Jun-2018	martin	Pull up following revision(s) (requested by jakllsch in ticket #832): sys/dev/pci/pcivar.h: revision 1.112 sys/dev/pci/pci_map.c: revision 1.34,1.35 sys/arch/x86/x86/efi.c: revision 1.15 Enable the appropriate memory or I/O space decode in the PCI Command/Status Register upon mapping a BAR. This should fix PR #53286. It's also possible there are other similar PRs that might be fixed by this. - Refine previous change to enable PCI window decoding in Command Register upon mapping; conditionalize on a global variable, that is set to true on x86 machines booting under EFI. For now, initialize the global variable at compile time to false. This is intended to limit potential problems for other NetBSD ports, should this changeset be pulled up to netbsd-8. Related to PR #53286.
1.36.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.36.2.1	10-Jun-2019	christos	Sync with HEAD
1.40.2.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.12	17-Oct-2018	jmcneill	PCI_QUIRK_HASEXTCNF is no longer required for AMD Seattle PCHB
1.11	16-Oct-2018	jmcneill	Set PCI_QUIRK_HASEXTCNF flag for AMD Seattle PCI host bridges
1.10	28-Feb-2018	msaitoh	branches: 1.10.2; 1.10.4; - Add new PCI quirk PCI_QUIRK_HASEXTCNF and PCI_QUIRK_NOEXTCNF. Some devices' extended configuration area may be broken or violate spec. If an extended configuration space is strange but it really exist, use PCI_QUIRK_HASEXTCNF. If an extended configuration space is plausible to exist but it really doesn't exist, use PCI_QUIRK_NOEXTCNF. - Add PCI_PRODUCT_INTEL_XEOND_MEM_0_TTR_1(0x6fa8) and PCI_PRODUCT_INTEL_COREI76K_IMC_0(0x6f68) with PCI_QUIRK_HASEXTCNF. The document clearly states they violate spec and it support the extended configuration space.
1.9	19-Aug-2009	pgoyette	Use __arraycount() for readability
1.8	06-Jan-2007	jmcneill	branches: 1.8.44; Sync with Xbox Host-PCI bridge name change.
1.7	04-Jan-2007	jmcneill	Add quirk for Xbox nForce, from Andrew Gillham.
1.6	11-Dec-2005	christos	branches: 1.6.20; merge ktrace-lwp.
1.5	28-Jun-2005	thorpej	branches: 1.5.2; Use ANSI function decls and static.
1.4	10-Jul-2004	nathanw	Remove SKIP_FUNCn quirks for SC1100 companion chip; the fix to SMM in i386/identcpu.c renders it counterproductive. (this was the only use and the reason for the invention of the SKIP_FUNCn quirk; it should probably be removed entirely).
1.3	15-Aug-2003	itojun	- check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.2	13-Nov-2001	lukem	branches: 1.2.10; 1.2.16; add RCSID
1.1	31-May-1998	cgd	branches: 1.1.26; 1.1.28; dd a PCI 'quirks' mechanism, meant to be used as the mechanism of last resort when trying to communicate information about bogus behaviour of PCI devices to the MI autoconfiguration code. In general, bogus behaviour should be handled by drivers, but there are some types of bogons which can't be addressed that way. The only quirk currently defined is one which indicates that the device is multi-function even though the device's header says otherwise. (Mmm, Intel 82371FB PCI-to-ISA Bridge (PIIX); you'd think that at least Intel would have gotten it right...)
1.1.28.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.26.1	14-Nov-2001	nathanw	Catch up to -current.
1.2.16.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.16.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.16.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.16.1	03-Aug-2004	skrll	Sync with HEAD
1.2.10.1	15-Aug-2003	tron	Pull up revision 1.3 (requested by itojun in ticket #1412): - check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.5.2.1	26-Feb-2007	yamt	sync with head.
1.6.20.1	12-Jan-2007	ad	Sync with head.
1.8.44.1	16-Sep-2009	yamt	sync with head
1.10.4.1	10-Jun-2019	christos	Sync with HEAD
1.10.2.1	20-Oct-2018	pgoyette	Sync with head
1.7	03-Mar-2025	riastradh	pci_resource(9): Handle multiple bus ranges too. Seen on orion o6. Prompted by: PR port-amd64/59118: Thinkpad T495s - iwm PCI BAR is zero
1.6	03-Mar-2025	riastradh	pci_resource(9): Handle multiple ranges of the same type. Rather than having the caller fill in an array indexed by range type of a single start/end pair for each type, the caller will now do: struct pci_resource_info info; memset(&info, 0, sizeof(info)); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart1, mend1); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart2, mend2); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart3, mend3); pci_resource_add_range(&info, PCI_RANGE_PMEM, pstart1, pend1); pci_resource_add_range(&info, PCI_RANGE_IO, iostart1, ioend1); ... While here, fix a format string -- 0x% or %# is enough to say that the output is hexadecimal; we don't need to belabour the point by printing a `0x0x' prefix! XXX While this handles multiple io/mem/prefetchable-mem ranges, it doesn't yet handle multiple bus ranges, which we have seen in practice, e.g. on orion o6. TBD. ok jmcneill@ PR port-amd64/59118: Thinkpad T495s - iwm PCI BAR is zero
1.5	30-Jun-2024	jmcneill	pci_resource: Make unexpected bus numbers in bridges non-fatal. Firmware bugs happen. Log a warning and continue instead of panicing.
1.4	12-Nov-2023	jmcneill	branches: 1.4.2; pci: Improve resource allocation for non-prefetchable BARs. When allocating resources for PCI devices, use the following criteria: - Prefetchable memory must be allocated from the prefetchable range of the parent bridge. - For 64-bit MMIO, try the prefetchable range first, and fallback to the non-prefetchable range. The idea here is to preserve 32-bit resources for 32-bit BARs. - For 32-bit MMIO, try the non-prefetchable range first. If that fails, make one last attempt at allocating from the prefetchable range, in the event that it has resources below 4GB.
1.3	15-Oct-2022	riastradh	branches: 1.3.2; pci_resource(9): Fix whitespace.
1.2	15-Oct-2022	riastradh	pci_resource(9): vmem_create and vmem_add never fail with VM_SLEEP. Prune dead error branches.
1.1	14-Oct-2022	jmcneill	Add a PCI resource manager and use it on Arm ACPI platforms. The Arm ACPI code relied on PCI_NETBSD_CONFIGURE to configure devices that were not enabled by system firmware. This is not safe to do unless the firmware explicitly permits it using a device specific method defined in the PCI firmware spec. Introduce a new PCI resource manager that discovers what has already been configured by firmware and allocates from the remaining space. This will ensure that devices setup by firmware are untouched and only will program BARs of devices that are not enabled at boot time. The current implementation assumes that the parent PCI-PCI bridge's are already configured. A worthwhile improvement in the future would be to support programming windows for bridges that are not fully configured.
1.3.2.2	03-Jul-2024	martin	Pull up following revision(s) (requested by jmcneill in ticket #735): sys/dev/pci/pci_resource.c: revision 1.5 sys/arch/arm/pci/pci_msi_machdep.c: revision 1.10 sys/arch/aarch64/include/vmparam.h: revision 1.21 sys/dev/acpi/acpi_resource.c: revision 1.43 pci_resource: Make unexpected bus numbers in bridges non-fatal. Firmware bugs happen. Log a warning and continue instead of panicing. acpi: Ignore producer/consumer bit for fixed memory resources. The requirement to honour the producer/consumer bit in fixed memory resource descriptors was dropped at some point in a revision to the ACPI 2.0 specification because too many firmware implementations got it wrong. aarch64: Bump VM_PHYSSEG_MAX to match DRAM_BANKS / FDT_MEMORY_RANGES. On aarch64 there is a single free list, so VM_PHYSSEG_MAX needs to be the same as FDT_MEMORY_RANGES (which needs to be the same as DRAM_BANKS). Future cleanup should be done to fold these into a single define. arm: pci: Fix ITS ID lookup for MSIs. pci_get_frameid expects a BDF requestor ID as input, not a Device ID. Fixes MSI/MSI-X support on Ampere Altra systems.
1.3.2.1	29-Nov-2023	martin	Pull up following revision(s) (requested by jmcneill in ticket #483): sys/dev/pci/pci_resource.c: revision 1.4 pci: Improve resource allocation for non-prefetchable BARs. When allocating resources for PCI devices, use the following criteria: - Prefetchable memory must be allocated from the prefetchable range of the parent bridge. - For 64-bit MMIO, try the prefetchable range first, and fallback to the non-prefetchable range. The idea here is to preserve 32-bit resources for 32-bit BARs. - For 32-bit MMIO, try the non-prefetchable range first. If that fails, make one last attempt at allocating from the prefetchable range, in the event that it has resources below 4GB.
1.4.2.2	02-Aug-2025	perseant	Sync with HEAD
1.4.2.1	01-Jul-2024	perseant	Sync with HEAD.
1.2	03-Mar-2025	riastradh	pci_resource(9): Handle multiple ranges of the same type. Rather than having the caller fill in an array indexed by range type of a single start/end pair for each type, the caller will now do: struct pci_resource_info info; memset(&info, 0, sizeof(info)); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart1, mend1); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart2, mend2); pci_resource_add_range(&info, PCI_RANGE_MEM, mstart3, mend3); pci_resource_add_range(&info, PCI_RANGE_PMEM, pstart1, pend1); pci_resource_add_range(&info, PCI_RANGE_IO, iostart1, ioend1); ... While here, fix a format string -- 0x% or %# is enough to say that the output is hexadecimal; we don't need to belabour the point by printing a `0x0x' prefix! XXX While this handles multiple io/mem/prefetchable-mem ranges, it doesn't yet handle multiple bus ranges, which we have seen in practice, e.g. on orion o6. TBD. ok jmcneill@ PR port-amd64/59118: Thinkpad T495s - iwm PCI BAR is zero
1.1	14-Oct-2022	jmcneill	branches: 1.1.8; Add a PCI resource manager and use it on Arm ACPI platforms. The Arm ACPI code relied on PCI_NETBSD_CONFIGURE to configure devices that were not enabled by system firmware. This is not safe to do unless the firmware explicitly permits it using a device specific method defined in the PCI firmware spec. Introduce a new PCI resource manager that discovers what has already been configured by firmware and allocates from the remaining space. This will ensure that devices setup by firmware are untouched and only will program BARs of devices that are not enabled at boot time. The current implementation assumes that the parent PCI-PCI bridge's are already configured. A worthwhile improvement in the future would be to support programming windows for bridges that are not fully configured.
1.1.8.1	02-Aug-2025	perseant	Sync with HEAD
1.8	24-Jun-2018	jdolecek	provide pci_intr_establish_xname() on x86 independantly from MSI, so it's available on XEN too; change also the stub to use weak symbol instead #ifdef
1.7	01-Jun-2017	chs	branches: 1.7.8; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.6	25-Nov-2016	knakahara	provide all PCI MSI/MSI-X manipulation stub functions. "#ifdef __HAVE_PCI_MSI_MSIX" workaround such as nvme_pci(4) is not required any more. http://mail-index.netbsd.org/source-changes/2016/09/17/msg077799.html
1.5	11-Jul-2016	knakahara	branches: 1.5.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.4	22-Oct-2015	knakahara	add pci_intr_alloc related stubs to reduce ifdef from device drivers.
1.3	24-Aug-2015	pooka	add some _KERNEL_OPT as the finishing touch
1.2	17-Aug-2015	knakahara	Add kernel code to support intrctl(8).
1.1	24-Aug-2011	dyoung	branches: 1.1.12; 1.1.30; Add to pci_bus_devorder() an argument that tells the number of slots available in the devs array. Change the type of the devs array from char to uint8_t. Treat the return value of pci_bus_devorder() as the number of slots that it filled. Don't use the __PCI_BUS_DEVORDER #definition to configure the kernel but let the linker do it. Make pci_bus_devorder() available on all architectures by adding a default implementation that will DTRT on all architectures but hpcmips, the only architecture to #define __PCI_BUS_DEVORDER. On hpcmips, adapt the implementation to the new calling convention. XXX I can compile an hpcmips GENERIC kernel, but I don't have a XXX hpcmips box to test it on.
1.1.30.5	28-Aug-2017	skrll	Sync with HEAD
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	05-Oct-2016	skrll	Sync with HEAD
1.1.30.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1.30.1	22-Sep-2015	skrll	Sync with HEAD
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.7.8.1	25-Jun-2018	pgoyette	Sync with HEAD
1.244	19-Apr-2024	andvar	s/Resorce/Resource/ in comment and log message.
1.243	10-Aug-2023	andvar	fix typos in comments.
1.242	01-Feb-2022	msaitoh	Fix typo. Found by rillig. s/vendor specific/vendor-specific/
1.241	31-Jan-2022	msaitoh	Decode SCSI programming interface. Whitespace fix.
1.240	31-Jan-2022	msaitoh	Add three new extended capabilities: - Physical Layer 32.0 GT/s - Alternate Protocol - System Firmware Intermediary
1.239	08-Jan-2022	tnn	place additional parens around multiline string constant to appease clang
1.238	07-Jan-2022	msaitoh	Two fixes: - Fix previous again. The missing comma is intended to split the long line. - Add missing right parenthesis.
1.237	07-Jan-2022	msaitoh	Add missing comma. Found by tnn@.
1.236	01-Jan-2022	msaitoh	Print Physical Layer 16.0 GT/s and Lane Margining at the Receiver extended cap. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Rename pcie_link_compliance_preset_deemphasis to pcie_link_preset_preshoot_deemphasis because the table is referenced from multiple places. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Rename PCI_EXTCAP_PYSLAY_16GT to PCI_EXTCAP_PL16G.
1.235	28-Dec-2021	msaitoh	%hhx -> %x
1.234	28-Dec-2021	msaitoh	Decode link control2's Compliance Preset/De-emphasis more. Fix typo.
1.233	03-Dec-2021	andvar	fix various typos in comments, log messages and documentation.
1.232	10-Oct-2021	msaitoh	Use PCI-SIG official acronyms: - RP stands for Root Port. - RC stands for Root Complex. - RCIEP stands for Root Complex Integrated End Point.
1.231	10-Oct-2021	msaitoh	Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet)
1.230	11-Sep-2021	mrg	add 32.0GT/s to the list of pcie speeds (PCIe 5.x.)
1.229	17-Aug-2021	andvar	fix multiplei repetitive typos in comments, messages and documentation. mainly because copy paste code big amount of files are affected.
1.228	17-Aug-2021	andvar	s/Windw/Window/ in log message.
1.227	12-Jul-2021	msaitoh	s/Precision Time Management/Precision Time Measurement/
1.226	29-Jun-2021	pgoyette	Rework the xxxVERBOSE option to share the common module-hook-based verbose mechanism with MIIVERBOSE. This reduces some duplicated code and allows us to once again permit auto-unload of MIIVERBOSE. Change details: * Update dev/devlist2h.awk to accomodate miidevs, including generation of MII_STR_oui_model definitions and use of oui and model rather than vendor and product. This also changes the compressed data in the xxxdevs_data.h files to uint32_t (since mii oui's are up to 6 hex digits long) * Update a couple of phy drivers to use new calls to get verbose data * Regen all of the xxxdevs{,_data}.h files (separate commit, coming very soon) * Update mii/mii_verbose.[ch] and mii/mii_physubr.c to use the various DEV_VERBOSE_xxx macros * Update the pci, usb, and hdaudio code as needed, to #include the xxxdevs.h files (in order to get the proper printf format strings) * Since dev/dev_verbose.c now uses non-literal printf format strings, (to deal with the vendor/product vs oui/model issue), we need to make sure it gets compiled with -Wno-error=format-nonliteral, even in userland's libpci and librumpdev! * Bump kernel version for the change in module interfaces Welcome to 9.99.86! XXX It might be useful in the future to extend the MII_STR_oui_model XXX definitions to PCI as well (and perhaps USB and HDAUDIO). This XXX would allow for a single centralized location for the products' XXX descriptions, rather than being dispersed among individual XXX drivers' xxx_match tables.
1.225	27-Jan-2021	thorpej	branches: 1.225.4; Introduce autoconfiguration helpers based around device_compatible_entry: - pci_compatible_match(): matches against the PCI ID. - pci_compatible_match_subsys(): matches against PCI SUBSYS ID. - pci_compatible_lookup(): look up entry by PCI ID. - pci_compatible_lookup_subsys(): look up entry by PCI SUBSYS ID. - pci_compatible_lookup_id(): look up entry by an arbitrary ID using the PCI ID code conventions. - Define PCI_COMPAT_EOL as a compat data array sentinel.
1.224	30-May-2020	jdolecek	branches: 1.224.2; change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack
1.223	10-Apr-2020	christos	Fix all the random __SHIFTOUT casts by introducing an intermediate macho, and then deleting and fixing the rest.
1.222	21-Mar-2020	thorpej	branches: 1.222.2; When parsing Enhanced Allocation entries, use the correct calulation for finding the next entry.
1.221	25-Jan-2020	msaitoh	Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet.
1.220	25-Jan-2020	msaitoh	- Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Avoid using magic number.
1.219	17-Jan-2020	msaitoh	Fix a bug that the virtual channel extended configuraton's arbitration phase register can't be decoded correcty. Found by jmcneill.
1.218	11-Dec-2019	msaitoh	branches: 1.218.2; s/enalbe/enable/
1.217	20-Sep-2019	msaitoh	Print some DPC register values not with %04x but with %08x because those are 32bit.
1.216	21-Aug-2019	msaitoh	Whitespace fixes. No functional change.
1.215	18-Jul-2019	msaitoh	branches: 1.215.2; Remove whitespace for consistency.
1.214	09-Jul-2019	msaitoh	Print extend capability from 0x25 to 0x29 correctly.
1.213	09-Jul-2019	msaitoh	Identify 16GT/s.
1.212	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.211	28-Feb-2019	khorben	Typo (s/vaule/value/)
1.210	30-Nov-2018	msaitoh	Add new PCIE_HAS_LINKREGS(pcie_devtype) and use it. No functional change.
1.209	30-Nov-2018	msaitoh	Add new macro PCIE_HAS_ROOTREGS(pcie_devtype) and use it. No functional change.
1.208	30-Nov-2018	msaitoh	The downstream port of PCIe switch is not a root port, so don't print root port related register. For example, Intel 63xxESB controller's downstream port device was printed by pcictl(8) with this bug:
1.207	05-Nov-2018	msaitoh	Decode PCI Enhanced Allocation.
1.206	04-Oct-2018	msaitoh	Don't print extra newline if the PCI extended configuration space doesn't exist.
1.205	03-Oct-2018	msaitoh	- Don't print TPH requester's ST Table Size if the ST table location field is not PCI_TPH_REQ_STTBLLOC_TPHREQ because the size field is only applicaple for PCI_TPH_REQ_STTBLLOC_TPHREQ case. - Add comment.
1.204	27-Sep-2018	msaitoh	Root Complex Event Collector Bus Number Association ECN. - If capability version is 2 (or greater), decode RCEC Associated Bus Numbers register.
1.203	12-Sep-2018	msaitoh	Add ATS Relaxed Ordering supported bit described in Address Translation Relaxed Ordering ECN.
1.202	03-Jul-2018	msaitoh	- Print Power Management Control/status register in 32bit. - Simplify.
1.201	09-May-2018	msaitoh	branches: 1.201.2; Fix typo. s/TPL/TLP/
1.200	01-Feb-2018	msaitoh	branches: 1.200.2; - Add PCie Link Activation ECN. - Use macro. - KNF.
1.199	01-Feb-2018	msaitoh	Cleanup: - Don't pass a capability pointer as a argument of pci_conf_find_cap() and determine the first pointer in the pci_conf_find_cap() function. - Don't pass a capability pointer as a argument of pci_conf_find_extcap() because it's not used. - Remove unsed code.
1.198	01-Feb-2018	msaitoh	"s/above 300W/greater than 300W/" in pci_conf_print_pcie_power(). From PCIe Base Spec 3.1a Errata 2017-12-13.
1.197	18-Dec-2017	msaitoh	Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified.
1.196	25-Oct-2017	msaitoh	Whitespace fix.
1.195	24-Oct-2017	msaitoh	Print Error Source Identification register correctly.
1.194	19-Oct-2017	msaitoh	Fix a bug that the TPH ST table is decoded even if it's not in the TPH Requester extended capability structure.
1.193	12-Oct-2017	msaitoh	IOMMU cap dump fixes: - Print Capability Register's value. - Indent output correctly. - s/cahced/cached/ - Print MSI Message number with 0x%02x
1.192	10-Oct-2017	msaitoh	Decode IOMMU capability of PCI secure device capability. From "AMD I/O Virtualization Technology(IOMMU) Specification (#48882) Revision 3.00".
1.191	05-Oct-2017	msaitoh	- Add PCI_MAPREG_ROM_ADDR_MASK macro and PCI_MAPREG_ROM_ADDR() macro. - print PCI_MAPREG_ROM_ENABLE bit. - Decode Expansion ROM Validation ECN. - Add Native PCIe Enclosure Management ECN's extended capability type. Not decoded yet.
1.190	13-Jul-2017	msaitoh	- Official shortname of LN Requester is LNR, so change PCI_EXTCAP_LN_REQ to PCI_EXTCAP_LNR - Use macro.
1.189	15-Jun-2017	flxd	Fix printf format/argument.
1.188	15-Jun-2017	msaitoh	Fix a bug that LTR's latency in L1 PM Substates capability and Latency Tolerance Reporting capability isn't printed correctly.
1.187	15-Jun-2017	msaitoh	For Attention Indicator Control bit and Power Indicator Control bit, it's allowed to be a read only value 0 if corresponding capability register bit is 0.
1.186	08-Jun-2017	msaitoh	For the Target Link Speed in LCSR2, 0 is allowed for a device which supports 2.5GT/s only (and this check also works for devices which compliant to versions of the base specification prior to 3.0. Tested with BCM5709: - Target Link Speed: unknown value (0) + Target Link Speed: 2.5GT/s
1.185	07-Jun-2017	msaitoh	Fix to print the following bit fields correctly. - Supported Link Speeds Vector in LCAP2 - Lower SKP OS Generation Supported Speed Vector in LCAP2 - Lower SKP OS Reception Supported Speed Vector in LCAP2 - Enable Lower SKP OS Generation Vector in LCTL3 Note that the above bitfields start from 0 and the follwing bitfields start from 1: - Maximum Link Speed in LCAP - Current Link Speed in LCSR - Target Link Speed in LCSR2
1.184	05-Jun-2017	msaitoh	Add missing return to print the Slot Power Limit Value correctly.
1.183	29-May-2017	msaitoh	branches: 1.183.2; Print MSI Message data in 32bits when the Extended Message Data Capable bit is set.
1.182	24-May-2017	msaitoh	Decode TPH Requester Control register.
1.181	22-May-2017	msaitoh	- The Captured Slot Power Limit value is only for device which has upsteam port. - The following three registers are the same encoding, so use the same function: - the Captured Slot Power Limit in the Device Capability register of PICe capability. - the Slot Power Limit in Slot Capability register of PCIe capability. - the Base Power and Data Scale in the Data Register of Power Budgeting capability. - Fix the alternative encoding check of power limit as document. Add check if scale equals to 0.
1.180	09-May-2017	msaitoh	A device except Root Complex integrated has a link, so print link related registers on device except Root Complex Integrated Endpoint and Root Complex Event Collector.
1.179	27-Apr-2017	msaitoh	Check slot registers if a device is PCI/PCI-X to PCI Express Bridge. Tested with Pericom Semiconductors(Diodes) PI7C9X111SL PCIe to PCI Reverse Bridge.
1.178	21-Apr-2017	kre	branches: 1.178.2; For want of a comma, the compiler was lost.
1.177	21-Apr-2017	msaitoh	Add Flattening Portal Bridge capability ID and Hierarchy ID extended capability ID.
1.176	20-Apr-2017	msaitoh	Add Downstream Port Containment (DPC) ECN and Enhanced DPC(eDPC) ECN.
1.175	20-Apr-2017	msaitoh	Add 0x for register output.
1.174	18-Apr-2017	msaitoh	Add missing ':'
1.173	17-Apr-2017	msaitoh	Use macro. No functional change.
1.172	06-Apr-2017	msaitoh	Fix typos: - s/Readness/Readiness/ - s/Presense/Presence/ - s/supportted/supported/ - s/Rquester/Requester/ - s/Check Enab/Check Enable/
1.171	30-Mar-2017	msaitoh	Fix a bug that SR-IOV's system page size was doubled. ffs()'s bit are numbered starting from 1.
1.170	28-Mar-2017	msaitoh	Use macro. No functional change.
1.169	28-Mar-2017	msaitoh	Use correct macro "PCI_SATA_REV" instead of PCI_MSIX_CTL for SATA revision register. This is not a real bug because PCI_SATA_REV and PCI_MSIX_CTL have the same value.
1.168	17-Mar-2017	msaitoh	Don't print BAR's window size on pcictl pciN dump again because there is no way to know it's correct value from userland now.
1.167	17-Mar-2017	msaitoh	- Don't try to sizing a BAR not if a device is host bridge but a device's PCI_COMMAND_MEM_ENABLE is set. - Fix 17 years old bug that all effective BARs are marked "not sized" on pcictl pciN dump. - Print BAR's window size on pcictl pciN dump. - remove extra "(?)"
1.166	14-Mar-2017	msaitoh	ECN: Change Root Complex Event Collector Class Code - Old RCEC has subclass 0x06. It's the same as IOMMU. Read the type in PCIe extend capability to know whether it's RCEC or IOMMU.
1.165	14-Mar-2017	msaitoh	Pint PCI Interface name if available.
1.164	14-Mar-2017	msaitoh	Add two new PCI classes: - processing accelerators - non-essential instrumentation
1.163	14-Mar-2017	msaitoh	No functional change: - modify variable names - KNF
1.162	27-Feb-2017	christos	fix printf formats (clang build)
1.161	27-Feb-2017	msaitoh	Decode AGP capability.
1.160	26-Feb-2017	msaitoh	- Fix the base power's value in the power budget capability. - Fix typo (s/Maximun/Maximum/)
1.159	24-Feb-2017	msaitoh	Use macro. KNF. No functional change.
1.158	24-Feb-2017	msaitoh	Fix 0x%u...
1.157	15-Feb-2017	msaitoh	- Print Data Select and Data Scale in pci_conf_print_pcipm_cap(). - The Message Data register of MSI cap is not 32bit but 16bit. - When the PCIE_LCAP_MAX_SPEED bitfiled is 0, it means it supports 2.5GT/s only. - Print link de-emphasis value by "-X dB". - Print Completion Timeout Ranges Supported filed with alphabets. - Print TPH Completer Supported fileld's meaning. - Print PCIE_DCAP2_MAX_EETLP correctly. 0 means 4 End-End TLP Prefixes. - If the Supported Link Speed Vector is 0, the Link Capability 2 register is not implemented. Don't decode LCAP2 when the vector is 0. - The ACS's Egress Control Vector is 32bit, so print with 0x%08x. - Print SR-IOV's device ID. - Use __SHIFTOUT() to avoid using magic number. - Prefix "0x" for hexadecimal value.
1.156	28-Dec-2016	msaitoh	branches: 1.156.2; The Power Controller Control bit (PCIE_SLCSR_PCC) in the Slot Control & Status Register is 0 on power on. Print "Power <on\|off>" instead of "<on\|off>".
1.155	02-Nov-2016	pgoyette	Protect against buffer overflow.
1.154	31-Oct-2016	martin	Stopgap fix for in-kernel compilation (differences between humanize_number(3) and humanize_number(9)), ok: msaitoh
1.153	31-Oct-2016	msaitoh	Decode Resizable BAR.
1.152	20-Oct-2016	msaitoh	- pci_conf_print_pwrbdgt_base_power(): From 0xf3 to 0xff of the Base power are reserved above 300W (PCI 3.0 Errata). - Emergency Power Reduction mechanism with PWRBRK signal ECN. - Extended Message Data for MSI ECN. - Fix typo in comment.
1.151	19-Oct-2016	msaitoh	Add VF Resizable BARs ECN.
1.150	22-Aug-2016	msaitoh	Don't follow an incorrect pointer in an extended capability header. Tested with Super Micro X10SDV-8C-TLN4F bus 255, device 12, function 0 (BIOS 1.1b).
1.149	17-Aug-2016	msaitoh	Add Dynamic Power Allocation (DPA) ECN support.
1.148	20-Jun-2016	msaitoh	branches: 1.148.2; B260 "Slot Power Limit" ECN. PCIe 3.1 ECN. September 18, 2015.
1.147	11-May-2016	msaitoh	Add Precision Time Management (PTM) ECN.
1.146	18-Nov-2015	msaitoh	- Add the Auto Slot Power Limit Disable bit in Slot Control register and the Completion Timeout Prefix/Header Log Capable bit in the AER capability and control register (ECN: Downstream Port Containment (DPC)). - Add the Poisoned TLP Egress Block bit (ECN: Enhanced DPC). - Update Link Capabilities 2 register and Link Control 3 register (ECN: Separate Refclk Independent SSC Architecture (SRIS)) - ECN: Readiness Notifications (RN) - Add the Retimer Presence Detect Supported bit in the Link Capabilities 2 register and the Retimer Presence Detected bit in the Link Status 2 register (ECN: Extension Devices)
1.145	17-Nov-2015	msaitoh	- ARI's function group is not bit 32-24 but 22-20. - Add the Structure Length field in AF capability register. - Add Enhanced Allocation extended capability ID (ECN: Enhanced Allocation (EA) for Memory and I/O Resources). - Add LN System CLS (ECN: Lightweight Notification (LN) Protocol). - Add ST Upper and Lower bit definitions (ECN: TLP Processiong Hints). - Add the Global Invalidate bit in the ATS capability register and the PRG Response PASID Required bit in the Page Request status register (ECN: PASID Translation) - Decode ASPM support bit more (ECN: ASPM Optionally) - Use __BITS()
1.144	17-Nov-2015	msaitoh	No functional change: - Add comments. - Remove obsolete comment. - Move definitions to better location. - Rename bit definition. - KNF. - Indent.
1.143	16-Nov-2015	msaitoh	Define PCIE_XCAP_{VER,TYPE}(x) and use them.
1.142	13-Nov-2015	msaitoh	Fix register offset to print HyperTransport registers correctly.
1.141	12-Nov-2015	msaitoh	- Restore pci_subr.c rev. 1.135's change in pci_conf_print_caplist(). As wrote in the comment, HyperTransport capability appears multiple times. pci_conf_cap() reruns only the first entry, so it can't be used here. - Try to decode HyperTransport capability. Currently, the capability type of each HyperTransport capability is printed and only the MSI Mapping capability is decoded. - Style change.
1.140	30-Oct-2015	msaitoh	- Move PCI_INTRSTR_LEN from pcireg.h to pcivar.h. - In PCI-X cap, print 2nd bus's PCI-X mode, error protection type, Max clock frequency and Max clock period. - In SATA cap, print register location correctly. - In Virtual Channel cap, print reference clock with "ns". - In Root Complex Link Declaration, print Link Entry number.
1.139	21-Oct-2015	msaitoh	Fix compile error...
1.138	21-Oct-2015	msaitoh	Decode SATA Capability and Multicast Extendeded Capability.
1.137	03-Oct-2015	joerg	Fix format string.
1.136	02-Oct-2015	msaitoh	Fix a bug that the TPH ST Table is wrongly decoded. Found by llvm.
1.135	02-Oct-2015	msaitoh	PCI Extended Configuration stuff written by nonaka@: - Add PCI Extended Configuration Space support into x86. - Check register offset of pci_conf_read() in MD part. It returns (pcireg_t)-1 if it isn't accessible. - Decode Extended Capability in PCI Extended Configuration Space. Currently the following extended capabilities are decoded: - Advanced Error Reporting - Virtual Channel - Device Serial Number - Power Budgeting - Root Complex Link Declaration - Root Complex Event Collector Association - Access Control Services - Alternative Routing-ID Interpretation - Address Translation Services - Single Root IO Virtualization - Page Request - TPH Requester - Latency Tolerance Reporting - Secondary PCI Express - Process Address Space ID - LN Requester - L1 PM Substates The following extended capabilities are not decoded yet: - Root Complex Internal Link Control - Multi-Function Virtual Channel - RCRB Header - Vendor Unique - Configuration Access Correction - Multiple Root IO Virtualization - Multicast - Resizable BAR - Dynamic Power Allocation - Protocol Multiplexing - Downstream Port Containment - Precision Time Management - M-PCIe - Function Reading Status Queueing - Readiness Time Reporting - Designated Vendor-Specific
1.134	27-Jul-2015	msaitoh	Add NVMe.
1.133	24-Nov-2014	msaitoh	branches: 1.133.2; Add PCIe CRS Software Visibility bit.
1.132	23-Oct-2014	msaitoh	- Cleanup pci_conf_print_caplist. Use table. The reason why it loops twice is that some capabilities appear multiple times (e.g. HyperTransport cap). - Print the specification revision of Power Management and AGP not in the capability list part but in the detail part.
1.131	23-Oct-2014	msaitoh	Fix a bug that the specification revision of the Power Management function was incorrectly printed in the output of capability "list". The value is also printed in the detail output and it has no bug.
1.130	23-Oct-2014	msaitoh	Fix typo in comment.
1.129	06-Oct-2014	msaitoh	s/genric/generic/
1.128	06-Oct-2014	msaitoh	- Add some PCI subclass and interfaces. - The interface of PCI_SUBCLASS_BRIDGE_RACEWAY is not decoded yet. - Fix typo in a message. - Add comment. - Modify comment.
1.127	22-Sep-2014	msaitoh	Always print the Slot implemented bit in the PCIe Capabilities Register using with onoff().
1.126	21-Sep-2014	christos	Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.125	05-Sep-2014	matt	Don't use class or typename as a variable name.
1.124	09-Jun-2014	msaitoh	branches: 1.124.2; Add IOMMU and the Root Complex Event Collector.
1.123	30-May-2014	msaitoh	- Remove some obsoleted comments. - KNF.
1.122	30-May-2014	msaitoh	- Add PCI-X capability stuff. - remove extra ':' in pci_conf_print_pcie_cap() - Add comments.
1.121	27-May-2014	msaitoh	Print "range: not set" if the decode window isn't set.
1.120	27-May-2014	msaitoh	- Add some register definition for MSI and MSI-X - print MSI-X capability
1.119	25-May-2014	njoly	Do not crash if subclassp == NULL, seen while attaching rstx(4) which match PCI_CLASS_UNDEFINED class.
1.118	24-May-2014	msaitoh	Print some PCI Capabilities: - Vendor specific (ID:0x09) - Debugport (ID:0x0a) - Subsystem (ID:0x0d) - PCI Advanced Features (ID:0x13)
1.117	24-May-2014	msaitoh	- Decode the programming interface field in the Class Code register and print it. - Print the cache line size in bytes. - Print the Link Status 2 register itself. - Some bits were not printed if the bit is 0. Always print them using with onoff() macro. - Print more bits. - KNF. - Use macro. - Add comments.
1.116	24-May-2014	msaitoh	- Fix calculation of supported max payload size in PCIe device capability register. - Fix for PCIE_SLCSR_DLLSCE flag. - invert PCIE_SLCSR_PCC
1.115	23-May-2014	msaitoh	No functional change: - KNF - Sort in PCI capability ID order. - Add comments.
1.114	23-May-2014	msaitoh	- Add some register definitions (subclass, power management, etc.) - Print some information (subclass, power management) - Use macro.
1.113	23-May-2014	msaitoh	PME# clock is not bit 2 but bit 3. Use the macro!
1.112	15-May-2014	msaitoh	Use onoff() macro.
1.111	13-May-2014	msaitoh	Print 32bit I/O region flag and 64bit memory region flag.
1.110	12-May-2014	msaitoh	Print the range of I/O, non-prefechable memory and prefechable memory.
1.109	12-May-2014	msaitoh	Use macro.
1.108	12-May-2014	msaitoh	Fix newline problem which was added in previous commit.
1.107	09-May-2014	msaitoh	Print the CRS Software Visibility Enable bit and the Crosslink Supported bit.
1.106	05-Aug-2013	msaitoh	branches: 1.106.2; One more Gb/s -> GT/s fix.
1.105	21-Apr-2013	msaitoh	branches: 1.105.4; - Print PCIe 2.0 or higher capability registers. - Print Link related registers only if the device is PCI Express Endpoint, Legacy PCI Express Endpoint or Root Port of PCI Express Root Complex. - Don't print Root related registers if the device is Root Complex Integrated Endpoint and print if the device is Root Complex Event Collector. - Not Gb/s but GT/s.
1.104	21-Apr-2013	msaitoh	- Add some PCIe 2.0 or higher capability register definitions. - Rename some registers. - Add comments. - Indent.
1.103	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.102	17-Apr-2013	msaitoh	Don't check whether PCIe Slot Control Register is all 0 or not. For example, 82801I PCI Express Port #1 (devid 0x2940) is really Root Port and it has the Root Control Register and the default value is 0 (the document say so and really 0 (Tested with my machine)).
1.101	17-Apr-2013	msaitoh	- Add slot related registers - Add root port related registers - Fix the definition of PCI_PCIE_SLCAP_PSN - Cleanup
1.100	17-Apr-2013	msaitoh	Add PCI_CAP_SUBVENDOR (= 0x0d).
1.99	16-Apr-2013	msaitoh	Decode some PCIe capability register bits.
1.98	16-Apr-2013	msaitoh	Use macro.
1.97	16-Apr-2013	msaitoh	Fix a bug that IRQ(MSI) bits in PCIe capability register is incorrectly decoded. The bit field is not 0x4e000000 but 0x3e000000.
1.96	15-Apr-2013	msaitoh	Print some PCIE device type.
1.95	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.94	20-Oct-2012	matt	Add NVM to the pci mass storage subclass
1.93	23-Sep-2012	chs	add an entry for subclass HD audio.
1.92	24-Apr-2012	drochner	branches: 1.92.2; print correct link speed for PCIexpress Gen2+ (the decoding code needs to be rewritten, sorry for only adding to the mess)
1.91	21-Mar-2012	matt	Use an unsigned value to hold the pci_class
1.90	29-Jan-2012	drochner	branches: 1.90.2; extend the pci_aprint_devinfo slightly to cover the cases commonly used by drivers: a short name for the quiet/naive case and a string to override the "pcidevs" based name by one provided by the driver, ride on yesterday's kernel minor version bump
1.89	26-Jan-2012	drochner	put printing of the pci_devinfo into its own function (not inlined by purpose) - this is a stack hog, and with this change my uTCA amd64 system boots again a lot of similar code can be eliminated from pci device drivers this way, but before doing so (and making the new function part of the module API) I'd like to consider a modification to make it work with drivers which prefer to print names from other sources (like pciide)
1.88	17-Aug-2011	dyoung	branches: 1.88.2; 1.88.6; Redefine PCI_MSI_* and PCI_PCIE_* constants in terms of bits(3). Use named constants and more conventional variable names in pci_msi_establish() and pci_msi_disestablish(). Fix a couple of bugs: pci_msi_establish() returned a pointer to the struct intrhand instead of to the struct msi_hdl as it was intended to, and pci_msi_disestablish() did not free(9) the msi_hdl.
1.87	06-Jun-2011	msaitoh	Add two new capabilities(PCI_CAP_SATA and PCI_CAP_PCIAF).
1.86	11-Dec-2010	matt	branches: 1.86.6; Add code to display MSI capabilities.
1.85	21-Aug-2010	pgoyette	Update the various xxx_verbose modules to conform to the module subsystem's new locking protocol.
1.84	25-Jul-2010	pgoyette	Move setting of pci_verbose_loaded flag into the module's init routine. This ensures that the flag is set even if the module was manually loaded by the user rather than just auto-loaded.
1.83	06-Jun-2010	pgoyette	Update pciverbose module to use module_autoload() rather than module_load(). Load the module right before each attempt to use its features, and let the module subsystem handle unloading.
1.82	26-May-2010	martin	Do not try to unload the pciverbose module if it had never been loaded.
1.81	25-May-2010	pgoyette	Rework the pciverbose module dispatch vectors to avoid renaming the externally-visible entrypoint name. Also this avoids a potential need to bump kernel version. Requested by dyoung@ and mrg@
1.80	24-May-2010	pgoyette	Extract the vendor/product tables and related access routines into a separate kernel module. Update pci bus attach routine to load the module (if available) when we're about to start scanning the bus, and unload the module after the scan is finished. On architectures which support loading of modules by the boot loader, the 'pciverbose' module can be loaded and executed without needing to rebuild the kernel. On all architectures, using 'options PCIVERBOSE' in the kernel configuration file will create a 'builtin' module which is functionally equivalent to previous behavior. XXX Although not nearly as large as the vendor and product tables, XXX the PCI class and subclass tables might also be offloaded into XXX the module at a future time. XXX Cardbus (and possibly other) drivers should also be modified to XXX load the module before scanning/attaching devices.
1.79	04-Mar-2010	dyoung	branches: 1.79.2; Extract both instances of the code that prints the PCI Secondary Status Register into a subroutine. Use terminology similar to that which PCI System Architecture (4th Edition) uses. For comparison with PCI reference books, specify flags and fields with __BIT(n) and __BITS(m, n) instead of hexadecimal constants.
1.78	13-Jan-2010	drochner	branches: 1.78.2; fix some bitmask in PCI Express capabilities decoding being here, fix some typos and add a comma for clarity
1.77	20-Jan-2009	jmcneill	Decode and display PCI Power Management registers when available.
1.76	17-Nov-2008	matt	Store pcidevs in a more compact manner. Instead of many string with duplicate words, store each word individual in a long string and then store offsets in the string to that word. This reduces the space needed by half.
1.75	29-Apr-2008	jmcneill	branches: 1.75.6; 1.75.8; 1.75.10; Oh, so that's how you spell Legacy -- thanks spellcheck!
1.74	10-Apr-2008	cegger	branches: 1.74.2; 1.74.4; use aprint_*_dev and device_xname
1.73	19-Oct-2007	ad	branches: 1.73.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.72	07-Oct-2007	joerg	branches: 1.72.2; Add some basic parsing of the PCI Express capabilities register. This is enough to help detect the source of interrupt storms at least for my Thinkpad.
1.71	16-Nov-2006	christos	branches: 1.71.8; 1.71.22; 1.71.24; 1.71.26; __unused removal on arguments; approved by core.
1.70	08-Nov-2006	drochner	while we are here, also print the interrupt disable bit in VERBOSE/pcictl
1.69	17-Oct-2006	dogcow	now that we have -Wno-unused-parameter, back out all the tremendously ugly code to gratuitously access said parameters.
1.68	13-Oct-2006	christos	more __unused
1.67	13-Oct-2006	dogcow	more unused variable fallout.
1.66	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.65	03-Sep-2006	christos	branches: 1.65.2; 1.65.4; add missing initializers
1.64	31-May-2006	drochner	-add 2 subclasses new in rev. 3.0 of the spec, and fix a pasto in another -get power management rev printing right
1.63	11-Dec-2005	christos	branches: 1.63.4; 1.63.6; 1.63.8; 1.63.14; merge ktrace-lwp.
1.62	30-Oct-2005	simonb	Only include <sys/systm.h> if _KERNEL is defined.
1.61	28-Jun-2005	thorpej	branches: 1.61.2; 1.61.4; Use ANSI function decls and static.
1.60	27-Feb-2005	perry	nuke trailing whitespace
1.59	02-Aug-2004	mycroft	branches: 1.59.4; 1.59.6; For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.58	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.57	04-Feb-2004	soren	Use the right bits for the AGP version.
1.56	29-Jun-2003	fvdl	branches: 1.56.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.55	29-Jun-2003	jdolecek	fix typo in comment
1.54	24-Nov-2002	scw	Quell an uninitialised variable warning.
1.53	21-Sep-2002	drochner	call the capability list ptr by its name
1.52	21-Sep-2002	drochner	correct caplist head for Cardbus bridges
1.51	21-Sep-2002	drochner	-corrected "MSI" capability -added some new subclasses and capabilities -move capability list printing into a separate function and call it for each header type (not type 0 only)
1.50	19-May-2002	sommerfeld	Fortunately for us, all interrupt controllers are not 8259's
1.49	03-May-2002	nathanw	branches: 1.49.2; Make pci_conf_print() dump the device-dependent header in userspace, too.
1.48	22-Mar-2002	drochner	iadd subclass codes from the 2.3 spec
1.47	13-Nov-2001	lukem	add RCSID
1.46	13-Nov-2001	enami	Include stdio.h when compiled in userland to pull in declaration of printf etc.
1.45	13-Sep-2001	thorpej	Allow this file to be used in userspace, as well.
1.44	13-Sep-2001	thorpej	ANSI'ify.
1.43	09-Sep-2001	enami	Fix sample code in comment; pci_conf_print() takes three arguments.
1.42	18-Jan-2001	jdolecek	branches: 1.42.2; 1.42.4; 1.42.6; adapt to pci_knowndevs[] being const, make local constant arrays const
1.41	21-Nov-2000	soren	s/iRDA/IrDA/
1.40	07-Oct-2000	cgd	Update for current PCI device class/subclass and capability codes. (also, tweak the I2O subclass string to be "standard" -- the removal of version info didn't extend that far.)
1.39	02-Oct-2000	ad	I2O subclasses currently have nothing to do with the protocol version.
1.38	02-Sep-2000	cgd	don't attempt to size Host Bridge BARs. (it's a good way to die.) provide a clue about 16-bit vs. 32-bit I/O BARs.
1.37	03-Aug-2000	nathanw	When performing pci_config_dump: - Display the full value of 64-bit BARs. - Avoid displaying the upper 32 bits of 64-bit BARs as a separate 32-bit BAR.
1.36	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.35	22-Mar-2000	cgd	branches: 1.35.4; opt_pciverbose.h -> opt_pci.h (it's gonna be used by more things, soon)
1.34	25-Jan-2000	drochner	use BUS_SPACE_MAP_PREFETCHABLE instead BUS_SPACE_MAP_CACHEABLE where the PCI BAR bit is referred to
1.33	15-Dec-1999	kleink	Report the Capability List support bit of the SR and, if set, print the the Capability list pointer register as such as well as the capabilities implemented in the list.
1.32	27-Sep-1999	cgd	branches: 1.32.2; 1.32.8; add classes/subclasses new in PCI 2.2. Needs a bit of cleanup, but then, so does everything involving configuration space headers and if i don't get this out of my source tree i'll go insane.
1.31	21-Dec-1998	drochner	use a symbolic definition for the PCI subsystem ID
1.30	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.29	12-Jul-1998	augustss	Add USB support. Supported so far: * UHCI and OHCI host controllers on PCI * Hubs * HID devices withe special drivers for mouse and keyboard * Printers
1.28	26-Jun-1998	cgd	factor out some common code, and enhance functionality slightly: allow the BAR-printing function to print a name for the register, factor out a common register-bits function which can handle the fact that type 2 headers have a different size than is usual, and actually do something useful with the rest of the bits in the type 2 header.
1.27	28-May-1998	cgd	Clean up slightly, abstract pci_conf_print family's base address register printing into a function, add a bit more pretty-printing of existing stuff. Implement pretty-printers for type 1 and type 2 headers. (Right now, these are just quick stabs based on some on-line bridge docs that I have handy on my laptop. Mmmm, meetings. I'll check the bits when I get back within reach of my official docs.)
1.26	18-May-1998	cgd	largely reimplement pci_conf_print(): * print all configuration space registers. Then, where possible, interpret them. (That is, PRESENT ALL THE DATA, then interpret it -- don't hide data behind interpretation. Also, when interpreting fields, try to print out the specific value that's being interpreted.) * handle different header types. * allow caller to specify a function which can interpret the device-dependent header and is responsible for pretty-printing it. It spews (use 'options MSGBUFSIZE=...' 8-), but when you want the data, you really want _all_ of it. Still needs some cleanup and additional code (e.g. interepretation of PCI-PCI (type 1) and PCI-Cardbus (type 2(?)) bridge headers).
1.25	03-May-1998	thorpej	In pci_conf_print(), don't print a mapping register if the size is 0.
1.24	03-May-1998	thorpej	Properly compute the region sizes in pci_conf_print(). Bug pointed out by Chris Demetriou.
1.23	20-Apr-1998	drochner	Avoid "unused variable" warning without PCIVERBOSE. From mynhier@cs.utk.edu (PR kern/5328).
1.22	14-Apr-1998	thorpej	Add pci_conf_print(), a function to dump the PCI configuration space, useful in driver debugging. From Zubin D. Dittia <zubin@clouseau.arl.wustl.edu>, PR #4249.
1.21	13-Sep-1997	enami	Declare PCIVERBOSE by defopt in files.pci. Include opt_pciverbose.h in pci_subr.c.
1.20	19-Mar-1997	cgd	branches: 1.20.4; fix multi-function device support, add new known classes/subclasses, and clean up class/subclass printing. From brb@brig.com via PR 3359.
1.19	13-Oct-1996	christos	backout kprintf changes
1.18	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.17	02-Mar-1996	cgd	remove some colons to be slightly more friendly to things trying to parse autoconfiguration output.
1.16	02-Mar-1996	cgd	oops; spaces -> tabs
1.15	02-Mar-1996	cgd	if not PCIVERBOSE, don't say that vendor/product is "unknown," just say what it is.
1.14	28-Feb-1996	cgd	make PCI bus match/attach and sub-device attachment machine-independent.
1.13	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.12	16-Aug-1995	cgd	check for null subclass pointer. From Todd Mummert
1.11	21-Jun-1995	cgd	fix reference to incorrect struct (paste-o)
1.10	18-Jun-1995	cgd	define and call "pci_devinfo()", which creates a string from the vendor & product IDs and class information, which is printed if device isn't found. Optionally (via "PCIVERBOSE" option) does table lookup to try to see if it knows what the device really is (informational only...).
1.9	17-Jun-1995	cgd	more appropriate name, as this only has PCI support subroutines.
1.8	23-May-1995	cgd	split single-subdevice lookup & attachment into a subroutine (pci_attach_subdev()). remove pciattach() function and the pcicd cfdriver struct, the former because thre are a lot of attachment actions which really are machine-dependent (perhaps even "most"), and the latter because now that both pcimatch() and pciattach() are machine-dependent it's bad style to declare them here and it gains nothing.
1.7	27-Jan-1995	cgd	include files from the correct places.
1.6	04-Nov-1994	mycroft	Make a wrapper match function to check the bus and device numbers, rather than insisting that every driver do it.
1.5	03-Nov-1994	mycroft	Rename pciprobe() to pcimatch(), and move it to pci_machdep.c.
1.4	03-Nov-1994	mycroft	Always use direct configuration.
1.3	27-Oct-1994	cgd	new RCS ID format.
1.2	10-Aug-1994	mycroft	Update some comments.
1.1	09-Aug-1994	mycroft	branches: 1.1.2; Add PCI autoconfiguration support.
1.1.2.1	10-Aug-1994	mycroft	update from trunk
1.20.4.1	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.32.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.32.2.3	11-Feb-2001	bouyer	Sync with HEAD.
1.32.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.32.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.35.4.2	25-Oct-2001	he	Pull up revisions 1.39-1.40 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.35.4.1	10-Aug-2000	soda	Pull up to netbsd-1-5 branch Approved by: thorpej When performing pci_config_dump: Display the full value of 64-bit BARs. Avoid displaying the upper 32 bits of 64-bit BARs as a separate 32-bit BAR. by Nathan J Williams <nathanw@netbsd.org> Revisions pulled up: > cvs rdiff -r1.36 -r1.37 syssrc/sys/dev/pci/pci_subr.c
1.42.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.42.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.42.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.42.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.42.4.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.42.2.6	11-Dec-2002	thorpej	Sync with HEAD.
1.42.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.42.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.42.2.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.42.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.42.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.49.2.1	30-May-2002	gehenna	Catch up with -current.
1.56.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.56.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.56.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.56.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.56.2.1	03-Aug-2004	skrll	Sync with HEAD
1.59.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.59.4.1	29-Apr-2005	kent	sync with -current
1.61.4.1	02-Nov-2005	yamt	sync with head.
1.61.2.3	27-Oct-2007	yamt	sync with head.
1.61.2.2	30-Dec-2006	yamt	sync with head.
1.61.2.1	21-Jun-2006	yamt	sync with head.
1.63.14.1	19-Jun-2006	chap	Sync with head.
1.63.8.2	03-Sep-2006	yamt	sync with head.
1.63.8.1	26-Jun-2006	yamt	sync with head.
1.63.6.1	01-Jun-2006	kardel	Sync with head.
1.63.4.1	09-Sep-2006	rpaulo	sync with head
1.65.4.2	10-Dec-2006	yamt	sync with head.
1.65.4.1	22-Oct-2006	yamt	sync with head
1.65.2.1	18-Nov-2006	ad	Sync with head.
1.71.26.1	14-Oct-2007	yamt	sync with head.
1.71.24.1	06-Nov-2007	matt	sync with HEAD
1.71.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.71.8.2	23-Oct-2007	ad	Sync with head.
1.71.8.1	09-Oct-2007	ad	Sync with head.
1.72.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.73.16.2	17-Jan-2009	mjf	Sync with HEAD.
1.73.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.74.4.5	09-Oct-2010	yamt	sync with head
1.74.4.4	11-Aug-2010	yamt	sync with head.
1.74.4.3	11-Mar-2010	yamt	sync with head
1.74.4.2	04-May-2009	yamt	sync with head.
1.74.4.1	16-May-2008	yamt	sync with head.
1.74.2.1	18-May-2008	yamt	sync with head.
1.75.10.1	24-Feb-2009	snj	branches: 1.75.10.1.4; 1.75.10.1.8; Pull up following revision(s) (requested by jmcneill in ticket #485): sys/dev/pci/pci_subr.c: revision 1.77 Decode and display PCI Power Management registers when available.
1.75.10.1.8.1	07-Jan-2011	matt	Add/define some MSI support
1.75.10.1.4.2	05-Nov-2013	matt	Pull in support for pci_aprint_devinfo_fancy
1.75.10.1.4.1	26-Dec-2011	matt	Merge -HEAD changes (except for PCIVERBOSE chagnes).
1.75.8.2	03-Mar-2009	skrll	Sync with HEAD.
1.75.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.75.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.78.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.78.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.78.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.79.2.4	12-Jun-2011	rmind	sync with head
1.79.2.3	05-Mar-2011	rmind	sync with head
1.79.2.2	03-Jul-2010	rmind	sync with head
1.79.2.1	30-May-2010	rmind	sync with head
1.86.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.88.6.3	29-Apr-2012	mrg	sync to latest -current.
1.88.6.2	05-Apr-2012	mrg	sync to latest -current.
1.88.6.1	18-Feb-2012	mrg	merge to -current.
1.88.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.88.2.3	30-Oct-2012	yamt	sync with head
1.88.2.2	23-May-2012	yamt	sync with head.
1.88.2.1	17-Apr-2012	yamt	sync with head
1.90.2.1	05-Aug-2013	martin	Pullup sys/dev/pci/pcireg.h 1.74-1.82 and 1.84 via patch sys/dev/pci/pci_subr.c 1.92-1.102, 1.104-1.105 via patch Add some PCI(e) register and bit definitions in pcireg.h. Fix the definition of PCI_PCIE_SLCAP_PSN. Fix a bug that IRQ(MSI) bits in PCIe capability register is incorrectly decoded. Print more registers in "pcictl dump". Fix bug in comment. Requested by msaitoh in ticket #928
1.92.2.4	03-Dec-2017	jdolecek	update from HEAD
1.92.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.92.2.2	23-Jun-2013	tls	resync from head
1.92.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.105.4.2	18-May-2014	rmind	sync with head
1.105.4.1	28-Aug-2013	rmind	sync with head
1.106.2.1	10-Aug-2014	tls	Rebase.
1.124.2.3	09-Dec-2016	snj	Pull up following revision(s) (requested by pgoyette in ticket #1296): sys/dev/pci/pci_subr.c: revision 1.155 via patch Protect against buffer overflow.
1.124.2.2	14-Dec-2014	martin	branches: 1.124.2.2.4; Pull up following revision(s) (requested by msaitoh in ticket #325): sys/dev/pci/pcireg.h: revision 1.100 sys/dev/pci/pci_subr.c: revision 1.133 Add PCIe CRS Software Visibility bit.
1.124.2.1	12-Dec-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #312): sys/dev/pci/pci_subr.c: revision 1.130 sys/dev/pci/pci_subr.c: revision 1.131 sys/dev/pci/pci_subr.c: revision 1.132 sys/dev/pci/pcireg.h: revision 1.97 sys/dev/pci/pcireg.h: revision 1.98 sys/dev/pci/pcireg.h: revision 1.99 sys/dev/pci/pci_subr.c: revision 1.127 sys/dev/pci/pci_subr.c: revision 1.128 sys/dev/pci/pci_subr.c: revision 1.129 Always print the Slot implemented bit in the PCIe Capabilities Register using with onoff(). - Add some PCI subclass and interfaces. - The interface of PCI_SUBCLASS_BRIDGE_RACEWAY is not decoded yet. - Fix typo in a message. - Add comment. - Modify comment. s/genric/generic/ Add comment. Fix typo in comment. Fix a bug that the specification revision of the Power Management function was incorrectly printed in the output of capability "list". The value is also printed in the detail output and it has no bug. - Cleanup pci_conf_print_caplist. Use table. The reason why it loops twice is that some capabilities appear multiple times (e.g. HyperTransport cap). - Print the specification revision of Power Management and AGP not in the capability list part but in the detail part. Add some HyperTransport related defines. It's required for the MSI.
1.124.2.2.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.133.2.8	28-Aug-2017	skrll	Sync with HEAD
1.133.2.7	05-Feb-2017	skrll	Sync with HEAD
1.133.2.6	05-Dec-2016	skrll	Sync with HEAD
1.133.2.5	05-Oct-2016	skrll	Sync with HEAD
1.133.2.4	09-Jul-2016	skrll	Sync with HEAD
1.133.2.3	29-May-2016	skrll	Sync with HEAD
1.133.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.133.2.1	22-Sep-2015	skrll	Sync with HEAD
1.148.2.4	26-Apr-2017	pgoyette	Sync with HEAD
1.148.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.148.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.148.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.156.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.178.2.2	11-May-2017	pgoyette	Sync with HEAD
1.178.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.183.2.14	29-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1728: sys/dev/pci/pci_subr.c 1.232-1.239 via patch sys/dev/pci/pcireg.h 1.62-1.63 - Decode link control2's Compliance Preset/De-emphasis more. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Fix typo.
1.183.2.13	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1714: sys/dev/pci/pcireg.h 1.148-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.217-1.222, 1.224, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158 sys/dev/pci/ppb.c 1.74 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. - Change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Remove unused shift and mask definitions. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.183.2.12	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1388: sys/dev/pci/pcireg.h 1.147 via patch sys/dev/pci/pci_subr.c 1.212, 1.215-1.217 via patch - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Remove whitespace for consistency. - Use macro. - Whitespace fixes.
1.183.2.11	17-Jul-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1295): sys/dev/pci/pci_subr.c: revision 1.213 sys/dev/pci/ppb.c: revision 1.69 Identify 16GT/s.
1.183.2.10	17-Jul-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1292): sys/dev/pci/pci_subr.c: revision 1.211 sys/dev/pci/pci_subr.c: revision 1.214 Typo (s/vaule/value/) Print extend capability from 0x25 to 0x29 correctly.
1.183.2.9	04-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1118): sys/dev/pci/pci_subr.c: revision 1.210 sys/dev/pci/pci_subr.c: revision 1.207 sys/dev/pci/pcireg.h: revision 1.143 sys/dev/pci/pci_subr.c: revision 1.208 sys/dev/pci/pcireg.h: revision 1.144 sys/dev/pci/pci_subr.c: revision 1.209 sys/dev/pci/pcireg.h: revision 1.145 sys/dev/pci/pcireg.h: revision 1.146 Decode PCI Enhanced Allocation. The register offset of the mask and pending register is depend on the 64bit address capable bit, so fix the definition of PCI MSI vector mask and pending register. This problem was not a real bug because PCI_MSI{MASK,PENDING} were not used from anywhere. The downstream port of PCIe switch is not a root port, so don't print root port related register. For example, Intel 63xxESB controller's downstream port device was printed by pcictl(8) with this bug:
1.183.2.8	30-Oct-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1074): sys/dev/pci/pci_subr.c: revision 1.204-1.206 sys/dev/pci/pcireg.h: revision 1.141-1.142 Root Complex Event Collector Bus Number Association ECN. - If capability version is 2 (or greater), decode RCEC Associated Bus Numbers register. - Don't print TPH requester's ST Table Size if the ST table location field is not PCI_TPH_REQ_STTBLLOC_TPHREQ because the size field is only applicable for PCI_TPH_REQ_STTBLLOC_TPHREQ case. - Add comment.
1.183.2.7	23-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1028): sys/dev/pci/pci_subr.c: revision 1.203 sys/dev/pci/pcireg.h: revision 1.140 Add ATS Relaxed Ordering supported bit described in Address Translation Relaxed Ordering ECN.
1.183.2.6	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #933): sys/dev/pci/pci_subr.c: revision 1.202 sys/dev/pci/pcireg.h: revision 1.139 sys/dev/pci/ppbreg.h: revision 1.9 VGA 16 bit decode bit is not bit 3 but bit 4. -- - Print Power Management Control/status register in 32bit. - Simplify.
1.183.2.5	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #930): sys/dev/pci/pci_subr.c: revision 1.201 sys/dev/pci/pcireg.h: revision 1.138 Fix typo. s/TPL/TLP/
1.183.2.4	26-Feb-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #576): sys/dev/pci/pci_subr.c: 1.197-1.200 sys/dev/pci/pcireg.h: 1.136-1.137 sys/dev/pci/ppbreg.h: 1.8 Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified. "s/above 300W/greater than 300W/" in pci_conf_print_pcie_power(). From PCIe Base Spec 3.1a Errata 2017-12-13. Cleanup: - Don't pass a capability pointer as a argument of pci_conf_find_cap() and determine the first pointer in the pci_conf_find_cap() function. - Don't pass a capability pointer as a argument of pci_conf_find_extcap() because it's not used. - Remove unsed code. - Add PCie Link Activation ECN. - Use macro. - KNF.
1.183.2.3	22-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #371): sys/dev/pci/pci_subr.c: revision 1.195 sys/dev/pci/pci_subr.c: revision 1.196 Print Error Source Identification register correctly. Whitespace fix.
1.183.2.2	21-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/pcireg.h: revision 1.133 sys/dev/pci/pcireg.h: revision 1.134 sys/dev/pci/pcireg.h: revision 1.135 sys/dev/pci/pci_subr.c: revision 1.190 sys/dev/pci/pci_subr.c: revision 1.191 sys/dev/pci/pci_subr.c: revision 1.192 sys/dev/pci/pci_subr.c: revision 1.193 sys/dev/pci/pci_subr.c: revision 1.194 sys/dev/pci/pcireg.h: revision 1.132 - Official shortname of LN Requester is LNR, so change PCI_EXTCAP_LN_REQ to PCI_EXTCAP_LNR - Use macro. - Add PCI_MAPREG_ROM_ADDR_MASK macro and PCI_MAPREG_ROM_ADDR() macro. - print PCI_MAPREG_ROM_ENABLE bit. - Decode Expansion ROM Validation ECN. - Add Native PCIe Enclosure Management ECN's extended capability type. Not decoded yet. Decode IOMMU capability of PCI secure device capability. From "AMD I/O Virtualization Technology(IOMMU) Specification (#48882) Revision 3.00". IOMMU cap dump fixes: - Print Capability Register's value. - Indent output correctly. - s/cahced/cached/ - Print MSI Message number with 0x%02x Fix a bug that the TPH ST table is decoded even if it's not in the TPH Requester extended capability structure.
1.183.2.1	04-Jul-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #80): sys/dev/pci/pci_subr.c: revision 1.184 sys/dev/pci/pci_subr.c: revision 1.185 sys/dev/pci/pci_subr.c: revision 1.186 sys/dev/pci/pci_subr.c: revision 1.187 sys/dev/pci/pci_subr.c: revision 1.188 sys/dev/pci/pci_subr.c: revision 1.189 sys/dev/pci/pcireg.h: revision 1.131 Add missing return to print the Slot Power Limit Value correctly. Fix to print the following bit fields correctly. - Supported Link Speeds Vector in LCAP2 - Lower SKP OS Generation Supported Speed Vector in LCAP2 - Lower SKP OS Reception Supported Speed Vector in LCAP2 - Enable Lower SKP OS Generation Vector in LCTL3 Note that the above bitfields start from 0 and the follwing bitfields start from 1: - Maximum Link Speed in LCAP - Current Link Speed in LCSR - Target Link Speed in LCSR2 For the Target Link Speed in LCSR2, 0 is allowed for a device which supports 2.5GT/s only (and this check also works for devices which compliant to versions of the base specification prior to 3.0. Tested with BCM5709: - Target Link Speed: unknown value (0) + Target Link Speed: 2.5GT/s For Attention Indicator Control bit and Power Indicator Control bit, it's allowed to be a read only value 0 if corresponding capability register bit is 0. Fix a bug that LTR's latency in L1 PM Substates capability and Latency Tolerance Reporting capability isn't printed correctly. Fix printf format/argument.
1.200.2.6	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.200.2.5	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.200.2.4	20-Oct-2018	pgoyette	Sync with head
1.200.2.3	30-Sep-2018	pgoyette	Ssync with HEAD
1.200.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.200.2.1	21-May-2018	pgoyette	Sync with HEAD
1.201.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.201.2.1	10-Jun-2019	christos	Sync with HEAD
1.215.2.6	29-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1412: sys/dev/pci/pci_subr.c 1.232-1.239 via patch sys/dev/pci/pcireg.h 1.62-1.63 - Decode link control2's Compliance Preset/De-emphasis more. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Fix typo.
1.215.2.5	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1384: sys/dev/pci/pcireg.h 1.152-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.222, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158, 1.163 sys/dev/pci/ppb.c 1.74 - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.215.2.4	26-Jul-2020	martin	Pull up following revision(s) (requested by jdolecek in ticket #1024): sys/dev/pci/pci_subr.c: revision 1.224 change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack
1.215.2.3	19-Mar-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #782): sys/dev/pci/pcireg.h: revision 1.150 sys/dev/pci/pcireg.h: revision 1.151 sys/dev/pci/pci_subr.c: revision 1.220 sys/dev/pci/pci_subr.c: revision 1.221 sys/dev/pci/pcireg.h: revision 1.149 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Avoid using magic number. Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. Remove unused shift and mask definitions. Add comment.
1.215.2.2	21-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #629: sys/dev/pci/pcireg.h 1.148 sys/dev/pci/pci_subr.c 1.218-1.219 - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - Fix typo.
1.215.2.1	26-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #243): sys/dev/pci/pci_subr.c: revision 1.216 sys/dev/pci/pci_subr.c: revision 1.217 Whitespace fixes. No functional change. - Print some DPC register values not with %04x but with %08x because those are 32bit.
1.218.2.2	25-Jan-2020	ad	Sync with head.
1.218.2.1	17-Jan-2020	ad	Sync with head.
1.222.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.224.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.225.4.1	01-Aug-2021	thorpej	Sync with HEAD.
1.31	05-Sep-2021	mrg	make sure PCI_IOC_DRVNAMEONBUS doesn't access NULL pointers. hopefully fixes PR#55744. it's not entirely clear why this does not affect most systems, as there's usually some PCI devices that do not have attached drivers, eg: 006:00:0: AMD Family17h/7xh Reserved SPP (non-essential instrumentation, subclass 0x00) on my system i'm testing on. XXX: pullup-8, pullup-9.
1.30	24-Sep-2016	mrg	branches: 1.30.6; introduce a better pci_drvname() and PCI_IOC_DRVNAME as pciio_drvnameonbus() and PCI_IOC_DRVNAMEONBUS. the new ones also take a (autoconf) PCI bus number, which allows lookups for any device on any pci bus node. use this in pcictl which current reports the wrong values sometimes. up next: use these in libpciaccess.
1.29	24-Aug-2015	pooka	branches: 1.29.2; add some _KERNEL_OPT as the finishing touch
1.28	25-Jul-2014	dholland	branches: 1.28.4; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.27	25-Jul-2014	mrg	introduce a new pci_drvname(3) to libpci. implement it in the backend via PCI_IOC_DRVNAME. update manual and set lists (and remove a couple of doubled entries.)o this will be used in libpciaccess() to implement the has_kernel_driver() method.
1.26	28-May-2014	riastradh	Reject unaligned PCI config register ioctl requests before we kassert.
1.25	16-Mar-2014	dholland	branches: 1.25.2; Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.24	13-Mar-2012	elad	branches: 1.24.2; 1.24.4; Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.23	10-Feb-2011	jmcneill	branches: 1.23.4; 1.23.8; 1.23.10; 1.23.14; 1.23.16; pcimmap: if the requested page is marked prefetchable in a child device's BAR, pass the BUS_SPACE_MAP_PREFETCHABLE flag down to bus_space_mmap
1.22	30-Jul-2009	macallan	branches: 1.22.4; 1.22.6; 1.22.8; do as elad suggested and call kauth instead of using INSECURE in pcimmap()
1.21	14-Jul-2009	macallan	nuke PCI_ALLOW_MMAP, instead allow mmap(/dev/pci*) with options INSECURE now the x86 camp should be happy again
1.20	09-Jul-2009	macallan	enable mmap() support for /dev/pci* with options PCI_ALLOW_MMAP
1.19	07-Jul-2009	christos	simplify previous.
1.18	04-Jul-2009	cegger	pciioctl(), pci_devioctl(): simplify implementations of PCI_IOC_BDF_CFGREAD, PCI_IOC_BDF_CFGWRITE, PCI_IOC_CFGREAD, PCI_IOC_CFGWRITE No functional changes.
1.17	04-Jul-2009	cegger	remove useless parenthesis
1.16	11-Jun-2008	cegger	branches: 1.16.6; 1.16.10; 1.16.14; remove unused variable
1.15	23-Mar-2008	cube	branches: 1.15.2; 1.15.4; 1.15.6; 1.15.8; Split device_t and softc for pci(4).
1.14	04-Mar-2007	christos	branches: 1.14.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.13	16-Nov-2006	christos	branches: 1.13.4; __unused removal on arguments; approved by core.
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	03-Sep-2006	christos	branches: 1.11.2; 1.11.4; add missing initializers
1.10	11-Dec-2005	christos	branches: 1.10.4; 1.10.8; merge ktrace-lwp.
1.9	28-Jun-2005	thorpej	branches: 1.9.2; Use ANSI function decls and static.
1.8	29-Jun-2003	fvdl	branches: 1.8.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.7	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.6	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.5	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.4	17-Mar-2002	atatat	branches: 1.4.4; Convert ioctl code to use EPASSTHROUGH instead of -1 or ENOTTY for indicating an unhandled "command". ERESTART is -1, which can lead to confusion. ERESTART has been moved to -3 and EPASSTHROUGH has been placed at -4. No ioctl code should now return -1 anywhere. The ioctl() system call is now properly restartable.
1.3	13-Nov-2001	lukem	branches: 1.3.2; add RCSID
1.2	13-Sep-2001	thorpej	branches: 1.2.2; 1.2.4; Oops, didn't intend to enable pcimmap() yet.
1.1	13-Sep-2001	thorpej	Add an ioctl interface to the PCI bus. Add ioctls to read/write PCI configuration space registers, and to fetch bus info.
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	13-Sep-2001	fvdl	file pci_usrreq.c was added on branch thorpej-devvp on 2001-10-01 12:45:58 +0000
1.2.2.6	11-Nov-2002	nathanw	Catch up to -current
1.2.2.5	17-Sep-2002	nathanw	Catch up to -current.
1.2.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	13-Sep-2001	nathanw	file pci_usrreq.c was added on branch nathanw_sa on 2001-09-21 22:35:59 +0000
1.3.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	13-Nov-2001	thorpej	file pci_usrreq.c was added on branch kqueue on 2002-01-10 19:56:55 +0000
1.4.4.1	16-May-2002	gehenna	Add the character device switch.
1.8.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.8.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.8.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.8.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.9.2.4	24-Mar-2008	yamt	sync with head.
1.9.2.3	03-Sep-2007	yamt	sync with head.
1.9.2.2	30-Dec-2006	yamt	sync with head.
1.9.2.1	21-Jun-2006	yamt	sync with head.
1.10.8.1	03-Sep-2006	yamt	sync with head.
1.10.4.1	09-Sep-2006	rpaulo	sync with head
1.11.4.2	10-Dec-2006	yamt	sync with head.
1.11.4.1	22-Oct-2006	yamt	sync with head
1.11.2.1	18-Nov-2006	ad	Sync with head.
1.13.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.14.36.2	29-Jun-2008	mjf	Sync with HEAD.
1.14.36.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.8.1	18-Jun-2008	simonb	Sync with head.
1.15.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.15.4.3	19-Aug-2009	yamt	sync with head.
1.15.4.2	18-Jul-2009	yamt	sync with head.
1.15.4.1	04-May-2009	yamt	sync with head.
1.15.2.1	17-Jun-2008	yamt	sync with head.
1.16.14.1	21-Apr-2010	matt	sync to netbsd-5
1.16.10.1	23-Jul-2009	jym	Sync with HEAD.
1.16.6.4	27-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1922): sys/dev/pci/pci_usrreq.c 1.26 via patch Fix to make pci(4) reject unaligned configuration register reads and writes before feeding them to a kassert in pci_conf_read/write or to a trap in the hardware itself.
1.16.6.3	15-Sep-2009	snj	branches: 1.16.6.3.2; 1.16.6.3.6; Pull up following revision(s) (requested by mrg in ticket #952): sys/dev/pci/pci_usrreq.c: revision 1.22 via patch do as elad suggested and call kauth instead of using INSECURE in pcimmap()
1.16.6.2	15-Sep-2009	snj	Pull up following revision(s) (requested by mrg in ticket #952): sys/dev/pci/pci_usrreq.c: revision 1.21 via patch nuke PCI_ALLOW_MMAP, instead allow mmap(/dev/pci*) with options INSECURE now the x86 camp should be happy again
1.16.6.1	15-Sep-2009	snj	Pull up following revision(s) (requested by mrg in ticket #952): sys/dev/pci/pci_usrreq.c: revision 1.20 via patch enable mmap() support for /dev/pci* with options PCI_ALLOW_MMAP
1.16.6.3.6.1	27-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1922): sys/dev/pci/pci_usrreq.c 1.26 via patch Fix to make pci(4) reject unaligned configuration register reads and writes before feeding them to a kassert in pci_conf_read/write or to a trap in the hardware itself.
1.16.6.3.2.1	27-Aug-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1922): sys/dev/pci/pci_usrreq.c 1.26 via patch Fix to make pci(4) reject unaligned configuration register reads and writes before feeding them to a kassert in pci_conf_read/write or to a trap in the hardware itself.
1.22.8.1	17-Feb-2011	bouyer	Sync with HEAD
1.22.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.22.4.1	05-Mar-2011	rmind	sync with head
1.23.16.1	13-Sep-2014	snj	Pull up following revision(s) (requested by riastradh in ticket #1124): sys/dev/pci/pci_usrreq.c: revision 1.26 Reject unaligned PCI config register ioctl requests before we kassert.
1.23.14.1	13-Sep-2014	snj	Pull up following revision(s) (requested by riastradh in ticket #1124): sys/dev/pci/pci_usrreq.c: revision 1.26 Reject unaligned PCI config register ioctl requests before we kassert.
1.23.10.1	13-Sep-2014	snj	Pull up following revision(s) (requested by riastradh in ticket #1124): sys/dev/pci/pci_usrreq.c: revision 1.26 Reject unaligned PCI config register ioctl requests before we kassert.
1.23.8.1	05-Apr-2012	mrg	sync to latest -current.
1.23.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.23.4.1	17-Apr-2012	yamt	sync with head
1.24.4.1	18-May-2014	rmind	sync with head
1.24.2.2	03-Dec-2017	jdolecek	update from HEAD
1.24.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.25.2.1	10-Aug-2014	tls	Rebase.
1.28.4.2	05-Oct-2016	skrll	Sync with HEAD
1.28.4.1	22-Sep-2015	skrll	Sync with HEAD
1.29.2.4	04-Nov-2016	pgoyette	Sync with HEAD
1.29.2.3	26-Jul-2016	pgoyette	Rename LOCALCOUNT_INITIALIZER to DEVSW_MODULE_INIT. This better describes what we're doing, and why.
1.29.2.2	19-Jul-2016	pgoyette	Instead of repeatedly typing the conditional initialization of the .d_localcount members in the various {b,c}devsw, define an initializer macro and use it. This also removes the need for defining new symbols for each 'struct localcount'. As suggested by riastradh@
1.29.2.1	18-Jul-2016	pgoyette	Rump drivers are always installed via devsw_attach() so we need to always allocate a 'struct localcount' for these drivers whenever they are built as modules.
1.30.6.2	29-Apr-2017	pgoyette	Remove more unnecessary #include for sys/localcount.h
1.30.6.1	27-Apr-2017	pgoyette	Restore all work from the former pgoyette-localcount branch (which is now abandoned doe to cvs merge botch). The branch now builds, and installs via anita. There are still some problems (cgd is non-functional and all atf tests time-out) but they will get resolved soon.
1.9	21-Sep-2014	christos	Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.8	29-Aug-2011	jmcneill	branches: 1.8.12; have pci register itself as a module, and make PCI driver modules depend on it
1.7	25-Jul-2010	pgoyette	branches: 1.7.2; 1.7.4; Move setting of pci_verbose_loaded flag into the module's init routine. This ensures that the flag is set even if the module was manually loaded by the user rather than just auto-loaded.
1.6	06-Jun-2010	pgoyette	Update pciverbose module to use module_autoload() rather than module_load(). Load the module right before each attempt to use its features, and let the module subsystem handle unloading.
1.5	28-May-2010	pgoyette	branches: 1.5.2; Remove opt_pci.h - it's not needed for anything.
1.4	27-May-2010	pgoyette	Remove a left-over debugging call to aprint_normal() Thanks njoly@
1.3	27-May-2010	pgoyette	Extract pci_verbose interface definitions into their own header file so we don't need to include pcivar.h in the module. pcivar.h tries to pull in machine/pci_machdep.h which doesn't exist on all machine/architecture combos. Keeping track and building the module only for those that work would have been a maintenance headache; this change allows us to build the module on all systems, regardless of whether the system has pci support or not.
1.2	25-May-2010	pgoyette	Rework the pciverbose module dispatch vectors to avoid renaming the externally-visible entrypoint name. Also this avoids a potential need to bump kernel version. Requested by dyoung@ and mrg@
1.1	24-May-2010	pgoyette	Extract the vendor/product tables and related access routines into a separate kernel module. Update pci bus attach routine to load the module (if available) when we're about to start scanning the bus, and unload the module after the scan is finished. On architectures which support loading of modules by the boot loader, the 'pciverbose' module can be loaded and executed without needing to rebuild the kernel. On all architectures, using 'options PCIVERBOSE' in the kernel configuration file will create a 'builtin' module which is functionally equivalent to previous behavior. XXX Although not nearly as large as the vendor and product tables, XXX the PCI class and subclass tables might also be offloaded into XXX the module at a future time. XXX Cardbus (and possibly other) drivers should also be modified to XXX load the module before scanning/attaching devices.
1.5.2.4	05-Mar-2011	rmind	sync with head
1.5.2.3	03-Jul-2010	rmind	sync with head
1.5.2.2	30-May-2010	rmind	sync with head
1.5.2.1	28-May-2010	rmind	file pci_verbose.c was added on branch rmind-uvmplock on 2010-05-30 05:17:36 +0000
1.7.4.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.7.4.1	25-Jul-2010	uebayasi	file pci_verbose.c was added on branch uebayasi-xip on 2010-08-17 06:46:28 +0000
1.7.2.2	11-Aug-2010	yamt	sync with head.
1.7.2.1	25-Jul-2010	yamt	file pci_verbose.c was added on branch yamt-nfs-mp on 2010-08-11 22:53:51 +0000
1.8.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5	25-Oct-2016	pgoyette	Increase max string length for PCI Product names. Affects only kernels with PCIVERBOSE (or corresponding module). We currently have a few product names that exceed the old limit, and this is triggering an SSP check in pci_devinfo(). This commit doesn't directly address the SSP issue, but pushes the can down the road...
1.4	21-Sep-2014	christos	branches: 1.4.2; 1.4.4; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.3	06-Jun-2010	pgoyette	branches: 1.3.2; 1.3.4; 1.3.22; Update pciverbose module to use module_autoload() rather than module_load(). Load the module right before each attempt to use its features, and let the module subsystem handle unloading.
1.2	28-May-2010	pgoyette	branches: 1.2.2; Fix symbol name in multiple-inclusion protection. (missing trailing '_')
1.1	27-May-2010	pgoyette	Extract pci_verbose interface definitions into their own header file so we don't need to include pcivar.h in the module. pcivar.h tries to pull in machine/pci_machdep.h which doesn't exist on all machine/architecture combos. Keeping track and building the module only for those that work would have been a maintenance headache; this change allows us to build the module on all systems, regardless of whether the system has pci support or not.
1.2.2.3	03-Jul-2010	rmind	sync with head
1.2.2.2	30-May-2010	rmind	sync with head
1.2.2.1	28-May-2010	rmind	file pci_verbose.h was added on branch rmind-uvmplock on 2010-05-30 05:17:36 +0000
1.3.22.1	03-Dec-2017	jdolecek	update from HEAD
1.3.4.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.3.4.1	06-Jun-2010	uebayasi	file pci_verbose.h was added on branch uebayasi-xip on 2010-08-17 06:46:28 +0000
1.3.2.2	11-Aug-2010	yamt	sync with head.
1.3.2.1	06-Jun-2010	yamt	file pci_verbose.h was added on branch yamt-nfs-mp on 2010-08-11 22:53:51 +0000
1.4.4.1	04-Nov-2016	pgoyette	Sync with HEAD
1.4.2.1	05-Dec-2016	skrll	Sync with HEAD
1.3	11-Dec-2005	christos	merge ktrace-lwp.
1.2	24-May-2005	lukem	branches: 1.2.2; add __KERNEL_RCSID
1.1	30-Aug-2004	drochner	branches: 1.1.2; add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.1.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	03-Sep-2004	skrll	Sync with HEAD
1.1.2.1	30-Aug-2004	skrll	file pcibusprint.c was added on branch ktrace-lwp on 2004-09-03 12:45:28 +0000
1.2.2.1	21-Jun-2006	yamt	sync with head.
1.55	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.54	13-Aug-2022	jmcneill	pciconf: Skip callbacks for reserved ranges if resource allocation fails. In the event that a BAR write is ignored, no need to notify the callback of any changes.
1.53	30-Aug-2021	jmcneill	If an allocation cannot be satisfied by the prefetchable range, retry with the non-prefetchable range.
1.52	03-Jan-2021	skrll	Misc whitespace
1.51	29-Dec-2020	skrll	Add support for Enhanced Allocations as seen in the Cavium ThunderX based GIGABYTE MT30-GS2-00 From thorpej@. Thanks!
1.50	20-Oct-2020	jmcneill	branches: 1.50.2; Instead of trying to prevent pciconf from reconfiguring the firmware's framebuffer, instead allow MD code to register callbacks. If a resource is changed, the driver can unmap the old resource and remap the new. Do this with simplefb so the console doesn't explode when the VGA device is (potentially) reconfigured at boot.
1.49	10-Oct-2020	jmcneill	Add pciconf_resource_reserve. This allows MD code to mark specific memory and I/O regions as in use. When pciconf finds a device already configured to use one of these regions, the device config is left as-is.
1.48	08-Jul-2020	thorpej	Turn off pci_conf_debug -- it was enabled by mistake in previous commit.
1.47	07-Jul-2020	thorpej	Overhaul the interface to pci_configure_bus(): - Don't expose how PCI bus configuration resource management is implemented. Provide a new resource provider API: ==> pciconf_resource_init() -- Initialize a PCI configuration resources container. ==> pciconf_resource_add() -- Add a PCI configuration resource to the container (I/O, MEM, or prefetchable MEM). Multiple resources of each type may be added. ==> pciconf_resource_fini() -- Tear down the PCI configurtation resources container once the bus has been configured. This is much easier to use than the previous method of providing an extent map for each kind of resource, and works better for e.g. ACPI platforms that provide potentially multiple PCI resources in tables provided by firmware. - Re-implement PCI configuration resource management using vmem arenas, rather than extent maps.
1.46	02-Feb-2020	jmcneill	Fix null deref for busses with no IO port resources.
1.45	20-Jan-2020	skrll	don't attempt to extent_alloc_subregion outside the parent extent ranges
1.44	18-Jan-2020	thorpej	Fix several problems with pci_configure_bus(): - Track the 64-bit range capability of prefetchable and non-prefetchable memory separately. Probe the extent maps provided by the caller to initialize these values. Without this, we never get 64-bit range capablity on the root bus, and thus are never able to forward it along to downstream busses. - Always prefer allocating space for a 64-bit memory BAR > 4GB. We will fall back on a 32-bit range if no space above 4GB is available. - Constrain allocation of 32-bit memory BARs (including expansion ROM BARs) to be below 4GB, even if the window has a larger range available. - When allocating non-prefetchable memory space for a PCI-PCI bridge, ensure it falls below 4GB, since a bridge cannot forward a 64-bit non-prefetchable range. - Account for expansion ROMs as non-prefetchable memory rather than prefetchable memory; expansion ROMs have 32-bit BARs, and if a device with an expansion ROM is downstream of a brige, a 32-bit prefetchable range might not be available. Tested by jmcneill@ on an Arm Neoverse N1 SDP, where the previous code failed to configure all devices correctly.
1.43	05-Dec-2019	msaitoh	branches: 1.43.2; KNF. No functional change.
1.42	01-Oct-2019	chs	in many device attach paths, allocate memory with KM_SLEEP instead of KM_NOSLEEP and remove code to handle failures that can no longer happen.
1.41	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.40	01-Mar-2019	msaitoh	KNF. No functional change.
1.39	01-Mar-2019	msaitoh	No functional change: - u_int* -> uint* - KNF.
1.38	29-Jan-2019	msaitoh	Fix a bug that PCI bridge's prefetchable memory area's configuration didn't work correctly in pciconf.c::configure_bridge(). Note that this function is only for "options PCI_NETBSD_CONFIGURE" machines.
1.37	05-Sep-2014	matt	branches: 1.37.20; Don't use class or typename as a variable name.
1.36	20-Oct-2012	matt	Enforce alignments for buses.
1.35	20-Oct-2012	matt	If exceeding io/mem windows, show the number that got exceeded.
1.34	27-Jan-2012	para	branches: 1.34.6; converting extent(9) from malloc(9) to kmem(9) preceding kmem-vmem-pool-uvm patch releng@ acknowledged
1.33	24-Aug-2011	dyoung	branches: 1.33.2; 1.33.6; Add to pci_bus_devorder() an argument that tells the number of slots available in the devs array. Change the type of the devs array from char to uint8_t. Treat the return value of pci_bus_devorder() as the number of slots that it filled. Don't use the __PCI_BUS_DEVORDER #definition to configure the kernel but let the linker do it. Make pci_bus_devorder() available on all architectures by adding a default implementation that will DTRT on all architectures but hpcmips, the only architecture to #define __PCI_BUS_DEVORDER. On hpcmips, adapt the implementation to the new calling convention. XXX I can compile an hpcmips GENERIC kernel, but I don't have a XXX hpcmips box to test it on.
1.32	11-Dec-2010	matt	Switch to kmem from malloc. Don't actually set the PREFETCH64 base/limit registers unless they really are in use.
1.31	02-Aug-2009	gavan	branches: 1.31.4; If we're going to complain about exceeding the PCI bandwidth, we may as well say how much we think we're using.
1.30	24-May-2007	briggs	branches: 1.30.32; 1.30.52; 1.30.56; The number of max_lat units in 1 sec is 4000000 instead of 40000000 (an extra 0 snuck into the code). Found when looking at this with dyoung.
1.29	27-Feb-2006	gdamore	branches: 1.29.24; 1.29.26; Add PCI_MAPREG_TYPE_ROM and allow it to be used with pci_mapreg_map(). Fix to configure (but do not enable) BARs for expansion ROMs. Reviewed by briggs@
1.28	10-Feb-2006	gdamore	PCI_NETBSD_CONFIGURE should allocate (but not map) address space expansion ROMS by default. Full discussion at http://mail-index.netbsd.org/tech-kern/2005/12/16/0023.html Closes PR kern/32467 Reviewed by briggs@
1.27	11-Dec-2005	christos	branches: 1.27.2; 1.27.4; 1.27.6; merge ktrace-lwp.
1.26	25-Mar-2005	tsutsui	branches: 1.26.2; Allow MD PCI code to specify or disable PCI_COMMAND_PARITY_ENABLE and PCI_CMMAND_SERR_ENABLE in MI pci_configure_bus(9). Patch from PR kern/27423 by KIYOHARA Takashi with some changes by me, approved by Allen Briggs.
1.25	27-Feb-2005	perry	branches: 1.25.2; nuke trailing whitespace
1.24	11-Feb-2005	simonb	White space nit- don't put a space before/after increment/decrement operators.
1.23	17-Mar-2004	scw	branches: 1.23.8; 1.23.10; Use PRIu64 and PRIx64 in place of %llu and %llx format strings for printing variables of type u_int64_t.
1.22	02-Dec-2003	briggs	Configure PCI-Cardbus bridges, too. Patch from KIYOHARA Takashi on current-users.
1.21	31-Mar-2003	augustss	branches: 1.21.2; Add an option to disable fast back-to-back transfers since this breaks certain hardware combinations.
1.20	27-Nov-2002	simonb	In configure_bus() return early if there are no devices on the bus to configure.
1.19	25-Nov-2002	thorpej	Avoid some "comparison always false" warnings by only compiling the code in question if ULONG_MAX > 0xffffffff.
1.18	08-Nov-2002	simonb	Add a prototype for set_busreg(), fix a couple of white space nits.
1.17	30-Jul-2002	augustss	Number PCI busses using a simple pre-order numbering instead of some strange binary split numbering that doesn't work with multiple bridges on the same bus, nor with deeply nested bridges.
1.16	27-Jun-2002	briggs	Change a 'continue' to a 'break' in switch configuring 64-bit devices. Add a little more debugging code.
1.15	21-Feb-2002	kleink	branches: 1.15.8; 1.15.10; As discussed with Jason Thorpe: while the `cacheline_size\' argument is a byte-size quantity, the CacheLine Size configuration register specifies the system cacheline size in units of 32-bit words; per PCI Local Bus Specification, rev. 2.1, section 6.2.4.
1.14	28-Nov-2001	thorpej	Add a "cacheline_size" argument to pci_configure_bus(). It is used to set the cacheline size in the BHLC register. This should be the size of the largest D-cache line on a system.
1.13	13-Nov-2001	lukem	add RCSID
1.12	09-Nov-2001	thorpej	Add a "firstbus" argument to pci_configure_bus(), indicating the first bus number to use, rather than always assuming that we should start at bus #0.
1.11	09-Nov-2001	thorpej	In query_bus(), use the parent pciconf_bus_t's PCI chipset tag, not the new bus's (which is potentially uninitialized when some calls are made).
1.10	03-Sep-2001	thorpej	branches: 1.10.4; Bump MAX_CONF_DEV to 32, add some missing newlines in error messages, and add a TODO item (allocate all metadata dynamically).
1.9	31-Aug-2001	briggs	Fix use of uninitialized variable. Missed by gcc, spotted by enami.
1.8	30-Aug-2001	briggs	Respect __PCI_BUS_DEVORDER. If all devices can handle 66MHz, then use 66MHz. Triple the number of configured I/O ranges that we can handle on a bus (8 was insufficient--originally didn't consider multifunction devices) Allow one to specify which types of memory to configure, I/O, ROM, or MEM--for example, one could configure only ROM or only non-ROM. Ensure that the ROM is disabled if we're not configuring it. Only set PCI_COMMAND_IO_ENABLE/PCI_COMMAND_MEM_ENABLE if there are valid memory ranges configured.
1.7	28-Aug-2001	thorpej	Add a TODO item: Do this in 2 passes. One pass that configures the bus (and optionally maps expansion ROMs), and an optional second pass to disable expansion ROMs that are mapped. This would allow MD code to possibly execute the expansion ROMs (possibly in an x86 emulator) to configure a device (e.g. a VGA card, which pretty much needs to be configured by its ROM).
1.6	28-Aug-2001	thorpej	If MD code doesn't provide pci_conf_hook(), then default to not mapping the expansion ROMs on cards, since address decoders may be shared between the ROM and PCI memory space on some cards (i.e. "only map the ROM if you're going to use it, and then unmap it when you're done" is the intended usage).
1.5	14-Jun-2001	thorpej	branches: 1.5.2; It's pretty questionable to printf a PCI tag, but if we're going to do so, at least make the printf compile across the board.
1.4	13-Jun-2001	simonb	Add a port to IBM's PPC405GP Reference Board (the "walnut") by Eduardo Horvath and Simon Burge of Wasabi Systems. IBM 4xx series CPU features: - New pmap and revised trap handler. - Support on-chip timers, PCI controller, UARTs - Framework for on-chip ethernet and watchdog timer. General PowerPC features: - Add in-kernel PPC floating point emulation - New in{,4}_cksum that is between 1.5 and 5 times faster than the old version depending on CPU type. General changes: - Kernel support for generic dbsym-style symbols.
1.3	27-May-2001	thorpej	Handle non-contiguous BARs. Needed to be able to configure e.g. IDE and USB on a PIIX3.
1.2	12-Feb-2001	briggs	branches: 1.2.2; Deal with a lack of the proper type of memory. For example, if a bridge only supports 16-bit I/O addresses, do not configure it with addresses having the upper 16-bits non-zero.
1.1	09-Feb-2001	briggs	branches: 1.1.2; Add an option (defopt) PCI_NETBSD_CONFIGURE that provides PCI bus configuration (assignment of bus numbers, BARs, timer values, interrupt lines, etc.). The interface must be called from m.d. code prior to probing the bus. It is meant to be called once for each primary (bus == 0) PCI bus in the system. It will configure any busses behind PCI-PCI bridges. Section 9 man page for pci_configure_bus() will come soon. In the meantime, sample usage is in arch/sandpoint/sandpoint/mainbus.c. [ Reviewed by thorpej ]
1.1.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.1.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.2.1	09-Feb-2001	bouyer	file pciconf.c was added on branch thorpej_scsipi on 2001-02-11 19:15:59 +0000
1.2.2.8	11-Dec-2002	thorpej	Sync with HEAD.
1.2.2.7	11-Nov-2002	nathanw	Catch up to -current
1.2.2.6	01-Aug-2002	nathanw	Catch up to -current.
1.2.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.2.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.2.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.5.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.5.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.5.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.5.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.10.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.15.10.1	27-Jun-2002	lukem	Pull up revision 1.16 (requested by briggs in ticket #382): Change a 'continue' to a 'break' in switch configuring 64-bit devices. Add a little more debugging code.
1.15.8.2	29-Aug-2002	gehenna	catch up with -current.
1.15.8.1	15-Jul-2002	gehenna	catch up with -current.
1.21.2.6	01-Apr-2005	skrll	Sync with HEAD.
1.21.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.21.2.4	15-Feb-2005	skrll	Sync with HEAD.
1.21.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.21.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.21.2.1	03-Aug-2004	skrll	Sync with HEAD
1.23.10.3	26-Mar-2005	yamt	sync with head.
1.23.10.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.23.10.1	12-Feb-2005	yamt	sync with head.
1.23.8.1	29-Apr-2005	kent	sync with -current
1.25.2.1	27-Mar-2005	tron	Pull up revision 1.26 (requested by tsutsui in ticket #67): Allow MD PCI code to specify or disable PCI_COMMAND_PARITY_ENABLE and PCI_CMMAND_SERR_ENABLE in MI pci_configure_bus(9). Patch from PR kern/27423 by KIYOHARA Takashi with some changes by me, approved by Allen Briggs.
1.26.2.2	03-Sep-2007	yamt	sync with head.
1.26.2.1	21-Jun-2006	yamt	sync with head.
1.27.6.1	22-Apr-2006	simonb	Sync with head.
1.27.4.1	09-Sep-2006	rpaulo	sync with head
1.27.2.2	01-Mar-2006	yamt	sync with head.
1.27.2.1	18-Feb-2006	yamt	sync with head.
1.29.26.1	11-Jul-2007	mjf	Sync with head.
1.29.24.1	27-May-2007	ad	Sync with head.
1.30.56.1	07-Jan-2011	matt	Use kmem. Fix prefetchable setting.
1.30.52.1	24-Dec-2011	matt	Fix problem when alignment of a device is > then the alignment of the bus bus spaces provided.
1.30.32.1	19-Aug-2009	yamt	sync with head.
1.31.4.1	05-Mar-2011	rmind	sync with head
1.33.6.1	18-Feb-2012	mrg	merge to -current.
1.33.2.2	30-Oct-2012	yamt	sync with head
1.33.2.1	17-Apr-2012	yamt	sync with head
1.34.6.2	03-Dec-2017	jdolecek	update from HEAD
1.34.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.37.20.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.37.20.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.37.20.1	10-Jun-2019	christos	Sync with HEAD
1.43.2.2	29-Feb-2020	ad	Sync with head.
1.43.2.1	25-Jan-2020	ad	Sync with head.
1.50.2.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.16	20-Oct-2020	jmcneill	Instead of trying to prevent pciconf from reconfiguring the firmware's framebuffer, instead allow MD code to register callbacks. If a resource is changed, the driver can unmap the old resource and remap the new. Do this with simplefb so the console doesn't explode when the VGA device is (potentially) reconfigured at boot.
1.15	10-Oct-2020	jmcneill	Add pciconf_resource_reserve. This allows MD code to mark specific memory and I/O regions as in use. When pciconf finds a device already configured to use one of these regions, the device config is left as-is.
1.14	07-Jul-2020	thorpej	Overhaul the interface to pci_configure_bus(): - Don't expose how PCI bus configuration resource management is implemented. Provide a new resource provider API: ==> pciconf_resource_init() -- Initialize a PCI configuration resources container. ==> pciconf_resource_add() -- Add a PCI configuration resource to the container (I/O, MEM, or prefetchable MEM). Multiple resources of each type may be added. ==> pciconf_resource_fini() -- Tear down the PCI configurtation resources container once the bus has been configured. This is much easier to use than the previous method of providing an extent map for each kind of resource, and works better for e.g. ACPI platforms that provide potentially multiple PCI resources in tables provided by firmware. - Re-implement PCI configuration resource management using vmem arenas, rather than extent maps.
1.13	17-Jun-2020	thorpej	Explcitly include <sys/extent.h> here.
1.12	08-Sep-2012	matt	Fix comment
1.11	10-Feb-2006	gdamore	branches: 1.11.104; PCI_NETBSD_CONFIGURE should allocate (but not map) address space expansion ROMS by default. Full discussion at http://mail-index.netbsd.org/tech-kern/2005/12/16/0023.html Closes PR kern/32467 Reviewed by briggs@
1.10	11-Dec-2005	christos	branches: 1.10.2; 1.10.4; 1.10.6; merge ktrace-lwp.
1.9	25-Mar-2005	tsutsui	branches: 1.9.2; Allow MD PCI code to specify or disable PCI_COMMAND_PARITY_ENABLE and PCI_CMMAND_SERR_ENABLE in MI pci_configure_bus(9). Patch from PR kern/27423 by KIYOHARA Takashi with some changes by me, approved by Allen Briggs.
1.8	04-Feb-2005	perry	branches: 1.8.4; de-__P
1.7	28-Sep-2002	scw	branches: 1.7.6; 1.7.14; 1.7.16; Before declaring pci_conf_interrupt(), check to see if it's already defined as a CPP macro, which is the case for SH5 at least.
1.6	11-Feb-2002	kleink	Move the declaration of pci_conf_hook down to <machine/pci_machdep.h>; a port may choose to implement it not as a function but i.e. as a pci_chipset_tag_t function vector.
1.5	28-Nov-2001	thorpej	Add a "cacheline_size" argument to pci_configure_bus(). It is used to set the cacheline size in the BHLC register. This should be the size of the largest D-cache line on a system.
1.4	09-Nov-2001	thorpej	Add a "firstbus" argument to pci_configure_bus(), indicating the first bus number to use, rather than always assuming that we should start at bus #0.
1.3	30-Aug-2001	briggs	branches: 1.3.4; Respect __PCI_BUS_DEVORDER. If all devices can handle 66MHz, then use 66MHz. Triple the number of configured I/O ranges that we can handle on a bus (8 was insufficient--originally didn't consider multifunction devices) Allow one to specify which types of memory to configure, I/O, ROM, or MEM--for example, one could configure only ROM or only non-ROM. Ensure that the ROM is disabled if we're not configuring it. Only set PCI_COMMAND_IO_ENABLE/PCI_COMMAND_MEM_ENABLE if there are valid memory ranges configured.
1.2	13-Jun-2001	simonb	branches: 1.2.2; Add a port to IBM's PPC405GP Reference Board (the "walnut") by Eduardo Horvath and Simon Burge of Wasabi Systems. IBM 4xx series CPU features: - New pmap and revised trap handler. - Support on-chip timers, PCI controller, UARTs - Framework for on-chip ethernet and watchdog timer. General PowerPC features: - Add in-kernel PPC floating point emulation - New in{,4}_cksum that is between 1.5 and 5 times faster than the old version depending on CPU type. General changes: - Kernel support for generic dbsym-style symbols.
1.1	09-Feb-2001	briggs	branches: 1.1.2; 1.1.4; Add an option (defopt) PCI_NETBSD_CONFIGURE that provides PCI bus configuration (assignment of bus numbers, BARs, timer values, interrupt lines, etc.). The interface must be called from m.d. code prior to probing the bus. It is meant to be called once for each primary (bus == 0) PCI bus in the system. It will configure any busses behind PCI-PCI bridges. Section 9 man page for pci_configure_bus() will come soon. In the meantime, sample usage is in arch/sandpoint/sandpoint/mainbus.c. [ Reviewed by thorpej ]
1.1.4.6	18-Oct-2002	nathanw	Catch up to -current.
1.1.4.5	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.4	08-Jan-2002	nathanw	Catch up to -current.
1.1.4.3	14-Nov-2001	nathanw	Catch up to -current.
1.1.4.2	21-Sep-2001	nathanw	Catch up to -current.
1.1.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.1.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.2.1	09-Feb-2001	bouyer	file pciconf.h was added on branch thorpej_scsipi on 2001-02-11 19:16:00 +0000
1.2.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.2.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.3.4.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.7.16.2	26-Mar-2005	yamt	sync with head.
1.7.16.1	12-Feb-2005	yamt	sync with head.
1.7.14.1	29-Apr-2005	kent	sync with -current
1.7.6.2	01-Apr-2005	skrll	Sync with HEAD.
1.7.6.1	04-Feb-2005	skrll	Sync with HEAD.
1.8.4.1	27-Mar-2005	tron	Pull up revision 1.9 (requested by tsutsui in ticket #67): Allow MD PCI code to specify or disable PCI_COMMAND_PARITY_ENABLE and PCI_CMMAND_SERR_ENABLE in MI pci_configure_bus(9). Patch from PR kern/27423 by KIYOHARA Takashi with some changes by me, approved by Allen Briggs.
1.9.2.1	21-Jun-2006	yamt	sync with head.
1.10.6.1	22-Apr-2006	simonb	Sync with head.
1.10.4.1	09-Sep-2006	rpaulo	sync with head
1.10.2.1	18-Feb-2006	yamt	sync with head.
1.11.104.1	30-Oct-2012	yamt	sync with head
1.1531	27-Oct-2025	andvar	Add IDs for one more VT82C596B PM Controller and VX11 PCI-ISA bridge. Will be needed later to add SMBus support.
1.1530	21-Oct-2025	pgoyette	add MediaTek MT7927
1.1529	21-Oct-2025	pgoyette	Add MediaTek
1.1528	17-Oct-2025	andvar	Add VIA VT8233C PCI-ISA Bridge.
1.1527	03-May-2025	rin	branches: 1.1527.2; pcidevs: Add EXAR XR17V358 -- octal-channel Universal PCIe UART
1.1526	01-May-2025	andvar	make whitespace and tab usage more consistent, at least for the same vendor. remove trailing whitspaces. remove leading whitespace before tabs. fix typo s/MIS/MSI/ for MSI 3090 description. No definition changes expected.
1.1525	01-May-2025	andvar	Add Etron vendor and its xHCI controllers (used in some AMD/Intel boards). Add various VIA devices, either documented in publicly available datasheets or some from dmesg in personally owned devices.
1.1524	28-Apr-2025	andvar	fix typo in previous commit s/FHC/FCH/ for AMD RAID mode SATA controller.
1.1523	28-Apr-2025	andvar	Add AMD X370/X399/A320/A520/600 Series SATA/xHCI devices, 600 Series and Navi PCIe Switch Upstream/Downstream Ports, Raven Ridge Vega GPUs and GeForce 6200A.
1.1522	20-Apr-2025	andvar	viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad.
1.1521	05-Apr-2025	andvar	viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID.
1.1520	03-Apr-2025	andvar	viaide(4): Add support for VIA VT6415/VT6330 single-channel IDE controllers. Beyond adding the usual identification code, several adjustments were required: 1) Introduced 'single_channel` to set 'nchannels` to 1 for this controller. 2) Added 'APO_IDECONF_ALWAYS_EN` definition and the 'no_ideconf` flag to mimic enabled channel bits, as this controller's enable chip register is unset. Applied the same to VT6410, following OpenBSD and Linux, since some controllers are known to not set this register as well. 3) Combined all VIA controllers identified by PCI ID (not ISA bus) using switch fallthrough, as they currently share the same UDMA setting. Consequently, the VX900 name printing was moved to the device description and adjusted to better reflect the device. 4) Moved setting interface bits for RAID controllers under the RAID capability check, enabling the above fallthrough and paving the way for support of other RAID-capable controllers in upcoming changes. 5) The VT6330 is a combo FireWire/IDE controller. Added its FireWire controller PCI ID to pcidevs. Tested on ASRock P5B-DE. Reviewed by jak and bad. Addresses PR kern/45917
1.1519	03-Mar-2025	andvar	Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more).
1.1518	23-Jan-2025	andvar	Add VIA VT8261 southbridge SATA controller IDs and PCIB.
1.1517	23-Jan-2025	mrg	add several more intel meteor lake and crucial ssd ids
1.1516	18-Jan-2025	mrg	add a bunch of intel arrow lake devices, and several more NPU devices
1.1515	03-Jan-2025	mrg	add a bunch of missing nvidia gpu ids. these are between 0xf00 and 0xfff, and mostly are the GT 400/500/600/700 series, and their variants.
1.1514	12-Dec-2024	msaitoh	pcidevs: Add some Oxford Semiconductor devices from FreeBSD.
1.1513	11-Nov-2024	msaitoh	Add many Brainboxes devices. Repoted in PR/kern 55824 by Cameron Williams.
1.1512	19-Oct-2024	mlelstv	Add Raspberry Pi5 Ethernet.
1.1511	16-Oct-2024	msaitoh	pcidevs: Add some AMD devies.
1.1510	20-Aug-2024	dogcow	Add Areca ARC-1224
1.1509	17-Jul-2024	mrg	add the intel "arc" a-series of cards.
1.1508	11-Jul-2024	msaitoh	Add Realtek RTL8126.
1.1507	23-Jun-2024	jakllsch	branches: 1.1507.2; Add Brainboxes and devices From Cameron Williams in PR 58358.
1.1506	12-May-2024	msaitoh	Add some Intel Rapotr Lake graphics.
1.1505	12-May-2024	msaitoh	Add Phison PS5021 and PS5026.
1.1504	06-Apr-2024	andvar	Add Zhaoxin devices, mainly from my TVS-675 NAS device, and some additional to fill the gaps from pci-ids database.
1.1503	28-Feb-2024	jakllsch	Add (Realtek/)Killer E2600 Ethernet
1.1502	18-Jan-2024	msaitoh	Add Intel I219-{LM,V}(20,21)
1.1501	28-Dec-2023	msaitoh	Update Intel Raptor Lake devices.
1.1500	28-Dec-2023	msaitoh	Add Intel C26[26] eSPI.
1.1499	13-Nov-2023	jnemeth	add some newer Areca boards in prep for updating driver
1.1498	26-Oct-2023	msaitoh	pcidevs: Add Mellanox ConnectX-6 Lx.
1.1497	11-Oct-2023	msaitoh	pcidevs: Add Intel I219{V,LM}({22,23})
1.1496	10-Oct-2023	abs	Add entry for ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier
1.1495	08-Oct-2023	msaitoh	Add Pericom(Diodes) PCIe switches.
1.1494	04-Oct-2023	rin	pcidevs: Add missing I225/I226 series devices
1.1493	04-Oct-2023	msaitoh	Add Intel 82599 LS.
1.1492	10-Aug-2023	andvar	rename PCI to PCIE for naming consistency.
1.1491	09-Aug-2023	msaitoh	Simplify descriptions of Snow Ridge internal Ethernet devices.
1.1490	09-Aug-2023	msaitoh	Add some Intel Xeon D-2100 devices. Fix description of Intel 0x37d0.
1.1489	05-Aug-2023	andvar	fix duplicate definition.
1.1488	05-Aug-2023	andvar	Add device ids of the VX800 chipset and s3 chrome 500 series GPU. Descriptions mainly aligned with VX800/820 Series System Programming Manual.
1.1487	04-Aug-2023	msaitoh	Add Intel I226 devices.
1.1486	04-Aug-2023	msaitoh	Fix some descriptions of Intel Snow Ridge.
1.1485	04-Aug-2023	msaitoh	Add Intel Snow Ridge devices.
1.1484	31-Jul-2023	tnn	pcidevs: add ASMedia ASM106x
1.1483	28-Jul-2023	msaitoh	Add AMD 19h/7xh Root Complex.
1.1482	23-Jul-2023	msaitoh	Add Intel 700 series chipset devices.
1.1481	22-Jul-2023	msaitoh	Add Alder Lake-N and Raptor Lake devices.
1.1480	06-Apr-2023	msaitoh	The SAATA device ID for Apollo Lake is not 0x5ae0 but 0x5ae3.
1.1479	14-Feb-2023	msaitoh	Add devices from PPR for AMD Family 19h Model 61h Revision B1 processors.
1.1478	03-Feb-2023	msaitoh	Add Samsung SM990.
1.1477	29-Jan-2023	thorpej	Add ID for the EXAR XR17V354 PCIe UART. From Håkan Engvall. PR kern/57202
1.1476	18-Jan-2023	msaitoh	Add AMD Family 19h/1xh devices.
1.1475	14-Jan-2023	ryo	add Aquantia (Marvell) AQC113 ethernet devices and the variants
1.1474	04-Jan-2023	msaitoh	Fix typo. s/SSC/SCC/.
1.1473	30-Dec-2022	msaitoh	Add Elkhart Lake Shared SRAM.
1.1472	28-Dec-2022	msaitoh	Update Intel Elkhart Lake devices.
1.1471	22-Nov-2022	msaitoh	branches: 1.1471.2; Add some Xeon Scalable devices from OpenBSD.
1.1470	26-Oct-2022	msaitoh	Add AMD F17/Axh devices.
1.1469	26-Oct-2022	msaitoh	Add some AMD 17h/9xh devices from OpenBSD.
1.1468	07-Oct-2022	snj	add NVIDIA GeForce GTX 770
1.1467	02-Oct-2022	martin	Add AMD FCH SATA Controller D
1.1466	01-Oct-2022	msaitoh	Add AMD 19h/6xh Root Complex.
1.1465	28-Sep-2022	msaitoh	Sort by number. No functional change.
1.1464	28-Sep-2022	msaitoh	Add Intel Core 8G (8core, H, Halo) Host Bridge, DRAM.
1.1463	28-Sep-2022	msaitoh	Jasper Lake Intel Trace Hub on Compute Die is not 0x4da6 but 0x4e29.
1.1462	13-Sep-2022	msaitoh	Add more Alder Lake devices.
1.1461	11-Sep-2022	mrg	add several samsung nvme entries
1.1460	27-Jun-2022	msaitoh	Add some MegaRAID devices.
1.1459	17-Jun-2022	msaitoh	Update some descriptions of Intel 700 series Ethernet devices.
1.1458	02-Jun-2022	mrg	add some AMD zen2/3 devices.
1.1457	11-May-2022	markd	Add some Broadcom / LSI RAID cards.
1.1456	04-May-2022	martin	Rename one of the XEONSC_RAS entries slightly to avoid conflicts
1.1455	04-May-2022	nia	More Xeon Scalable / Skylake-E devices
1.1454	27-Apr-2022	msaitoh	Update some AMD devices: - Rename AMD 15h/6xh devices. - Add some devices. - Use OHCI for OHCI device. - Use EHCI for EHCI device. - and some minor changes.
1.1453	22-Apr-2022	skrll	Add some Broadcom devices. From OpenBSD.
1.1452	24-Mar-2022	yamaguchi	Added Intel V710 for 5000BaseT
1.1451	24-Mar-2022	yamaguchi	Change description about Intel X710 10GBaseT device
1.1450	16-Mar-2022	yamaguchi	Added Intel X710 devices and rename X710_10G_T to X710_10G_T_1
1.1449	15-Feb-2022	msaitoh	Add VMware AHCI and NVMe.
1.1448	02-Feb-2022	msaitoh	Add some devices from AMD BKDG.
1.1447	02-Feb-2022	msaitoh	Fix AMD F16_HB from 0x1568 to 0x1538. AMD BKDG says the host bridge's device ID is 0x1538. Before pcidevs rev. 1.1261, F16_HB was located next to 0x1537, so I suspect 0x1568 was typo of 0x1538.
1.1446	25-Jan-2022	msaitoh	Add Alder Lake devices and Intel 600 Series PCH devices.
1.1445	01-Jan-2022	msaitoh	s/sytem/system/
1.1444	30-Nov-2021	msaitoh	Add Intel Rocket Lake devices.
1.1443	03-Nov-2021	msaitoh	- Rename EHL_TRACE* macros - Update Jasper Lake's Processor Transaction Routers.
1.1442	27-Oct-2021	msaitoh	Add Intel Elkhart Lake devices.
1.1441	27-Oct-2021	msaitoh	Add Intel Gemini Lake TXE HECI 1.
1.1440	13-Oct-2021	msaitoh	Add Intel I219V 15-19 and I219LM 16-19.
1.1439	13-Oct-2021	msaitoh	Add Intel I219's version number.
1.1438	12-Oct-2021	msaitoh	Add Intel 400 Series PCH-V devices.
1.1437	11-Oct-2021	msaitoh	Add Intel 500 Series PCH-H devices.
1.1436	11-Oct-2021	msaitoh	Remove duplicated entries.
1.1435	02-Oct-2021	mrg	add intel 660p ssd, and expand the 760p description.
1.1434	23-Sep-2021	msaitoh	Add some Intel devices: - Tiger Lake - I225V and I225LM - WiFi 6 AX201
1.1433	06-Sep-2021	mrg	the CPU ident and AMD both say this is "R2 Graphics", not R3.
1.1432	03-Sep-2021	mrg	add a whole bunch of radeon devices. now all the radeons i have are here (and dozens i don't :)
1.1431	30-Aug-2021	andvar	add RDC R6022 PCI-Host bridge.
1.1430	30-Aug-2021	mrg	add a bunch of nvidia devices.
1.1429	30-Aug-2021	mrg	add radeon x1550.
1.1428	17-Jul-2021	mrg	add current generation NVIDIA graphics cards (3050-3090 etc.)
1.1427	12-Jul-2021	msaitoh	Add Intel Jasper Lake devices and Reealtek RTL8821CE.
1.1426	21-Apr-2021	jakllsch	branches: 1.1426.2; Fix Realtek RTL8125 description
1.1425	02-Apr-2021	rin	PDC20265 is Ultra/100, not 66.
1.1424	30-Jan-2021	jmcneill	branches: 1.1424.2; Add some more product IDs for mcx(4)
1.1423	29-Dec-2020	skrll	Sort Cavium devices
1.1422	29-Dec-2020	skrll	Trailing whitespace
1.1421	20-Nov-2020	msaitoh	Add newer Intel PCH internal devices.
1.1420	19-Nov-2020	reinoud	Add modern QUMRANET/Red Hat VIRTIO range PCI devices
1.1419	14-Jul-2020	jdolecek	branches: 1.1419.2; add newer ciss(4) devices, list obtained from FreeBSD
1.1418	09-Jul-2020	msaitoh	Remove duplicated entry.
1.1417	24-Jun-2020	thorpej	Add a couple of additional device IDs for the AMD Cryptographic Coprocessor.
1.1416	13-Jun-2020	ryo	add ASIX AX99100 Multi I/O (Serial,Parallel,I2C,SPI,LocalBus,GPIO) Controller
1.1415	11-Jun-2020	jdolecek	add Intel XMM 7360 LTE Modem
1.1414	26-May-2020	martin	Add RTL8192EE Wireless LAN 802.11n PCI-E NIC
1.1413	01-May-2020	jmcneill	Add IDs for Ampere eMAG PCIe Root Ports
1.1412	23-Apr-2020	msaitoh	Remove duplicated entry.
1.1411	22-Apr-2020	christos	Add AQUANTIA AQC100 from Andrius V.
1.1410	22-Apr-2020	prlw1	Add an eclectic mix of GeForce GTX 680, RTS522A PCI-E Card Reader, intel wifi 9560 + whitespace
1.1409	22-Apr-2020	msaitoh	Add Intel Whiskey Lake U and Amber Lake Y devices.
1.1408	22-Apr-2020	msaitoh	Add Aquantia AQC100, AQC100S and D100.
1.1407	18-Apr-2020	simonb	Add ATI Radeon R5/R6/R7 Graphics.
1.1406	16-Apr-2020	msaitoh	- Add some Intel UHD Graphics devices. Mainly taken from OpenBSD. - Add Comet Lake devices. - Modify description of Intel 0x591e from HD Graphics to UHD Graphics. - Add Western Digital WD Blue SN550 NVMe SSD
1.1405	10-Apr-2020	christos	Fix typo
1.1404	07-Apr-2020	christos	branches: 1.1404.2; Add more RDC products (Andrius V)
1.1403	01-Apr-2020	msaitoh	Add NVIDIA Quadro NVS 295.
1.1402	11-Mar-2020	msaitoh	Add some GeForce devices.
1.1401	11-Mar-2020	fcambus	Add device IDs of the VIA VX900 chipset.
1.1400	08-Mar-2020	msaitoh	NVIDIA 0x036[0-7] are nForce MCP55 LPC Bridge.
1.1399	05-Mar-2020	msaitoh	Add Farallon PN9000SX Ethernet.
1.1398	19-Feb-2020	mrg	add a couple of GPUs i have: product ATI RADEON_HD4290 0x9714 Radeon HD4290 Graphics product NVIDIA GF_440 0x0de0 GeForce GT 440
1.1397	28-Jan-2020	msaitoh	Remove duplicated entries.
1.1396	21-Jan-2020	msaitoh	Add Intel I219 LM10-LM15 and V10-V14 from OpenBSD.
1.1395	01-Jan-2020	ryo	branches: 1.1395.2; add Aquantia AQC 10G network adapters
1.1394	29-Dec-2019	sevan	Killer E3000 from OpenBSD
1.1393	18-Dec-2019	msaitoh	Add AMD Family14h PCIe.
1.1392	27-Nov-2019	msaitoh	- Add Xeon D-1500 NTB-secondary - Add Xeon D NS QuickData DMA channel 0-7
1.1391	14-Nov-2019	msaitoh	Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.1390	08-Nov-2019	msaitoh	- Update Intel's NVMe SSDs. - Modify 0x0953's description to "750 or DC P3[567]00 SSD" - Add DC P4[56]00 - Add Apollo Lake TXE HECI.
1.1389	29-Oct-2019	msaitoh	Add Realtek RTL8125.
1.1388	12-Oct-2019	skrll	Spell controller correctly
1.1387	21-Sep-2019	msaitoh	Add Killer E2400 and E2500.
1.1386	21-Sep-2019	msaitoh	Modify Attansic Ethernet devices' description to clarify.
1.1385	24-Aug-2019	jmcneill	Add Mellanox ConnectX-4, ConnectX-4 Lx, ConnectX-5, and ConnectX-5 Ex
1.1384	21-Aug-2019	mrg	add micron/crucial SM2263 nvme. add some device found on asus x570-p with ryzen 3200G cpu. spell it "PCIe' when using the name.
1.1383	28-Jul-2019	mlelstv	branches: 1.1383.2; Add BCM2711 PCIe Host bridge for RPI4
1.1382	24-Jul-2019	msaitoh	- Add some AMD devices mainly taken from OpenBSD. - Rename some device names for consistency. - Use _HDA for HD Audio.
1.1381	18-Jul-2019	msaitoh	- Intel C620 Virtual Switch Port for Termal Sensor is not 0x37c6 but 0x37c7. - Add Some AMD Ryzen 3xxx's devices.
1.1380	12-Jul-2019	msaitoh	Fix previous (s/1000/1100/).
1.1379	12-Jul-2019	msaitoh	Add some ServerWorks HT1100 devices from OpenBSD.
1.1378	09-Jul-2019	msaitoh	Add Phison PS5000 and PS5016 NVMe.
1.1377	05-Jul-2019	msaitoh	Add some Intel Whisky Lake devices.
1.1376	02-Jul-2019	msaitoh	Modify some descriptions of Virtio devices from OpenBSD to identify the functions.
1.1375	04-Jun-2019	msaitoh	Update some Intel ixgbe devices: - Add Xeon D XFI. - Add "(Hyper-V)" to 0x15a9's description. - Add C3000 QSFP devices.
1.1374	01-Jun-2019	jdolecek	rename VT8237S_SATA_2 to VT8237S_SATA_RAID and update description to have different one to VT8237S_SATA part of PR kern/43309
1.1373	30-May-2019	mlelstv	add IDs for SATA controllers VT8237S and VX855
1.1372	15-Apr-2019	msaitoh	Avoid duplication.
1.1371	15-Apr-2019	msaitoh	- Add some Intel Xeon Scalable processor's devices. - Add Some Qlogic devices from OpenBSD. - Rename Qlogic 0x3022's name from ISP3022 to ISP4022.
1.1370	12-Apr-2019	msaitoh	- Add some new C620 LPC devices. - Remove extra white spaces.
1.1369	11-Apr-2019	mrg	fix a couple of errors in the previous.
1.1368	11-Apr-2019	mrg	add a bunch of PCI devices on my systems, and a bunch of similar ones found in the same sources online (mostly, Intel, AMD, Nvidia and ASMedia web sites.) this includes: - bunch of xeon e5 v3 devices - bunch of radeon chipsets - bunch of nvidia chipsets - bunch of marvell chipsets - bunch of asmedia chipsets be more consistent with device naming within some groups.
1.1367	08-Mar-2019	msaitoh	- Add other two Core 8G host bridges. - Add Intel Xeon E devices.
1.1366	05-Mar-2019	msaitoh	Add Intel I219 variations for Cannon Lake.
1.1365	20-Feb-2019	msaitoh	Add SK-NET SK-9Mxx Gigabit Ethernet.
1.1364	20-Feb-2019	msaitoh	Add some Broadcom Ethernet devices from {Open,Free}BSD.
1.1363	14-Feb-2019	msaitoh	Add Tundra (now IDT) TSI381 and PEB383 from OpenBSD.
1.1362	08-Feb-2019	msaitoh	- Add Some Freescale MPC834x devices from OpenBSD. - Add Intel I210 SGMII Flash-less device from the spec update document.
1.1361	06-Feb-2019	msaitoh	From OpenBSD: - Add NetXen's vendor ID and devices. - Add Sun Neptune.
1.1360	28-Jan-2019	msaitoh	Add AMD Family 17h devices from OpenBSD.
1.1359	26-Jan-2019	msaitoh	Add some RADEON devices. Mainly taken from OpenBSD.
1.1358	21-Jan-2019	msaitoh	Marvell 0x4365 is 88E8070.
1.1357	26-Dec-2018	msaitoh	Add yet another Intel Core QuickPath Generic Non-Core Register.
1.1356	03-Dec-2018	bouyer	Add 2 Symbios products, from OpenBSD
1.1355	29-Nov-2018	jmcneill	Add Amazon.com UART, NVMe, and ENA PCI IDs.
1.1354	24-Nov-2018	skrll	Add Rockchip vendor ID and a device. Requested by jak@ From OpenBSD
1.1353	24-Nov-2018	skrll	Add a whole bunch of Cavium ThunderX devices
1.1352	12-Nov-2018	msaitoh	Add some Intel 100 series and C230 series devices.
1.1351	08-Nov-2018	msaitoh	"s/ are are / are /" in comment. No functional change.
1.1350	08-Nov-2018	msaitoh	Add two otehr I2C devices of Intel.
1.1349	16-Oct-2018	jmcneill	Add AMD Seattle Host Bridge and PCIE root port devices
1.1348	15-Oct-2018	uwe	Adaptec RAID 2445 and 2805
1.1347	23-Aug-2018	msaitoh	- Add C62x to the descriptions of Intel C620 series devices. - Add 0xa1ca for Intel C629 LPC or eSPI.
1.1346	30-Jul-2018	msaitoh	Add Toshiba's XG4 and XG5 NVMe SSD.
1.1345	30-Jul-2018	msaitoh	Add some Intel devices. Some of them are found in Core i7-8550U.
1.1344	06-Jul-2018	prlw1	Add Western Digital (Sandisk) WD Black NVMe SSD as per suggestion from thorpej@
1.1343	20-Jun-2018	msaitoh	branches: 1.1343.2; Add nSYS Technologies.
1.1342	14-Jun-2018	msaitoh	Add Mobility Radeon HD 4300(0x9552) and Mobility Radeon HD 4500(0x9553) from OpenBSD.
1.1341	12-Jun-2018	jdolecek	add two more Yukon variants (from OpenBSD)
1.1340	09-Jun-2018	jdolecek	small sync of Marvell Yukon PCI product ids with OpenBSD - add 8048 variant, rename YUKON_1 to YUKON_8070 also adjust description for YUKON_8055_2 to have "88E8055-2" just so that it shows up differently
1.1339	31-May-2018	mrg	add a bunch of nvidia devices from various generations.
1.1338	31-May-2018	msaitoh	- Add some Intel 82965 devices. - Change Intel 0x2a06 from IDE to IDER
1.1337	28-May-2018	msaitoh	Add some Intel 4 series chipset devices.
1.1336	26-May-2018	jakllsch	Add Intel 82Q45 IDER
1.1335	24-May-2018	msaitoh	Add some Marvell Yukon devices.
1.1334	21-May-2018	jnemeth	add Oxford Semiconductor OXPCIe954 PCIe 4-port serial card
1.1333	03-May-2018	maya	Add some broadcom 802.11ac wifi devices
1.1332	28-Apr-2018	mlelstv	Add Symbios/LSI RAID and SAS controllers.
1.1331	24-Apr-2018	maya	add realtek RT525A PCI-E card reader found in my Dell XPS 15 9550.
1.1330	17-Apr-2018	nonaka	Added some NVMe devices.
1.1329	16-Apr-2018	msaitoh	Add Intel SSD 760p.
1.1328	16-Apr-2018	msaitoh	Add some 8th Generation Intel Core Processor devices.
1.1327	09-Apr-2018	msaitoh	Add new I219 devices (using with 300 series chipset).
1.1326	09-Apr-2018	msaitoh	Add Intel 300 series devices.
1.1325	04-Apr-2018	msaitoh	Add Intel Coffee Lake internal graphics devices. From ark.intel.com.
1.1324	02-Apr-2018	msaitoh	Add Intel X299 and Z370 LPC.
1.1323	31-Mar-2018	sevan	What we have listed as the PowerBook G3 Firwire device is actually the UniNorth Firwire interface which is not exclusive to the PowerBook G3. It can be found in the G3 iMac DV for example http://dmesgd.nycbug.org/index.cgi?do=view&id=1124 Rename the PBG3_FW device UNINORTH_FW & set the previously listed UNINORTH_FW to UNINORTH2_FW which is what the device actually is. Verified using https://pci-ids.ucw.cz/read/PC/106b
1.1322	26-Mar-2018	msaitoh	Add two VIA xHCI devices from OpenBSD.
1.1321	20-Mar-2018	nonaka	Add Intel Wireless AC 9260.
1.1320	16-Mar-2018	msaitoh	Intel 0x15c4 is SFI based and 0x15ce is KR based.
1.1319	04-Mar-2018	jdolecek	branches: 1.1319.2; according to VT82C686A chip specs, the VIA Technologies device 0x3057 is more a Power Management controller, rename the pcidevs entry and device macro PR kern/31963 by Nicolas Joly
1.1318	04-Mar-2018	jdolecek	apply some updates and recommended corrections to pcidevs descriptions PR kern/45733 by Greg A.Woods
1.1317	02-Mar-2018	msaitoh	Add Intel Gemini Lake devices.
1.1316	24-Feb-2018	mlelstv	Add AMD KERNCZ chipset for Zen processors.
1.1315	22-Feb-2018	msaitoh	Add some Xeon E5 (v1) and E5 v2 devices.
1.1314	22-Feb-2018	msaitoh	s/C62x9/C62x/
1.1313	22-Feb-2018	msaitoh	Fix C62x9 sSATA Controller (RAID) device ID.
1.1312	22-Feb-2018	msaitoh	Add C620 devices.
1.1311	21-Feb-2018	msaitoh	Add some Intel Ethernet devices.
1.1310	21-Feb-2018	msaitoh	Add some ServerEngines (Emulex) products from OpenBSD.
1.1309	17-Feb-2018	sevan	grammar
1.1308	16-Feb-2018	sevan	Add Nvidia GeForce 9400M as found in the late 2009 13" MacBook (MacBook6,1)
1.1307	25-Jan-2018	sborrill	Add Marvell Yukon 88E8040T
1.1306	24-Jan-2018	sevan	Add Nvidia GeForce FX Go5200, as found in the PowerBook6,4
1.1305	21-Jan-2018	sevan	Add the GeForce4 MX440 Go, as found on the 17" flat panel G4 iMac (PowerMac4,5). Thanks to Jasper Wallace <jasper AT pointless DOT net> for the loan of the hardware.
1.1304	09-Jan-2018	msaitoh	Add Intel Apollo Lake devices.
1.1303	09-Jan-2018	msaitoh	- s/MMC/eMMC/ for Intel 0x0f14 (Bay Trail) - Add Atom E3800 (Bay Trail) eMMC 4.5
1.1302	18-Dec-2017	msaitoh	Add some Chelsio devices.
1.1301	09-Nov-2017	msaitoh	Add PEX 8605 4port 4lane PCIe Gen 2 switch.
1.1300	30-Oct-2017	sevan	Add Areca ARC1214 & ARC1880. From OpenBSD.
1.1299	20-Oct-2017	christos	update entry for Marvel Yukon 8058, and fix URL for pci ids (Rocky Hotas)
1.1298	15-Sep-2017	hikaru	Add Intel QuickAssist devices.
1.1297	17-Aug-2017	msaitoh	Add C3000 devices.
1.1296	02-Aug-2017	msaitoh	PCI device ID of Atom x5-Z8350's video is 0x22b0. ark.intel.com says 0x22b0 is Intel HD Graphics 400.
1.1295	26-Jul-2017	msaitoh	Add Intel 5520/X58 QuickPath from OpenBSD.
1.1294	22-Jul-2017	msaitoh	- Add some AMD devices. - Add Broadcom BCM5762.
1.1293	07-Jul-2017	msaitoh	Add Atom Z8000 devices from "Intel Atom Z8000 Processor Series Datasheet (Volume 2 of 2)".
1.1292	04-Jul-2017	msaitoh	Add Intel 82Q965 HECI.
1.1291	17-Jun-2017	jdolecek	add entry for Adaptec 1220SA
1.1290	06-Jun-2017	msaitoh	- Change VMware's 0x0770 from "Virtual USB" to "Virtual EHCI" - Add VMware's 0x0774(UHCI) and 0x0778(xHCI)
1.1289	23-May-2017	nonaka	branches: 1.1289.2; Add some AMD entries.
1.1288	08-May-2017	msaitoh	- Add RICOH R5C576 SD Card Controller. - Add some SiS devices. Taken from OpenBSD. - Change SiS 0x6325 from 6325 AGP VGA to 650 AGP VGA.
1.1287	27-Apr-2017	msaitoh	- Add some Pericom (Diodes) devices. - s/P17C9/PI7C9/
1.1286	19-Apr-2017	msaitoh	branches: 1.1286.2; Add BCM4401 B0 step. Old BCM4401 B0 entry is changed to B1. From OpenBSD.
1.1285	13-Apr-2017	msaitoh	Add Broadcom BCM577x5 SDMMC, Memstic and xD.
1.1284	31-Mar-2017	skrll	Fix BCM4727 entry
1.1283	31-Mar-2017	msaitoh	- Add BROADCOM BCM4313 802.11b/g/n - Add Intel's another Pineview Integrated Graphics Device - Add Realtek RTS5208 PCI-E Card Reader
1.1282	23-Mar-2017	msaitoh	Add Huawei Hi1710 BMC
1.1281	14-Mar-2017	nonaka	Added Intel Wireless 3168 entry.
1.1280	27-Feb-2017	msaitoh	Add some NVIDIA devices.
1.1279	16-Jan-2017	msaitoh	Add Intel SSD DC P3520.
1.1278	15-Jan-2017	msaitoh	branches: 1.1278.2; Add Intel 200 series chipset devices from "Table 2-2. PCH-H Device and Revision ID Table, Intel 200 Series Chipset Family Platform Controller Hub(PCI) Datasheet Volume 1 of 2 (335192-001)".
1.1277	15-Jan-2017	msaitoh	Add Core 7G devices from "7th Generation Intel Processor Families for {S, U/Y, H} Platforms Datasheel Volume 2 of 2" (S: 335196-001, U/Y: 334662-002, H: 335191-001). For 0x5906 (U processor, HD Graphics 610), it's taken from ark.inte.com.
1.1276	10-Jan-2017	nonaka	Add Intel Dual Band Wireless 8265 entry.
1.1275	27-Dec-2016	msaitoh	Add some Radeon devices. Mainly taken from OpenBSD.
1.1274	26-Dec-2016	msaitoh	0x10cb is not 82801H(ICH8) but 82801I(ICH9).
1.1273	01-Dec-2016	msaitoh	- Add more Intel 10G devices - indent
1.1272	23-Nov-2016	christos	one more atheros device
1.1271	28-Oct-2016	msaitoh	- Add some Core i7-800 and i5-700 devices from "Intel Core i7-800 and i5-700 Desktop Processor Series Datasheet - Volume 2" (Document Number 322910-003). - Change some descriptions of Core i5-600 and i3-500 devices.
1.1270	26-Oct-2016	msaitoh	i82567V-3 is not ICH9 but ICH8.
1.1269	24-Oct-2016	msaitoh	Add Xeon E7 v4 devices from "Intel Xeon Processor E7 v4 Product Famliy Datasheet Volume 2: Registers".
1.1268	20-Oct-2016	msaitoh	Add newer I219 devices.
1.1267	19-Oct-2016	msaitoh	Add Xeon 5[56]00 devices from "Intel Xeon Processor 5500 Series Datasheet, Volume 2" and "Intel Xeon Processor 5600 Series Datasheet, Volume 2"
1.1266	04-Oct-2016	kiyohara	Add Marvell 88AP510.
1.1265	16-Sep-2016	msaitoh	- Add ASMEDIA ASM1083/1085/1142 - Add AMD RADEON 34xx HD Audio - Add some Intel Core i7-6xxxK devices from "Intel Core i7 Processor Family for LGA2011-v3 Socket Datasheet - Volume 2 of 2: Registers" - Add "Core i7-6xxxK" to some Intel devices' description.
1.1264	01-Sep-2016	msaitoh	Add some 1st gen Core PCI devices. Mainly taken from "Intel Core, Pentium Desktop Processor Series: Datasheet, Vol. 2"
1.1263	23-Aug-2016	msaitoh	Fix: - ATI 0x4394 is AHCI SATA. - ATI 0x4395 is Storage mode SATA.
1.1262	23-Aug-2016	msaitoh	Fix XE5_V3_IMC1_DDRIO_[34]'s values.
1.1261	23-Aug-2016	msaitoh	- Fix entries. The following problems were found while sorting entries: - Remove the following entries because the value is duplicated: CMDTECH_240 (CMDTECH_AAR_1210SA has the same value) INTEL_C610_SATA_RAID (INTEL_82801H_SATA_RAID has the same vaule) INTEL_C610_PCIE_1_3 (82801BA_HPB has the same value) GEFORCE_610M3 (GEFORCE_610M2 has the same value) - RADEON_RV350_NQ is not 0x4e41 but 0x4e51 - 82801JD_SATA_RAID is not 0x3a02 but 0x3a05. - The following two entries have the same number. Fix one of them: DH89XXCL_USB_2 to 0x23b5 (was 0x23b4 (== DH89XXCL_USB_1)) 5500_HB to 0x3404 and change the desc (was 0x3403 (== 3400_HB)) 82X58_PCIE_0_1 to 0x3421 (was 0x3420 (== 82X58_PCIE_0_0)) 82801JI_THERMAL to 0x3a32 (was 0x3a30 (== 82801JI_SMB)) - Sort entries in device ID's order.
1.1260	19-Aug-2016	msaitoh	Add more Xeon D devices.
1.1259	17-Aug-2016	msaitoh	Update Skylake devices.
1.1258	17-Aug-2016	msaitoh	Add Xeon D devices.
1.1257	08-Aug-2016	msaitoh	Add some Xeon E5 v3 devices.
1.1256	29-Jul-2016	nonaka	Add MegaRAID SAS3008
1.1255	28-Jul-2016	nonaka	Add Super Micro Computer Inc
1.1254	28-Jul-2016	nonaka	sort by vendor id.
1.1253	06-Jul-2016	christos	branches: 1.1253.2; more ral products
1.1252	27-Jun-2016	pgoyette	Add a couple more devices, from PR kern/40987
1.1251	01-Jun-2016	pgoyette	Add Intel 82597EX_SR, from PR kern/47750
1.1250	30-May-2016	nonaka	Added some Intel wireless devices from OpenBSD.
1.1249	27-Apr-2016	christos	more ral devices from OpenBSD
1.1248	26-Apr-2016	msaitoh	Add devices from "6th Generation Intel Processor Families I/O Platform Datasheet (U/Y-Processor platform) Volume 1 of 2"
1.1247	12-Feb-2016	mrg	add several NVIDIA geforce and NVS cards.
1.1246	01-Feb-2016	msaitoh	- s/82081/82801/ from OpenBSD. - 0x2829 is SATA. 0x282a is RAID.
1.1245	22-Jan-2016	msaitoh	Add anoter new I219-LM.
1.1244	01-Dec-2015	msaitoh	Add Intel 100 series chipset devices from "Table 2-1. PCH-H Device and Revision ID Table, Intel 100 Series Chipset Family Platform Controller Hub(PCI) Datasheet".
1.1243	01-Dec-2015	msaitoh	Add Intel Core 6G devices from datasheets.
1.1242	16-Nov-2015	msaitoh	Change macro name of Braswell PCU SMBus.
1.1241	16-Nov-2015	msaitoh	Add Braswell devices from "Table 23-4 PCI Devices and Functions, N-Series Intel Pentium Processorts and Intel Celeron Processors Datasheet (Document Number: 332093-001)"
1.1240	06-Nov-2015	nonaka	Add some ATI/AMD devices.
1.1239	22-Oct-2015	msaitoh	Add some entries from Intel 965 Express Chipset Family Datasheet.
1.1238	21-Oct-2015	nonaka	Add Samsung SM951 M.2 NVMe SSD.
1.1237	19-Oct-2015	tnn	fix I219_LM2 entry
1.1236	19-Oct-2015	tnn	add I219 family Ethernet devices
1.1235	28-Aug-2015	nonaka	Add Intel Dual Band Wireless AC 3165.
1.1234	27-Aug-2015	nonaka	Added Realtek RTL8192CE.
1.1233	26-Aug-2015	msaitoh	Add Intel PCIe NVMe SSD.
1.1232	23-Aug-2015	jakllsch	Add another ID for Intel 82Q45 KT
1.1231	14-Aug-2015	nonaka	Added ASMedia ASM1042A USB 3.0 Host Controller.
1.1230	14-Aug-2015	knakahara	Add Marvell Armada38x.
1.1229	13-Aug-2015	msaitoh	Add some Red Hat devices.
1.1228	26-Jun-2015	msaitoh	Add PLX Technology PEX 8111.
1.1227	02-Jun-2015	nonaka	fix typo. s/Wirelsss/Wireless/
1.1226	16-May-2015	msaitoh	Add some Intel Core 5G devices and Realtek RTL8188EE.
1.1225	13-May-2015	nonaka	Added Qualcomm Atheros QCA986x/988x.
1.1224	11-May-2015	msaitoh	Add Samsung SM951.
1.1223	09-May-2015	christos	PR/49889: Maxim Tsyplakov: support for D-Link DFE520TX
1.1222	04-May-2015	ryo	add SystemBase SB16C1050 UARTs
1.1221	03-May-2015	msaitoh	Add some NVIDIA devices.
1.1220	02-Apr-2015	tnn	Mobile 5th Generation Intel Core devices, from intel doc #330837-004
1.1219	20-Mar-2015	msaitoh	Add MegaRAID SAS3108
1.1218	20-Mar-2015	msaitoh	Add Xeon E5 v3 and C61x devices.
1.1217	18-Mar-2015	msaitoh	Add some Intel 9 series devices.
1.1216	04-Mar-2015	knakahara	add nForce Host Bridges.
1.1215	15-Feb-2015	msaitoh	Add some devices: - Radeon HD 54xx Audio - Dell DARC 4 devices - Some Intel IOP332 devices - Intel 63xxESB AC'97 Audio Controller - Some Intel E7520 devices
1.1214	25-Jan-2015	msaitoh	- Add another HP vendor ID (was 3PAR) - Add two PLX PCI Bridges.
1.1213	18-Jan-2015	christos	PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.1212	16-Jan-2015	msaitoh	Add some devices - Add COMPAQ iLOs. - Add HP IPMI device. - Add some Intel devices. - Fix product ID of INTEL 5000_PCIE_4. - Add some VIATECH devices.
1.1211	14-Jan-2015	msaitoh	Sort VIATECH's entries in the device ID's order.
1.1210	31-Dec-2014	veego	Add two Intel 9 Series devices
1.1209	23-Dec-2014	msaitoh	Add some Intel 10/40G devices.
1.1208	17-Dec-2014	hikaru	Add QuickAssist Virtual Function Device IDs
1.1207	15-Dec-2014	msaitoh	- Rename DH89xxCC's names from DH89XX_ to DH89XXCC_. - Add some DH89xxCC's devices. - Add DH89XXCL's devices.
1.1206	15-Dec-2014	msaitoh	Change Intel 0x0434 entry: - Rename DH89XX_QA to DH89XXCC_IQIA - Modify the description to DH89xxCC PCIe Endpoint and QuickAssist (include typo fix)
1.1205	30-Oct-2014	nonaka	branches: 1.1205.2; Add some Intel Wi-Fi devices.
1.1204	29-Oct-2014	nonaka	Add Realtek RTL8402 PCI-E Card Reader.
1.1203	01-Oct-2014	msaitoh	Add Intel Atom Z36xx and Z37xx devices from the datasheet.
1.1202	25-Sep-2014	msaitoh	- Add some Intel gigabit Ethernet devices. - Remove trailing white spaces.
1.1201	16-Sep-2014	msaitoh	Add some new Intel I218 devices.
1.1200	01-Sep-2014	msaitoh	Add Samsung Electronics XP941 M.2 SSD
1.1199	30-Jul-2014	msaitoh	branches: 1.1199.2; Add Intel Xeon E3-1200 v3 Host Bridge, DRAM.
1.1198	25-Jul-2014	msaitoh	Add some Intel Core internal devices.
1.1197	08-Jul-2014	msaitoh	Add some RDC devices.
1.1196	03-Jul-2014	msaitoh	Add some Intel X38 devices.
1.1195	02-Jul-2014	msaitoh	Add Intel DH89xx devices.
1.1194	23-Jun-2014	msaitoh	Add some HP iLO devices. Add Matrox MGA G200eH.
1.1193	09-Jun-2014	msaitoh	Add some devices: - Add Intel Bay Trail devices from the document. - Add Intel EP80579 devices from OpenBSD. - Add Intel DH89xx's SMBus controller. - Add ASPEED AST1150 PCIe-to-PCI bridge.
1.1192	30-May-2014	msaitoh	Add Intel 3200 Host and PCIe.
1.1191	28-May-2014	msaitoh	- Add some Intel devices (Sandy Bridge, 6702PXH, X38 and Pineview) - Add ITE IT8213
1.1190	28-May-2014	msaitoh	Add Intel X540-AT2.
1.1189	25-May-2014	njoly	Add GeForce GT 640M.
1.1188	26-Apr-2014	mrg	add some new Attansic ethernet devices (but no driver)
1.1187	09-Apr-2014	msaitoh	Add yet another 82599 device
1.1186	09-Apr-2014	msaitoh	- Add some Ivy Bridge devices from the document. - sort by device ID.
1.1185	07-Apr-2014	njoly	Add Intel Ivy Bridge PCI Express Root Port
1.1184	15-Mar-2014	kiyohara	branches: 1.1184.2; Add some Armada 370 IDs.
1.1183	14-Feb-2014	njoly	Add GeForce 210 High Definition Audio Controller.
1.1182	26-Jan-2014	msaitoh	Add some OXPCIe952 devices.
1.1181	21-Jan-2014	msaitoh	Rename MARVELL2's 0x91a3 entry to 88SE91XX.
1.1180	10-Jan-2014	msaitoh	Rework for Marvell 88SE9128. Change the description of 0x9123 to 88SE912[38]. For 0x91a3, add '(unclear)' into the description. FreeBSD, Linux and http://pci-ids.ucw.cz have no such device.
1.1179	09-Jan-2014	christos	revert unintended changes
1.1178	09-Jan-2014	msaitoh	Remove 88SE9128(0x913a) entry. At least one of 88SE9128 chip's product ID is 0x9123. I have this one. Add new 88SE912X entry with 0x9123. OK'ed by jakllsch.
1.1177	04-Jan-2014	msaitoh	Add Intel Z68 LPC
1.1176	01-Jan-2014	msaitoh	Change description from ASPEED2000 to ASPEED Graphics Family because the IP is used on AST1000,2050 and 2100, too.
1.1175	30-Dec-2013	msaitoh	Add ASPEED AST2000.
1.1174	27-Dec-2013	msaitoh	Add some Marvell 88SE9xxx devices.
1.1173	25-Dec-2013	msaitoh	Add DIGI Neo 8-port serial(PCIe) and OXFORD OXPCIe952 Parallel.
1.1172	23-Dec-2013	msaitoh	Add Intel C2000 devices.
1.1171	22-Dec-2013	nonaka	Added some Realtek devices.
1.1170	21-Dec-2013	msaitoh	Add some Intel E600 and EG20T devices.
1.1169	15-Dec-2013	pgoyette	Add ASMedia ASM1042 xHCI USB3 controller
1.1168	13-Dec-2013	pedro	Add entries for Atheros AR9462 and a new flavour of Intel Haswell Integrated Graphics Device; OK rmind@ mbalmer@
1.1167	12-Dec-2013	nonaka	Added Radeon HD6320 Graphics.
1.1166	08-Dec-2013	msaitoh	Add devices of Atom S1200 series from the datasheet (Table 2-1 Local PCI Hierarchy, datasheet Volume 2 of 2).
1.1165	13-Nov-2013	msaitoh	Add Intel I347-AT4
1.1164	07-Nov-2013	msaitoh	Add Quark X1000 devices.
1.1163	07-Nov-2013	msaitoh	Add some Intel devices. Mainly taken from "Mobile 4th Generation Intel Core Processor Family I/O" datasheet (Document number: 329003)
1.1162	14-Sep-2013	soren	Update some AMD/ATI chipset names. Add Renesas USB 3.0 controllers.
1.1161	04-Sep-2013	christos	PR/48180: Christian Groessler: Support for Oxford Semi OXuPCI952 dual serial PCI card
1.1160	23-Aug-2013	christos	PR/48150: Noriyuki Koizumi: Add support for Intel Centrino Advanced-N 6235 Wi-Fi controller
1.1159	17-Jul-2013	soren	Extend support for Nanjing QinHeng Electronics puc(4) devices.
1.1158	13-Jun-2013	matt	branches: 1.1158.2; 1.1158.4; Add ATI RADEON_HD7340
1.1157	05-Jun-2013	msaitoh	Add some Intel devices from the following datasheets: - Desktop 4th Generation Intel Core Processor Family Datasheet Volume 2 of 2 - Mobile 4th Generation Intel Core Processor Family Datasheet Volume 2 of 2 - Intel Xeon Processor E3-1200 v3 Product Family Datasheet Volume 2 of 2
1.1156	04-Jun-2013	msaitoh	Add some Intel devices from document (Intel 8 Series / C220 Chipset Family Platform Controller Hub (PCH) Datasheet)
1.1155	03-Jun-2013	msaitoh	Add Intel 8 Series KT.
1.1154	20-Apr-2013	rkujawa	Add IDs for Marvell Armada XP. Obtained from Marvell, Semihalf.
1.1153	18-Apr-2013	msaitoh	Fix cut & paste bug in some I210 series.
1.1152	18-Apr-2013	msaitoh	Add Intel I21[0178] Ethernet.
1.1151	06-Mar-2013	msaitoh	Add ALTIMA AC1003, BROADCOM BCM57782 and BCM57786.
1.1150	03-Mar-2013	msaitoh	BCM5785F is not gigabit Ethernet.
1.1149	27-Feb-2013	msaitoh	- Add APPLE BCM5701 - Add BROADCOM BCM57766 - BCM5719 is not 10G but 1G
1.1148	20-Feb-2013	msaitoh	- Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM5719. - Change BCM5720's device ID from 0x1658 to 0x165f. 0x1658 is only for the pre-product and not released to public.
1.1147	30-Jan-2013	msaitoh	Change from "3400 USB" to "3400 USB EHCI" for EHCI devices.
1.1146	06-Dec-2012	msaitoh	Update some Intel LPC devices - Add 82801E LPC - Add 82801GH LPC - Update Intel 5 series and 3400 series LPC from document (5-and-3400-chipset-specification-update) - 0x3b14 is not 3400 but 3420
1.1145	05-Dec-2012	christos	Intel Atom E600 PCI-LPC bridge, adds a watchdog + HPET support. Tested on a Soekris net6501. (jmcneill)
1.1144	29-Nov-2012	msaitoh	Update SMBus devices from Document(c600-series-chipset-spec-update and x79-express-chipset-specification-update)
1.1143	21-Oct-2012	matt	Add more ASMEDIA 1061 variants.
1.1142	20-Oct-2012	msaitoh	Add SB600_USB_OHCI0 again. This was (accidentally) removed in rev. 1.1105
1.1141	20-Oct-2012	matt	Pull in changes from matt-nb5-mips64 (RMI & NETLOGIC).
1.1140	19-Oct-2012	prlw1	Add Intel Centrino Wireless-N 2230
1.1139	29-Sep-2012	gsutre	Add Intel Ivy Bridge host bridge and integrated graphics device IDs, from the Linux Intel AGP driver.
1.1138	25-Sep-2012	gson	Add Lava Computers SSERIAL-PCI single serial port adapter.
1.1137	23-Sep-2012	chs	add a bunch of devices.
1.1136	18-Sep-2012	matt	Fix tpyo.
1.1135	17-Sep-2012	matt	Add ASMEDIA ASM1061
1.1134	17-Sep-2012	tsutsui	Add Broadcom BCM57762 Gigabit Ethernet, per PR kern/46961.
1.1133	13-Sep-2012	nonaka	Added ALTERA EP4CGX15BF14C8N entry.
1.1132	10-Aug-2012	bouyer	branches: 1.1132.2; Add some PCI AHCI controllers, from linux
1.1131	07-Aug-2012	msaitoh	Add some Intel's KT (Serial over LAN) devices.
1.1130	06-Aug-2012	msaitoh	Add 82GM45_KT for puc(4)
1.1129	05-Aug-2012	msaitoh	Add entries for puc(4) devices.
1.1128	05-Aug-2012	bouyer	Add LSI MegaRAID SAS2208
1.1127	27-Jul-2012	matt	Add EG20T PCH. Closes PR/45567.
1.1126	26-Jul-2012	uwe	Add VirtualBox ids.
1.1125	25-Jul-2012	msaitoh	Add some Intel devices from document (Intel 7 Series / C216 Chipset Family Platform Controller Hub (PCH) Family Specification Update) Modify some devices to clarify
1.1124	23-Jul-2012	msaitoh	Add some ATI and SIS devices. Fixes PR#39580
1.1123	23-Jul-2012	msaitoh	Add some Intel 7 series devices.
1.1122	22-Jul-2012	matt	Add more FREESCALE devices Fix RICOH RU5230 description
1.1121	20-Jul-2012	msaitoh	Add some Intel's devices.
1.1120	19-Jul-2012	msaitoh	Add XGI Technology's devices.
1.1119	12-Jul-2012	kiyohara	Add Marvell SoC 88F6282.
1.1118	26-Jun-2012	drochner	fix pci id for the Intel H61 LPC bridge to fit actual hardware (and the chipset's documentation), and add some PCI-to-legacy bridges found on recent boards
1.1117	25-May-2012	msaitoh	Add some Intel I350 Ethernet products.
1.1116	18-Apr-2012	bouyer	Add various newer LSI RAID/SAS controllers, from OpenBSD.
1.1115	13-Apr-2012	cegger	Add AMD Family15h ids
1.1114	09-Apr-2012	mlelstv	Add Intel Sandybridge integrated graphics
1.1113	08-Apr-2012	jakllsch	Make the RS690 entries more closely match the terminology of the RS690 RRG/RPR/BDG documents.
1.1112	08-Apr-2012	jakllsch	- NEC uPD720100A is USB2. - Add NEC uPD720200, a USB3 Host Controller
1.1111	08-Apr-2012	jakllsch	Add IDs for Fresco Logic and their FL1000 and FL1009 xHCI chips. Derived from OpenBSD.
1.1110	07-Apr-2012	christos	more broadcom stuff
1.1109	04-Apr-2012	uwe	Add Marvell Yukon 88e8040 ethernet.
1.1108	28-Mar-2012	pgoyette	Add JMB 362 (as found on ASUS M5A99X motherboard)
1.1107	21-Mar-2012	sborrill	Add some Symbios SAS2108 devices. Add new Symbios SAS2008 device (as used by IBM ServeRAID M1015).
1.1106	18-Mar-2012	tsutsui	Add VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda.
1.1105	03-Mar-2012	nonaka	Added some AMD/ATI devices.
1.1104	02-Mar-2012	nonaka	Added Ricoh 5U822. Rename Ricoh 5CE823 to 5U823.
1.1103	02-Mar-2012	nonaka	Added Ricoh 5CE823.
1.1102	15-Feb-2012	tsutsui	branches: 1.1102.2; Add VIA VX900 Host Bridge, to handle its quirk. PR/46018 Ok releng@
1.1101	18-Oct-2011	jmcneill	branches: 1.1101.2; 1.1101.6; add a couple vmware product IDs
1.1100	06-Oct-2011	jmcneill	add GeForce GTX 550 Ti
1.1099	04-Oct-2011	njoly	Add 3WARE 9750 Series RAID.
1.1098	23-Sep-2011	njoly	Add NVIDIA GeForce 210.
1.1097	05-Sep-2011	msaitoh	Change Intel 0x1229 product name from 82557 to 8255x
1.1096	02-Sep-2011	msaitoh	Add some fxp devices.
1.1095	15-Aug-2011	msaitoh	Sort lines.
1.1094	11-Aug-2011	bouyer	Add SiS 315 Pro VGA
1.1093	09-Aug-2011	mbalmer	Add the QUANCOM PWDOG1.
1.1092	06-Aug-2011	jakllsch	Add Marvell 88SE9128.
1.1091	06-Aug-2011	jakllsch	Adjust inconsistent whitespace in MARVELL products.
1.1090	06-Aug-2011	mbalmer	It's QUANCOM Electronic GmbH, not Quancm.
1.1089	04-Aug-2011	jakllsch	Add Conexant CX23885
1.1088	23-Jul-2011	matt	Add MARVELL2 vendor and 88SE9123 device.
1.1087	28-Jun-2011	sborrill	Add Nvidia GeForce 8500 GT
1.1086	27-Jun-2011	nonaka	Added some JMicron PCI-E SD/MMC/MS/xD Host Contollers.
1.1085	07-Jun-2011	cegger	add ATI Radeon Cypress and Lucent IEEE 1394b Controller
1.1084	06-Jun-2011	msaitoh	Rename 82801DB LPC Interface Bridge for consisitency.
1.1083	28-May-2011	phx	branches: 1.1083.2; Renamed VT6410 to VT6410_RAID.
1.1082	28-May-2011	phx	Add VIA VT6410 ATA RAID controller.
1.1081	28-May-2011	ryo	add SystemBase SB16C1058 8 port serial cards
1.1080	27-May-2011	matt	Add ALI M5288 SATA/Raid Controller
1.1079	22-May-2011	mrg	add some fujitsu ethernet id's and a new vendor id. partly from Michael Moll <kvedulv@kvedulv.de>, taken from openbsd. XXX: move the #if 0'd usage of PCI_VENDOR_CITICORP inside the #if 0 code XXX: region, as this id conflits with the (4th) fijutsi id.
1.1078	21-May-2011	msaitoh	Add some iwn(4) devices.
1.1077	20-May-2011	msaitoh	Add Intel 82567V, 82580 Quad-1000BaseX and Hanksvill with ICH10.
1.1076	15-May-2011	msaitoh	Add entries for Intel 6 series' devices.
1.1075	14-May-2011	martin	Add Nanjing QinHeng Electronics, from PR kern/44956.
1.1074	13-May-2011	msaitoh	add entries for 82579LM and 82579V
1.1073	02-May-2011	manu	Don't conflict SUNIX new vendor Id with the previous one
1.1072	02-May-2011	manu	Add new PCI vendor Id for Sunix, and the corresponding device Id.
1.1071	13-Apr-2011	wiz	+ ATI Radeon HD4650.
1.1070	04-Apr-2011	bouyer	Add RDC PCI-ISA bridge
1.1069	04-Apr-2011	bouyer	Add rdcide(4), a driver for the IDE controller found in RDC's vortex86/PMX-1000 system-on-chip.
1.1068	16-Mar-2011	matt	Add Freescale P2020/P2010 device ids
1.1067	05-Mar-2011	matt	Add ATHEROS AR2427 and AR9287
1.1066	25-Feb-2011	jmcneill	add AMD (ATI) Radeon HD 6310 Graphics ID
1.1065	25-Feb-2011	jmcneill	add AMD Family14h host bridge IDs
1.1064	24-Feb-2011	matt	Update desc for RT3090. Add Intel NM10_LPC
1.1063	24-Feb-2011	matt	Add RT3090
1.1062	23-Feb-2011	jmcneill	add Realtek RTL8188CE Wireless LAN 802.11n PCI-E NIC
1.1061	22-Feb-2011	jmcneill	Add Atheros AR8132, AR8151, AR8151 v2.0, AR8152 v1.1, and AR8152 v2.0
1.1060	21-Feb-2011	njoly	Add two more Intel 82Q45 devices.
1.1059	20-Feb-2011	msaitoh	Add MosChip(Netmos) 9904 and 9922. Add MosChip(Netmos) 9865 (already supported by puc(4))
1.1058	18-Feb-2011	jmcneill	add Adaptec RAID 3405
1.1057	03-Feb-2011	njoly	Add a few Nvidia devices.
1.1056	26-Jan-2011	bouyer	Add RDC Semiconductor R6040 10/100 Ethernet
1.1055	11-Jan-2011	dyoung	branches: 1.1055.2; 1.1055.4; Add Intel 82599 product IDs.
1.1054	16-Dec-2010	cegger	Brdige -> Bridge
1.1053	11-Dec-2010	matt	Add some more devices that I've encountered.
1.1052	11-Nov-2010	cegger	add AMD RD880, ATI Radeon HD5600, ATI Radeon HD4200 and Atheros AR9285
1.1051	06-Nov-2010	jakllsch	Add Intel SCH IDE Controller, from OpenBSD via Tomokazu HARADA in PR#42310.
1.1050	14-Oct-2010	taca	Remove duplicate entries: * "vendor SYBATECH" exists as vendor SYBA. * "product INTEL 82915G_IGD" simply duplicated.
1.1049	26-Aug-2010	dogcow	Add several Broadcom BCM43xx devices.
1.1048	01-Aug-2010	msaitoh	Add VScom PCI-200Li
1.1047	30-Jul-2010	njoly	Update ICH10 devices, from Intel I/O Controller Hub 10 datasheet errata.
1.1046	29-Jul-2010	njoly	Adjust ICH10 watchdog timer entries which are really LPC interface bridges. While here consistently use `xxx LPC Interface Bridge' for Intel products.
1.1045	29-Jul-2010	njoly	Add Matrox MGA G200eW. From OpenBSD.
1.1044	27-Jul-2010	msaitoh	Add Quatech, System Base, Pcre Systems and Syba Tech.
1.1043	13-Jul-2010	msaitoh	Add Advantech PCI-1604
1.1042	13-Jul-2010	kiyohara	Add Marvell MV78100/MV78200 SoC. (Discovery Innovation)
1.1041	08-Jul-2010	msaitoh	- Add "vendor SYBA" and two cards. - Add two Avlab cards. - Add two Moxa cards. - Add some VScom cards.
1.1040	08-Jul-2010	msaitoh	Add Advantech PCI-1610 and PCI-1612
1.1039	07-Jul-2010	msaitoh	0x1409 is SUNIX
1.1038	06-Jul-2010	msaitoh	Fix the name of PCI-1620A-AE.
1.1037	06-Jul-2010	msaitoh	Add Advantech and Decision Computer's vendor IDs and their products.
1.1036	02-Jul-2010	mrg	add NetMOS 9901 PCI-E uart
1.1035	16-Jun-2010	riz	Add a bunch of Intel host bridge and integrated graphics device IDs, from the Linux Intel AGP driver and PR#42906 from Henry Bent.
1.1034	08-Jun-2010	pgoyette	Add a couple entries for ATI Radeon 4250 HD
1.1033	08-Jun-2010	mrg	add two pciids for the satalink driver, from the linux driver, for PR#32577.
1.1032	15-May-2010	jakllsch	Add NetChip (now PLX) vendor ID, as well as NET228[02] device IDs.
1.1031	12-May-2010	christos	Use device names from the OpenBSD iwn driver, adding and modifying device names as needed. From Sverre Froyen
1.1030	10-May-2010	matt	Add MARVELL 88SB2211 PCIe-PCI bridge
1.1029	08-May-2010	pgoyette	Add a couple more SMBus controllers. From FreeBSD.
1.1028	07-May-2010	kiyohara	Add Marvell MV64460. Rename Marvell GT64360 to MV64360.
1.1027	17-Apr-2010	christos	add support for the 6000. From sverre froyen
1.1026	02-Apr-2010	jakllsch	Add Intel 82855GM Host-AGP Bridge.
1.1025	14-Mar-2010	mrg	branches: 1.1025.2; add several core i7 devices, and some more ricoh card reader controllers.
1.1024	25-Feb-2010	macallan	add 3Dlabs Wildcat 5110, also known as Sun XVR-500
1.1023	16-Feb-2010	msaitoh	Add entries for 82576 and 82580
1.1022	09-Feb-2010	hubertf	Add entry for Juniper Networks Experimental Clock Version 0 Fixes PR kern/42742
1.1021	09-Feb-2010	msaitoh	Make some bge's name close to the real product names.
1.1020	08-Feb-2010	msaitoh	Add some MegaRAID Entries
1.1019	03-Feb-2010	macallan	branches: 1.1019.2; add Apple PowerBook G3 Firewire controllers
1.1018	30-Jan-2010	jakllsch	Add Qumranet vendor and Virtio device IDs. Also, sort some vendor entries.
1.1017	24-Jan-2010	msaitoh	Add BCM5717,5718,5723,5724,5787F,5761,5761E,5764,5761S,5761SE,5760,57788, 57780,57790,5784M,5785F,5785G,57761,57781,57791,57765,57785 and 57795.
1.1016	21-Jan-2010	hubertf	add/improve a few devices
1.1015	19-Jan-2010	mbalmer	Add two IBM RSS devices.
1.1014	14-Jan-2010	sborrill	Fix typo in (unused) vendor name
1.1013	12-Jan-2010	msaitoh	add i82567LM-2 i82567LM-4 i82567V-3 LAN controller
1.1012	12-Jan-2010	msaitoh	rename some wm devices to know PHY type
1.1011	12-Jan-2010	nonaka	Added RENESAS SH7780/SH7785 PCI Controller.
1.1010	11-Jan-2010	msaitoh	fix for i82577
1.1009	07-Jan-2010	msaitoh	Add PCH LAN and i82575
1.1008	07-Jan-2010	jdc	Add Sun Cassini/NS Saturn devices.
1.1007	03-Jan-2010	jnemeth	PR/42558 - Ryo ONODERA -- add entry for IO-DATA TV tuner card
1.1006	29-Dec-2009	msaitoh	add entry for i82583V
1.1005	13-Dec-2009	matt	Add RMI (NetLogic)
1.1004	05-Dec-2009	dsl	Correct what are clearly typos leading to duplicate vendor:product pairs. Where possible values have been checked against the pcidatabase and x.org. In some cases I've only commented on some dubious entries! Values will be picked up by the next update. Info from PR/36487
1.1003	30-Nov-2009	sborrill	Add Intel 3400 devices
1.1002	21-Nov-2009	jakllsch	Add CX700 and VX800/VX820 LPC bridge IDs.
1.1001	18-Nov-2009	bouyer	Complete BROADCOM BCM5709 family, and add BCM5716 family.
1.1000	18-Nov-2009	jakllsch	Drop unused/duplicate VIATECH VT8237_PCIB entry. While here, drop "(Apollo KT600)" from VIATECH VT8237, and s/ISA/LPC/. Additionally, add LPC Bridge ID for VT8251 from PR/37517.
1.999	13-Oct-2009	cegger	- replace 'Advanced Micro Devices' with 'AMD'. Requested by jmcneill@ - remove duplicate printing of 'AMD' with the AMD HUDSON-2 chipset.
1.998	13-Oct-2009	cegger	AMD SB900 renamed to AMD HUDSON-2
1.997	05-Oct-2009	cegger	add SB900 IDE/SATA ids
1.996	20-Sep-2009	christos	PR/42100: Dave J. Barnes: Support for old toshiba PICCOLO IDE controllers.
1.995	01-Aug-2009	matt	Add Silicon Motion SM502
1.994	07-Jul-2009	cegger	correct product ids for RD890 chipset
1.993	29-Jun-2009	hubertf	add a few sdhc(4) devices
1.992	12-Jun-2009	rjs	Add some more Marvell devices.
1.991	11-Jun-2009	cegger	add ADP2, Marvell and Triones devices. Taken from mvsata diff from Kiyohara Takashi, resolved apply conflicts.
1.990	29-May-2009	skrll	Add a bunch of HP stuff. From OpenBSD.
1.989	18-May-2009	skrll	Add more devices in my c3700.
1.988	13-May-2009	cegger	add RX780/RX790 chipset host bridge and Radeon X1600
1.987	09-May-2009	sborrill	Add alternative product ID for VIA VT8237A Integrated SATA Controller
1.986	30-Apr-2009	skrll	Add another NS product found in hppa machines.
1.985	30-Apr-2009	skrll	Add the National Semiconductor PC87560 Legacy IO chip on systems with PA-8500 and later 64-bit CPUs.
1.984	28-Apr-2009	ginsbach	Add Mellanox Technologies Infiniband devices.
1.983	25-Apr-2009	tsutsui	Tweak some Realtek re(4) device names.
1.982	07-Apr-2009	msaitoh	add 82801J_D_BM_LF (ICH10)
1.981	10-Mar-2009	msaitoh	add i82567LF-3 LAN Controller
1.980	28-Feb-2009	schwarz	correct entry for Lucent 0x5802 into 2-port USB Host Controller
1.979	26-Feb-2009	mjf	Add Initio INIC-1622 SATA Controller.
1.978	19-Feb-2009	sborrill	Add Intel i82574L 1000baseT Ethernet as used in Pro/1000 CT Desktop adapter
1.977	11-Feb-2009	markd	Add more ICH9 SATA controller variants
1.976	06-Feb-2009	jakllsch	branches: 1.976.2; Add TI PCIxx12-series product IDs.
1.975	27-Jan-2009	markd	Add Intel 82567LM_3 ethernet
1.974	21-Jan-2009	jnemeth	add a comment explaining how to regenerate pcidevs.h and pcidevs_data.h
1.973	19-Jan-2009	jmcneill	Attansic L1E Gigabit Ethernet Adapter
1.972	07-Jan-2009	jnemeth	fix typo in LUCENT ET1301
1.971	27-Dec-2008	tsutsui	Fix tyop in previous.
1.970	27-Dec-2008	jnemeth	PR/39094 - Kaspar Brand -- Add et (Agere ET1310/ET1301) netowrk driver
1.969	21-Dec-2008	nonaka	Add VIA VT8237S Integrated SATA Controller ID.
1.968	16-Dec-2008	christos	add more intel iwn devices
1.967	29-Nov-2008	christos	Add support for the Intel G45 AGP. From Arnaud Lacombe
1.966	18-Nov-2008	matt	Make more consistent (change capitalization) etc. Shrinks by 33 tokens and 300 bytes.
1.965	08-Nov-2008	christos	Support for Intel G35 as found on Asus P5E-VM HDMI motherboard from Milos Negovanovic
1.964	04-Nov-2008	cegger	fix typo in previous: prodcut -> product
1.963	04-Nov-2008	cegger	add ATI RD790 and RD890 chipset ids
1.962	15-Oct-2008	bouyer	branches: 1.962.2; 1.962.4; Fix entry description.
1.961	15-Oct-2008	bouyer	Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.
1.960	01-Oct-2008	christos	ICH10 support from Brad du Plessis
1.959	28-Sep-2008	bouyer	Add JMicron JMC250 Gigabit Etheret and JMC260 Fast Ethernet Controllers. While here, fix other JMicron device IDs (product ID is 4 digits, not five).
1.958	18-Sep-2008	cegger	correct C51 host bridge numbers. Spotted by njoly@
1.957	18-Sep-2008	cegger	add NVIDIA C51 products
1.956	13-Sep-2008	njoly	Add Nvidia GeForce 6150 devices.
1.955	08-Sep-2008	cegger	Serverworks: correct product id for LPC, add XIOAPIC and Watchdog Timer
1.954	08-Sep-2008	cegger	correct product id for serverworks usb controllers
1.953	06-Sep-2008	rmind	PR/39119: Jan Thorsson: Support for VIA CX700M2 IDE controller.
1.952	06-Sep-2008	rmind	PR/38012: David Dudley: Support for B&B Electronics MIPort serial cards.
1.951	05-Sep-2008	cegger	Add some ServerWorks products
1.950	27-Aug-2008	christos	Add one more cardbus bridge
1.949	23-Aug-2008	tsutsui	- add a Realtek 8139D variant (from FreeBSD) - remove a duplicated entry of Realtek 0x8138
1.948	22-Aug-2008	tnn	Add Intel 82945GME Host Bridge Intel 82945GME Integrated Graphics Device
1.947	19-Aug-2008	matthias	Typo
1.946	19-Aug-2008	matthias	Add pci ids for intel 82946GZ_{HB,IG} (used in my AcerPower 2000)
1.945	06-Aug-2008	dyoung	Define O2 Micro OZ711E0. From OpenBSD.
1.944	29-Jul-2008	tron	Add nVidia GeForce 7300 LE.
1.943	15-Jul-2008	tonnerre	Add some PCI IDs from the Thinkpad T61/X61/R61 series as observed in PR 37988 and on my own T61.
1.942	25-May-2008	chs	branches: 1.942.2; 1.942.4; add devices found in the last model of powermac G5. from openbsd.
1.941	20-May-2008	cegger	add some ATI products (SB600 modem, RD790 and RS690)
1.940	13-May-2008	rjs	Add SIS 756 Host Bridge.
1.939	08-May-2008	sborrill	Add support for IBM ServeRAID 8k (and 8k-l)
1.938	30-Apr-2008	joerg	branches: 1.938.2; + 3ware 9690
1.937	17-Apr-2008	christos	branches: 1.937.2; 1.937.4; sync with openbsd; fixes observed corruption issues.
1.936	15-Apr-2008	cegger	add devices found on HP Pavilion dv9700 Notebook
1.935	09-Apr-2008	joerg	Add PCI ID of 3Ware 9650 family.
1.934	03-Apr-2008	riz	Digi International Neo 4 and 8 port serial boards.
1.933	01-Apr-2008	cegger	add XenSource and its virtual product - only found on Xen HVM guests
1.932	26-Mar-2008	jmcneill	PR#38301: Add support for SMBus controllers for ATI SB600, SB700, SB800 chipsets, from Christoph Egger
1.931	24-Feb-2008	xtraeme	+ Dell PERC 6.
1.930	22-Feb-2008	tsutsui	Add I-O DATA RSA-PCI 2 port serial board.
1.929	13-Feb-2008	tron	branches: 1.929.2; 1.929.6; Import PCI ids for the Matrox MGA G200e and the Intel E7230 and 82975X chipsets from OpenBSD.
1.928	10-Feb-2008	ichiro	add few devices PM965 interface
1.927	05-Feb-2008	ichiro	added some devices of "PCI-Cardbus bridge"
1.926	11-Jan-2008	tnn	Add ServeRAID 6/7
1.925	11-Jan-2008	ragge	Add Intel quad 1000baseT copper PCI express card.
1.924	03-Jan-2008	jnemeth	Symbios Logic 0x0030 is used for both 53c1020/53c1030 from Dave Huang
1.923	30-Dec-2007	jdc	Add Apple Shasta GMAC (From FreeBSD).
1.922	28-Dec-2007	dholland	typo fix: brige -> bridge
1.921	24-Dec-2007	hamajima	Add Intel 82801GB Ethernet Controller
1.920	26-Nov-2007	bjs	branches: 1.920.2; 1.920.6; Add PCI IDs for the ATI Radeon X800 GTO (R430). Most sources I have consulted indicated that this ID is an R430 (including xf86-video-ati). PCI_PRODUCT_ATI_RADEON_R430_554F (primary) PCI_PRODUCT_ATI_RADEON_R430_556F (secondary) vga0 at pci1 dev 0 function 0: ATI Technologies product 0x554f (rev. 0x00) ATI Technologies product 0x556f (miscellaneous display) at pci1 dev 0 function 1 not configured
1.919	24-Nov-2007	markd	Add INTEL Q33 and Q35 chipsets
1.918	24-Nov-2007	markd	s/product\tINTEL/product INTEL/
1.917	23-Nov-2007	xtraeme	Add some Intel 5000 Series Chipset devices.
1.916	19-Nov-2007	ad	Add a few more pcibs
1.915	15-Nov-2007	manu	It's 88C055, not 88CO55.
1.914	14-Nov-2007	xtraeme	Add NVIDIA MCP7[37] devices, from OpenBSD.
1.913	12-Nov-2007	joerg	Merge a PCI IDs of Intel 82G33 and 82965PM from jmcneill-pm.
1.912	30-Oct-2007	xtraeme	branches: 1.912.2; Erm fix previous, it's Intel IOP333 not 332.
1.911	30-Oct-2007	xtraeme	Add Intel IOP332 A/B Segments.
1.910	30-Oct-2007	xtraeme	Areca ARC SATA RAID Controllers, from OpenBSD.
1.909	30-Oct-2007	jnemeth	PR/37201 - Yasushi Oshima -- Intel 82965G chipset support
1.908	24-Oct-2007	xtraeme	- Add support for the ServerWorks HT-1000 IDE Controller in rccide. - Add support for another ServerWorks HT-1000 SATA Controller. From OpenBSD.
1.907	22-Oct-2007	manu	Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.906	22-Oct-2007	nisimura	PCI ID 10b5.9030 and 10b5.5201 are for PLX I/O accelerator product model 9030 and 9050/9052. They work a functional bridge to host any kind of I/O devices. The exact product type can be identified via VPD (vital product ID) PCI configuration registers if they are programmed by the particular PLX I/O accelerator applications. Otherwise, it's not possible to distinguish which kind of function is implemented inside.
1.905	14-Oct-2007	xtraeme	branches: 1.905.2; IC Plus Corp IP100A support for ste(4), submitted by Pedro Alejandro Lopez-Valencia in PR kern/36605.
1.904	09-Oct-2007	garbled	Add a bunch of IBM PCI dev id's for graphics cards and PCI bridges
1.903	20-Sep-2007	xtraeme	branches: 1.903.2; + ATI Radeon HD2600 XT
1.902	17-Sep-2007	msaitoh	addd SYMBIOS SAS1068_2, SAS1064E_2 and SAS1068E_2
1.901	11-Sep-2007	jnemeth	sort CMDTECH and add CMDTECH SiI3531 SATALink
1.900	09-Sep-2007	xtraeme	Another ICH8M SATA Controller found in jmcneill's laptop.
1.899	08-Sep-2007	ober	Add definitions for Intel 4965AGN.
1.898	07-Sep-2007	cube	Add EVE.
1.897	02-Sep-2007	ober	Add Intel LPC 82801HEM and 82801IO entries
1.896	01-Sep-2007	ober	ICH8M LPC Interface bridge. ok xtraeme
1.895	31-Aug-2007	xtraeme	+ VIA CX700 IDE Controller
1.894	30-Aug-2007	xtraeme	+ Intel 82801HBM IDE Controller
1.893	28-Aug-2007	msaitoh	add ICH9's Ethers
1.892	26-Aug-2007	xtraeme	branches: 1.892.2; Add two ICH8 LPC Interface Bridges.
1.891	19-Aug-2007	riz	Add 27 Intel ICH9 device IDs, from: http://download.intel.com/design/chipsets/specupdt/31697301.pdf
1.890	14-Aug-2007	jnemeth	PR/36759 - Christoph Egger -- add some MARVELL devices
1.889	11-Aug-2007	jnemeth	PR/36759 - Christoph Egger -- AMD product updates
1.888	08-Aug-2007	tron	Add some Attansis and VIA devices, correct order. Patch provided by Christoph Egger on "tech-kern" mailing list.
1.887	06-Aug-2007	martin	branches: 1.887.2; Add a serverworks south bridge
1.886	04-Aug-2007	kiyohara	Add support agp_amd64. Imported from FreeBSD.
1.885	07-Jul-2007	jklos	branches: 1.885.2; 1.885.6; Added Chelsio vendor plus 10 gig hardware (and friends).
1.884	06-Jul-2007	hpeyerl	ATI updates from Christoph Egger
1.883	27-Jun-2007	xtraeme	Improve the NS CS5535 entries and remove the NS part, it's redundant.
1.882	26-Jun-2007	xtraeme	Add National Semiconductor CS5535 devices, found in the decTOP.
1.881	28-May-2007	mrg	add ALi M5261 ethernet (tulip clone)
1.880	19-May-2007	tsutsui	Add IBM GXT250P video.
1.879	16-May-2007	mrg	add Nvidia MPC04 SMBus
1.878	02-May-2007	bouyer	Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers.
1.877	29-Apr-2007	bouyer	Remove extra "Intel" from previous commit
1.876	29-Apr-2007	bouyer	Add: Intel i82801H IFE (GT) LAN Controller Intel i82801H IFE (G) LAN Controller from FreeBSD
1.875	27-Mar-2007	pooka	add GeForce Go 7600
1.874	25-Mar-2007	isaki	Add ATI ES1000.
1.873	17-Mar-2007	dyoung	Add Infineon ADM5120 PCI host bridge.
1.872	10-Mar-2007	christos	branches: 1.872.2; 1.872.4; PR/35970: Todd Kover: Add support for HP ML110 G2 / Adaptec 2610SA
1.871	09-Mar-2007	xtraeme	branches: 1.871.2; +ATI Radeon X1300
1.870	03-Mar-2007	jmcneill	Add pcHDTV HD-5500 HDTV video capture
1.869	01-Mar-2007	jmcneill	Add Conexant CX2388X Audio/Video Decoder
1.868	16-Feb-2007	markd	branches: 1.868.2; Add i965Q
1.867	10-Feb-2007	mlelstv	Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.866	30-Jan-2007	msaitoh	add some msk devices
1.865	28-Jan-2007	chs	add another Realtek wifi, some SIIG serial cards and another nForce chipset. remove some duplicate entries.
1.864	27-Jan-2007	isaki	Fix typo. s/Seria ATA/Serial ATA/
1.863	21-Jan-2007	isaki	Add AMD Geode LX and CS5536. Reviewed by tsutsui@ on port-i386.
1.862	14-Jan-2007	rpaulo	Add new Parallels video id.
1.861	06-Jan-2007	jmcneill	Add Xbox SMBus controller, and rename "Xbox nForce" to accurately reflect that it is the Host-PCI controller.
1.860	04-Jan-2007	jmcneill	It's nForce Ethernet, not Ethetnet
1.859	04-Jan-2007	jmcneill	Add Xbox nForce video, from Andrew Gillham.
1.858	31-Dec-2006	ichiro	add RCF-CBA2 CF-Adapter (WORKBIT-CF32A)
1.857	10-Dec-2006	bouyer	Add another PCI ID for Dell PERC 5 (from OpenBSD)
1.856	09-Dec-2006	manu	Add support for 4 and 8 ports modem boards from Perle systems
1.855	26-Nov-2006	tsutsui	branches: 1.855.2; - add more Broadcom product IDs, and fix two alternative IDs (from FreeBSD and OpenBSD) - sort by product ID
1.854	26-Nov-2006	enami	Add product id of yet another i82557 variant.
1.853	18-Nov-2006	tsutsui	Add an ID of Broadcom BCM5752M. From Matt Kolb on tech-kern.
1.852	18-Nov-2006	tsutsui	- add Realtek RTL8100 (from FreeBSD) - fix entry of RTL8169SC - remove vague comment in Realtek entry
1.851	17-Nov-2006	tsutsui	Add and note more Realtek Ethernet chips.
1.850	13-Nov-2006	xtraeme	Fix typo in previous (hi bouyer).
1.849	13-Nov-2006	bouyer	Add NetMos 9855 Quad UART and 1284 Printer port
1.848	08-Nov-2006	rpaulo	Add AMD PCI-X IOAPIC.
1.847	08-Nov-2006	martti	Fix aligment.
1.846	08-Nov-2006	martti	List MEGARAID_SAS and MEGARAID_VERDE_ZCR as SYMBIOS products.
1.845	07-Nov-2006	martti	Added some MegaRAID SAS controllers.
1.844	03-Nov-2006	abs	Add Broadcom BCM5754 (0x167a)
1.843	31-Oct-2006	manu	Add support for Perle systems PCI-RAS 4 modem ports See http://www.perle.com/products/Multimodem-Cards.shtml for product details
1.842	25-Oct-2006	bouyer	Add VIA VT8237A Integrated SATA Controller, from FreeBSD.
1.841	23-Oct-2006	rjs	Add Motorola MPC5200B host bridge.
1.840	21-Oct-2006	bouyer	Add support for the Intel 80003 Gigabit Ethernet controller (found e.g. in newer server chipsets) to wm(4), from the FreeBSD em(4) driver. While there, add a few other Intel Ethernet controller that should work as is. Properly update the RX error and TX collision counters. Add ikphy(4), a driver for the Intel i82563 Kumeran 10/100/1000 Ethernet PHYs
1.839	09-Oct-2006	cube	Add one puc device by Moschip (formerly NetMos). Not the one I have, but as I went through the datasheets...
1.838	02-Oct-2006	cube	Oops, pasto.
1.837	02-Oct-2006	cube	Add an NVIDIA MCP55 device I forgot (ISA bridge).
1.836	29-Sep-2006	cube	Add ids of devices found in the nForce 570 chipset. Add GeForce 6200LE.
1.835	27-Sep-2006	cube	Add the Intel PRO/100 variant found on the D946GZIS motherboard.
1.834	27-Sep-2006	sketch	Add TI ACX100B and rename existing (unused) TI ACX100 to ACX100A.
1.833	25-Sep-2006	heas	Add entries for Datum Inc. (Bancomm) & the BC635PCI_U card.
1.832	24-Sep-2006	briggs	Add IDs for the devices in the last-generation PowerBook G4.
1.831	07-Sep-2006	riz	branches: 1.831.2; Add the SK-9Sxx Gigabit Ethernet, from OpenBSD.
1.830	07-Sep-2006	itohy	branches: 1.830.2; Add Workbit NinjaATA-32Bi and NPATA-32 busmastering PIO IDE controllers.
1.829	03-Sep-2006	xtraeme	Add missing PDC20265 device... removed accidentally in previous.
1.828	03-Sep-2006	xtraeme	Add some missing devices from Promise and sort them.
1.827	03-Sep-2006	xtraeme	Add some ICH8 devices... from obsd.
1.826	03-Sep-2006	xtraeme	Erm forgot to add MCP61/65 SATA Controllers and sort them.
1.825	03-Sep-2006	xtraeme	- Add all pciids for nvidia MCP61/65 controllers, from obsd. - Improve description in all them.
1.824	03-Sep-2006	xtraeme	- Add nForce MCP61/65 LAN controllers... from obsd. - Improve description in some nForce controllers as well.
1.823	27-Aug-2006	christos	PR/34289: St�phane Witzmann: Add some missing PCI devices in pcidevs
1.822	24-Aug-2006	riz	Add a bunch of Marvell product IDs, from OpenBSD.
1.821	23-Aug-2006	cube	Add Broadcom NetXtreme II devices.
1.820	22-Aug-2006	riz	Rename vendor GALILEO -> vendor MARVELL . Marvell bought Galileo years ago, we have little existing stuff in the tree, and this will make it easier to keep certain network drivers in sync with other OSes. OK matt@, uwe@
1.819	13-Aug-2006	dsainty	Add the Promise PDC20771, present on the Promise FastTrak TX2300
1.818	10-Aug-2006	simonb	Fix typo in 3945 Intel wireless card name. Pointed out in private mail by Jean-Baptiste Campesato.
1.817	09-Aug-2006	jnemeth	add some Token Ring devices
1.816	08-Aug-2006	cube	Add device IDs for Intel's 631xESB/632xESB southbridges, and start the list for the 5000X northbridge. I'll add the rest later, the spec is currently unavailable so I'll collect them from another source.
1.815	06-Aug-2006	jmcneill	branches: 1.815.2; Add entry for IBM ThinkPad audio. From Stephane Witzmann.
1.814	03-Aug-2006	bsh	Device 10de:026c is not AC-97 controller but nForce430 HD Audio controller. Pointed out by Brad. Thanks.
1.813	01-Aug-2006	xtraeme	Add nForce3 250 SATA2.
1.812	01-Aug-2006	bsh	add an another device ID for AC-97 controller in nForce430 (MCP51). XXX: I found this ID on MSI K8NGM-V, which has nForce410. Maybe 0x26b is for 410 and 0x26c for 430?
1.811	30-Jul-2006	jmcneill	Add VIA VT3314 (CN900) UniChrome integrated graphics
1.810	30-Jul-2006	oster	Recognize another Intel PRO/100 VM Network Controller.
1.809	30-Jul-2006	simonb	Add 82945P_IGD. Should this be "82945G_IGD", and other 945 device IDs have G instead of P suffix?
1.808	21-Jul-2006	gdamore	Update ADMtek pci device ids to match data sheets. While here, add the 9513 device (although I've never seen one, I have a datasheet for it. :-)
1.807	18-Jul-2006	itohy	Identify vendor 0x1317, product 0x9511 as ADMtek AN983B, since OpenSolaris says it is "ADMTek Comet 983b".
1.806	13-Jul-2006	kent	- add VIA VT8237A/8251 High Definition Audio Controller - add some tabs
1.805	07-Jul-2006	simonb	Add another Intel 3945 wireless minipci device id. Small part of kern/33778 by Jean-Baptiste Campesato.
1.804	30-Jun-2006	nisimura	Add kse(4) driver for Micrel KSZ8842/8841 Ethernet controller.
1.803	30-Jun-2006	xtraeme	Add ITE IT8211 IDE Controller.
1.802	30-Jun-2006	xtraeme	Add some missing ATI devices.
1.801	27-Jun-2006	kent	improve the description of NVIDIA MCP55_HDA
1.800	26-Jun-2006	xtraeme	Add SIS 660, 760, 761 IDE Controllers and SIS 181/182 SATA controllers (from obsd).
1.799	26-Jun-2006	xtraeme	Add PROMISE PDC20775 and PDC40718 SATA controllers (from obsd).
1.798	25-Jun-2006	soren	Add quirk for the not quite standard PCI bus in Parallels Desktop for Mac.
1.797	18-Jun-2006	christos	PR/33761: dieter roelants: NetBSD doesn't recognize a RealTek 8168B nic
1.796	18-Jun-2006	xtraeme	Add ALI M1533 PCI-ISA Bridge
1.795	15-Jun-2006	heinz	Added information about LSI's SAS, parallel SCSI and FC adapters (from LSI's Linux driver package and http://pciids.sourceforge.net/).
1.794	14-Jun-2006	grant	branches: 1.794.2; QLogic is the correct writing, not "Q Logic".
1.793	14-Jun-2006	grant	add QLogic QLA200 fibre channel card.
1.792	08-Jun-2006	kent	add 23 ICH8 devices. http://download.intel.com/design/chipsets/specupdt/31305701.pdf No regen follows because no one refers them.
1.791	06-Jun-2006	rpaulo	Attachment framework for the rt2561.c and rt2661.c drivers. From OpenBSD.
1.790	05-Jun-2006	martin	Add D-Link Air DWL-610
1.789	05-Jun-2006	msaitoh	add 82541ER_LOM, 82547EI_MOBILE, 82546GB_PCIE(PRO/100MT), 82546GB_QUAD_COPPER(PRO/1000MT QP) and 82572EI(PRO/1000PT).
1.788	31-May-2006	riz	Add a couple DLINK product IDs, from OpenBSD.
1.787	19-May-2006	markd	branches: 1.787.2; Add some devices found on a Dell Latitude D620.
1.786	30-Apr-2006	xtraeme	PR kern/32849: pcidevs update for Intel 82801CA AC'97 Modem Controller. Apply patch from Nicolas Joly.
1.785	30-Apr-2006	xtraeme	PR port-amd64/33393: viaide driver patch for VT8237R SATA.
1.784	29-Apr-2006	gdamore	Add Audigy4 device id.
1.783	25-Apr-2006	wrstuden	Tweak 3ware 9500 ID, and add 3ware 9550 ID.
1.782	20-Apr-2006	gdamore	Add a bunch of additional Radeon entries, and correct/cleanup a few of them for consistency and to match Xorg. I am assuming that Xorg contains the correct information, since it actually makes use of this information.
1.781	15-Apr-2006	jmcneill	Add entry for Turtle Beach Santa Cruz, and Cirrus Logic CS4615 while we're here.
1.780	06-Apr-2006	simonb	Sort some Intel GigE nics by product ID.
1.779	05-Apr-2006	nonaka	Added "Faraday Technology" in vendor list.
1.778	26-Mar-2006	jonathan	Fix cut-and-paste typo in human-readable description of bcm5780S (used to say bcm5782s, should be bcm5780s; make it so).
1.777	25-Mar-2006	rpaulo	Add Conexant LANfinity. From Frank Wille <frank@phoenix.owl.de> on port-i386.
1.776	21-Mar-2006	he	Add a bunch of IDs for Compaq/HP Smart Array controllers supported by the soon to be added ciss(4) driver.
1.775	12-Mar-2006	chs	branches: 1.775.2; fix name of Radeon 9000 Pro. add Radeon 9200 Secondary ID. add Comtrol, Magma and SIIG multi-port serial cards. add more Dolphin SCI cards. add IBM ServeRAID 4/5. add Mitsubishi Tornado 3000. add or rename various NVIDIA IDs to match openbsd. add SafeNet SafeXcel.
1.774	09-Mar-2006	he	Add vendor code for Nortel and Baystack 21 id. Part of info submitted in PR#33044.
1.773	09-Mar-2006	hamajima	Add EXAR devices. Fix duplicated NVIDIA product id 0x0057.
1.772	07-Mar-2006	bouyer	Also support Serverwork Frodo4, Frodo8 and HT-1000 in svwsata. From OpenBSD.
1.771	06-Mar-2006	bouyer	branches: 1.771.2; Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.770	06-Mar-2006	martin	Add MOXA devices, from Olivier Cherrier
1.769	01-Mar-2006	dogcow	branches: 1.769.2; add a bunch of Nvidia nForce ethernet IDs (via OpenBSD.)
1.768	28-Feb-2006	gdamore	Add Radeon 7000/VE (QZ).
1.767	21-Feb-2006	chuck	add entry for LINKSYS PCM200 10/100 ethernet card
1.766	16-Feb-2006	kochi	add two more ICH7-M LPC device ids from intel document (Intel I/O Controller Hub 7 (ICH7) Family Specification Update)
1.765	12-Feb-2006	tron	Fix inconsistencies in names for components for recent Intel chipsets.
1.764	12-Feb-2006	tron	Correct cut and paste error in last change.
1.763	12-Feb-2006	tron	Add entries for Intel i955X chipset.
1.762	07-Feb-2006	cube	I wonder how the leading junk arrive there...
1.761	07-Feb-2006	cube	Add PLX's 9054 I/O Accelerator.
1.760	06-Feb-2006	abs	Add SIS 190 Ethernet, as found in ASUS T2-AE1
1.759	02-Feb-2006	christos	branches: 1.759.2; PR/32695: Chuck Tuffli: incorrect PCI device id for CMD Technology (0x1095) SiI3132 SATALink (0x3132)
1.758	26-Jan-2006	bouyer	branches: 1.758.2; Add Promise PDC2057x SATA/150 controllers, from FreeBSD.
1.757	16-Jan-2006	christos	PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.756	10-Jan-2006	cdi	PR/32494: Add ID 0x1068 for Intel PRO/100 VE (LOM) found on Sony Vaio S4HRP laptops.
1.755	02-Jan-2006	manu	Add support for nForce430 ATA133 and SATA controllers. My disks now run at a decent speed.
1.754	02-Jan-2006	tron	Fix cut and paste error. Noted by Bernhard Moellemann.
1.753	02-Jan-2006	tron	- Spell NVIDIA as "NVIDIA" instead of "Nvidia". This matches their website and viaide(4). - Add entry for builtin ethernet of NVIDIA nForce4 chipset. Information provided by Bernhard Moellemann.
1.752	01-Jan-2006	xtraeme	Fix VIA VT8231 product: it's a PCI-ISA Bridge not an IDE Controller.
1.751	29-Dec-2005	jonathan	branches: 1.751.2; Add pcidevs entry for the (Serverworks?) PCI-Express to PCI-X bridge chip integral to Broadcom bcm5714 and bcm5715 PCI-Express dual-port gigabit Ethernet chips.
1.750	19-Dec-2005	jonathan	Add an entry for S&K (Schneider&koch, also SysKonnect) PCI-Express gigabit ethernet boards (SK-9E21D single-port, SK-9E22 dual-port) based on the Marvell Yukon-2 chipset.
1.749	19-Dec-2005	jonathan	Add ID 0x3132 for Silicon Image, PCI-Express dual-port SATA-II version of SiI 3124 (PCI-X quad SATA).
1.748	18-Dec-2005	sekiya	Add device information for Linksys IPN2200. From ndiswrapper list.
1.747	11-Dec-2005	jonathan	Remove extraneous //// markers of potential conflict in last commit.
1.746	11-Dec-2005	jonathan	Add pcidevs entries and support to amr(4) for newer LSI MegaRAID devices, including the LSI MegaRAID SCSI 320-0x, 320-2x, 320-4x, 320-1E, 320-2E, the LSI MegaRAID SATA 300-6x and 300-8x. Tested on an LSI MegaRAID SATA 300-8x, which confirms private advice that after PCI-IDs to our amr(4) driver, the newer LSI MegaRAID cards will Just Work The Linux megaraid2 driver accepts various Intel cards with the same pci-device IDs as the LSI MegaRAID cards, treating them exaclty as the above-listed LSI cards. Rework our amr(4) driver to also match and attach those device-ID if seen with an Intel vendor-ID. Add the Intel-vendor PCI decice-ids to pcidevs, so that PCI_VERBOSE will correctly identify the Intel cards. Update src/share/man/man4/amr.4 to list the newly-supported cards. (NB: I don't have the LSI SCSI or any Intel cards to test, just the LSI SATA, which works fine for me.)
1.745	10-Dec-2005	jonathan	Add PCI device entries for Broadcom Ethernet devices: PCI-Express gigabit bcm5714 (copper), bcm5714S (serdes), bcm5715 (copper), bcm5715S (SERDES), also bcm5780 (copper) and bcm5780S (SERDES). Extracted from Broadcom-supplied Linux driver and from prerelease patches for Linux-2.6.15.
1.744	09-Dec-2005	xtraeme	Add two VMware SVGA device ids.
1.743	08-Dec-2005	xtraeme	Add three Rhine ethernet adapters.
1.742	07-Dec-2005	jonathan	Add PCI device-ids to sys/dev/pci/pcidevs for newer Broadcom gigabit ethernet devices: bcm5714, bcm5752 (previously the 5789, but fvdl committed that whilst I was musing). Add definitions to sys/dev/pci/if_bgereg.h for the Broadcom 57xx-family ASIC revisions on these newer chips. Add entries to the PCI-device-version and bge-internal-asic-revision tables in sys/dev/pci/if_bge.c to use these new devices. Pending further information, follow the lead of FreeBSD's if_bge.c driver and, pending further info, treat these new chips as we do the 5750.
1.741	06-Dec-2005	fvdl	Add ID for yet another 57xx-compatible Broadcom chip, the 5789.
1.740	04-Dec-2005	christos	PR/32181: Matthias-Christian Ott: Add support for the Belkin F5D6001 pci device
1.739	04-Dec-2005	christos	PR/32238: Ryo Shimizu: Add support for Promise SATA300TX4 SATA Controller (PDC40718)
1.738	27-Nov-2005	jdolecek	change the description of the 3ware raid controllers to 'ATA RAID Controller' rather then 'IDE RAID' fix the ID of the 9xxx series controller and rename to ATARAID_9k, since 3ware doesn't refer to it using the 'Escalade' family name
1.737	22-Nov-2005	rpaulo	Fix entry for TI FlashMedia Controller.
1.736	18-Nov-2005	rpaulo	It's really PCI7x21/7x11 Cardbus Controller.
1.735	18-Nov-2005	rpaulo	Add a bunch of devices: o Intel 82801FBM ICH6M LPC Interface Bridge o Intel 82801FB/FR PCI Express Port #2 o TI PCIxx21/x515 Cardbus Controller o TI PCI7x21/7x11 IEEE 1394 Host Controller o TI PCIxx11/21 Integrated FlashMedia Controller
1.734	01-Nov-2005	abs	branches: 1.734.2; Add Radeon 9000 PCI (0x4966)
1.733	05-Oct-2005	jonathan	branches: 1.733.2; Add PCI ids for Intel Pro/1000 PT PCI-Express gigabit ethernet controllers: 82571EB 2-port, 82572EI single-port, both in copper, fibre, and SERDES versions, and 82573 embedded chip.
1.732	04-Sep-2005	tron	Add PCI ids for ATI Radeon 9600 LE.
1.731	21-Aug-2005	chs	add nvidia devices found on an ASUS K8N-DL motherboard, plus various other cards.
1.730	21-Aug-2005	gavan	Add BCM5751M
1.729	30-Jul-2005	grant	use same form for secondary display as other ATI devices
1.728	30-Jul-2005	grant	add some device ids for the ATI Radeon 9200 Pro.
1.727	29-Jul-2005	cube	New Dell day: add missing device IDs, and order one correctly.
1.726	25-Jul-2005	christos	PR/29925: Mihai CHELARU: Fix for stge 1023.
1.725	01-Jul-2005	drochner	branches: 1.725.2; add ralink dev IDs, from FUKAUMI Naoki per PR kern/30449
1.724	01-Jul-2005	augustss	Add the ATI SB400 chip set and the Radeon X850 XT graphics card.
1.723	26-Jun-2005	seb	Add ATI Radeon Mobility X600. Add Mobile Intel 915PM/GM/GMS and 910GML Express Chipset.
1.722	24-Jun-2005	seb	Add Texas Instruments PCI6515A Cardbus Controller with Smart Card Controller. Approved by gimpy@
1.721	20-Jun-2005	briggs	Add Intel 6300ESB SATA RAID Controller.
1.720	19-Jun-2005	chs	add some misc cards and merge the two mylex sections.
1.719	15-Jun-2005	bouyer	Add Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern@
1.718	14-Jun-2005	bouyer	Add SIS 965 bridge. From FreeBSD.
1.717	13-Jun-2005	tron	Add support for the LinkSys PCMPC200 Cardbus ethernet card.
1.716	13-Jun-2005	nonaka	Add PCI ids for HITACHI SH7751{,R} PCI Controller
1.715	13-Jun-2005	nonaka	Add PCI id for "ATI Mobility Radeon 9700".
1.714	07-Jun-2005	skrll	Add some Intel 802.11 cards. From OpenBSD.
1.713	28-May-2005	kent	Add the followings: - 6300ESB AC'97 Audio - 6300ESB AC'97 Modem - ICH6 HD-Audio - ICH6 AC'97 Modem - All of ICH7 devices
1.712	27-May-2005	augustss	Add the TM8000 integrated northbridge.
1.711	18-May-2005	riz	Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.710	12-May-2005	cube	Two re(4) candidates. From Francois Romieu in private mail.
1.709	09-May-2005	martin	PR kern/30173: add Corega CG-LAPCIGT
1.708	04-May-2005	cube	Add a few devices seen on a Dell PowerEdge SC1420, and while there make the ordering of devices proper.
1.707	02-May-2005	rh	Add IC Ensemble / VIA Technologies Envy24 Multichannel Audio controllers.
1.706	20-Apr-2005	bouyer	Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.705	06-Apr-2005	martti	Added support for Dell CERC SATA RAID 1.5/6ch
1.704	24-Mar-2005	tron	Add ids for ATI Radeon 9700 Pro and 9800 Pro.
1.703	19-Mar-2005	tron	Add PCI id for Intel PRO/1000 GT.
1.702	19-Mar-2005	tron	Add missing PCI id for Adaptec ASR-2410SA.
1.701	16-Mar-2005	xtraeme	branches: 1.701.2; Add VIATECH KT880 CPU to PCI Bridges.
1.700	16-Mar-2005	uwe	ATI Radeon 9100 IGP chipset.
1.699	15-Mar-2005	xtraeme	Add entries for VIA Apollo KM400 chipsets, provided by Thilo Manske in PR kern/29549.
1.698	15-Mar-2005	xtraeme	Add entry for "S3 Prosavage KM133", provided by Nicolas Joly in PR kern/20572.
1.697	15-Mar-2005	uwe	Fix product ids of ENE CB710 and CB720 CardBus controllers.
1.696	07-Mar-2005	christos	PR/29614: Andrzej Kukula: Missing PCI IDs of ATI IXP 300 and 400 ATA controller
1.695	27-Feb-2005	perry	nuke trailing whitespace
1.694	20-Feb-2005	jdolecek	change the VIATECH VT61XX to VT612X
1.693	02-Feb-2005	christos	PR/29196: Robert Elz: Add support for Broadcom BCM4401-B0 ethernet
1.692	02-Feb-2005	abs	Reminder not to trust cheap no-name packaging. VIA product 0x3249 is a VT6421 not VT642. Also add to list in viaide manpage and note that drives need to be configured into RAID/JOBS sets via its BIOS. http://www.via.com.tw/en/products/peripherals/serial-ata_raid/vt6421/
1.691	01-Feb-2005	abs	Add 'VT642 SATA RAID Controller'. It looks like a viaide, it smells like a viaide... but you can only address the drives once you have put two or more of them into a RAID or JBOD.
1.690	27-Jan-2005	jdolecek	add id for nForce2 Ultra 400 audio
1.689	27-Jan-2005	kent	add some nForce 4 devices
1.688	25-Jan-2005	matt	branches: 1.688.2; Add additional PCI Device IDs for ATI Radeon 9200
1.687	23-Jan-2005	fredb	Recognize Belkin Gigabit Desktop Network PCI card.
1.686	18-Jan-2005	tsubai	Add some Apple devices.
1.685	15-Jan-2005	skrll	branches: 1.685.2; Add Intel PRO/Wireless LAN 2915ABG Mini-PCI Adapter
1.684	15-Jan-2005	jdolecek	Fix Geforce 6800GT entry
1.683	15-Jan-2005	jdolecek	add bunch of GeForce 6800 entries, from Andrew Gillham
1.682	15-Jan-2005	jdolecek	use full 'nForce2 Ultra 400' tag for the NFORCE2_400* enties
1.681	13-Jan-2005	jdolecek	add bunch of NVIDIA nForce2 Ultra 400 chipset devices
1.680	13-Jan-2005	jdolecek	add VIA Velocity ID
1.679	08-Jan-2005	fredb	Recognize the Broadcom 5721 Gigabit Ethernet, as found on SuperMicro P8SCi motherboards.
1.678	08-Jan-2005	ginsbach	Use the more common name, Radeon U1, for the IGP 320M.
1.677	06-Jan-2005	he	Add entry for LSI Logic FC919X, as found on LSI Logic LSI7102XP-LC controllers.
1.676	27-Dec-2004	reinoud	Addition of description for : /* ENE Technology products */ product ENE MCR510 0x510 MCR510 PCI Memory Card Reader Controller
1.675	18-Dec-2004	xtraeme	+ ATI Technologies Radeon 9600 XT
1.674	16-Dec-2004	heas	Add entry for the Radeon Mobility T2.
1.673	11-Dec-2004	reinoud	Add ATI IXP audio controllers
1.672	07-Dec-2004	jdarrow	Add a new vendor and product ID for a cardbus NIC I've got Belkin 0x1799 F5D6020v3 0x6020
1.671	06-Dec-2004	cube	Add ATI IXP IDE Controller.
1.670	05-Dec-2004	bouyer	Cleanup in VIA chipset names. From Lloyd Parkes in kern/15466.
1.669	30-Nov-2004	grant	add ITE 8212 IDE controller.
1.668	22-Nov-2004	rumble	kern/28384: VT3606 -> VT6306 typo
1.667	19-Nov-2004	tacha	add Broadcom BCM5705K.
1.666	17-Nov-2004	jdarrow	Add PCI IDs for two wireless cards I've got. Ralink Technologies (vendor 0x1814) RT2460 (product 0x0101). Texas Instruments (vendor 0x104c) ACX111 (product 0x9066).
1.665	17-Nov-2004	junyoung	Revert ATI to ATI Technologies. Remove trailing Corp., Inc., Ltd., and etc. for consistency.
1.664	16-Nov-2004	skrll	Add Intel PRO/Wireless LAN 2200BG Mini-PCI Adapter
1.663	14-Nov-2004	soren	Add Mylex ExtremeRAID 1100. Shorten ATI Technologies to ATI. Simplify names of Intel mobile USB controllers.
1.662	06-Nov-2004	xtraeme	Add Nvidia nForce3 250 chipset products, from Kouichirou Hiratsuka in PR 27843.
1.661	28-Oct-2004	martti	Align all SYMBIOS entries
1.660	28-Oct-2004	martti	Added Dell PERC 4/SC
1.659	26-Oct-2004	augustss	Add a vendor and some products.
1.658	22-Oct-2004	cube	Add two Broadcom devices (575x chips), correct and add a few Intel entries.
1.657	20-Oct-2004	martti	Added support for DELL PERC 4e/Si RAID controller
1.656	13-Oct-2004	augustss	Add ALi M1563 bridge.
1.655	13-Oct-2004	drochner	recognize ADDI-DATA APCI-7800 8-port serial adapter
1.654	11-Oct-2004	jmmv	Add ATI Technologies Radeon 9200SE Secondary identifier.
1.653	05-Oct-2004	augustss	Add some ALI devices.
1.652	30-Sep-2004	cube	Add a bunch of devices found on the SuperMicro P4SCi motherboard.
1.651	18-Sep-2004	martin	Add LSI FC929 (aka FC929X); from PR kern/26836.
1.650	17-Sep-2004	xtraeme	Add GeForceFX 5700 LE and regen.
1.649	16-Sep-2004	ragge	Add S2io 10 Gigabit ethernet adapter.
1.648	10-Sep-2004	darrenr	Add Sierra Wireless AirCard 580 (1xEV-DO)
1.647	26-Aug-2004	itohy	Add Workbit NinjaSCSI-32 PCI/CardBus products.
1.646	05-Aug-2004	mycroft	Fix an Aironet product name, and normalize the usage of "Mini-PCI".
1.645	05-Aug-2004	cube	Add the GeForce 5900XT. Totally not related to Doom3 release. Honest.
1.644	05-Aug-2004	lukem	consistently use "product<SPACE>"
1.643	04-Aug-2004	mycroft	Normalize more names.
1.642	04-Aug-2004	mycroft	Normalize some product names.
1.641	04-Aug-2004	mycroft	Remove vendor names from product strings.
1.640	02-Aug-2004	mycroft	Add a device I missed.
1.639	02-Aug-2004	mycroft	Merge in data from cardbusdevs.
1.638	02-Aug-2004	mycroft	Mmmm, Microsoft products.
1.637	29-Jul-2004	cube	Add a bunch of devices that are in the Dell Dimension 8400 sitting on my desk: ATI Radeon X300, Broadcom BMC5751, Intel 915 chipset family (ICH6).
1.636	27-Jul-2004	christos	add compaq advanced systems management controller.
1.635	27-Jul-2004	rumble	Fix typos in the SC1100 IDE and AUDIO ids.
1.634	25-Jul-2004	christos	Add viatech's VT8231 Power controller. (Sergio Jimenez)
1.633	21-Jul-2004	mycroft	Add SIS 741.
1.632	17-Jul-2004	mycroft	Fix Intersil product names.
1.631	15-Jul-2004	junyoung	Add 3Com 3c920B-EMB-WNM Integrated Fast Ethernet.
1.630	05-Jul-2004	mycroft	Add the CSB5 LPC bridge.
1.629	04-Jul-2004	mycroft	Update Serverworks chip IDs, per: http://www.plasma-online.de/english/identify/picture/serverworks.html This seems remarkably more sane.
1.628	15-Jun-2004	bouyer	Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.627	27-May-2004	christos	PR/25731: Adam K kirchhoff: Audigy support with the emuxki driver.
1.626	27-May-2004	abs	Add ALI 'M5257 PCI Software Modem'
1.625	12-May-2004	tron	Add support for Broadcom BCM5788 GbE. Patches supplied by FUKAUMI Naoki in PR kern/25166.
1.624	11-May-2004	thorpej	Add i6300ESB IDE / SATA controller PCI IDs.
1.623	10-May-2004	lukem	add Nvidia GeForce4 Ti 4200 Go AGP 8x (0x286)
1.622	10-May-2004	gendalia	Add support for Adaptec AAR 2810SA raid controller.
1.621	07-May-2004	sekiya	Add support for Silicon Image 3512 chip to satalink driver -- treat it as a 3112 variant (without RAID support), which does the right thing with the controller on my K8NNXP.
1.620	22-Apr-2004	skd	Added support for new sis south bridge and 180 sata controller.
1.619	17-Apr-2004	matt	Add devid of Marvell MV6436x System Controller
1.618	03-Apr-2004	martti	Added another SB Live! 5.1 soundcard and joystick version
1.617	03-Apr-2004	mycroft	Add RealTek RT8180.
1.616	01-Apr-2004	tron	Correct entry for PCI vendor id 0x11af. It is assigned to Avid Technology Inc. and not to Pro-Log Corporation. Problem reported by Chris Tribo in PR kern/25004 and verified at http://www.pcidatabase.com/.
1.615	26-Mar-2004	gendalia	branches: 1.615.2; Add Adaptec 2200S Vulcan 2M / Dell PERC 320/DC
1.614	25-Mar-2004	drochner	recognize both parts of the AMD8151 AGP tunnel (see PR kern/24838 by Nicolas Joly), and improve AMD8131
1.613	12-Mar-2004	drochner	add ICH5 builtin ethernet
1.612	03-Mar-2004	gavan	Add Creative Labs SB Audigy LS
1.611	26-Feb-2004	drochner	add intel e7501 host bridge
1.610	28-Jan-2004	chs	add some JNI and Emulex fibre-channel cards. add all the built-in stuff from my Asus A7N8X-E Deluxe motherboard. add some other misc cards. update the URL at the top.
1.609	26-Jan-2004	blymn	Added note for correct update procedure for pcidevs.
1.608	26-Jan-2004	blymn	Add amd8151 agp bridge id.
1.607	19-Jan-2004	ragge	Add "Intel PRO/10GbE LR Server Adapter". Regen.
1.606	06-Jan-2004	matt	Add HP Tachyon TL and TS device IDs.
1.605	05-Jan-2004	matt	Add Agilent Technologies vendor ID Add Agilent Tachyon_DX2 product ID Add HP Tachyon_XL2 product ID
1.604	20-Dec-2003	thorpej	Add product ID for the SiI 3124 SATALink.
1.603	19-Dec-2003	thorpej	Add product ID for the Silicon Image 3114 SATALink.
1.602	17-Dec-2003	tls	Add support for some more Marvell-based adapters to the sk driver: D-Link DGE-530T Linksys EG1032v2 Linksys EG1064v2 The Linux driver treats all of these the same. Tested with a DGE-530T; it works, though performance on at least one of my systems is atrocious. Fix setting of ethernet capabilities into interface capabilites word in attach routine. Manual page: note that jumbo support doesn't work and checksum support doesn't exist.
1.601	14-Dec-2003	tsutsui	Add ATI Radeon 9200SE.
1.600	09-Dec-2003	ad	Add second ID for Intel i960 RM.
1.599	08-Dec-2003	pooka	Broadcom 5703 A3
1.598	08-Dec-2003	scw	Add an entry for Intersil's Prism-54 (ISL3890) WLAN mini-PCI card.
1.597	06-Dec-2003	jdolecek	add NetMos NM9805 and NM9815
1.596	05-Dec-2003	grant	add Adaptec AAR-1210SA Serial ATA RAID controller.
1.595	28-Nov-2003	chs	add Tandem ServerNet II and various HPT IDE controllers.
1.594	28-Nov-2003	cube	Add 3c910 OfficeConnect 10/100B Ethernet.
1.593	26-Nov-2003	matt	Add pcHDTV HD-2000 HDTV video capture card.
1.592	22-Nov-2003	fvdl	Clean up the Promise IDE controller defines.
1.591	03-Nov-2003	cube	Add some IDs for the EPIA-M 10000 motherboard.
1.590	25-Oct-2003	chs	add some misc cards I've got.
1.589	23-Oct-2003	fvdl	Add a few more broadcom 57xx ids. From FreeBSD.
1.588	23-Oct-2003	briggs	Add some ENE Cardbus bridges from FreeBSD.
1.587	22-Oct-2003	fvdl	Add: * an id for the Promise FastTrak 378 * several ids for NVidia nForce3 devices
1.586	18-Oct-2003	matt	Add Philips SAA7146AH
1.585	18-Oct-2003	lukem	Remove http://members.hyperlink.net.au/~chart/pci.htm link; it doesn't exist anymore. Add some more ServerWorks devices
1.584	15-Oct-2003	itojun	add Atheros product IDs from athhal_devid.h
1.583	14-Oct-2003	martti	Added another Dell Perc 4/DI
1.582	10-Oct-2003	thorpej	Add several more Intel Gig-E product IDs.
1.581	08-Oct-2003	jonathan	Add Hifn 7954/7955 and 7956. Put adjacent to the very similar hifn 7951. From Rajesh Vaidyanath <RVaidyanath@hifn.com>, as submitted to FreeBSD.
1.580	26-Sep-2003	matt	Add Dell Perc 4/DI
1.579	25-Sep-2003	joda	more nvidia devices
1.578	21-Sep-2003	nisimura	Redo, again.
1.577	21-Sep-2003	nisimura	Redo the previous fix.
1.576	21-Sep-2003	nisimura	Fix the company name. It seems these days the company calls itself "STMicroelectronics" anyway.
1.575	15-Sep-2003	bouyer	Rename VT8237_RAID to VT8237_SATA, as this is really what it is. Suggested by Stephen Degler in kern/22727.
1.574	13-Sep-2003	jdolecek	fix description of Intel PRO/100 VM Network Controller with 82562ET/EZ PHY don't activate any quirks for VM_6 variant problem reported by Lars Heidieker on current-users also don't activate any quirks for VM_3 - VM_5 variants, neither VE_2 - VE_4, in anticipation that the more modern variants don't need the quirks
1.573	28-Aug-2003	thorpej	Hi/Fn is now spelled Hifn. Add some more Hifn products (from openbsd).
1.572	27-Aug-2003	kleink	While here, add the AMD8111's ACPI controller.
1.571	27-Aug-2003	kleink	Add AMD8111 USB and SMBus controller IDs; from Nicolas Joly in kern/22617.
1.570	26-Aug-2003	jdolecek	add SysKonnect SK9821
1.569	26-Aug-2003	tron	Add entry for Intel i865G Integrated Graphics Device.
1.568	23-Aug-2003	dogcow	add VT6105 support to vr(4)
1.567	20-Aug-2003	jonathan	Fix typo (s/5822/5823/) in line for bcm5823 device.
1.566	17-Aug-2003	chs	add the uninorth ATA found in the latest G4s, and a highpoint IDE.
1.565	17-Aug-2003	bouyer	Add 3com 3c940 Gigabit Ethernet, and VIA KT600 PCI devices. From Stephen Degler in PR port-i386/22453
1.564	14-Aug-2003	itojun	s,PCI/ISA,PCI-ISA,
1.563	14-Aug-2003	itojun	add National Semiconductors SC1100 (GEODE) PCI devices.
1.562	07-Aug-2003	wiz	Fix ID of Radeon 9100 Secondary; noted by Quentin Garnier.
1.561	01-Aug-2003	wiz	Avoid duplicate definitions. From Quentin Garnier in PR 22331.
1.560	01-Aug-2003	jonathan	Add product IDs for Sun's bcm582x-compatible crypto accelerators
1.559	29-Jul-2003	thorpej	Add the product ID used for the i82546EB chips present on the Intel PRO/1000 MT Quad Port Server Adapter.
1.558	29-Jul-2003	martin	Add Ricoh 5C551 CardBus + FireWire product ID. From FUKAUMI Naoki in PR 22292.
1.557	29-Jul-2003	martin	Intel 8255GM is GM, not PM (obviously a copy&pasto). From FUKAUMI Naoki in PR 22293.
1.556	27-Jul-2003	jonathan	Add PCI vendor Cavium and product code for their Nitrox XL crypto device.
1.555	27-Jul-2003	jonathan	Add PCI vendor code for BlueSteel Networks (now part of Broadcom), and PCI product codes for the Bluesteel/Broadcom family of security processors.
1.554	23-Jul-2003	kleink	Add a couple of Philips A/V broadcast decoder device IDs.
1.553	17-Jul-2003	hannken	Add support for the BCM5705: - Only BCM5705M asic rev A1 was tested. Thanks to Bill Paul (wpaul@freebsd.org) for help and support. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.552	14-Jul-2003	jmmv	Add ATI Radeon 9200 chip.
1.551	14-Jul-2003	wiz	Add MGA G550, from Kouichirou Hiratsuka in PR 22133.
1.550	12-Jul-2003	hubertf	Recognize Toshiba Smart Media and Secure Digital cards
1.549	06-Jul-2003	dyoung	Add driver atw for PCI/Mini-PCI/Cardbus 802.11b NICs based on the ADMtek ADM8211. Read the man page for bugs and other outstanding issues.
1.548	06-Jul-2003	simonb	Trailing white-space and TABs nits.
1.547	06-Jul-2003	simonb	Fill out E7505 devices. Fix E7500 HI_D descriptions.
1.546	06-Jul-2003	tron	Add Intel 82547EI gigabit chip.
1.545	06-Jul-2003	simonb	Add an entry for the Intel E7505 Chipset Memory Controller Hub.
1.544	06-Jul-2003	tron	Correct names and descriptions for i875 entries to be consistent with other Intel chipsets.
1.543	06-Jul-2003	tron	Add entries for Intel 865PE chipset.
1.542	28-Jun-2003	bouyer	branches: 1.542.2; Add some intel devices, from Quentin Garnier in private mail.
1.541	27-Jun-2003	christos	PR/21498: URA Hiroshi: Add S3 Savage/IXC
1.540	27-Jun-2003	drochner	add some missing 3c990 variants
1.539	26-Jun-2003	ichiro	add product Aironet 350 miniPCI Wireless LAN Adapter
1.538	14-Jun-2003	ichiro	delete duplicate line
1.537	13-Jun-2003	kent	Add some ICH5/ICH5R device IDs.
1.536	08-Jun-2003	ichiro	add product ID PRO/Wireless LAN 2100 3B Mini PCI Adapter
1.535	07-Jun-2003	wiz	Add some Brooktree device IDs.
1.534	07-Jun-2003	wiz	Add ATI Radeon 9100, 9600TX, 9500/9700.
1.533	07-Jun-2003	ichiro	add productID of ATI
1.532	07-Jun-2003	ichiro	Upps, mistaken fixed sufix
1.531	07-Jun-2003	ichiro	add deviceID - Intel 82855PM/GM chipset
1.530	28-Apr-2003	haya	Add some Texas Instruments' CardBus bridges. Device ID are taken from data sheets provided by TI web site and 'PCI Vendor and Device list' web page <http://www.yourvote.com/pci>.
1.529	28-Apr-2003	haya	Add some O2Micro CardBus bridges regarding PR/20084. Device ID are taken from FreeBSD and cross-checked by the web page 'PCI Vendor and Device list' <http://www.yourvote.com/pci>
1.528	24-Apr-2003	nathanw	Add a few Opteron and AMD8000 chipset device IDs.
1.527	22-Apr-2003	fvdl	Add AMD 8131 PCI-X tunnel and IO apic.
1.526	15-Apr-2003	augustss	Another VIA device.
1.525	15-Apr-2003	augustss	Add another VIA CPU-AGP bridge.
1.524	15-Apr-2003	rjs	Add ID for ITE IT8152 companion chip.
1.523	13-Apr-2003	explorer	typo, s/TX4/TX2/
1.522	12-Apr-2003	explorer	add an older pci id for SATA150TX2PLUS, too
1.521	12-Apr-2003	christos	PR/8754: Andreas Wrede: Add SIIG devices
1.520	11-Apr-2003	explorer	add Promise ATA/150 TX4 id
1.519	05-Apr-2003	lha	add Radeon Mobility M7 LW (AGP)
1.518	03-Apr-2003	kent	- Forte Media PCI joystick - Intel 82540EP chips - Intel ICH4M IDE controller - Modify descriptions of some ICH4 devices which ICH4M also has: 82801DB xxx -> 82801DB/DBM xxx
1.517	29-Mar-2003	jdc	Add support for USRobotics Wireless Access PCI adapter (model 132415). From Sergio Jimenez in PR kern/20642.
1.516	25-Mar-2003	tsubai	Add new Apple devices.
1.515	20-Mar-2003	thorpej	Add some more LSI MPT IDs.
1.514	18-Mar-2003	thorpej	Add Silicon Image SATALink 3112.
1.513	13-Mar-2003	joda	more nvidia devices
1.512	06-Mar-2003	bouyer	Remove duplicate in previous.
1.511	06-Mar-2003	bouyer	Add several SiS host bridges, from FreeBSD.
1.510	03-Mar-2003	soren	Add ATI Radeon M9.
1.509	02-Mar-2003	salo	Actually, there is no space in Silicon Motion Lynx* names (e.g. Lynx EM -> LynxEM), the already existing records misguided me.
1.508	28-Feb-2003	enami	Add product id of ServerWorks CSB6 IDE controller.
1.507	26-Feb-2003	salo	Add few more Silicon Motion, Inc. cards (Lynx EM, Lynx 3DM, Lynx 3D, Lynx) Print various [company], Inc. consistently as (surprise) [company], Inc.
1.506	18-Feb-2003	grant	add Intel PRO/100M.
1.505	03-Feb-2003	pooka	add GeForce2 Go
1.504	03-Feb-2003	joda	correct previous, and add some more devices found in the Shuttle SS51G.
1.503	02-Feb-2003	bouyer	Add SiS 651 host bridge.
1.502	01-Feb-2003	matt	Add Tundra (Newbridge) PowerSpan/PowerSpanII/QSpanII PCI bridges
1.501	24-Jan-2003	thorpej	Tweak a Promise description.
1.500	24-Jan-2003	thorpej	Add some more Promise products.
1.499	24-Jan-2003	thorpej	Sort the Promise entries.
1.498	24-Jan-2003	thorpej	Add Nvidia nForce2 ATA133 IDE product ID, and shuffle some Nvidia entries around a little.
1.497	17-Jan-2003	jonathan	Add pcidevs entries for more Broadcom 570x family variants.
1.496	16-Jan-2003	gendalia	Add 3com 3c905CX-TX and Nvidia nForce2 MCP-T ac-97, on the ASUS A7N8X Deluxe motherboard, Chipset: NVIDIA nForce2 MCP-T
1.495	10-Jan-2003	augustss	Add two Intel PCI-PCI bridges.
1.494	30-Dec-2002	explorer	add atheros communications, Inc. AR5001 reference goop. This is what the 2.4 Ghz / 5 Ghz linksys cardbus card has inside.
1.493	26-Dec-2002	matt	Add Promise Serial ATA/150 TX2plus
1.492	26-Dec-2002	matt	Add some more BCM570x variants.
1.491	11-Dec-2002	jdolecek	add Madge Smart 16/4 PCI Ringnode Mk2 entry
1.490	03-Dec-2002	explorer	add Ricoh 5C552 Cardbus + Firewire product ID
1.489	26-Nov-2002	thorpej	Add IBM 133 PCI-X Bridge.
1.488	26-Nov-2002	thorpej	Add Intel i80321 I/O processor.
1.487	19-Nov-2002	bsh	Two more vendor IDs for Samsung Electronics Co. Ltd.
1.486	15-Nov-2002	tsutsui	Remove an entry for AMD PCNETS_PCI. AMD Am79C974 PCnet-SCSI Ethernet and SCSI conrtoller is a multi-function PCI device which has two device IDs of Am53C974(PCscsi-PCI) and Am79C970(PCnet-PCI), and there is no its own ID.
1.485	15-Nov-2002	ad	Remove dup product.
1.484	09-Nov-2002	grant	add ATI Rage Mobility M1. from FUKAUMI Naoki in kern/18987.
1.483	08-Nov-2002	msaitoh	add Adaptec 200[05]S (DPT)
1.482	07-Nov-2002	thorpej	Add Intel i31244 Serial ATA controller.
1.481	07-Nov-2002	martin	Add Myson MTD803 3-in-1 Fast Ethernet Controller.
1.480	04-Nov-2002	wiz	Fix typo reported by Nicolas Joly in PR 18921.
1.479	25-Oct-2002	kent	Correct a typo in the previous change.
1.478	25-Oct-2002	kent	Add vender IDs and product IDs: - ENE CB1410 (http://mail-index.netbsd.org/tech-kern/2002/10/24/0038.html) - iCompression iTVC15 Correct ID of AMD PBC8111_AC.
1.477	24-Oct-2002	grant	COREGA are OEM NDC products. add NDC NCP130 rev A2 wireless NIC. From T. M. Pederson in kern/18609.
1.476	21-Oct-2002	soren	Correct name of TI 802.11 chip.
1.475	21-Oct-2002	grant	add D-Link 520+ wireless ethernet.
1.474	18-Oct-2002	yamt	fix typos; "wiseman" is 82542, not 82452.
1.473	16-Oct-2002	leo	Add Fritz!PCI v2.0 ISDN Interface
1.472	05-Oct-2002	kent	- Broadcom BCM4401 - VIA VT8235 - Modify description of VT8233_AC97: "VT8233 AC-97 ..." -> "VT8233/VT8235 AC-97 ..."
1.471	24-Sep-2002	pooka	Fix Asus ISDN card ID to be the actual device ID instead of the vendor ID
1.470	20-Sep-2002	pooka	Add vendor Dynalink and a few devices by other manufacturers for use in the iwic ISDN driver.
1.469	19-Sep-2002	nathanw	Add AMD 761 system controller and ppb.
1.468	11-Sep-2002	bouyer	Add 2 accard pciide controllers. From linux.
1.467	06-Sep-2002	uwe	Add product ids for Integraphics Systems CyberPro 2000 and 2010.
1.466	23-Aug-2002	bouyer	Add Promise Ultra133TX2 and Ultra133TX2v2 PCI ID.
1.465	16-Aug-2002	heinz	Add Lucent/Agere FW322/323 IEEE 1394 controller
1.464	10-Aug-2002	toshii	Add SiS745 host bridge.
1.463	08-Aug-2002	drochner	addanother ppb
1.462	04-Aug-2002	thorpej	Add product ID for the ADI i80200 Big Endian Companion Chip.
1.461	03-Aug-2002	kent	Add the following: - SiS 7012 AC-97 (kern/17716) - NVIDIA nForce MCP AC-97 - AMD 8111 AC-97
1.460	02-Aug-2002	soren	Add O2 Micro Cardbus bridge ID from PR kern/17469 by Nicolas Joly.
1.459	02-Aug-2002	ad	PR 17719 from Mark Davies: Additional Adaptec RAID controller types.
1.458	31-Jul-2002	bouyer	Ops, remove duplicate entry for 82801DB_LPC, pointed out by geoff wing.
1.457	30-Jul-2002	bouyer	Add Intel 845G PCI IDs, from kern/17743 by Andreas Wrede <andreas@planix.com>
1.456	29-Jul-2002	augustss	Add a couple of bridge devices.
1.455	26-Jul-2002	onoe	Add Silicon Image (CMDTECH) 0680
1.454	22-Jul-2002	msaitoh	Add Intel 82801E Communications I/O Controller Hub (C-ICH)
1.453	12-Jul-2002	enami	Add product id of Altima AC1001.
1.452	09-Jul-2002	matt	Add Apple Xserve PCI devices (more UniNorth) stuff.
1.451	09-Jul-2002	thorpej	Clean up Intel i8254x product names, and add i82540, i82545, and i82546 product IDs.
1.450	07-Jul-2002	augustss	Add some Sandburst products.
1.449	05-Jul-2002	enami	Add vendor id of Avlab technology, Inc. and product id of Low Profile PCI 4 Serial (but not used actually).
1.448	04-Jul-2002	rjs	Add Peak System Technik CAN Controller.
1.447	03-Jul-2002	msaitoh	add ServerWorks CMIC LE, CMIC SL and CIOB-X2
1.446	27-Jun-2002	pooka	two AVM isdn interfaces
1.445	26-Jun-2002	cjs	Add Altima AC9100 Gigabit Ethernet ID. This may not be the correct name. However, it's the chip on the NetGear GA302T card.
1.444	26-Jun-2002	cjs	Add a couple more TI IEEE 1394 host controllers.
1.443	22-Jun-2002	fvdl	Add Altima PCI Ids.
1.442	22-Jun-2002	cjs	For new Intel network controllers, update descriptions and fix numbering.
1.441	22-Jun-2002	cjs	Add new Intel Ethernet cards.
1.440	21-Jun-2002	thorpej	Add product ID for Intel 82452KX/GX PCI-Host bridge. PR 17353, from MOCHIDA Shuji.
1.439	17-Jun-2002	jdolecek	add VIA Technologies VT8653 and VT8633 entries move VT82C686A_MC97 entry to where it numerically belongs
1.438	10-Jun-2002	cjs	Add Viatech VT8233A PCI-ISA Bridge.
1.437	09-Jun-2002	taca	Add VIA Technologies's VT8231 IDE Controller.
1.436	04-Jun-2002	fvdl	Add AMD 8111 IDE id.
1.435	03-Jun-2002	briggs	Add Motorola's MPC8245 (Kahlua II) host bridge.
1.434	02-Jun-2002	seb	Add product Titan PCI-800L aka VSCom PCI-800L, a 8 serial puc card, from PR 17140. Reviewed/approved by veego@ and martin@.
1.433	28-May-2002	christos	add action tec
1.432	28-May-2002	tron	Add entries for Intel i850 and i860 chipsets.
1.431	27-May-2002	drochner	add one more of our own products
1.430	25-May-2002	ichiro	add 21555 Non-Transparent PCI-PCI Bridge
1.429	24-May-2002	lukem	add NetMos Quad UART and 1284 Printer port
1.428	19-May-2002	bouyer	branches: 1.428.2; Add HPT372 and 74 PCI IDE controllers.
1.427	19-May-2002	sommerfeld	Some E7500 device ids
1.426	18-May-2002	sommerfeld	Some P64H2 device ids
1.425	18-May-2002	matt	Add SYMBIOS FC9[12]9 and 1030R
1.424	15-May-2002	augustss	branches: 1.424.2; Give a more suitable name to the DEC 21554 bridge.
1.423	14-May-2002	augustss	Add a Heuricon host bridge.
1.422	03-May-2002	augustss	Add Sandburst.
1.421	29-Apr-2002	nonaka	Added two IBM devices.
1.420	26-Apr-2002	ad	Oops, add more IDs used by aac_pci.c.
1.419	18-Apr-2002	thorpej	Add Symbios 53c875A product ID.
1.418	18-Apr-2002	tshiozak	add PCI/USB vendor ID for Kurusugawa Electronics, Inc. PCI-ID: 0x6809, USB-ID: 0x0b7e.
1.417	16-Apr-2002	drochner	add some C&T VGA chip
1.416	15-Apr-2002	mycroft	Use chip numbers for most of the new Adaptec IDs, not board numbers.
1.415	15-Apr-2002	ad	Add a bunch of Adaptec, Dell and HP products. From OpenBSD.
1.414	05-Apr-2002	mycroft	Update some ServerWorks and VIA IDs. Add the ServerWorks CSB5 southbridge and the Apollo ProMedia 133 chipset.
1.413	27-Mar-2002	abs	Add 'CONEXANT 56KFAXMODEM 0x1085 HW 56K Fax Modem'
1.412	19-Mar-2002	bouyer	Add AMD768 devices, from http://members.hyperlink.net.au/~chart/download/pcidevs.txt
1.411	11-Mar-2002	joda	some devices found in a dell machine
1.410	01-Mar-2002	kleink	Add Motorola Raven, Falcon and Hawk product IDs.
1.409	28-Feb-2002	junyoung	Oops, readd an ID for Rage XL which was removed in previous commit.
1.408	28-Feb-2002	junyoung	Better fix of duplicate ATI Rage Pro IDs. `_B' and `_P' mean BGA and PQFP, respectively.
1.407	27-Feb-2002	martin	Avoid duplicate define for ATI Rage Pro PCI
1.406	27-Feb-2002	junyoung	Fully revise product IDs for ATI Mach64 and Rage graphics cards. XXX Rage 128 not revised.
1.405	24-Feb-2002	kleink	Add IBM 82660 PowerPC to PCI Bridge and Memory Controller.
1.404	20-Feb-2002	simonb	Use BCM prefix instead of SB prefix in previous.
1.403	20-Feb-2002	simonb	Add the SiByte/Broadcom SB1250 PCI and LDT bridge IDs.
1.402	20-Feb-2002	simonb	Rename the SIBYTE entry to "Broadcom Corp. (SiByte)".
1.401	16-Feb-2002	thorpej	Add the AMD Elan SC520 System Controller.
1.400	04-Feb-2002	junyoung	Add some more product ID's for ATI Mach64 graphics cards.
1.399	02-Feb-2002	tsutsui	Add Initio INIC-850/1060 SCSI and variants.
1.398	02-Feb-2002	tsutsui	Add HiNT Corp. HB1 PCI-PCI bridge.
1.397	30-Jan-2002	tron	Fix vendor entry for "Rendition Inc.". This patch was supplied by Paul Shupak in PR kern/15367.
1.396	29-Jan-2002	junyoung	- s/INTERGRAPHICS/INTEGRAPHICS/ - Add two product IDs for Integraphics Systems graphics cards.
1.395	28-Jan-2002	thorpej	Add product ID for the Intel i82544GC 64-bit variant.
1.394	26-Jan-2002	tron	Add missing names for devices supported by puc(4). This patch was supplied by Paul Shupak in PR kern/15368.
1.393	26-Jan-2002	ichiro	OCHI -> OHCI
1.392	22-Jan-2002	junyoung	Rename RAGEXL to MACH64_GR.
1.391	22-Jan-2002	junyoung	Add ATI Mach64 GS.
1.390	21-Jan-2002	ichiro	fixed some products name
1.389	21-Jan-2002	eeh	Add the device id for something that claims to be an ATI Rage XL.
1.388	21-Jan-2002	thorpej	* Add Broadcom Corporation vendor ID. * Add 3Com 3c996 product ID. * Add Alteon "Tigon 3" (Broadcom BCM570x) product IDs. * Add Broadcom Corporation BCM570x product IDs.
1.387	15-Jan-2002	thorpej	Add the Intel i82544GC Gig-E.
1.386	14-Jan-2002	augustss	Fix typo.
1.385	14-Jan-2002	augustss	Change some names that I added earlier.
1.384	13-Jan-2002	augustss	Add 3Com 3crwe777a AirConnect.
1.383	09-Jan-2002	augustss	Add Intel 82801 and 82830.
1.382	05-Jan-2002	junyoung	Remove Mach64 B that conflicts with Mach64 GP.
1.381	04-Jan-2002	augustss	Add a bunch of PCI-wireless PCMCIA adapters. They are probably all based on a PLX chip. (For which we have no driver yet.)
1.380	03-Jan-2002	rafal	Add Rainbow Technologies vendor ID & Rainbow CryptoSwift 200 PKI accelerator (CS200/PCI).
1.379	01-Jan-2002	augustss	Add a radio card.
1.378	30-Dec-2001	ichiro	add TSB43AA22 OHCI IEEE 1394 Host Controller
1.377	29-Dec-2001	ichiro	add products VT6305 OCHI IEEE 1394 Controller VT3606 OCHI IEEE 1394 Controller
1.376	20-Dec-2001	thorpej	Add the product ID for the SiS 630 PCI-Host bridge.
1.375	16-Dec-2001	thorpej	Add Middle Digital, Inc. vendor ID. Add product IDs for the Middle Digital PCI-Weasel Virtual VGA, Serial Port, and Control registers.
1.374	15-Dec-2001	soren	Add Cologne Chip Designs ISDN controller.
1.373	13-Dec-2001	bouyer	Add some promise controllers, from FreeBSD
1.372	11-Dec-2001	ad	Add some Intel/ICP-Vortex/Compaq products.
1.371	11-Dec-2001	kleink	Add product ID for the Motorola MPC107.
1.370	07-Dec-2001	onoe	Add HPB and LPC functions of 82801BAM.
1.369	07-Nov-2001	thorpej	Correct i80312 description.
1.368	07-Nov-2001	thorpej	Add product ID for the Cyclone PCI-700 (IQ80310).
1.367	25-Oct-2001	augustss	Add some VIA KT266 numbers.
1.366	24-Oct-2001	thorpej	Add product IDs for the Powerhouse Systems PowerPro and PowerTop PowerPC system controllers.
1.365	24-Oct-2001	thorpej	More Intel Gig-E ID fixups, and add Intel i82544 Gig-E IDs.
1.364	16-Oct-2001	tron	branches: 1.364.2; Add host bridges of SiS 645, 650 and 730 chipsets.
1.363	13-Oct-2001	ichiro	add vendor & product ID Intersil PRISM2.5 Mini-PCI WLAN
1.362	11-Oct-2001	tron	Add host bridge of the SiS 735 chipset.
1.361	03-Oct-2001	simonb	Minor correction to the name of the 405GP PCI bridge.
1.360	01-Oct-2001	thorpej	Add product ID for the RealTek 8139[BC].
1.359	28-Sep-2001	thorpej	More corrections to Intel Gig-e product IDs.
1.358	27-Sep-2001	itojun	add 4DWAVE and SiS sound devices.
1.357	15-Sep-2001	chris	Add voodoo 4/5 entry for 3DFX.
1.356	10-Sep-2001	eeh	Add Sun GEM gigabit network card.
1.355	05-Sep-2001	drochner	branches: 1.355.2; we (research center Juelich) are a PCISIG member now - add the vendor ID and some products
1.354	04-Sep-2001	thorpej	Add product ID for the Intel IXP1200 Network Processor.
1.353	04-Sep-2001	pooka	add Allegro-1 ID also from Jared D. McNeill
1.352	04-Sep-2001	pooka	add another device ID for the ESS Maestro3 from Jared D. McNeill
1.351	30-Aug-2001	soren	Add some more Hifn chips.
1.350	28-Aug-2001	tsubai	Add more Apple devices.
1.349	27-Aug-2001	enami	Sort Intel products by id and kill duplicated entries.
1.348	27-Aug-2001	ad	Add a DPT PCI-PCI bridge.
1.347	23-Aug-2001	thorpej	Take part numbers out of the PCnet-PCI product descriptions.
1.346	23-Aug-2001	thorpej	Add Promise ATA/100 TX2 device ID.
1.345	20-Aug-2001	fvdl	Add some entries for the AMD 76x chipset for completeness.
1.344	20-Aug-2001	gehenna	Add NEC USB2.0 Host Controller
1.343	18-Aug-2001	thorpej	Add Avance Logic ALS4000 Audio.
1.342	17-Aug-2001	matt	Add Apple Uni-N (as used in QuickSilver) product IDs.
1.341	13-Aug-2001	cjs	Add Transmeta and products, and an ALI integrated sound/modem device.
1.340	31-Jul-2001	matt	Add pci ids for the Intel 80312 I/O companion chip for the Intel 80200 XScale processor.
1.339	28-Jul-2001	pooka	3c556 V.90 MiniPCI modem
1.338	25-Jul-2001	thorpej	Add the Antares Microsystems, Inc. vendor ID, and the product ID for their Gigabit Ethernet board (a TC9021 board targeted at the Sun market).
1.337	23-Jul-2001	bouyer	Add TITAN PCI100H.
1.336	30-Jun-2001	thorpej	branches: 1.336.2; Change the DLINK ST-201 entry, add more Sundance/DLINK Gig-E products.
1.335	28-Jun-2001	thorpej	Add Tamarack Microelectronics, Inc. vendor ID, and product ID for Tamarack TC9021 Gigabit Ethernet.
1.334	27-Jun-2001	mcr	added Solidum PAX.ware 1100 and Lucent ORCA PCI targets
1.333	20-Jun-2001	thorpej	Add more/update existing 3Com 3CR990 product IDs.
1.332	20-Jun-2001	thorpej	Add SysKonnect SKNET-GE (GEnesis Gigabit Ethernet).
1.331	20-Jun-2001	imp	Add O2Micro PCI-{CardBus,PCMCIA} bridges: OZ6729, OZ6730, OZ6832/OZ6833, OZ6860 and OZ6812/OZ6872
1.330	18-Jun-2001	onoe	Add Cisco/Aironet 350
1.329	17-Jun-2001	tsubai	Add device IDs on the new iBook.
1.328	13-Jun-2001	simonb	Add a port to IBM's PPC405GP Reference Board (the "walnut") by Eduardo Horvath and Simon Burge of Wasabi Systems. IBM 4xx series CPU features: - New pmap and revised trap handler. - Support on-chip timers, PCI controller, UARTs - Framework for on-chip ethernet and watchdog timer. General PowerPC features: - Add in-kernel PPC floating point emulation - New in{,4}_cksum that is between 1.5 and 5 times faster than the old version depending on CPU type. General changes: - Kernel support for generic dbsym-style symbols.
1.327	11-Jun-2001	thorpej	Add Level One product ID, and LXT-1001 10/100/1000 Ethernet product ID.
1.326	11-Jun-2001	tron	Add support for the D-Link DFE-530TX+. Patch supplied by Gabriel Rosenkoetter in PR kern/13161.
1.325	31-May-2001	nisimura	Add product IDs for IDT RC32234 and RC32334 system controllers.
1.324	21-May-2001	thorpej	Add ICH3 network controller IDs.
1.323	18-May-2001	enami	Add product IDs of internal devices found in NEC VRC4173 and vendor/product ID of MediaQ MQ200.
1.322	17-May-2001	drochner	add some stuff I got between my fingers
1.321	14-May-2001	thorpej	Add product ID for the National Semiconductor DP83820 Gigabit Ethernet chip.
1.320	06-May-2001	fvdl	Add some IDs for the AMD760 chipset, as specified in the datasheet.
1.319	25-Apr-2001	simonb	Spell "portege" correctly - it's a computer, not a car.
1.318	21-Apr-2001	tsutsui	Correct an entry for Acard UDMA IDE controller. The chip is actually marked as ATP860-A, not ATP860R.
1.317	13-Apr-2001	hubertf	Move the Trident CyberClade entry where it belongs.
1.316	12-Apr-2001	itojun	change symbol for LMC T1 card. part of PR 12331.
1.315	12-Apr-2001	matt	Add UltraSparc IIe PCI interface.
1.314	09-Apr-2001	ad	Add Tekram DC-315/DC-395.
1.313	31-Mar-2001	minoura	Add Cyrix products (MediaGX).
1.312	21-Mar-2001	soren	Lower-case a few hex digits for easier searching.
1.311	21-Mar-2001	hubertf	Add Trident CyberBlade as found in Toshiba Satellite Pro 4600
1.310	20-Mar-2001	mrg	add sun device ids found on sunblade1000 systems.
1.309	17-Mar-2001	thorpej	Fixup some Intel Gig-E entries.
1.308	13-Mar-2001	bouyer	Add Titan Electronics Inc vendor ID (from linux) and Titan VS-800H product.
1.307	07-Mar-2001	matt	Add Galileo GT-64260.
1.306	05-Mar-2001	matt	branches: 1.306.2; Add Winbond W83C533F functions.
1.305	04-Mar-2001	msaitoh	Add an entry for NetMos
1.304	02-Mar-2001	thorpej	Add Mylex AcceleRAID and eXtremeRAID product and sybsystem IDs.
1.303	02-Mar-2001	thorpej	Add product ID for Xircom WinGlobal modem.
1.302	18-Feb-2001	bouyer	Add Intel 82801BAM IDE controller, found in laptops. From OpenBSD
1.301	18-Feb-2001	matt	Add Symbios FC909
1.300	12-Feb-2001	tshiozak	add CMI8738B.
1.299	11-Feb-2001	ad	Add another 3ware product.
1.298	04-Feb-2001	briggs	Add the MPC8240 (Kahlua) Host Bridge
1.297	02-Feb-2001	hubertf	Add M5243 PCI-AGP Bridge Fixes PR 11466 by Jared D. McNeill <jmcneill@invisible.yi.org>
1.296	31-Jan-2001	thorpej	Add the SEGA Dreamcast Broadband Adapter.
1.295	26-Jan-2001	rh	Add Toshiba Protege product code
1.294	26-Jan-2001	rh	Add product code for the Compaq Armada M700 Notebook
1.293	22-Jan-2001	jdolecek	add US Robotics (3Com) 56K Voice Internal PCI Modem (WinModem) entry I wish I'd been less foolish and not bought this device ...
1.292	22-Jan-2001	mjl	Add stallion products.
1.291	17-Jan-2001	thorpej	Update the Intel Gigabit Ethernet entries.
1.290	15-Jan-2001	thorpej	Add the 66MHz version of the Symbios 53c1010.
1.289	12-Jan-2001	itojun	add product ID for LMC T1 card. from freebsd.
1.288	09-Jan-2001	rh	Add COMPAL vendor and COMPAL 38W2 product (OEM for Dell Inspiron 5000e)
1.287	09-Jan-2001	thorpej	Add Symbios 53c1510D.
1.286	08-Jan-2001	thorpej	Add product ID for the Accton EN2242 10/100 Ethernet (an ADMtek AN985).
1.285	08-Jan-2001	thorpej	Add product ID for DEC SWXCR RAID (a Mylex v5 controller).
1.284	08-Jan-2001	rh	Add IDs for the NEC Versa LX and Versa Pro LX
1.283	03-Jan-2001	mycroft	Add Chrysalis-ITS LunaVPN card.
1.282	31-Dec-2000	ad	Fix Mylex products.
1.281	30-Dec-2000	eeh	Fix typo.
1.280	30-Dec-2000	eeh	Add Sun psycho bus controller.
1.279	27-Dec-2000	tsutsui	Fix entries of Acard UDMA IDE controllers.
1.278	27-Dec-2000	augustss	Add a VIA chipset and Nvidia GeForce2 MX.
1.277	23-Dec-2000	tsutsui	Add Acard UDMA IDE controllers.
1.276	14-Dec-2000	onoe	add Aironet 4500/4800
1.275	05-Dec-2000	drochner	add some chips I met in the last weeks
1.274	20-Nov-2000	onoe	CXD32222 -> CXD3222 (Sony OHCI Controller)
1.273	18-Nov-2000	thorpej	Add some 3D Labs products, from XFree86, via groo@netbsd.org
1.272	14-Nov-2000	bouyer	Add Alteon and netgear 1000baseT variants of the gigabit ethernet adapter. From FreeBSD
1.271	04-Nov-2000	thorpej	Add a couple NeoMagic products. From OpenBSD.
1.270	03-Nov-2000	ad	Oops
1.269	03-Nov-2000	ad	Pequr Technology is now called ServerWorks.
1.268	27-Oct-2000	thorpej	Add a bunch of Intel PCI chipsets. From OpenBSD.
1.267	27-Oct-2000	thorpej	Add a bunch of Pequr (ServerWorks) products. From OpenBSD.
1.266	19-Oct-2000	thorpej	Add the ICS Advent PCI-WDT500/501 watchdog timer boards.
1.265	14-Oct-2000	simonb	Add ATI Mobility M3 (both PCI and AGP).
1.264	12-Oct-2000	itojun	pci device ID for hifn 7751-based crypto cards. from openbsd.
1.263	03-Oct-2000	thorpej	Add ADMtek AN985 10/100 Ethernet.
1.262	27-Sep-2000	augustss	Add some ATI Rage cards and some Intel chips.
1.261	20-Sep-2000	thorpej	Add National Semiconductor DP83815 10/100 Ethernet.
1.260	19-Sep-2000	fvdl	Add IDs for the 3c555, 556 and 556B 3Com MinPCI Ethernet cards.
1.259	13-Sep-2000	ad	Add product IDs for AMI MegaRAID boards.
1.258	13-Sep-2000	tron	Remove incomplete (and incorrect) entry for D-Link 530.
1.257	13-Sep-2000	tron	Use "VT6102 (Rhine II) 10/100 Ethernet" as description to be consistent with existing entry for "VT86C100A (Rhine-II) 10/100 Ethernet".
1.256	13-Sep-2000	tron	Add VT6102 Rhine II 10/100BaseTX ethernet chip used on e.g. the D-Link 530TX NIC.
1.255	07-Sep-2000	soren	s/AT&T Microelectronics/Lucent Technologies/
1.254	04-Sep-2000	soren	Add Intel 440GX host bridge.
1.253	04-Sep-2000	itojun	Brooktree Bt8474 Multichannel Synchronous Communictaions Controller (MUSYCC)
1.252	02-Sep-2000	cgd	after discussion with veego, rename constant for 3D Labs Permedia 2 from PERMEDI2 to PERMEDIA2. (substance over style. 8-) Also, fix a typo I noticed in a comment.
1.251	29-Aug-2000	wiz	add vendors HAUPPAUGE, ASKEY, AVERMEDIA
1.250	21-Aug-2000	enami	Add a product id for the promise ata-100 controller found on MS-6321 (694D Pro). Reported in the PR#10756 by Kazuki Sakamoto.
1.249	21-Aug-2000	haya	Add several Texas Instrument's PCI-CardBus bridges.
1.248	10-Aug-2000	soren	s/Toshiba America Info Systems/Toshiba/ since they are the main PCI issuer. Add new version of the Oboe FIR device.
1.247	07-Aug-2000	billc	Added PCI IDs for 3Com's Home Connect (3c450) and Office Connect (3cSOHO100). They both are register compat with later versions of the 3c905 (tornado/ vortex) series.
1.246	04-Aug-2000	ad	Vortex Computer Systems changes, from OpenBSD: - Add product ID for GDT6518RS. - Fix bogus product IDs.
1.245	23-Jul-2000	jonathan	* Add entries for older Yamaha YMF chips (724, 740) plus the newer steppings (740C, 724F) which use the same microcode as the DS-1S/DS-1E
1.244	20-Jul-2000	ad	Add what is apparently ESS's old vendor ID and corresponding product ID for the Maestro 1.
1.243	10-Jul-2000	ad	Add ICP Vortex RAID controllers.
1.242	09-Jul-2000	mycroft	Add ID for i82559ER.
1.241	08-Jul-2000	mrg	add microSPARC IIep PCI controller
1.240	04-Jul-2000	enami	Add device ID of promise ultra100/ata controller.
1.239	03-Jul-2000	gmcgarry	Devices found in NEC Versa Note VX
1.238	24-Jun-2000	joda	add Intel 82801BA devices
1.237	21-Jun-2000	bouyer	Add CMD PCI0649 product ID.
1.236	16-Jun-2000	soren	branches: 1.236.2; Add some S3 chips.
1.235	13-Jun-2000	tron	Add CMD PCI0648.
1.234	12-Jun-2000	nathanw	Add AMD 79c978 PCnet-PCI Home.
1.233	12-Jun-2000	bouyer	Shorter description for the HPT366
1.232	12-Jun-2000	bouyer	Add HPT366/370 device ID. XXX Is 'TRIONES' the rigth vendor name ?
1.231	09-Jun-2000	augustss	Add some PCI devices found in the IOpener (commited from the IOpener itself :).
1.230	07-Jun-2000	ad	Add vendor/device ID for 3ware Escalade IDE RAID controller.
1.229	07-Jun-2000	mrg	fix a couple of the sun entries.
1.228	07-Jun-2000	soren	Add Trident 9397DVD.
1.227	06-Jun-2000	kleink	Add the product ID for the ESS Maestro 3 modem function as well.
1.226	05-Jun-2000	kleink	Add the ESS Maestro 3 product ID as well.
1.225	05-Jun-2000	kleink	ESSTECH: * add what's allegedly the Maestro 1's product ID. * align product names with data sheets.
1.224	30-May-2000	matt	Add the TI 1394 controllers. Makes all the 1394 controller descriptions similar.
1.223	28-May-2000	drochner	-correct description for 3c905b-FX (100 Mb only) -add Neomagic 256ZX
1.222	27-May-2000	scw	Add entries for a couple of OPTi PCI-IDE controllers.
1.221	27-May-2000	nathanw	Whitespace police.
1.220	27-May-2000	soren	branches: 1.220.2; Add a few devices.
1.219	26-May-2000	matt	Add NEC 1394 OHCI chips
1.218	18-May-2000	drochner	fix typo in last commit
1.217	18-May-2000	garbled	Add the device ID for the IBM MPIC-II.
1.216	14-May-2000	soren	Add SGI IOC3.
1.215	12-May-2000	thorpej	Add product ID for 3Com 3c980C-TXM 10/100 Ethernet. From Paul J. Lavoie, kern/10037.
1.214	07-May-2000	jonathan	Product codes for Yamaha YMF744B (DS-1S) YMF754 (DS-1E) audio chips.
1.213	30-Apr-2000	dante	Add ASB-3940U3W host adpeter (AdvanSys)
1.212	28-Apr-2000	augustss	Add 2 more eap audio devices.
1.211	26-Apr-2000	thorpej	Add product ID for the PLX Technology 9060ES PCI bus controller. Note this ID isn't normally seen, as just about every device with a PLX on it loads a different ID from an EEPROM, but there's no harm in documenting it.
1.210	22-Apr-2000	augustss	Add some CMedia audio devices. From T.SHIOZAKI <AoiMoe@imou.to>
1.209	19-Apr-2000	bouyer	Add some Symbios devices.
1.208	18-Apr-2000	cgd	SiByte, Inc. is vendor 0x166d.
1.207	17-Apr-2000	thorpej	Add product ID for the S3 Trio3D/2X AGP graphics adapter.
1.206	03-Apr-2000	joda	add some NVIDIA chipsets
1.205	31-Mar-2000	tsarna	Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.204	30-Mar-2000	joda	add the device id for the VIA 686A modem controller, continuing the trend of mislabelling VIA's south bridges
1.203	27-Mar-2000	ad	Add a couple of DPT devices.
1.202	23-Mar-2000	joda	add two Intel 82810E devices (from IWAMOTO Toshihiro)
1.201	23-Mar-2000	ad	Back out rev 1.197.
1.200	21-Mar-2000	joda	add two VIA VT8371 devices
1.199	16-Mar-2000	thorpej	Add product ID for the Intel InBusiness 10/100 Ethernet adapter (which is an i82559 with a different product ID).
1.198	14-Mar-2000	tsarna	Add product ids for some VIA Technologies VT82C686A southbridge devices.
1.197	13-Mar-2000	ad	Add a bunch of Compaq array controllers.
1.196	11-Mar-2000	jdolecek	Add id for Alliance AT25, as sent by Dave Sainty in kern/9596.
1.195	09-Mar-2000	jdolecek	Add Conexant SoftK56 PCI Software Modem, Sony CXD3222 OHCI i.LINK(IEEE 1394) PCI controller, Sony Memory Stick I/F Controller Ids send by John Hawkinson in kern/9588
1.194	06-Mar-2000	mark	Add SYMPHONY 83C553 PCI-ISA bridge. Add alternative ID for Winbond W89C940F.
1.193	04-Mar-2000	tron	Add AMD Athlon chipset related devices. Patch supplied by Dave Sainty in PR kern/9533.
1.192	16-Feb-2000	soren	Add Toshiba ISA bridge.
1.191	08-Feb-2000	bouyer	Add US Robotics vendor ID, and USR 3CP5609 PCI 16550 Modem product ID. From Kent Polk.
1.190	03-Feb-2000	dante	Add AdvanSys U2W boards
1.189	03-Feb-2000	joda	2 Micro -> O2 Micro
1.188	31-Jan-2000	thorpej	Add SiS 7016 10/100 Ethernet.
1.187	25-Jan-2000	haya	Add TI PCI 1451 PCI-CardBus bridge.
1.186	24-Jan-2000	joda	TI PCI2030 is a PCI-PCI bridge
1.185	23-Jan-2000	tsubai	Add Apple KeyLargo and UniNorth.
1.184	12-Jan-2000	haya	Add an entry for Toshiba ToPIC100 CardBus-PCI bridge.
1.183	11-Jan-2000	matt	Add CompuModules DVB receiver and transmitter boards.
1.182	09-Jan-2000	soren	Add MPC105.
1.181	05-Jan-2000	drochner	add some stuff which was mentioned on the mailing lists recently
1.180	30-Dec-1999	hubertf	Add an entry for the Trident Microsystems CYBER 9525 (as found in an Toshiba Satellite 2650 XDVD).
1.179	17-Dec-1999	drochner	add Intel 450NX stuff
1.178	17-Dec-1999	augustss	Change the Galileo entries so the generated code can be compiled.
1.177	17-Dec-1999	soren	Fix comment.
1.176	17-Dec-1999	soren	Add Galileo Technology system controllers. Remove obsolete ID database URL.
1.175	15-Dec-1999	kleink	* Add ESS Maestro2 product ID. * Add TerraTec vendor ID.
1.174	04-Dec-1999	castor	Add vendor ID for Lava Semiconductor (http://www.lavalink.com)
1.173	04-Dec-1999	thorpej	Add some of the Compaq RAID controller IDs.
1.172	01-Dec-1999	drochner	add Cyclades 4Y and 8Y
1.171	30-Nov-1999	drochner	add 3c905B-COMBO
1.170	17-Nov-1999	augustss	Add the SBLive! joystick device.
1.169	12-Nov-1999	is	Identification of DOMEX PCI-SCSI board, by Thilo Manske, PR 8772.
1.168	07-Nov-1999	cgd	back out rev 1.166: (1) the list added isnt' complete and i'd call it incorrect. (There are other devices with these names, where were omitted. I've never seen the devices with these IDs, I _have_ seen the other devices that go by these names.) (2) the PR (8754) had not been updated to note that the problem was fixed, and i'd preiously asked the submitter for more information and/or a corrected list.
1.167	06-Nov-1999	rh	Add vendor/product entry for Auravision VxP524 PCI Video Processor.
1.166	06-Nov-1999	augustss	Add SIIG devices. Closes PR kern/8754 from Andreas Wrede <andreas@planix.com>
1.165	04-Nov-1999	tsarna	Add Aureal AU8820
1.164	04-Nov-1999	augustss	Add SBLive!
1.163	04-Nov-1999	joda	add Xircom X3201-3 Fast Ethernet Controller
1.162	01-Nov-1999	augustss	Add FM801 audio device driver. From Witold J. Wnuk ww181302@zodiac.mimuw.edu.pl Closes PR kern/8729.
1.161	29-Oct-1999	matt	Add AHA-1480 (adaptec cardbus scsi)
1.160	29-Oct-1999	haya	Maintainance the name of Ricoh PCI-CardBus bridges.
1.159	11-Oct-1999	hannken	branches: 1.159.2; 1.159.4; Add an entry for the Adaptec 2940UW Pro Scsi adapter.
1.158	29-Sep-1999	thorpej	branches: 1.158.2; - Add Sundance Technolgy vendor ID. - Add Sundance Technolgy ST201 10/100 Ethernet product ID. - Add D-Link DFE-550TX product ID (uses ST201 chip).
1.157	29-Sep-1999	thorpej	Add adaptec AIC-6915 10/100 Ethernet.
1.156	27-Sep-1999	ad	s@SmartCache/Raid@SmartCache/SmartRAID@
1.155	11-Sep-1999	soren	Add ENI SpeedStream.
1.154	10-Sep-1999	augustss	Add more TI CardBus bridges. Taken from the CardBus patches.
1.153	08-Sep-1999	thorpej	Add some more Tulip clones.
1.152	08-Sep-1999	thorpej	Add Davicom Semiconductor vendor ID, and Davicom DM9102 10/100 Etherent product ID (Yet Another Tulip Clone).
1.151	08-Sep-1999	thorpej	LITEON_PNIC -> LITEON_82C168
1.150	01-Sep-1999	drochner	-add some incarnations of the Alteon gigabit Ethernet adapter -sort some entries which were put in by legasthenics
1.149	28-Aug-1999	augustss	Add a bunch of i810 devices. From Paul Shupak <paul@plectere.com>.
1.148	27-Aug-1999	thorpej	Add Samsung Semiconductor KS8920 10/100 Ethernet.
1.147	24-Aug-1999	soren	Add Intel i740.
1.146	17-Aug-1999	drochner	-mention a location where the old PCI database can still be found -add Alteon Gigabit Ethernet
1.145	16-Aug-1999	mcr	Added Solidum Systems PCI vendor ID and Device IDs
1.144	12-Aug-1999	bouyer	Add Promise Ultra66, change entry of Ultra33 to reflect the fact that can "only" do ultra/33.
1.143	01-Aug-1999	veego	Add an entry for the TI PCI1211 PCI-CardBus Bridge. From the cardbus patches.
1.142	29-Jul-1999	thorpej	Add some Initio SCSI controller IDs.
1.141	28-Jul-1999	soren	Correct Acard device names. Add and update a few IBM devices. Add Toshiba and NEC devices in Portege 3K's.
1.140	27-Jul-1999	castor	Add vendor ID for Geocast
1.139	27-Jul-1999	thorpej	Add product ID for the SMC 83c175 Fast Ethernet chip. This is a cousin of the SMC 83c170 which also supports CardBus.
1.138	20-Jul-1999	drochner	add TI TVP4020 Permedia 2
1.137	15-Jul-1999	ross	product 3COM 3C905CTX 0x9200
1.136	15-Jul-1999	augustss	Add Cinemaster C 3.0 DVD Decoder.
1.135	14-Jul-1999	fvdl	Add/change a few Adaptec controller IDs.
1.134	12-Jul-1999	kleink	Change the ESS Solo-1 description string to match the data sheet.
1.133	30-Jun-1999	fair	A reference to http://members.hyperlink.net.au/~chart/pci.htm in comments; this new effort at collecting PCI data outside the PCI SIG replaces the one that was referenced in here before. No regeneration required.
1.132	27-Jun-1999	drochner	add some realtek 8139 clones
1.131	19-Jun-1999	augustss	Add Sony CXD 1947A FireWrire Controller and Neomagic 128XD.
1.130	18-Jun-1999	tsubai	Add some Apple and MOT products.
1.129	05-Jun-1999	tron	Add Matrox MGA 400, 3Dfx Banshee and Voodoo 3.
1.128	05-Jun-1999	tron	Add entry for Riva TNT and correct entry for Riva 128. Fixes first half of PR kern/7699 by Bjoern Labitzke.
1.127	02-Jun-1999	drochner	add SIS 6326 low-end AGP graphics fix CL_PD6832 device id, PR kern/7691 from Paul Dokas
1.126	29-May-1999	mrg	fix error in previous.
1.125	29-May-1999	mrg	s/bus/bridge/ for the sun pci bridges. my bad.
1.124	29-May-1999	fvdl	Add 3c900B-TPC product ID. From Robert Elz.
1.123	27-May-1999	mrg	add the UltraSPARC IIi PCI interface
1.122	23-May-1999	veego	Add several entries.
1.121	20-May-1999	augustss	Change vendor id 0x11c1 to Lucent since that seems to be what it is now. Add a Lucent USB controller.
1.120	17-May-1999	drochner	add NeoMagic 256AV Video/Audio chip
1.119	07-May-1999	hwr	Added CIRRUS CL_PD6832 from "Johnny C. Lam" <lamj@stat.cmu.edu> in kern/7119.
1.118	07-May-1999	hwr	Added TI PCI1251 from joda@pdc.kth.se (Johan Danielsson) in kern/6873.
1.117	07-May-1999	drochner	add another AGP graphics chip and a PCI-VME interface
1.116	07-May-1999	matt	Add TI LYNX FireWire Controller (as used in the Apple B&W G3)
1.115	06-May-1999	mrg	add pci devices found in U5's: - ati weird mach64 - ebus bus - `network' hme compatible ethernet - `simba' PCI bus
1.114	12-Apr-1999	jonathan	Add entries for TI, Ricoh, Toshiba PCI-to-cardbus bridge chips, so we can identify them as cardbus chips supported by the cardbus patches. Add entry for OPTI chipsets whose interrupts arent properly set up by some BIOSes. From cardbus patches<ftp://nandra.iri.co.jp/pub/NetBSD/CardBus by HAYAKAWA Koichi <haya@tcad.ulsi.sony.co.jp>.
1.113	25-Mar-1999	explorer	branches: 1.113.2; Add LAN Media Corporation info
1.112	19-Mar-1999	mycroft	Add AudioPCI 97 (a.k.a. Creative Ensoniq AudioPCI).
1.111	17-Mar-1999	thorpej	Add Silicon Integrated Systems SiS900 10/100 Ethernet.
1.110	16-Mar-1999	mjacob	add Qlogic 1080 && 1240 product IDs
1.109	11-Mar-1999	sakamoto	Add several VIA Technologies chipsets.
1.108	09-Mar-1999	drochner	add 2 ISDN cards, from Martin Husemann per PR kern/7110
1.107	19-Feb-1999	thorpej	Improve descriptions for DEC Ethernet chips.
1.106	19-Feb-1999	thorpej	Improve the descriptions of 3Com Ethernet cards, and add the 3c905B-FX and 3c980 Server Adapter products.
1.105	19-Feb-1999	thorpej	Add several Tulip clone chips.
1.104	14-Feb-1999	tron	Add 3Dfx's Voodoo II. Fixes PR kern/7000 by Brett Lymn.
1.103	05-Feb-1999	thorpej	Add some Olicom Ethernet device IDs, from FreeBSD's tl driver.
1.102	28-Jan-1999	augustss	Add S3 86C365 Trio3D.
1.101	16-Jan-1999	drochner	add another NCR chip, from Dave Sainty <dave@dtsp.co.nz> per PR kern/6819
1.100	30-Dec-1998	hwr	Add NeoMagic MagicGraph 128ZV graphic chip.
1.99	21-Dec-1998	drochner	change name of 82443BX according to Intel's data sheet correct "no AGP" case to real topology of the chip clarification: s/AGP Device/AGP Interface/
1.98	20-Dec-1998	drochner	add a Zoran multimedia controller chip
1.97	18-Dec-1998	thorpej	Add product ID for the Yamaha 724 Audio device. PR #6605, Dave Sainty.
1.96	09-Dec-1998	thorpej	Some more touchup on the VIA Rhine chip model number, and so some sorting.
1.95	05-Dec-1998	matt	Add Sigma Designs REALmagic Hollywood Plus MPEG-2 decoder
1.94	30-Nov-1998	jtk	Per PR/6172 (Johnny Lam, lamj@stat.cmu.edu), fix pcidevs identification of CL-PD6729. See <http://www.cirrus.com/products/overviews/pd6729.html>.
1.93	29-Nov-1998	thorpej	Fix the original Rhine's part number (thanks to Greg Oster).
1.92	29-Nov-1998	thorpej	Garbage-collect some bogus RealTek entries.
1.91	29-Nov-1998	thorpej	Add product IDs for the VIA Rhine and Rhine-II 10/100 Ethernet chips. (XXX What is the original Rhine's part number?!)
1.90	27-Nov-1998	drochner	add Powerstorm 4D50T graphics found in some Alphas
1.89	23-Nov-1998	mcr	added 82443BX_AGP_NOT as per kern/5847
1.88	20-Oct-1998	thorpej	Add RealTek 8129 and 8139 10/100Mb Ethernet IDs.
1.87	19-Oct-1998	bouyer	Yet another attemp at getting the VIA tech. product name/ID's rigth. It seems that different products have the same name (sigh).
1.86	05-Oct-1998	tron	Correct vendor 0x1191 to "ACARD", add missing Matrox boards. Fixes PR kern/6251 by Soren S. Jorvang.
1.85	04-Oct-1998	tron	Add S3 Virge/DX, fixes PR kern/6245 by Steve Allen.
1.84	01-Oct-1998	fvdl	Fix typo in 905BT4 product name.
1.83	26-Sep-1998	dante	Change product ADVSYS 2300 to ADVSYS WIDE
1.82	22-Sep-1998	mark	Add ALI M1531 Host-PCI bridge. Add Cirrus Logic CS4610 and CS4280 Audio chips. Add Cirrus Logic CS4400 communications controller.
1.81	18-Sep-1998	drochner	Add PCI-VME interfaces: Tundra Universe and Bit3 2706 Note that Newbridge is now Tundra Add Adaptec 7890 & Co.
1.80	18-Sep-1998	mark	Add yet another vendor ID for Fujitsu. Add vendor ID for ESS Technology Inc. Add product ID for ESS Technology ES1938 SOLO-1 AudioDrive Add product ID for Fujitsu K56flex LTMODEM Add product ID for TI PCI1250 PC Card bridge Add product ID for Trident Cyber 9397 Video controller.
1.79	06-Sep-1998	veego	Add the Brooktree 878 and 879 id's and fix the comment of the 849 entry.
1.78	26-Aug-1998	drochner	add the SIS 5597/5598 all-in-one PCI chipset
1.77	26-Aug-1998	dante	Changed ADVSYSPROD to ADVSYS vendor ID Added 1200A, 1200B, ULTRA, 2300 AdvanSys products ID
1.76	07-Aug-1998	thorpej	Add product IDs for the 3Com Cyclone-based PCI Ethernet devices, the 3c90xB DMA-only cards.
1.75	28-Jul-1998	thorpej	Add the IMS TwinTurbo 128M graphics board.
1.74	27-Jul-1998	tron	"Brooktree Brooktree Bt848" -> "Brooktree Bt848"
1.73	27-Jul-1998	tron	Add Brooktree Bt848 and BT849 video capture chips.
1.72	27-Jul-1998	augustss	Add S3 SonicVibes.
1.71	25-Jul-1998	explorer	define RISCom and RISCom N2 product
1.70	22-Jul-1998	drochner	add IDs for ACC Micro 2051 "PCI Single Chip Solution"
1.69	19-Jul-1998	tron	Fix product ID for Matrox Mystique, add NEC PowerVR PCX2.
1.68	13-Jul-1998	mark	Add ID for ALI M1541 Host-PCI bridge.
1.67	10-Jul-1998	tron	Add Intel's BX chipset, 3Dfx Interactive and their Voodoo board.
1.66	30-Jun-1998	augustss	Add a 2 Micro CardBus controller.
1.65	15-Jun-1998	bouyer	Argh, comments have to be in /* */, not # ! Sorry for the one-in-tree commit.
1.64	15-Jun-1998	bouyer	Hum the "VT82C595 (Apollo VP2) Host-PCI Bridge" has 2 devices entries (in FreeBSD too). Make it VT82C595 and VT82C595_2.
1.63	15-Jun-1998	bouyer	Add some VIA tech. products.
1.62	25-May-1998	mark	Added TriTech Microelectronics vendor ID. Added TriTech Pyramid3D 25202 product ID. Added Promise Ultra33 UltraDMA IDE controller product ID. Added ALI product ID's for M1543 'Southbridge' PCI devices.
1.61	14-May-1998	augustss	Add another USB controller.
1.60	13-May-1998	augustss	Add OPTi OHCI USB host controller.
1.59	05-May-1998	thorpej	Add a few more NE-2000 compatible PCI Ethernet interfaces. XXX Some of these may need further revision to get vendor and product names exactly right.
1.58	04-May-1998	thorpej	Add STB Systems' second PCI vendor ID, and their Velocity128 AGP product. kern/5362, Paul Goyette <paul@whooppee.com>
1.57	30-Apr-1998	mycroft	Add the Apple Bandit chip.
1.56	16-Apr-1998	drochner	add 2 ATM chips, correct entry for Toshiba MIPS bridge
1.55	28-Mar-1998	thorpej	Add the Contaq/Cypress 82C693 PCI-ISA bridge.
1.54	28-Mar-1998	cgd	fix bogus produce name tags for 82371FB_ISA and 82371FB_IDE (were 82471FB_*). Pointed out by Soren S. Jorvang. Rename "BAD" vendor id to "INVALID".
1.53	07-Mar-1998	augustss	Add Matrox Millenium II AGP.
1.52	04-Mar-1998	cgd	fix comment bogon in last commit
1.51	04-Mar-1998	cgd	add CMD PCI0643, PCI0647, USB0670, USB0673. Remove the 'A' from the 640A entry (because the same product id is used for the 640B, as well). Note that a few of the entries (PCI0642, PCI0650A) no longer have data to be found on the CMD web site, and note that PCI0650A should probably have its "A" trimmed as well.
1.50	27-Feb-1998	mycroft	Add Macronix MX98713.
1.49	07-Feb-1998	augustss	Fix typo, and commit in the right order.
1.48	07-Feb-1998	augustss	Add the AGP device in the 82443LX.
1.47	27-Jan-1998	thorpej	Add product ID for the Packet Engines G-NIC gigabit Ethernet interface.
1.46	22-Jan-1998	mycroft	Add the Millenium II.
1.45	16-Nov-1997	christos	PR/4510: William Coldwell: Add TI thunderlan support.
1.44	27-Oct-1997	thorpej	Fix the description of the Winbond NE2000-compatible Ethernet ASIC. It's a W89C940F.
1.43	14-Oct-1997	thorpej	branches: 1.43.2; Pull down from marc-pcmcia branch.
1.42	08-Oct-1997	mycroft	Correct spelling of Millennium.
1.41	06-Oct-1997	lonhyn	Added entries for GLINT 3D chips.
1.40	01-Oct-1997	drochner	Correct a number of Intel PCI devices, sync with names used in data sheet. Add newer Intel / VIA i386 chipsets. Correct one S3 chip, add one. Mostly from Carl Shapiro <css@samsara.dialup.access.net> per PR kern/4169 and kern/4170 (identical).
1.39	23-Sep-1997	mycroft	Add the Ensoniq AudioPCI.
1.38	15-Sep-1997	bouyer	Add some of Compaq's network products.
1.37	12-Sep-1997	drochner	another graphics chip...
1.36	28-Aug-1997	thorpej	Add an entry for the SMC 83C170 "EPIC/100" Fast Ethernet chip.
1.35	23-Aug-1997	drochner	Add one more graphics chip.
1.34	16-Aug-1997	mjacob	Add in PCI_PRODUCT_QLOGIC_ISP2100.
1.33	14-Aug-1997	drochner	Include some more stuff.
1.32	03-Aug-1997	fvdl	Add a few VIA Apollo chipset IDs
1.31	19-Jul-1997	cgd	branches: 1.31.2; clean up a few entries, fix the entry for the TI PCI1130 so the device is properly named, add entries for the TI PCI1131 and PCI2030. also note that you shouldn't always believe everything you read.
1.30	24-Jun-1997	thorpej	Correct Intel Triton MX PCI IDs, from Tom Yu <tlyu@mit.edu>, PR #3478.
1.29	24-Jun-1997	kml	Added definitions for Essential Communications RoadRunner HIPPI and Gigabit Ethernet interfaces.
1.28	13-May-1997	jtk	correct the identification of Intel Orion PCIset chips
1.27	04-Apr-1997	augustss	Two more PCI devices.
1.26	20-Mar-1997	chuck	- add Adaptec ATM ID's (0x5900, 0x5905). still trying to find out how they correspond to the ANA numbers... certainly includes ANA-5940. - add Efficient Nets product 0x0000. according to linux pci.h the 0x0000 is an FPGA version of the midway card and the 0x0002 is the ASIC version.
1.25	16-Feb-1997	veego	Fix errors in the last commit.
1.24	15-Feb-1997	tls	greatly expanded pcidevs file courtesy of Carl Shapiro <cshapiro@sparky.ic.sunysb.edu>.
1.23	01-Jan-1997	veego	branches: 1.23.4; Add the Matrox Millenium and regen the header files.
1.22	28-Nov-1996	jtk	add Cirrus CL-7541, 7543 and ACC 2188; regen .h files
1.21	28-Nov-1996	thorpej	Fill out the Intel section of this a bit, and note the URL of a periodically updated list of PCI vendor/product codes.
1.20	19-Oct-1996	jonathan	Add productId for RealTek 8029 PCI ethernet board.
1.19	15-Oct-1996	christos	Use the official 3com part names; from Curt Sampson
1.18	13-Oct-1996	christos	Apply PR/2840: Recognize more 3com models...
1.17	08-Oct-1996	gibbs	dev/microcode/aic7xxx/aic7xxx.seq, dev/microcode/aic7xxx_seq.h, dev/ic/aic7xxxreg.h: Remove intrinsic knowledge about SDTR and WDTR messages and replace it with a generic message system that allows the kernel driver to handle SDTR, WDTR and any other type of extended message it chooses too. This makes the sequencer code much simpler, makes extended message handling debuggable since the bulk of the work is in the kernel driver, and saves lots of instruction space. Regen microcode header file. dev/ic/aic7xxx.c, dev/ic/aic7xxxvar.h: Add code to handle WDTR and SDTR negotiation in light of the changes in the message interface to the sequencer. Don't reject targets that negotiate async by sending an SDTR with a 0 offset. Use an sdtr message with 0,0 to negotiate async when a target suggests a period that is too long for us to handle. Some tape and cdrom drives don't like us doing the message reject that we did in the past. Fix a problem with handing the QUEUE FULL condition. Fix a race condition (most likely the cause of the SCB paging problems) that might allow the sequencer to get unpaused before the condition that caused it to be paused (a SEQINT) was handled. Race condition pointed out by Doug Ledford <dledford@dialnet.net> and by "Dan Willis" <dan@plutotech.com>. dev/pci/ahc_pci.c: Add support for the 2940AU, an aic7860 based controller. dev/pci/pcidevs.h, dev/pci/pcidevs_data.h: Add product IDs for the 2940AU, aic7860 and aic7855. Regen data file. scsi/scsi_message.h: Add MSG_EXT_SDTR_LEN and MSG_EXT_WDTR_LEN - the length of bytes in these extended messages. Thanks to Chuck Cranor <chuck@maria.wustl.edu> for testing these changes out for me.
1.16	24-Sep-1996	christos	New cyclades driver from Timo Rossi.
1.15	29-Aug-1996	cgd	add entry for the IBM 82351 PCI-PCI Bridge
1.14	26-Aug-1996	cgd	add ids for the PGXGB (TGA2), and DGLPB (OPPO).
1.13	23-Aug-1996	cgd	add an entry for the FORE PCA-200e
1.12	15-Jul-1996	cgd	clean up, sort a few entries. Add many new entries, based on information gleaned from the XFree86 3.1.2E release.
1.11	22-Jun-1996	chuck	add product code for efficient networks eni155p atm card and regenerate
1.10	16-May-1996	mycroft	branches: 1.10.4; Add another ID for BusLogic cards.
1.9	13-May-1996	thorpej	Add an LSI Logic vendor id, and an S3 product id. Provide more detail on a couple of S3 and UMC product ids. From Shelby Noonan <shelby@og.org>
1.8	07-May-1996	thorpej	Add ID for AMD PCnet-PCI Ethernet.
1.7	04-Apr-1996	mycroft	Add some NCR chips.
1.6	19-Feb-1996	christos	Add 3c595 id
1.5	31-Jan-1996	cgd	update/clean entries for Adaptec PCI scsi chips/boards.
1.4	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.3	10-Nov-1995	christos	- pcidevs: add 3c590 and adaptec ultra - aic7xxx.c: fix the probe routines to use pcidevs.h
1.2	20-Jun-1995	cgd	vendor IDs for Acer and SIS
1.1	18-Jun-1995	cgd	learn about some PCI vendor and product numbers. not yet used
1.10.4.1	04-Mar-1997	mycroft	Pull up from -current.
1.23.4.1	12-Mar-1997	is	Merge in changes from The Trunk
1.31.2.9	14-Oct-1997	thorpej	Sync w/ trunk.
1.31.2.8	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.7	14-Oct-1997	thorpej	Add the Winbond NE2000-compatible Ethernet interfaces. XXX A better description of this card is needed.
1.31.2.6	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.5	22-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.4	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.3	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.2	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.31.2.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.43.2.2	17-Nov-1997	thorpej	Sync w/ trunk.
1.43.2.1	27-Oct-1997	thorpej	Pull up from trunk: Fix the description of the Winbond NE2000-compatible Ethernet ASIC. It's a W89C940F.
1.113.2.12	11-Sep-2000	he	Pull up revision 1.247 (requested by billc): Recognize 3Com's Home Connect (3c450) and Office Connect (3cSOHO100) Ethernet cards.
1.113.2.11	23-Jul-2000	he	Pull up revisions 1.214,1.245 (requested by jonathan): Add product codes for various Yamaha audio chips.
1.113.2.10	07-Jul-2000	he	Apply patch (requested by bouyer): Add Intel i810 PCI ID, and a few PCI IDE controllers.
1.113.2.9	03-Jun-2000	he	Pull up revision 1.172 (requested by tsubai): Add support for PCI variants of Cyclom-4Y and -8Y cards. Also fixes PR#7175 (new serial controller).
1.113.2.8	22-May-2000	he	Pull up revision 1.198 (requested by bouyer): Add product IDs for some VIA Technologies devices.
1.113.2.7	10-Feb-2000	he	Pull up revision 1.191 (via patch, requested by bouyer): Add support for US Robotics 3CP5609 PCI 16550 Modem.
1.113.2.6	15-Jan-2000	he	Pull up revision 1.156 (requested by ad): Fix description of DPT SCSI boards.
1.113.2.5	20-Oct-1999	he	Pull up via patch (requested by bouyer): Back-port of the Realtek ethernet driver.
1.113.2.4	05-Oct-1999	he	Pull up revision 1.150 (via patch, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.113.2.3	29-Aug-1999	he	Pull up revision 1.137: Recognize 3c905C-TX card. (bouyer)
1.113.2.2	22-Jun-1999	perry	pullup 1.123->1.124 (fvdl): recognize another 390xB card
1.113.2.1	13-Apr-1999	jonathan	branches: 1.113.2.1.2; 1.113.2.1.4; Pull up sys/dev/pci/pcidevs 1.113->1.114 (CardBus bridge chips) to 1.4. pcidevs.h, pcidevs_data.h: regen.
1.113.2.1.4.2	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.113.2.1.4.1	28-Jun-1999	itojun	KAME/NetBSD 1.4 SNAP kit, dated 19990628. NOTE: this branch (kame) is used just for refernce. this may not compile due to multiple reasons.
1.113.2.1.2.3	02-Aug-1999	thorpej	Update from trunk.
1.113.2.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.113.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.158.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.159.4.1	15-Nov-1999	fvdl	Sync with -current
1.159.2.10	23-Apr-2001	bouyer	Sync with HEAD.
1.159.2.9	21-Apr-2001	bouyer	Sync with HEAD
1.159.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.159.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.159.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.159.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.159.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.159.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.159.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.159.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.220.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.236.2.27	07-Apr-2004	jmc	Pullup rev 1.563-1.564 (requested by itojun in ticket #71) Add National Semiconductors SC1100 (GEODE) PCI devices
1.236.2.26	06-Jun-2002	he	Pull up revision 1.324 (via patch, requested by mycroft): Add Intel ICH3 network controller IDs. Part of fix to fxp driver to work around a bug which can cause some machines to randomly hang, and to drag along performance improvements and improved hardware support.
1.236.2.25	25-Mar-2002	he	Pull up revision 1.412 (via patch, requested by bouyer): Add support for the pciide part of the AMD-768 chipset.
1.236.2.24	06-Feb-2002	he	Pull up revisions 1.277,1.279,1.318 (requested by tsutsui): Add support for Acard ATP-850/860 pciide controllers.
1.236.2.23	29-Jan-2002	he	Pull up revision 1.287 (requested by bouyer): Add support for the 53c1010-33, and bring documentation up to date.
1.236.2.22	24-Jan-2002	he	Pull up revisions 1.352-1.353 (requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.236.2.21	25-Oct-2001	he	Pull up revisions 1.282,1.285,1.299 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.236.2.20	18-Jun-2001	jhawk	Pull up revision 1.329 (requested by tsubai): Support the new dual USB iBook.
1.236.2.19	11-Jun-2001	he	Pull up revision 1.326 (requested by tron): Support the D-Link DFE-530TX+. Fixes PR#13161.
1.236.2.18	15-May-2001	he	Pull up revision 1.320 (via patch, requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.236.2.17	06-May-2001	he	Pull up revision 1.275 (via patch, requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.236.2.16	03-May-2001	he	Pull up revisions 1.244,1.284,1.288,1.294-1.295,1.319 (via patch, requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.236.2.15	23-Apr-2001	he	Pull up revisions 1.263,1.286 (requested by hubertf): Support Accton EN2242 and other AmdTek AN985 compatible cards with the tlp(4) driver.
1.236.2.14	22-Mar-2001	he	Pull up revision 1.11 (requested by hubertf): Make the Intel Pro/100 as e.g. found on Toshiba Satellite Pro 4600s work.
1.236.2.13	20-Mar-2001	he	Pull up revision 1.260 (via patch, requested by chuck): Add support for the 3Com 3c555, 3c556 and 3c556B MiniPCI Ethernet cards.
1.236.2.12	26-Feb-2001	he	Pull up revision 1.302 (requested by bouyer): Add support for the Intel 82801BAM IDE controller.
1.236.2.11	31-Dec-2000	jhawk	Pull up revision 1.272 via patch (requested by bouyer): For the ti driver, add hardware 802.1Q support, update firmware to 12.4.13+wpaul, and support the Netgear GA620T copper gigabit card, and the Alteon ACEnic 100baseSX and 1000baseT gigabit cards. Various small changes from FreeBSD.
1.236.2.10	30-Oct-2000	tv	Pullup 1.267 [thorpej]: Add a bunch of Pequr (ServerWorks) products. From OpenBSD.
1.236.2.9	19-Oct-2000	he	Pull up revision 1.247 (requested by billc): Add PCI IDs for 3Com Home Connect (3c450) and Office Connect (3cSOHO100), register compatible with later versions of 3c905.
1.236.2.8	21-Sep-2000	thorpej	Pull up rev. 1.261: Add National Semiconductor DP83815 10/100 Ethernet.
1.236.2.7	13-Sep-2000	tron	Pull up from current (approved by thorpej): Add support VIA VT6102 chip found e.g on the D-Link DFE530TX NIC. syssrc/sys/dev/pci/if_vr.c 1.36-1.37 syssrc/sys/dev/pci/pcidevs 1.256-1.258
1.236.2.6	22-Aug-2000	enami	Pullup rev. 1.250 (approved by jhawk and thorpej): Add a product id for the promise ata-100 controller found on MS-6321 (694D Pro). Reported in the PR#10756 by Kazuki Sakamoto.
1.236.2.5	23-Jul-2000	jonathan	Pull up rev 1.245 (Yamaha YMF 7xx audio product codes) to 1.5 branch.
1.236.2.4	16-Jul-2000	jhawk	Pullup pcidevs rev 1.242 and regen pcidevs_data.h, pcidevs.h, approved by jhawk: Add ID for i82559ER.
1.236.2.3	06-Jul-2000	enami	Pull up rev. 1.240 (approved by releng-1-5): Add device ID of promise ultra100/ata controller.
1.236.2.2	27-Jun-2000	thorpej	Pull up rev. 1.238: add Intel 82801BA devices (and a few others)
1.236.2.1	26-Jun-2000	bouyer	Pull up rev 1.237, approved by thorpej: Add product ID for the CMD PCI0649 PCIIDE controller
1.306.2.28	17-Jan-2003	thorpej	Sync with HEAD.
1.306.2.27	15-Jan-2003	thorpej	Sync with HEAD.
1.306.2.26	03-Jan-2003	thorpej	Sync with HEAD.
1.306.2.25	29-Dec-2002	thorpej	Sync with HEAD.
1.306.2.24	19-Dec-2002	thorpej	Sync with HEAD.
1.306.2.23	11-Dec-2002	thorpej	Sync with HEAD.
1.306.2.22	26-Nov-2002	thorpej	Sync with HEAD.
1.306.2.21	11-Nov-2002	nathanw	Catch up to -current
1.306.2.20	18-Oct-2002	nathanw	Catch up to -current.
1.306.2.19	17-Sep-2002	nathanw	Catch up to -current.
1.306.2.18	27-Aug-2002	nathanw	Catch up to -current.
1.306.2.17	13-Aug-2002	nathanw	Catch up to -current.
1.306.2.16	01-Aug-2002	nathanw	Catch up to -current.
1.306.2.15	20-Jun-2002	nathanw	Catch up to -current.
1.306.2.14	17-Apr-2002	nathanw	Catch up to -current.
1.306.2.13	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.306.2.12	28-Feb-2002	nathanw	Catch up again (ATI lossage).
1.306.2.11	28-Feb-2002	nathanw	Catch up to -current.
1.306.2.10	11-Jan-2002	nathanw	More catchup.
1.306.2.9	08-Jan-2002	nathanw	Catch up to -current.
1.306.2.8	14-Nov-2001	nathanw	Catch up to -current.
1.306.2.7	22-Oct-2001	nathanw	Catch up to -current.
1.306.2.6	08-Oct-2001	nathanw	Catch up to -current.
1.306.2.5	21-Sep-2001	nathanw	Catch up to -current.
1.306.2.4	24-Aug-2001	nathanw	Catch up with -current.
1.306.2.3	21-Jun-2001	nathanw	Catch up to -current.
1.306.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.306.2.1	05-Mar-2001	nathanw	file pcidevs was added on branch nathanw_sa on 2001-04-09 01:57:04 +0000
1.336.2.9	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.336.2.8	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.336.2.7	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.336.2.6	16-Mar-2002	jdolecek	Catch up with -current.
1.336.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.336.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.336.2.3	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.336.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.336.2.1	03-Aug-2001	lukem	update to -current
1.355.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.355.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.364.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.424.2.4	29-Aug-2002	gehenna	catch up with -current.
1.424.2.3	15-Jul-2002	gehenna	catch up with -current.
1.424.2.2	20-Jun-2002	gehenna	catch up with -current.
1.424.2.1	30-May-2002	gehenna	Catch up with -current.
1.428.2.40	26-Oct-2005	jmc	Pullup (via patch) requested by riz in ticket #5773) Add support for 82801FB 10/100 ethernet
1.428.2.39	01-Jun-2005	riz	Pull up revision 1.702 (requested by tron in ticket #5721): Add missing PCI id for Adaptec ASR-2410SA.
1.428.2.38	01-May-2005	tron	Pull up revision 1.706 (requested by bouyer in ticket #5756): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.428.2.37	19-Mar-2005	tron	Apply patch (requested by he in ticket #1755): Add support for Adaptec 2200S Vulcan 2M, Dell PERC 320/DC, Adaptec 2410SA and Adaptec 2810SA controllers.
1.428.2.36	11-Nov-2004	he	Pull up revision 1.647 (via patch, requested by itohy): Add support for the Workbit NinjaSCSI-32 PCI/Cardbus SCSI driver, njs(4).
1.428.2.35	12-Jul-2004	he	Pull up revision 1.514 (via patch, requested by grant in ticket #1720): Add support for SATALink 3112.
1.428.2.34	22-Jun-2004	tron	Pull up revision 1.628 (requested by bouyer in ticket #1718): Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.428.2.33	06-Apr-2004	jmc	Pullup patch (requested by briggs in ticket #1648) Pull in changes to support the Intel 82801EB/ER (ICH5) Controller. PR#25002
1.428.2.32	28-Mar-2004	jmc	Pullup rev 1.565 via patch (requested by bouyer in ticket #1631) Add 3com 3c940 Gigabit Ethernet, and VIA KT600 PCI devices. PR#22453
1.428.2.31	15-Mar-2004	jmc	Pullup patch (requested by kent in ticket #1612) Add support for ICH3, ICH4, ICH5, SiS 7012, nForce MCP, nForce2 MCP-T, and nForce3 MCP-T to auich driver
1.428.2.30	20-Oct-2003	cyber	Pull up revision 1.569 (requested by tron in ticket #1440): Add entry for Intel i865G Integrated Graphics Device.
1.428.2.29	10-Sep-2003	grant	Pull up revision 1.544 (requested by tron in ticket #1417): Correct names and descriptions for i875 entries to be consistent with other Intel chipsets.
1.428.2.28	10-Sep-2003	grant	Pull up revision 1.543 (requested by tron in ticket #1416): Add entries for Intel 865PE chipset.
1.428.2.27	15-Aug-2003	tron	Pull up revision 1.564 (requested by itojun in ticket #1414): s,PCI/ISA,PCI-ISA,
1.428.2.26	15-Aug-2003	tron	Pull up revision 1.563 (requested by itojun in ticket #1414): add National Semiconductors SC1100 (GEODE) PCI devices.
1.428.2.25	10-Jul-2003	tron	Pull up revision 1.542 (requested by bouyer in ticket #1365): Add some intel devices, from Quentin Garnier in private mail.
1.428.2.24	30-Jun-2003	grant	Apply patch (requested by drochner in ticket #1343): add Intel PRO/100M.
1.428.2.23	19-Jun-2003	grant	Pull up revision 1.516 (requested by hamajima in ticket #1323): Add new Apple devices.
1.428.2.22	16-Jun-2003	grant	Pull up revision 1.511 (requested by bouyer in ticket #1234): Add several SiS host bridges, from FreeBSD. Remove duplicate in previous.
1.428.2.21	02-Jun-2003	tron	Pull up revision 1.503 (requested by bouyer): Add SiS 651 host bridge.
1.428.2.20	28-Jan-2003	jmc	Pullup revisions 1.442-1.443,1.444-1.445,1.452-1.453,1.491-1.492,1.496-1.497 (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.428.2.19	27-Jan-2003	jmc	Pullup revisions 1.453-1.454 (required hand patching) (requested by masanobu in ticket #1109) Add Intel 82801E Communications I/O Controller Hub (C-ICH)
1.428.2.18	12-Dec-2002	he	Pull up revisions 1.483,1.485 (requested by ad in ticket #1030): Add support for Adaptec 2005S. Handle DPT ioctls() from native binaries. Fix SCSI probe problem.
1.428.2.17	28-Nov-2002	tron	Pull up revision 1.472 (requested by kent in ticket #1019): - Broadcom BCM4401 - VIA VT8235 - Modify description of VT8233_AC97: "VT8233 AC-97 ..." -> "VT8233/VT8235 AC-97 ..."
1.428.2.16	26-Nov-2002	tron	Pull up revision 1.489 (requested by thorpej in ticket #1018): Add IBM 133 PCI-X Bridge.
1.428.2.15	26-Nov-2002	tron	Pull up revision 1.488 (requested by thorpej in ticket #1017): Add Intel i80321 I/O processor.
1.428.2.14	21-Nov-2002	tron	Pull up revision 1.474 (requested by yamt in ticket #925): fix typos; "wiseman" is 82542, not 82452.
1.428.2.13	16-Nov-2002	he	Pull up revision 1.459 (requested by ad in ticket #646): Add additional Adaptec RAID controller types. Fixes PR#17719.
1.428.2.12	07-Nov-2002	tron	Pull up revision 1.451 (requested by thorpej in ticket #452): Clean up Intel i8254x product names, and add i82540, i82545, and i82546 product IDs.
1.428.2.11	01-Nov-2002	tron	Pull up revision 1.466 (requested by bouyer in ticket #724): Add Promise Ultra133TX2 and Ultra133TX2v2 PCI ID.
1.428.2.10	02-Aug-2002	lukem	pullup rev 1.438 (requested by bouyer in ticket #603): PCI_PRODUCT_VIATECH_VT8233A
1.428.2.9	01-Aug-2002	lukem	Pull up revision 1.458 (requested by bouyer in ticket #589): Ops, remove duplicate entry for 82801DB_LPC, pointed out by geoff wing.
1.428.2.8	01-Aug-2002	lukem	Pull up revision 1.457 (requested by bouyer in ticket #589): Add Intel 845G PCI IDs, from kern/17743 by Andreas Wrede <andreas@planix.com>
1.428.2.7	12-Jul-2002	thorpej	pullup-1-6 ticket #488 (matt). Original log message: Add Apple Xserve PCI devices (more UniNorth) stuff.
1.428.2.6	25-Jun-2002	lukem	Pull up revisions 1.441-1.442 (requested by cjs in ticket #362): revision 1.442 For new Intel network controllers, update descriptions and fix numbering. revision 1.441 Add new Intel Ethernet cards.
1.428.2.5	21-Jun-2002	lukem	Pull up revision 1.440 (requested by thorpej in ticket #344): Add product ID for Intel 82452KX/GX PCI-Host bridge. PR 17353, from MOCHIDA Shuji.
1.428.2.4	11-Jun-2002	lukem	Pull up revision 1.434 (requested by seb in ticket #249): Add product Titan PCI-800L aka VSCom PCI-800L, a 8 serial puc card, from PR 17140. Reviewed/approved by veego@ and martin@.
1.428.2.3	11-Jun-2002	lukem	Pull up revision 1.429 (requested by lukem in ticket #165): add NetMos Quad UART and 1284 Printer port
1.428.2.2	11-Jun-2002	lukem	Pull up revision 1.432 (requested by tron in ticket #139): Add entries for Intel i850 and i860 chipsets.
1.428.2.1	10-Jun-2002	tv	Pull up revision 1.437 (requested by taca in ticket #228): Add VIA Technologies's VT8231 IDE Controller.
1.542.2.18	11-Dec-2005	christos	Sync with head.
1.542.2.17	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.542.2.16	01-Apr-2005	skrll	Sync with HEAD.
1.542.2.15	08-Mar-2005	skrll	Sync with HEAD.
1.542.2.14	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.542.2.13	04-Feb-2005	skrll	Sync with HEAD.
1.542.2.12	24-Jan-2005	skrll	Sync with HEAD.
1.542.2.11	17-Jan-2005	skrll	Sync with HEAD.
1.542.2.10	18-Dec-2004	skrll	Sync with HEAD.
1.542.2.9	29-Nov-2004	skrll	Sync with HEAD.
1.542.2.8	14-Nov-2004	skrll	Sync with HEAD.
1.542.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.542.2.6	19-Oct-2004	skrll	Sync with HEAD
1.542.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.542.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.542.2.3	03-Sep-2004	skrll	Sync with HEAD
1.542.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.542.2.1	03-Aug-2004	skrll	Sync with HEAD
1.615.2.17	12-Nov-2004	jmc	branches: 1.615.2.17.2; Pullup rev 1.660-1.661 (requested by martti in ticket #953) Added Dell PERC 4/SC
1.615.2.16	12-Nov-2004	jmc	Pullup rev 1.657 (requested by martti in ticket #947) Added support for DELL PERC 4e/Si RAID controller
1.615.2.15	08-Oct-2004	jmc	Pullup rev 1.622 (requested by he in ticket #896) Add support for Adaptec AAR 2810SA raid controller.
1.615.2.14	18-Sep-2004	he	Pull up revision 1.651 (requested by martin in ticket #856): Add support for PCI-X version of LSI FC929, from PR#26836.
1.615.2.13	30-Aug-2004	tron	Backout previous two overeager pullups and apply patch (requested by itohy in ticket #800): Add Workbit NinjaSCSI-32 PCI/CardBus products.
1.615.2.12	30-Aug-2004	tron	Pull up revision 1.647 (requested by itohy in ticket #800): Add Workbit NinjaSCSI-32 PCI/CardBus products.
1.615.2.11	30-Aug-2004	tron	Pull up revision 1.639 (requested by itohy in ticket #800): Merge in data from cardbusdevs.
1.615.2.10	22-Aug-2004	tron	Pull up revision 1.631 (requested by junyoung in ticket #781): Add 3Com 3c920B-EMB-WNM Integrated Fast Ethernet.
1.615.2.9	28-Jul-2004	tron	Pull up revision 1.624 (requested by joda in ticket #721): Add i6300ESB IDE / SATA controller PCI IDs.
1.615.2.8	28-Jul-2004	tron	Pull up revision 1.635 (requested by rumble in ticket #718): Fix typos in the SC1100 IDE and AUDIO ids.
1.615.2.7	23-Jul-2004	he	Pull up revision 1.633 (requested by mycroft in ticket #691): Add support for SiS 741. Fixes PR#26394.
1.615.2.6	05-Jul-2004	he	Pull up revision 1.629 (requested by mycroft in ticket #606): Update ServerWorks chip IDs, per http://www.plasma-online.de/english/identify/picture/serverworks.html Seems remarkably more sane.
1.615.2.5	22-Jun-2004	tron	Pull up revision 1.628 (requested by bouyer in ticket #537): Apply patch from kern/25936 by Michael Blake: Add support for the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.615.2.4	13-May-2004	jdc	Pull up revision 1.625 (requested by tron in ticket #316) Add support for Broadcom BCM5788 GbE. Patches supplied by FUKAUMI Naoki in PR kern/25166.
1.615.2.3	11-May-2004	tron	Pull up revision 1.621 (requested by sekiya in ticket #284): Add support for Silicon Image 3512 chip to satalink driver -- treat it as a 3112 variant (without RAID support), which does the right thing with the controller on my K8NNXP.
1.615.2.2	05-Apr-2004	tron	Pull up revision 1.618 (requested by martti in ticket #54): Added another SB Live! 5.1 soundcard and joystick version
1.615.2.1	01-Apr-2004	jmc	Pullup rev 1.616 (requested by tron in ticket #38) Correct entry for PCI vendor id 0x11af. It is assigned to Avid Technology Inc. and not to Pro-Log Corporation. PR#25004
1.615.2.17.2.19	21-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #10749): sys/dev/pci/pucdata.c: revision 1.49 sys/dev/pci/pcidevs: revision 1.849 sys/dev/pci/pcidevs: revision 1.850 Add NetMos 9855 Quad UART and 1284 Printer port Fix typo in previous (hi bouyer).
1.615.2.17.2.18	11-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #10693): sys/dev/pci/pcidevs: revision 1.820 via patch Rename vendor GALILEO -> vendor MARVELL . Marvell bought Galileo years ago, we have little existing stuff in the tree, and this will make it easier to keep certain network drivers in sync with other OSes. OK matt@, uwe@
1.615.2.17.2.17	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #10635): sys/dev/pci/pcidevs: revision 1.788 Add a couple DLINK product IDs, from OpenBSD.
1.615.2.17.2.16	11-Apr-2006	riz	Pull up following revision(s) (requested by he in ticket #10409): sys/dev/pci/pcidevs: revision 1.776 via patch Add a bunch of IDs for Compaq/HP Smart Array controllers supported by the soon to be added ciss(4) driver.
1.615.2.17.2.15	17-Mar-2006	riz	Pull up following revision(s) (requested by hw in ticket #10364): sys/dev/pci/pcidevs: revision 1.774 Add vendor code for Nortel and Baystack 21 id. Part of info submitted in PR#33044.
1.615.2.17.2.14	01-Dec-2005	tron	Pull up following revision(s) (requested by riz in ticket #10158): sys/dev/pci/pcidevs: revision 1.710 Two re(4) candidates. From Francois Romieu in private mail.
1.615.2.17.2.13	01-Dec-2005	riz	Pull up following revision(s) (requested by cube in ticket #5990): sys/dev/pci/pcidevs: revision 1.727 via patch New Dell day: add missing device IDs, and order one correctly.
1.615.2.17.2.12	21-Jul-2005	snj	Pull up revision 1.703 (requested by tron in ticket #5507): Add PCI id for Intel PRO/1000 GT.
1.615.2.17.2.11	05-Jul-2005	riz	Pull up patch (requested by bouyer in ticket #2029): Add the SIS 964 and 965 host bridges
1.615.2.17.2.10	21-Jun-2005	riz	Pull up revision 1.719 via patch (requested by bouyer in ticket #2007): Add Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern@
1.615.2.17.2.9	19-May-2005	tron	Pull up revision 1.711 (requested by riz in ticket #1532): Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.615.2.17.2.8	19-May-2005	tron	Pull up revision 1.658 via patch (requested by riz in ticket #1531): Add two Broadcom devices (575x chips), correct and add a few Intel entries.
1.615.2.17.2.7	19-May-2005	tron	Pull up revision 1.637 (requested by riz in ticket #1531): Add a bunch of devices that are in the Dell Dimension 8400 sitting on my desk: ATI Radeon X300, Broadcom BMC5751, Intel 915 chipset family (ICH6).
1.615.2.17.2.6	11-May-2005	riz	Pull up revision 1.705 (requested by martti in ticket #1402): Added support for Dell CERC SATA RAID 1.5/6ch
1.615.2.17.2.5	10-May-2005	riz	Pull up revision 1.702 (requested by tron in ticket #1341): Add missing PCI id for Adaptec ASR-2410SA.
1.615.2.17.2.4	05-May-2005	riz	Pull up revisions 1.691-1.692 (requested by abs in ticket #1153): Add 'VT642 SATA RAID Controller'. It looks like a viaide, it smells like a viaide... but you can only address the drives once you have put two or more of them into a RAID or JBOD.
1.615.2.17.2.3	01-May-2005	tron	Pull up revision 1.706 (requested by bouyer in ticket #1438): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.615.2.17.2.2	17-Apr-2005	tron	Pull up revision 1.687 (requested by fredb in ticket #1120): Recognize Belkin Gigabit Desktop Network PCI card.
1.615.2.17.2.1	07-Apr-2005	tron	Pull up revision 1.662 (requested by hira in ticket #1103): Add Nvidia nForce3 250 chipset products, from Kouichirou Hiratsuka in PR 27843.
1.685.2.1	29-Apr-2005	kent	sync with -current
1.688.2.4	26-Mar-2005	yamt	sync with head.
1.688.2.3	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.688.2.2	12-Feb-2005	yamt	sync with head.
1.688.2.1	25-Jan-2005	yamt	file pcidevs was added on branch yamt-km on 2005-02-12 18:17:48 +0000
1.701.2.50	10-Nov-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #1976): sys/dev/pci/if_wm.c: revision 1.162 sys/dev/pci/pcidevs: revision 1.961, 1.962 via patch Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.
1.701.2.49	08-Apr-2008	jdc	Pull up revisions: src/sys/dev/pci/pcidevs 1.816 src/sys/dev/pci/piixide.c 1.28 (requested by bouyer in ticket #1909). Add device IDs for Intel's 631xESB/632xESB southbridges, and start the list for the 5000X northbridge. I'll add the rest later, the spec is currently unavailable so I'll collect them from another source. Add support for the 631xESB/632xESB.
1.701.2.48	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1863): sys/dev/pci/if_wm.c 1.137-1.138,1.141,1.144-1.146 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 via patch share/man/man4/wm.4 1.18-1.19 Fix typo. Fix EEPROM reload sequence. Add ICH9 support. Workaround for 82541 Errata 29 and 82547 Errata 28: These devices have to reset the PHY before reset the MAC. Reported and tested by salo.
1.701.2.47	19-Nov-2007	bouyer	Pull up following revision(s) via patch (requested by msaitoh in ticket #1862): sys/dev/pci/if_bge.c 1.89,1.95,1.97,1.101,1.103, 1.105,1.107,1.117-1.122, 1.128-1.131,1.133-1.135,1.139 via patch sys/dev/pci/if_bgereg.h 1.31,1.34-1.39,1.42-1.43 sys/dev/pci/pcidevs patch Some cleanups(typo, whitespace, debug message). Add TSO support to bge(4) Add support for BCM5715, BCM5752M, BCM5754, BCM5755, BCM5786 and 5787. Add support for BCM5789. Fixes PR/33828 from Pawel Chwalowskibg. Add support for BCM5753 and BCM5753M Fixes PR kern/36139 from Mustafa Dogan. Add and fix some BGE_CHIPIDs. From FreeBSD and OpenBSD. In bge_ioctl(), don't call bge_init() if the interface is already up (Makes ifwatchd work on this interface). BCM5701 B2 requires BGE_QUIRK_ONLY_PHY_1 quirk. From Brian Baird in kern/23778. Move TX ring full sanity check further up and check the number of DMA segments from the DMA map, instead of counting the DMA segments in the for loop and breaking out later. Unload the DMA map if encountering an error condition. Fix a NULL reference on failing mbuf allocation. In bge_start return if IFF_OACTIVE is set in combination with IFF_RUNNING, drop obscure check for length of interface queue. Remove message about failing bge_encap, it can happen just too easily because of full descriptor rings. Fix some watchdog timeout and "block failed to stop" problems on some cards. Fix "pcistate failed to revert" problem on many cards.
1.701.2.46	14-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1770): share/man/man4/wm.4: revision 1.17 sys/dev/pci/if_wmreg.h: revision 1.22 sys/dev/pci/if_wm.c: revision 1.139 via patch sys/dev/pci/if_wm.c: revision 1.140 via patch sys/dev/pci/pcidevs: patch Add support for ICH8 (i82801H) LAN, from FreeBSD. Should work on: i82801H IFE (GT) LAN Controller i82801H IFE (G) LAN Controller i82801H (M_AMT) LAN Controller i82801H (AMT) LAN Controller i82801H LAN Controller i82801H (IFE) LAN Controller i82801H (M) LAN Controller tested with: Intel i82801H (AMT) LAN Controller, rev. 2 Restore $NetBSD$ tag. Note ICH8 LAN support.
1.701.2.45	14-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1769): sys/dev/pci/pcidevs: revision 1.876 sys/dev/pci/pcidevs: revision 1.877 Add: Intel i82801H IFE (GT) LAN Controller Intel i82801H IFE (G) LAN Controller from FreeBSD Remove extra "Intel" from previous commit
1.701.2.44	13-Sep-2007	bouyer	Regen for ticket #1822
1.701.2.43	31-Mar-2007	bouyer	pullup the following revisions (requested by msaitoh in ticket 1681): sys/dev/pci/if_wm.c 1.104-1.105, 1.116-1.121, 1.127,1.133-1.134 via patch sys/dev/pci/if_wmreg.h 1.17-1.20 sys/dev/pci/pcidevs patch sys/dev/mii/igphy.c 1.11 sys/dev/mii/makphy.c 1.20, 1.23 sys/dev/mii/ikphy.c patch sys/dev/mii/ikphyreg.h patch sys/dev/mii/miidevs 1.68 sys/dev/mii/files.mii 1.39 sys/arch/i386/conf/GENERIC 1.788-1.789 via patch sys/arch/i386/conf/GENERIC.MPACPI patch sys/arch/i386/conf/GENERIC_LAPTOP 1.209 via patch sys/arch/i386/conf/INSTALL 1.291 via patch sys/arch/i386/conf/INSTALL_LAPTOP 1.104 via patch sys/arch/i386/conf/XEN2_DOM0 1.13 via patch share/man/man4/wm.4 1.14-1.16 Add support for many cards (include PCI-express based chips). Many bug fixes about auto negotiations (PR#30078, PR#30490, PR#30906, PR#33429 and PR#35386). Fix media link issues with fiber-based card (PR#35797).
1.701.2.42	03-Mar-2007	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1673): distrib/sets/lists/man/mi 1.844, 1.871 share/man/man4/Makefile 1.372, 1.379 via patch share/man/man4/re.4 1.8, 1.9, 1.11, 1.12 share/man/man4/rgephy.4 1.1 share/man/man4/rlphy.4 1.1, 1.2 sys/arch/amd64/conf/GENERIC 1.79, 1.89 sys/arch/amd64/conf/INSTALL 1.45, 1.49 sys/arch/i386/conf/GENERIC 1.717, 1.747 sys/arch/i386/conf/GENERIC_LAPTOP 1.164 sys/arch/i386/conf/INSTALL 1.277, 1.283 sys/arch/i386/conf/INSTALL_LAPTOP 1.96, 1.98 sys/arch/i386/conf/XEN2_DOM0 1.22 via patch sys/arch/macppc/conf/GENERIC 1.220, 1.246 sys/arch/macppc/conf/INSTALL 1.96, 1.100 sys/dev/cardbus/if_re_cardbus.c 1.10 sys/dev/cardbus/if_rtk_cardbus.c 1.29-1.31 sys/dev/ic/rtl8169.c 1.14, 1.20, 1.24, 1.25, 1.28-1.61, 1.63, 1.64-1.81 via patch sys/dev/ic/rtl81x9.c 1.52, 1.54-1.63, 1.65, 1.67-1.70 sys/dev/ic/rtl81x9reg.h 1.15-1.26 sys/dev/ic/rtl81x9var.h 1.19, 1.21-1.37 sys/dev/mii/files.mii 1.38 sys/dev/mii/miidevs 1.62, 1.64 sys/dev/mii/rgephy.c 1.10, 1.13-1.16 sys/dev/mii/rlphy.c 1.1, 1.6, 1.7, 1.11 via patch sys/dev/pci/if_re_pci.c 1.13, 1.15-1.17, 1.19-1.23 sys/dev/pci/if_rtk_pci.c 1.25, 1.28, 1.29, 1.31 sys/dev/pci/pcidevs 1.851, 1.852 on re(4): - improve stability (I believe ;-) - add a workaround for hardware ip4csum-tx bug - support newer chips (8169SB/SC, PCIe based 8168 etc.) - fix 8139C+ support - enable hardware VLAN - misc bus_dma(9) fix (which makes re(4) work on mips ports) on rtk(4): - fix kern/31348 - fix possible panic on dreamcast
1.701.2.41	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1609): sys/arch/i386/conf/INSTALL: revision 1.298 sys/arch/i386/conf/GENERIC: revision 1.802 sys/dev/pci/files.pci: revision 1.274 sys/arch/amd64/conf/GENERIC: revision 1.122 sys/dev/pci/mfi_pci.c: revision 1.1 via patch distrib/sets/lists/man/mi: revision 1.958 sys/arch/amd64/conf/INSTALL: revision 1.61 sys/dev/ic/mfireg.h: revision 1.1 share/man/man4/Makefile: revision 1.416 via patch sys/dev/ic/mfi.c: revision 1.1 sys/arch/i386/conf/XEN2_DOM0: revision 1.19 sys/dev/ic/mfivar.h: revision 1.1 sys/conf/files: revision 1.821 share/man/man4/mfi.4: revision 1.1 share/man/man4/mfi.4: revision 1.2 sys/dev/pci/pcidevs via patch Add mfi(4), a driver for LSI Logic & Dell MegaRAID SAS RAID controller. Ported from OpenBSD, tested on i386 and amd64. Punctuation issues.
1.701.2.40	23-Dec-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1608): sys/arch/i386/conf/INSTALL: revision 1.299 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.20 via patch sys/arch/i386/conf/GENERIC: revision 1.803 via patch sys/dev/pci/files.pci: revision 1.275 via patch sys/arch/i386/conf/GENERIC_LAPTOP: revision 1.218 via patch sys/arch/amd64/conf/GENERIC: revision 1.123 via patch sys/dev/pci/if_bnx.c: revision 1.1 via patch share/man/man4/bnx.4: revision 1.1-1.2 via patch distrib/sets/lists/man/mi: revision 1.959 via patch sys/arch/amd64/conf/INSTALL: revision 1.62 via patch sys/dev/microcode/bnx/bnxfw.h: revision 1.1-1.2 via patch sys/dev/pci/if_bnxreg.h: revision 1.1 via patch share/man/man4/Makefile: revision 1.417 via patch sys/arch/i386/conf/INSTALL_LAPTOP: revision 1.112 via patch sys/dev/pci/pcidevs: patch Add bnx(4), a driver for Broadcom NetXtreme II 10/100/1000 Ethernet device. Ported from OpenBSD by cube@, with some bus_dma fixes by me. Tested on i386 and amd64. Add a NetBSD RCS tag. Nx -> Ox; remove trailing whitespace.
1.701.2.39	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1584): sys/dev/pci/pucdata.c: revision 1.49 sys/dev/pci/pcidevs: revision 1.849 sys/dev/pci/pcidevs: revision 1.850 Add NetMos 9855 Quad UART and 1284 Printer port Fix typo in previous (hi bouyer).
1.701.2.38	27-Aug-2006	ghen	branches: 1.701.2.38.2; Pull up following revision(s) (requested by oster in ticket #1479): sys/dev/pci/pcidevs: revision 1.810 sys/dev/pci/if_fxp_pci.c: revision 1.44 sys/dev/pci/if_fxp_pci.c: revision 1.47 sys/dev/pci/pcidevs: revision 1.756 PR/32494: Add ID 0x1068 for Intel PRO/100 VE (LOM) found on Sony Vaio S4HRP laptops. PR/32494: Dennis Chernoivanov: Intel PRO/100 VE (LOM) not detected on Sony Vaio Recognize another Intel PRO/100 VM Network Controller.
1.701.2.37	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #1466): sys/dev/pci/pcidevs: revision 1.798, 1.819 sys/dev/pci/pdcsata.c: revision 1.8, 1.10 Add support for PDC20775 and PDC40518. Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.701.2.36	03-Aug-2006	tron	Pull up following revision(s) (requested by xtraeme in ticket #1446): sys/dev/pci/pcidevs: revision 1.813 Add nForce3 250 SATA2.
1.701.2.35	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.701.2.34	30-Jul-2006	tron	Pull up following revision(s) (requested by kent in ticket #1432): sys/dev/pci/pcidevs: revision 1.806 via patch - add VIA VT8237A/8251 High Definition Audio Controller - add some tabs
1.701.2.33	04-Jun-2006	tron	Apply patch (requested by jonathan in ticket #1351): Add support to bge(4) for bcm5714, bcm5715, bcm5780/HT-2000 and bcm5752 chip variants.
1.701.2.32	04-Jun-2006	tron	Pull up following revision(s) (requested by riz in ticket #1349): sys/dev/pci/pcidevs: revision 1.788 Add a couple DLINK product IDs, from OpenBSD.
1.701.2.31	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): share/man/man4/pciide.4: revision 1.60 share/man/man4/Makefile: revision 1.376 via patch sys/dev/pci/svwsata.c: revision 1.1 sys/dev/pci/files.pci: revision 1.243 sys/arch/i386/conf/INSTALL: revision 1.281 sys/dev/pci/pcidevs: revision 1.771 via patch sys/dev/pci/pciidevar.h: revision 1.33 sys/arch/i386/conf/XEN0: revision 1.34 sys/arch/i386/conf/GENERIC: revision 1.732 sys/dev/pci/pciide_svwsata_reg.h: revision 1.1 Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.701.2.30	30-Apr-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #1295): sys/dev/pci/pcidevs: revision 1.785 sys/dev/pci/viaide.c: revision 1.29 PR port-amd64/33393: viaide driver patch for VT8237R SATA.
1.701.2.29	20-Apr-2006	snj	Pull up following revision(s) (requested by tron in ticket #1266): sys/dev/pci/pcidevs: revision 1.775 via patch fix name of Radeon 9000 Pro. add Radeon 9200 Secondary ID. add Comtrol, Magma and SIIG multi-port serial cards. add more Dolphin SCI cards. add IBM ServeRAID 4/5. add Mitsubishi Tornado 3000. add or rename various NVIDIA IDs to match openbsd. add SafeNet SafeXcel.
1.701.2.28	08-Apr-2006	tron	Pull up following revision(s) (requested by he in ticket #1247): sys/dev/pci/pcidevs: revision 1.776 Add a bunch of IDs for Compaq/HP Smart Array controllers supported by the soon to be added ciss(4) driver.
1.701.2.27	17-Mar-2006	tron	Pull up following revision(s) (requested by he in ticket #1203): sys/dev/pci/pcidevs: revision 1.774 Add vendor code for Nortel and Baystack 21 id. Part of info submitted in PR#33044.
1.701.2.26	15-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1169): sys/dev/pci/pcidevs: revision 1.765 Fix inconsistencies in names for components for recent Intel chipsets.
1.701.2.25	15-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1169): sys/dev/pci/pcidevs: revision 1.763 Add entries for Intel i955X chipset.
1.701.2.24	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/agp_i810.c: revision 1.28 sys/arch/i386/pci/pchb.c: revision 1.57 sys/dev/pci/agpvar.h: revision 1.11 sys/dev/pci/agp_via.c: revision 1.9 sys/dev/pci/agp.c: revision 1.37 sys/dev/pci/agpreg.h: revision 1.10 sys/dev/pci/pcidevs: revision 1.757 sys/dev/pci/agp_amd.c: revision 1.14 sys/dev/pci/agp_intel.c: revision 1.17 sys/dev/pci/agp_sis.c: revision 1.8 sys/dev/pci/agp_ali.c: revision 1.9 sys/arch/i386/pci/pchb_rnd.c: revision 1.16 PR/32550: Michael Eriksson: No agp(4) support for i915 integrated graphics
1.701.2.23	14-Feb-2006	riz	Pull up following revision(s) (requested by tron in ticket #1168): sys/dev/pci/pcidevs: revision 1.723 Add ATI Radeon Mobility X600. Add Mobile Intel 915PM/GM/GMS and 910GML Express Chipset.
1.701.2.22	05-Feb-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1156): sys/dev/pci/pcidevs: revision 1.758 Add Promise PDC2057x SATA/150 controllers, from FreeBSD.
1.701.2.21	21-Jan-2006	snj	Pull up following revision(s) (requested by simonb in ticket #1126): sys/dev/pci/pcidevs: revision 1.739 sys/dev/pci/pdcsata.c: revision 1.4 PR/32238: Ryo Shimizu: Add support for Promise SATA300TX4 SATA Controller (PDC40718)
1.701.2.20	10-Jan-2006	snj	Pull up following revision(s) (requested by tron in ticket #1112): sys/dev/pci/pcidevs: revision 1.754 Fix cut and paste error. Noted by Bernhard Moellemann.
1.701.2.19	10-Jan-2006	snj	Pull up following revision(s) (requested by tron in ticket #1112): sys/dev/pci/pcidevs: revision 1.753 - Spell NVIDIA as "NVIDIA" instead of "Nvidia". This matches their website and viaide(4). - Add entry for builtin ethernet of NVIDIA nForce4 chipset. Information provided by Bernhard Moellemann.
1.701.2.18	05-Jan-2006	riz	Pull up following revision(s) (requested by manu in ticket #1086): sys/dev/pci/pcidevs: revision 1.755 sys/dev/pci/viaide.c: revision 1.27 Add support for nForce430 ATA133 and SATA controllers. My disks now run at a decent speed.
1.701.2.17	15-Dec-2005	tron	branches: 1.701.2.17.2; Pull up following revision(s) (requested by jonathan in ticket #1047): share/man/man4/amr.4: revision 1.11 sys/dev/pci/amr.c: revision 1.31 sys/dev/pci/pcidevs: revision 1.746 via patch Add pcidevs entries and support to amr(4) for newer LSI MegaRAID devices, including the LSI MegaRAID SCSI 320-0x, 320-2x, 320-4x, 320-1E, 320-2E, the LSI MegaRAID SATA 300-6x and 300-8x. Tested on an LSI MegaRAID SATA 300-8x, which confirms private advice that after PCI-IDs to our amr(4) driver, the newer LSI MegaRAID cards will Just Work The Linux megaraid2 driver accepts various Intel cards with the same pci-device IDs as the LSI MegaRAID cards, treating them exaclty as the above-listed LSI cards. Rework our amr(4) driver to also match and attach those device-ID if seen with an Intel vendor-ID. Add the Intel-vendor PCI decice-ids to pcidevs, so that PCI_VERBOSE will correctly identify the Intel cards. Update src/share/man/man4/amr.4 to list the newly-supported cards. (NB: I don't have the LSI SCSI or any Intel cards to test, just the LSI SATA, which works fine for me.)
1.701.2.16	24-Nov-2005	tron	Pull up following revision(s) (requested by riz in ticket #994): sys/dev/pci/pcidevs: revision 1.710 Two re(4) candidates. From Francois Romieu in private mail.
1.701.2.15	21-Nov-2005	tron	Pull up following revision(s) (requested by cube in ticket #957): sys/dev/pci/pcidevs: revision 1.727 New Dell day: add missing device IDs, and order one correctly.
1.701.2.14	08-Sep-2005	tron	Pull up following revision(s) (requested by gavan in ticket #752): sys/dev/pci/pcidevs: revision 1.730 Add BCM5751M
1.701.2.13	06-Sep-2005	riz	Pull up following revision(s) (requested by tron in ticket #748): sys/dev/pci/pcidevs: revision 1.732 Add PCI ids for ATI Radeon 9600 LE.
1.701.2.12	24-Jul-2005	riz	Pull up revision 1.713 (requested by tron in ticket #617): Add the followings: - 6300ESB AC'97 Audio - 6300ESB AC'97 Modem - ICH6 HD-Audio - ICH6 AC'97 Modem - All of ICH7 devices
1.701.2.11	28-Jun-2005	tron	Pull up revision 1.721 (requested by briggs in ticket #485): Add Intel 6300ESB SATA RAID Controller.
1.701.2.10	22-Jun-2005	tron	Pull up revision 1.718 (requested by bouyer in ticket #478): Add SIS 965 bridge. From FreeBSD.
1.701.2.9	21-Jun-2005	tron	Pull up revision 1.719 (requested by bouyer in ticket #476): Add Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern@
1.701.2.8	18-Jun-2005	riz	Pull up revision 1.717 (requested by tron in ticket #450): Add support for the LinkSys PCMPC200 Cardbus ethernet card.
1.701.2.7	28-May-2005	snj	Pull up revision 1.711 (requested by riz in ticket #323): Add support for the 82801FB (ICH6) 10/100 ethernet appearing on new Intel motherboards. From Brian Buhrow in PR#30025. Approved by jmcneill.
1.701.2.6	10-May-2005	tron	Pull up revision 1.709 (requested by martin in ticket #272): PR kern/30173: add Corega CG-LAPCIGT
1.701.2.5	21-Apr-2005	tron	Pull up revision 1.706 (requested by bouyer in ticket #180): Add "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller". Fix kern/29218 by A L Meyers.
1.701.2.4	07-Apr-2005	jmc	Pullup rev 1.704 (requested by tron in ticket #59) Add ids for ATI Radeon 9700 Pro and 9800 Pro.
1.701.2.3	07-Apr-2005	jmc	Pullup rev 1.703 (requested by tron in ticket #16) Add support for Intel PRO/1000 GT.
1.701.2.2	07-Apr-2005	jmc	Pullup rev 1.702 (requested by tron in ticket #15) Add missing PCI id for Adaptec ASR-2410SA.
1.701.2.1	07-Apr-2005	tron	Pull up revision 1.705 (requested by martti in ticket #116): Added support for Dell CERC SATA RAID 1.5/6ch
1.701.2.38.2.1	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #1822): sys/dev/ic/mpt_netbsd.c: revision 1.12 via patch sys/dev/ic/mpt_netbsd.h: revision 1.6 via patch sys/dev/ic/mpt_mpilib.h: revision 1.3 via patch sys/dev/ic/mpt.h: revision 1.6 via patch sys/dev/ic/mpt.c: revisions 1.9, 1.10 via patch sys/dev/ic/mpt_debug.c: patch sys/dev/pci/mpt_pci.c: revision 1.11 via patch sys/dev/pci/pcidevs: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/pcidevs_data.h: patch Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.701.2.17.2.1	04-Oct-2007	bouyer	Pull up following revision(s) (requested by tron in ticket #1822): sys/dev/ic/mpt_netbsd.c: revision 1.12 via patch sys/dev/ic/mpt_netbsd.h: revision 1.6 via patch sys/dev/ic/mpt_mpilib.h: revision 1.3 via patch sys/dev/ic/mpt.h: revision 1.6 via patch sys/dev/ic/mpt.c: revisions 1.9, 1.10 via patch sys/dev/ic/mpt_debug.c: patch sys/dev/pci/mpt_pci.c: revision 1.11 via patch sys/dev/pci/pcidevs: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/pcidevs_data.h: patch Add support for newer SAS and similar devices to mpt(4). Tested with the LSI SAS1064 in a Sun x4200 server. These enhancements were developed by Garrett D'Amore and contributed to NetBSD by the TELES AG.
1.725.2.10	27-Feb-2008	yamt	sync with head.
1.725.2.9	11-Feb-2008	yamt	sync with head.
1.725.2.8	21-Jan-2008	yamt	sync with head
1.725.2.7	07-Dec-2007	yamt	sync with head
1.725.2.6	15-Nov-2007	yamt	sync with head.
1.725.2.5	27-Oct-2007	yamt	sync with head.
1.725.2.4	03-Sep-2007	yamt	sync with head.
1.725.2.3	26-Feb-2007	yamt	sync with head.
1.725.2.2	30-Dec-2006	yamt	sync with head.
1.725.2.1	21-Jun-2006	yamt	sync with head.
1.733.2.1	02-Nov-2005	yamt	sync with head.
1.734.2.2	29-Nov-2005	yamt	sync with head.
1.734.2.1	22-Nov-2005	yamt	sync with head.
1.751.2.4	01-Mar-2006	yamt	sync with head.
1.751.2.3	18-Feb-2006	yamt	sync with head.
1.751.2.2	01-Feb-2006	yamt	sync with head.
1.751.2.1	15-Jan-2006	yamt	sync with head.
1.758.2.1	09-Sep-2006	rpaulo	sync with head
1.759.2.3	07-Jun-2006	kardel	Sync with head.
1.759.2.2	01-Jun-2006	kardel	Sync with head.
1.759.2.1	22-Apr-2006	simonb	Sync with head.
1.769.2.8	14-Sep-2006	yamt	sync with head.
1.769.2.7	03-Sep-2006	yamt	sync with head.
1.769.2.6	11-Aug-2006	yamt	sync with head
1.769.2.5	26-Jun-2006	yamt	sync with head.
1.769.2.4	24-May-2006	yamt	sync with head.
1.769.2.3	11-Apr-2006	yamt	sync with head
1.769.2.2	01-Apr-2006	yamt	sync with head.
1.769.2.1	13-Mar-2006	yamt	sync with head.
1.771.2.2	11-May-2006	elad	sync with head
1.771.2.1	19-Apr-2006	elad	sync with head.
1.775.2.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.775.2.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.787.2.3	22-Jun-2006	chap	Complete a sync sys/ with head.
1.787.2.2	19-Jun-2006	chap	Sync with head.
1.787.2.1	19-May-2006	chap	file pcidevs was added on branch chap-midi on 2006-06-19 04:01:36 +0000
1.794.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.815.2.8	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/pcidevs: revision 1.831 Add the SK-9Sxx Gigabit Ethernet, from OpenBSD.
1.815.2.7	15-Sep-2006	tron	Pull up following revision(s) (requested by riz in ticket #153): sys/dev/pci/pcidevs: revision 1.822 Add a bunch of Marvell product IDs, from OpenBSD.
1.815.2.6	08-Sep-2006	rpaulo	Pull up following revision(s) (requested by riz in ticket #139): sys/arch/cobalt/pci/pchb.c: revision 1.9 sys/dev/pci/if_sk.c: revision 1.30 sys/arch/cobalt/pci/pci_machdep.c: revision 1.24 sys/dev/pci/pcidevs: revision 1.820 sys/arch/evbmips/malta/pci/pchb.c: revision 1.9 Rename vendor GALILEO -> vendor MARVELL . Marvell bought Galileo years ago, we have little existing stuff in the tree, and this will make it easier to keep certain network drivers in sync with other OSes. OK matt@, uwe@ Vendor GALILEO is now MARVELL.
1.815.2.5	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #115): share/man/man4/pdcsata.4: revision 1.5 sys/dev/pci/pcidevs: revision 1.827 sys/dev/pci/pcidevs: revision 1.828 sys/dev/pci/pcidevs: revision 1.829 sys/dev/pci/pdcsata.c: revision 1.11 Add some ICH8 devices... from obsd. Add some missing devices from Promise and sort them. Add missing PDC20265 device... removed accidentally in previous. Add support for PDC40519 and PDC40779... from obsd. Mention PDC40519 and PDC40779... and bump date.
1.815.2.4	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #113): sys/dev/pci/viaide.c: revision 1.33 share/man/man4/viaide.4: revision 1.7 sys/dev/pci/pcidevs: revision 1.825 sys/dev/pci/pcidevs: revision 1.826 - Add all pciids for nvidia MCP61/65 controllers, from obsd. - Improve description in all them. Erm forgot to add MCP61/65 SATA Controllers and sort them. Add support for nvidia MCP61/65 IDE/SATA Controllers. Mention MCP61/65... and bump date.
1.815.2.3	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #112): sys/dev/pci/if_nfe.c: revision 1.4 sys/dev/pci/pcidevs: revision 1.824 - Add nForce MCP61/65 LAN controllers... from obsd. - Improve description in some nForce controllers as well. Add support for MCP61/65 controllers... from obsd.
1.815.2.2	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #28): sys/dev/pci/pcidevs: revision 1.819 sys/dev/pci/pcidevs.h: revision 1.819 sys/dev/pci/pcidevs_data.h: revision 1.818 (via patch) sys/dev/pci/pdcsata.c: revision 1.10 Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.815.2.1	11-Aug-2006	riz	Pull up following revision(s) (requested by cube in ticket #8): sys/dev/pci/piixide.c: revision 1.28 sys/dev/pci/pcidevs: revision 1.816 Add device IDs for Intel's 631xESB/632xESB southbridges, and start the list for the 5000X northbridge. I'll add the rest later, the spec is currently unavailable so I'll collect them from another source. Add support for the 631xESB/632xESB.
1.830.2.3	01-Feb-2007	ad	Sync with head.
1.830.2.2	12-Jan-2007	ad	Sync with head.
1.830.2.1	18-Nov-2006	ad	Sync with head.
1.831.2.3	18-Dec-2006	yamt	sync with head.
1.831.2.2	10-Dec-2006	yamt	sync with head.
1.831.2.1	22-Oct-2006	yamt	sync with head
1.855.2.25	31-Oct-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #1215): sys/dev/pci/if_wm.c: revision 1.162 sys/dev/pci/pcidevs: revision 1.961,1.962 Add support for more ethenet ICH9 devices. Tested by Pouya D. Tafti on a 82801I_IGP_M_AMT.' sys/dev/pci/if_wm.c sys/dev/pci/pcidevs
1.855.2.24	29-Sep-2008	jdc	Pull up revision 1.902 (requested by bouyer in ticket #1192). addd SYMBIOS SAS1068_2, SAS1064E_2 and SAS1068E_2
1.855.2.23	17-May-2008	bouyer	Pull up following revision(s) (requested by jdc in ticket #1145): src/share/man/man4/gem.4 1.5,1.6 via patch src/sys/arch/sparc64/conf/GENERIC 1.97 via patch src/sys/dev/ic/gem.c 1.59,1.61-1.68,1.70-1.73 via patch src/sys/dev/ic/gemreg.h 1.11 via patch src/sys/dev/ic/gemvar.h 1.16,1.18 via patch src/sys/dev/mii/bmtphy.c 1.24 via patch src/sys/dev/pci/if_gem_pci.c 1.25,1.27 via patch src/sys/dev/pci/pcidevs 1.12 via patch src/sys/dev/pci/pcidevs.h regenerate src/sys/dev/pci/pcidevs_data.h regenerate Add support for Sun PCI X1141A (SX fibre) cards. Add improvements to the driver. Disable HW UDP4 and RX TCP4 checksums because they are buggy (PR#34799) Add improvements to bmtphy.
1.855.2.22	17-May-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1143): sys/dev/pci/aac_pci.c: revision 1.25 via patch share/man/man4/aac.4: revision 1.13 via patch sys/dev/pci/pcidevs: revision 1.939 via patch Add support for IBM ServeRAID 8k (and 8k-l)
1.855.2.21	19-Apr-2008	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1128): sys/dev/pci/if_lii.c: revisions 1.1, 1.3 via patch sys/arch/i386/conf/ALL: revision 1.161 via patch share/man/man4/lii.4: revision 1.1 via patch sys/dev/pci/files.pci: revision 1.303 via patch sys/dev/DEVNAMES: revision 1.238 via patch sys/arch/i386/conf/GENERIC: revision 1.885 via patch distrib/sets/lists/man/mi: revisions 1.1062, 1.1063 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.43 via patch share/man/man4/Makefile: revision 1.460 via patch sys/arch/i386/conf/INSTALL: revision 1.323 via patch sys/dev/pci/if_liireg.h: revision 1.1 via patch sys/dev/pci/pcidevs: patch Add a driver for the Atheros/Attansic L2 Fast-Ethernet chip found on a series of hardware that includes Asus's famous EeePC.
1.855.2.20	09-Mar-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1097): sys/dev/pci/pucdata.c: revision 1.53 sys/dev/pci/pcidevs: revision 1.930 Add I-O DATA RSA-PCI 2 port serial board. Add an entry for I-O DATA RSA-PCI 2-port serial board. (works only on 16550A compatible mode)
1.855.2.19	23-Feb-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1061): sys/dev/pci/if_nfe.c: revision 1.22, 1.23, 1.25, 1.27 sys/dev/pci/pcidevs: revision 1.914 sys/dev/pci/if_nfereg.h: revision 1.5, 1.6 share/man/man4/nfe.4: revision 1.4 sys/dev/pci/if_nfevar.h: revision 1.5, 1.6 Add NVIDIA MCP7[37] devices, from OpenBSD. Attach to NVIDIA MCP67/73 Ethernet controllers. Add the MCP{67,73} products on nfe_attach(). Sync with reality. Handle a quirk of newer MCP6x chips on reading MAC address. Fix from Richy Kim in PR kern/36576. Wakeup some newer chips from powerdown mode. From FreeBSD/Linux driver via OpenBSD. Tested by wiz@ in PR kern/37868.
1.855.2.18	22-Feb-2008	bouyer	Pull up following revision(s) (requested by tron in ticket #1074): sys/dev/pci/pcidevs: revision 1.929 Import PCI ids for the Matrox MGA G200e and the Intel E7230 and 82975X chipsets from OpenBSD.
1.855.2.17	11-Jan-2008	xtraeme	Pull up following revision(s) (requested by christos in ticket #1037): sys/dev/pci/arcmsr.c: patch sys/dev/pci/arcmsrvar.h: patch sys/dev/pci/pcidevs: patch sys/dev/pci/files.pci: patch sys/arch/i386/conf/GENERIC: patch sys/arch/i386/conf/XEN2_DOM0: patch sys/arch/i386/conf/INSTALL_LARGE: patch sys/arch/amd64/conf/GENERIC: patch sys/arch/amd64/conf/INSTALL: patch share/man/man4/Makefile: patch share/man/man4/arcmsr.4: patch distrib/sets/lists/man/mi: patch Add the Areca Technology Corporation SATA RAID controller driver, ported from OpenBSD.
1.855.2.16	15-Nov-2007	pavel	Backout accidental commit to the release branch.
1.855.2.15	15-Nov-2007	manu	It's not 88CO55, it's 88C055
1.855.2.14	04-Nov-2007	xtraeme	Pull up following revision(s) (requested by manu in ticket #954): sys/dev/pci/pcidevs_data.h: patch sys/dev/pci/pcidevs.h: patch sys/dev/pci/if_msk.c: revision 1.11 sys/dev/pci/pcidevs: revision 1.907 Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.855.2.13	24-Oct-2007	pavel	Pull up following revisions (requested by msaitoh in ticket #935): sys/dev/pci/piixide.c: revision 1.38-1.41 via patch sys/dev/pci/pcidevs: revision 1.900 via patch Attach to the ICH9 SATA devices. Attach to the ICH8M IDE Controller. Proper ICH8M IDE Controller support (missed in previous). Another ICH8M SATA Controller found in jmcneill's laptop.
1.855.2.12	22-Oct-2007	pavel	Pull up following revisions (requested by xtraeme in ticket #931): sys/dev/pci/pcidevs: revision 1.905 sys/dev/pci/if_ste.c: revision 1.30 IC Plus Corp IP100A support for ste(4), submitted by Pedro Alejandro Lopez-Valencia in PR kern/36605.
1.855.2.11	07-Oct-2007	xtraeme	Pull up following revision(s) (requested by christos in ticket #923): sys/dev/pci/aac_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.872 PR/35970: Todd Kover: Add support for HP ML110 G2 / Adaptec 2610SA
1.855.2.10	31-Aug-2007	pavel	Pull up following revision(s) (requested by bouyer in ticket #678): distrib/sets/lists/man/mi: revision 1.1001 doc/CHANGES: revision 1.853 via patch share/man/man4/Makefile: revision 1.428 share/man/man4/jmide.4: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.143 sys/arch/amd64/conf/INSTALL: revision 1.69 sys/arch/i386/conf/GENERIC: revision 1.827 sys/arch/i386/conf/INSTALL_LARGE: revision 1.2 via patch sys/arch/i386/conf/XEN2_DOM0: revision 1.26 sys/conf/files: revision 1.844 via patch sys/dev/ic/ahcisata_core.c: revision 1.1 via patch sys/dev/ic/ahcisata_core.c: revision 1.3 sys/dev/ic/ahcisatareg.h: revision 1.1 sys/dev/ic/ahcisatavar.h: revision 1.1 sys/dev/pci/ahcisata.c: delete sys/dev/pci/ahcisatareg.h: delete sys/dev/pci/ahcisata_pci.c: revision 1.1 sys/dev/pci/files.pci: revision 1.287-1.288 sys/dev/pci/jmide.c: revision 1.1-1.2 sys/dev/pci/jmide_reg.h: revision 1.1 sys/dev/pci/pcidevs: revision 1.878 Add JMicron Technology vendor ID, and their current PCIe SATA/PATA controllers. Split the ahcisata driver in pci front-end and bus-independant back-end. add jmide(4), a driver for the JMicron Technology JMB36x PCIe to SATA II/PATA controllers. These controllers can be found on add-on PCIe cards, or on some motherboards to provide the PATA connectivity (e.g. some intel ICH8-based motherboards). Thanks to JMicron Technology for providing me documentation and different sample boards for this work. Move mapping of AHCI register so that jmide knows if it fails, and avoids calling ahci_intr() (which would cause a panic). Try to use the pciide function for SATA drives if attaching ahci fails (this doesn't seems to work though, it may be BIOS dependant). Thanks to Gary Duzan for testing multiple pacthes. We're not ready to handle ATAPI yet so just claim there's no drive to the upper layer. This should work around a NULL pointer dereference when an ATAPI device is detected on a AHCI device.
1.855.2.9	29-Aug-2007	liamjfoy	Pull up following (requested by xtraeme ticket #846): Import i386/gcscide(4). A driver for the IDE Controller of the AMD CS5535 Companion device found in the decTOP. gcscide0 at pci0 dev 15 function 2 gcscide0: National Semiconductor/AMD CS5535 IDE Controller (rev. 0x00) Supports Ultra DMA mode 4, Pio Mode 4 and MDMA mode 2. "Go for it" jmcneill@.
1.855.2.8	29-Aug-2007	liamjfoy	Pull up following revision(s) (requested by masanobu in ticket #850): sys/dev/pci/if_wm.c 1.137 sys/dev/pci/if_wm.c 1.138 sys/dev/pci/if_wm.c 1.141 sys/dev/pci/if_wm.c 1.144 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphy.c 1.12 sys/dev/mii/miidevs 1.72 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/wm.4 1.18 Fix for 82541 and 82547's reset bug. Modify wm_reset() to make some device stable. Add ICH9 support to wm. Add I82566 support to igphy. Pulled done via patch.
1.855.2.7	26-Aug-2007	liamjfoy	Pull up following revision(s) (requested by riz in ticket #816): sys/dev/pci/if_skreg.h: revision 1.10 sys/dev/pci/if_skreg.h: revision 1.11 sys/dev/pci/pcidevs: revision 1.866 sys/dev/pci/if_skreg.h: revision 1.9 sys/dev/pci/if_sk.c: revision 1.36 sys/dev/pci/if_msk.c: revision 1.4 sys/dev/pci/if_msk.c: revision 1.5 sys/dev/pci/if_msk.c: revision 1.6 sys/dev/pci/if_msk.c: revision 1.7 sys/dev/pci/if_mskvar.h: revision 1.2 sys/dev/pci/if_mskvar.h: revision 1.3 Apply OpenBSD's rev. 1.33 to shut up "msk0: phy write timed out" Original commit message: > In msk_marv_miibus_writereg, wait for busy flag to clear instead of > continuing when busy flag set. add some msk devices sync with OpenBSD-current (many bugfixes, add some devices) if_msk.c: rev. 1.42 if_mskvar.h rev. 1.3 if_skreg.h rev. 1.41 sync with if_skreg.h rev. 1.41 need more work? apply some patches from FreeBSD o fix device timeout o add some workaround o TX underrun bug (grr...) o and some bugs fix bit definitions for the RX FIFO Flush mode Apply OpenBSD's rev. 1.14 via patch: > remove another Yukon Lite workaround.
1.855.2.6	10-Jun-2007	bouyer	Pull up following revision(s) (requested by hira in ticket #713): sys/dev/pci/pcidevs: revision 1.863 sys/dev/pci/viaide.c: revision 1.38 share/man/man4/viaide.4: revision 1.8 Add PCI Ids for AMD Geode LX and CS5536. Add support for CS5536 IDE.
1.855.2.5	02-May-2007	snj	branches: 1.855.2.5.2; Pull up following revision(s) (requested by bouyer in ticket #597): sys/dev/pci/pcidevs: revision 1.876 sys/dev/pci/pcidevs: revision 1.877 Add: Intel i82801H IFE (GT) LAN Controller Intel i82801H IFE (G) LAN Controller from FreeBSD Remove extra "Intel" from previous commit
1.855.2.4	21-Feb-2007	tron	Pull up following revision(s) (requested by markd in ticket #446): sys/dev/pci/pcidevs: revision 1.868 Add i965Q
1.855.2.3	17-Feb-2007	tron	Pull up following revision(s) (requested by mlelstv in ticket #449): sys/dev/pci/pcidevs: revision 1.867 sys/dev/pci/viaide.c: revision 1.40 Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.855.2.2	17-Feb-2007	tron	Pull up following revision(s) (requested by chs in ticket #423): sys/dev/pci/pcidevs: revision 1.865 add another Realtek wifi, some SIIG serial cards and another nForce chipset. remove some duplicate entries.
1.855.2.1	12-Dec-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #267): sys/dev/pci/pcidevs: revision 1.857 Add another PCI ID for Dell PERC 5 (from OpenBSD)
1.855.2.5.2.4	03-Jun-2008	skrll	Sync with netbsd-4.
1.855.2.5.2.3	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.855.2.5.2.2	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.855.2.5.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.868.2.5	17-May-2007	yamt	sync with head.
1.868.2.4	07-May-2007	yamt	sync with head.
1.868.2.3	15-Apr-2007	yamt	sync with head.
1.868.2.2	24-Mar-2007	yamt	sync with head.
1.868.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.871.2.9	23-Oct-2007	ad	Sync with head.
1.871.2.8	12-Oct-2007	ad	Sync with head.
1.871.2.7	09-Oct-2007	ad	Sync with head.
1.871.2.6	20-Aug-2007	ad	Sync with HEAD.
1.871.2.5	15-Jul-2007	ad	Sync with head.
1.871.2.4	09-Jun-2007	ad	Sync with head.
1.871.2.3	27-May-2007	ad	Sync with head.
1.871.2.2	10-Apr-2007	ad	Sync with head.
1.871.2.1	13-Mar-2007	ad	Sync with head.
1.872.4.2	29-Mar-2007	reinoud	Pullup to -current
1.872.4.1	18-Mar-2007	reinoud	First attempt to bring branch in sync with HEAD
1.872.2.1	11-Jul-2007	mjf	Sync with head.
1.885.6.13	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.885.6.12	21-Nov-2007	joerg	Sync with HEAD.
1.885.6.11	14-Nov-2007	joerg	Sync with HEAD.
1.885.6.10	31-Oct-2007	joerg	Sync with HEAD.
1.885.6.9	28-Oct-2007	joerg	Sync with HEAD.
1.885.6.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.885.6.7	02-Oct-2007	joerg	Sync with HEAD.
1.885.6.6	06-Sep-2007	jmcneill	Add Intel 82G33 definitions.
1.885.6.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.885.6.4	27-Aug-2007	jmcneill	Add some more Intel 965 PCI IDs
1.885.6.3	16-Aug-2007	jmcneill	Sync with HEAD.
1.885.6.2	09-Aug-2007	jmcneill	Sync with HEAD.
1.885.6.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.885.2.3	10-Sep-2007	skrll	Sync with HEAD.
1.885.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.885.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.887.2.2	06-Aug-2007	martin	Add a serverworks south bridge
1.887.2.1	06-Aug-2007	martin	file pcidevs was added on branch matt-mips64 on 2007-08-06 22:39:32 +0000
1.892.2.3	23-Mar-2008	matt	sync with HEAD
1.892.2.2	09-Jan-2008	matt	sync with HEAD
1.892.2.1	06-Nov-2007	matt	sync with HEAD
1.903.2.1	14-Oct-2007	yamt	sync with head.
1.905.2.4	21-Nov-2007	bouyer	Sync with HEAD
1.905.2.3	18-Nov-2007	bouyer	Sync with HEAD
1.905.2.2	13-Nov-2007	bouyer	Sync with HEAD
1.905.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.912.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.912.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.912.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.912.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.920.6.3	11-Jan-2008	bouyer	Sync with HEAD
1.920.6.2	08-Jan-2008	bouyer	Sync with HEAD
1.920.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.920.2.1	26-Dec-2007	ad	Sync with head.
1.929.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.929.6.4	05-Oct-2008	mjf	Sync with HEAD.
1.929.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.929.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.929.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.929.2.1	24-Mar-2008	keiichi	sync with head.
1.937.4.9	09-Oct-2010	yamt	sync with head
1.937.4.8	11-Aug-2010	yamt	sync with head.
1.937.4.7	11-Mar-2010	yamt	sync with head
1.937.4.6	19-Aug-2009	yamt	sync with head.
1.937.4.5	18-Jul-2009	yamt	sync with head.
1.937.4.4	20-Jun-2009	yamt	sync with head
1.937.4.3	16-May-2009	yamt	sync with head
1.937.4.2	04-May-2009	yamt	sync with head.
1.937.4.1	16-May-2008	yamt	sync with head.
1.937.2.2	04-Jun-2008	yamt	sync with head
1.937.2.1	18-May-2008	yamt	sync with head.
1.938.2.4	10-Oct-2008	skrll	Sync with HEAD.
1.938.2.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.938.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.938.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.942.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.942.4.1	19-Oct-2008	haad	Sync with HEAD.
1.942.2.2	31-Jul-2008	simonb	Sync with head.
1.942.2.1	18-Jul-2008	simonb	Sync with head.
1.962.4.21	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.962.4.20	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1962): sys/dev/pci/if_wm.c: revisions 1.259-1.266, 1.269, 1.271, 1.273-1.274, 1.277-1.278, 1.282, 1.284-1.285, 1.287, 1.290, 1.294, 1.297-1.298, 1.300-1.301, 1.304-1.307, 1.310, 1.312-1.314, 1.316 via patch sys/dev/pci/if_wmreg.c: revisions 1.54-1.57, 1.59-1.60, 1.62, 1.64-1.66 via patch sys/dev/pci/if_wmvar.c: revisions 1.17, 1.20-1.21 via patch sys/dev/pci/pcidevs: revisions 1.1172, 1.1195, 1.1201 via patch - Fix "MDIC write error" bug for 82574 and 82583. For those chips, the semaphore must be released after chip reset. Found and tested by Mark Davies. - Fix BMC related bugs. - Fix yet another NVM bank detect problem in wm(4). Use bank 0 if the detect function failed. It's the same as FreeBSD. Observed and tested with Asus P8P67 Deluxe motherboard and tested by jnemeth. - Add support for I354 DH89xxCC and some new I218 devices. - Fix definition of CTRL_GIO_M_DIS bit. - Insert completion barrier between register write and delay(). - Bump max TX DMA size to avoid pathological condition with TSO. From dyoung. - Fix semaphore related bugs. - Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too. Same as FreeBSD and OpenBSD. - Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD. - Fix fiber link problem (PR#44776 and PR#30880). - WM_T_82545 is not 1000base-SX but 1000base-LX. Same as FreeBSD. - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - It's not required to print "failed to detect NVM bank" with aprint_error_dev(). Use DPRINTF(). Same as {Free,Open}BSD. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Initialize some hardware bits for 8257[1234], 82583, 80003, ICH* and PCH. Some of them are workaround code. From other BSDs, Linux and documents. - Fix a bug that wm_sgmii_writereg() function doesn't pass the "val" argument to the I2CCMD register. Reported by Bernard Merindol in PR#49789. - Delete 82580ER related code. It was from FreeBSD and was removed in r203049. - Remove extra debug message. - Remove unused variable. - Remove a duplicated error message. - Cleanup comments. - Fix debug message.
1.962.4.19	07-Sep-2013	bouyer	sys/dev/pci/if_wm.c 1.246-1.247, 1.249-1.258 sys/dev/pci/if_wmreg.h 1.51, 1.53 sys/dev/pci/if_wmvar.h 1.15-1.16 sys/dev/pci/pcidevs 1.1152-1.1153 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add I21[0178] support. Fix a bug that wm_attach() may fail on some PCH2 or newer system. wm_valid_nvm_bank_detect_ich8lan() misunderstood the NVM's bank number. Fixes PR#47878. Fix a bug that the check of reset complete fails on Intel 8 series with "wm_lan_init_done: lan_init_done failed to complete" message. The broken code was used for ICH8, 9... and PCH2. The wm_linkintr_gmii() function is called from interrupt. That's not tick, so call mii_pollstat() instead of mii_tick(). Add ECC support for the packet buffer. Only 82571 and I21[78] support ECC. Fix a bug that wrong semaphore is used in wm_gmii_hv_{read,write}reg. Add comments, fix some comments, use macros and remove trailing whitespaces. [msaitoh, ticket #1867]
1.962.4.18	19-Jun-2013	bouyer	Pullup the following revisions via patch, requested by msaitoh in ticket #1850: sys/dev/pci/if_wm.c 1.201, 1.203-1.204, 1.207-1.212, 1.215, 1.217-1.218, 1.220-1.223, 1.228, 1.232-245 sys/dev/pci/if_wmreg.h 1.40-1.45, 1.47-1.48 sys/dev/pci/if_wmvar.h 1.11-1.13 sys/dev/pci/pcidevs 1.1074, 1.1077, 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.6 sys/dev/mii/ihphy.c 1.1-1.2 sys/dev/mii/ihphyreg.h 1.1 sys/dev/mii/inbmphyreg.h 1.3 sys/dev/mii/files.mii 1.47 via patch sys/dev/mii/miidevs 1.97 and 1.100 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/arch/i386/conf/ALL 1.280 sys/arch/i386/conf/GENERIC 1.1001 sys/arch/i386/conf/INSTALL_FLOPPY 1.11 sys/arch/i386/conf/XEN2_DOM0 patch sys/arch/amd64/conf/GENERIC 1.293 sys/arch/amd64/conf/XEN3_DOM0 1.61 share/man/man4/wm.4 1.21-1.24 Apply almost all fixes and improvements from netbsd-6 except for the rev. 1.196's iqdrops' change. - Add the detach code. - Add code for WOL, ASF, IPMI and Intel AMT. WOL is disabled by default - Add Yet another workaround for ICH8. - 82576 is dual port, so check the FUNCID and increment the MAC address for the 2nd port. - Fix the names of 82577L[MC] LAN controllers (for mobile). - Fix CTRL_EXT_SWDPIN() and CTRL_EXT_SWDPIO() macros. The bit order of the SW definable pin is not 6543 but 3654!!! - Rewrite the code to read MAC address from eeprom. - Add 82580 support. - 82571 quirk. Only 82571 shares port 0 of EEMNGCTL_CFGDONE. - The document says that the TDH register must be set after TCL.EN is set on 82575 and newer devices. - Fix some register names. No functional change. - Omit U+00AE "REGISTERED SIGN" in a product name due to its non-ASCII nature. - Stop wm(4) from needlessly resetting when you add or delete a vlan(4). - Fix MAC address check on 8257[156] and 80003 case. Some cards have non 0xffff pointer but those don't use alternative MAC address in reality. So we check whether the broadcast bit is set or not like Intel's e1000 driver. Fixes PR kern/44072 reported by Jean-Yves Moulin. - Add PCH2(and 82579) support. Fixes PR#46487 - Add yet another 82567V support. - Add ICH10+HANKSVILL support. - Add support Intel I350 Ethernet. - Make vlan and all ip/ip6 checksum offload work for the I350. - Fix compile error with WM_DEBUG. - Fix a bug that PHY isn't set to low-power mode on PCH and PCH2. - Add WM_DEBUG_NVM. If WM_DEBUG_NVM is enabled, dump the FLASH ROM data. - Skip 64bit BAR correctly. - Fix RAL_TABSIZE for ICH8, 82576, 82580 and I350. - Use 82580(and I350) specific PHY read/write functions. Fixes PR#47542. - Style fix. Fix typo in comment. Fix comments. Add comments.
1.962.4.17	20-Jan-2013	bouyer	Apply patch, requested by msaitoh in ticket #1842: sys/dev/pci/pcidevs 1.969 via patch sys/dev/pci/viaide.c 1.58 Add VT8237S Integrated SATA Controller support. PR#47452.
1.962.4.16	17-Mar-2012	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1739): sys/dev/pci/piixide.c: revision 1.58 sys/dev/pci/pcidevs: patch Add support for Intel 6 Series Serial ATA Controller
1.962.4.15	22-Nov-2010	riz	branches: 1.962.4.15.2; Pull up following revision(s) (requested by macallan in ticket #1470): sys/dev/pci/pcidevs: revision 1.1019 sys/dev/pci/fwohci_pci.c: revision 1.35 do not attach to Powrbook G3 onboard Firewire controllers until =20 someone with access to the hardware can figure out why it deadlocks during device discovery add Apple PowerBook G3 Firewire controllers
1.962.4.14	20-Nov-2010	riz	Pull up following revision(s) (requested by jakllsch in ticket #1362): sys/dev/pci/pcidevs: revision 1.1026 sys/dev/pci/agp_intel.c: revision 1.35 sys/dev/pci/agp_i810.c: revision 1.67 Add Intel 82855GM Host-AGP Bridge. regen Add (non-IGP) AGP port support for Intel 82855GM. agp_intel tested with radeondrm(4). agp_i810 tested with i915drm(4).
1.962.4.13	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1359): sys/dev/pci/if_bge.c 1.170-1.181 sys/dev/pci/if_bgereg.h 1.53-1.56 sys/dev/pci/if_bgevar.h 1.2-1.5 sys/dev/pci/pcidevs 1.1017, 1.1021 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/miidevs 1.89, 1.93 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen sys/dev/mii/brgphy.c 1.49, 1.52 - Fix a bug of BGE_RDMA_MODE_FIFO_LONG_BURST definition. - Fix the bug that both BGE_PCIE and BGE_PCIX are set on PCI-Express devices. - Add support for BCM5705F,BCM5714,SBCM5715S,BCM5717,BCM5718,BCM5720,BCM5723, BCM5724,BCM5751F,BCM5753F,BCM5756,BCM5761,BCM5761E,BCM5761S,BCM5761SE, BCM5764,BCM5781,BCM5784M,BCM5903M,BCM57760,BCM57761,BCM57765,BCM57780, BCM57781,BCM57785,BCM57788,BCM57790,BCM57791 and BCM57795. Fixes PR#41694. - Fix misunderstanding the capability for TSO4 on some chips. - Use proplib for no EEPROM systems (e.g. sparc64 with onboard bge). - Add debug function to show some flags. - (brgphy(4)) Add support for BCM5461, BCM5464, BCM5784 and BCM5761. - (brgphy(4)) Enable brgphy_jumbo_settings(). - (brgphy(4)) Enable the Ethernet@Wirespeed function. - Add support for device properties to override the mac address and set the BGE_NO_EEPROM flag if we have an onboard device w/o SEEPROM. - Introduce IPMI and ASF related code from FreeBSD. It fixes some problems which occured in netboot on sparc64 and PR#32767 - move the code of disabling host interrput in bge_stop() like linux tg3 driver. - fix the return value of bge_eeprom_getbyte(). - Fixes the bug that error bits arent cleard because these bit are W2C (in other words W1C). - Many other bugfixes from FreeBSD and OpenBSD. - remove an unused structure. - ANSI C, KNF and tabify - fix typo in comment.
1.962.4.12	19-Nov-2010	riz	Pull up revisions (requested by msaitoh in ticket #1358): sys/dev/pci/if_wm.c 1.196-1.199,1.202,1.205 sys/dev/pci/if_wmvar.h 1.9 sys/dev/pci/if_wmreg.h 1.36-1.39 sys/dev/pci/pcireg.h 1.61-1.64 sys/dev/pci/pcidevs 1.1023 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen mii/igphy.c 1.21 mii/igphyvar.h 1.1 mii/inbmphyreg.h 1.2 - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems. - Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working. - (pcireg.h) Add PCIe config register definitions. - Note that the changes to count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops in rev. 1.196 of if_wm.c is not pulled up.
1.962.4.11	28-Mar-2010	snj	Apply patch (requested by msaitoh in ticket #1326): Add support for MFI gen2 devices. Mention newer devices, RAID 6 and RAID60 in mfi.4.
1.962.4.10	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.962.4.9	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1268): sys/dev/mii/miidevs: revision 1.90 sys/dev/pci/pcidevs: revision 1.1001 sys/dev/mii/brgphy.c: revision 1.51 sys/dev/microcode/bnx/bnxfw.h: revision 1.3 sys/dev/pci/if_bnx.c: revision 1.29 sys/dev/pci/if_bnxreg.h: revision 1.9 sys/dev/pci/if_bnx.c: revision 1.30 Add support for the Broadcom BCM5709 and BCM5716 chips. Upgrade the B06 firmware and add the B09 firmware required by the BCM5709. Replace TX dmamap array with a dynamically-grown list. Support associated BCM5709CAX and BCM5709C PHYs.
1.962.4.8	23-Jan-2010	bouyer	Pull up following revision(s) (requested by mbalmer in ticket #1254): sys/dev/pci/pcidevs: revision 1.1015 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen Add two IBM RSS devices.
1.962.4.7	18-Dec-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #1175): sys/dev/pci/pcidevs: revision 1.1003 Add Intel 3400 devices
1.962.4.6	28-Oct-2009	bouyer	Pull up the following revisions, requested by sborrill in ticket #1114: share/man/man4/sdhc.4 1.1-1.2 sys/dev/pci/sdhc_pci.c 1.1-1.3 distrib/sets/lists/man/mi patch share/man/man4/Makefile patch sys/arch/amd64/conf/GENERIC patch sys/arch/i386/conf/ALL patch sys/arch/i386/conf/GENERIC patch sys/dev/pci/files.pci patch sys/dev/pci/pcidevs patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add sdhc(4), a driver for SD controllers following the SD Host Controller Standard Simplified Specification.
1.962.4.5	26-Sep-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #1021): sys/dev/pci/pcidevs: revision 1.987 Add alternative product ID for VIA VT8237A Integrated SATA Controller
1.962.4.4	21-May-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #711): sys/dev/pci/pcidevs: revisions 1.975, 1.981, 1.982 via patch sys/dev/pci/if_wm.c: revisions 1.164, 1.167, 1.173, 1.174 via patch sys/dev/pci/if_wmreg.h: revisions 1.25, 1.27 via patch Add Intel 82567LM_3 ethernet -- Add i82567LM-3 -- add i82567LF-3 LAN Controller -- add an entry for 82567LF-3. fix the register access for ICH10DO. -- Fix about TBI mode. This fix doesn't influence MII mode. -- - Fix panic in mediachange. - Fix SWDPIN(1)'s polarity on some chips. - Fix flow control stuff (includes PR#32009). - Stop RXCFG storm. It ocours easily. - And more fix about autonego. -- add 82801J_D_BM_LF (ICH10) -- Reload sc_ctrl in wm_reset(). Add an ICH10 entry. Remove some obsolete comments.
1.962.4.3	05-May-2009	bouyer	Pull up following revision(s) (requested by snj in ticket #737): sys/arch/x86/pci/pchb.c: revisions 1.15 - 1.17 sys/dev/pci/agp.c: revisions 1.63 - 1.65 sys/dev/pci/agp_i810.c: revisions 1.57 - 1.64 sys/dev/pci/pcidevs: revisions 1.965, 1.967 via patch sys/dev/pci/agpreg.h: revision 1.20 Add AGP support for Intel G35, G45, and Q45.
1.962.4.2	03-May-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #722): sys/dev/pci/pcidevs: revision 1.973 Attansic L1E Gigabit Ethernet Adapter
1.962.4.1	20-Mar-2009	msaitoh	branches: 1.962.4.1.4; Pull up following revision(s) (requested by sborrill in ticket #587): sys/dev/pci/pcidevs: revision 1.978 sys/dev/pci/if_wm.c: revision 1.165 Add support for 82574L (as found on Desktop CT adaptor) based on 82573 support.
1.962.4.15.2.2	14-Oct-2011	matt	Add ALI M5288
1.962.4.15.2.1	07-Jan-2011	matt	Add some devices (not synced to HEAD).
1.962.4.1.4.8	05-Nov-2013	matt	Pull down xhci support from HEAD
1.962.4.1.4.7	03-Jan-2012	matt	Change XLP_RAID to XLP_DMA.
1.962.4.1.4.6	31-Dec-2011	matt	Add XLP RXE and SRIO product ids.
1.962.4.1.4.5	24-Dec-2011	matt	Add NetLogic(RMI) XLP devices
1.962.4.1.4.4	10-May-2010	matt	Add Marvell 88SB2211
1.962.4.1.4.3	20-Apr-2010	matt	sync with netbsd-5
1.962.4.1.4.2	07-Apr-2010	cliff	add RMI XLR PCI-X bridge
1.962.4.1.4.1	09-Nov-2009	cliff	in pcidevs: - add vendor entry for RMI - add product entry for RMI XLS integrated PCIe Interface regenerate (make) pcidevs.h and pcidevs_data.h
1.962.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.962.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.962.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.976.2.2	23-Jul-2009	jym	Sync with HEAD.
1.976.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1019.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.1019.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.1019.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1025.2.7	12-Jun-2011	rmind	sync with head
1.1025.2.6	31-May-2011	rmind	sync with head
1.1025.2.5	21-Apr-2011	rmind	sync with head
1.1025.2.4	06-Mar-2011	rmind	sync with head (and fix few botches with this)
1.1025.2.3	05-Mar-2011	rmind	sync with head
1.1025.2.2	03-Jul-2010	rmind	sync with head
1.1025.2.1	30-May-2010	rmind	sync with head
1.1055.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.1055.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.1055.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1083.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.1101.6.7	02-Jun-2012	mrg	sync to latest -current.
1.1101.6.6	29-Apr-2012	mrg	sync to latest -current.
1.1101.6.5	05-Apr-2012	mrg	sync to latest -current.
1.1101.6.4	06-Mar-2012	mrg	sync to -current
1.1101.6.3	06-Mar-2012	mrg	sync to -current
1.1101.6.2	04-Mar-2012	mrg	sync to latest -current.
1.1101.6.1	18-Feb-2012	mrg	merge to -current.
1.1101.2.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1101.2.4	16-Jan-2013	yamt	sync with (a bit old) head
1.1101.2.3	30-Oct-2012	yamt	sync with head
1.1101.2.2	23-May-2012	yamt	sync with head.
1.1101.2.1	17-Apr-2012	yamt	sync with head
1.1102.2.18	30-Apr-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1296): sys/dev/pci/pcidevs: revision 1.1195 via patch sys/dev/pci/if_wm.c: revisions 1.290, 1.304 via patch sys/dev/pci/if_wmvar.h: revision 1.20 via patch - Set the WM_F_ATTACHED flag if wm_attach() finished succesfully and check the flag in wm_detach(). It will avoid to panic in wm_detach(). Fixes PR#49102. - Support DH89xxCC device. - Add extra delay for 82580 and newer devices except DH89XXCC SGMII device. Same as FreeBSD.
1.1102.2.17	19-Apr-2015	riz	sys/dev/pci/pcidevs 1.1208-1.1220 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen - Add Intel QuickAssist Virtual Function Device IDs - Add some Intel 10/40G devices. - Add Some Intel E7520 devices - Add Some Intel IOP332 devices - Add Xeon E5 v3 and C61x devices. - Add Mobile 5th Generation Intel Core devices. - Add Intel 63xxESB AC'97 Audio Controller - Fix product ID of INTEL 5000_PCIE_4. - Add some Intel devices. - Sort VIATECH's entries in the device ID's order. - Add some VIATECH devices. - Add COMPAQ iLOs. - Add HP IPMI device. - Add another HP vendor ID (was 3PAR) - Add AR816x/AR817x chips (part of PR/49584) - Add two PLX PCI Bridges. - Add Radeon HD 54xx Audio - Add Dell DARC 4 devices - Add nForce Host Bridges. - Add MegaRAID SAS3108 [msaitoh, ticket #1293]
1.1102.2.16	07-Dec-2014	martin	Pull up the following, requested by msaitoh in #1207: sys/dev/pci/pcidevs 1.1148-1.1149 via patch
1.1102.2.15	04-Dec-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #1202): sys/dev/pci/pcidevs: revisions 1.1160-1.1161, 1.1163-1.1171, 1.1173-1.1177, 1.1183-1.1194, 1.1196, 1.1198-1.1205 via patch Add some PCI devices to pcidevs: - PR/48180: Christian Groessler: Add Oxford Semi OXuPCI952 dual serial PCI card. - Add DIGI Neo 8-port serial(PCIe) and OXFORD OXPCIe952 Parallel. - Add some Intel devices mainly taken from "Mobile 4th Generation Intel Core Processor Family I/O" datasheet. - Add some Intel Core internal devices. - Add Intel Quark X1000 devices. - Add devices of Intel Atom S1200 series from the datasheet. - Add some Intel E600 and EG20T devices. - Add some Intel X38 devices. - Add some Intel devices (Sandy Bridge, 6702PXH, X38 and Pineview) - Add some Intel Ivy Bridge devices. - Add Intel 3200 Host and PCIe. - Add Intel EP80579 devices from OpenBSD. - Add Intel Bay Trail devices. - Add Intel Atom Z36xx and Z37xx devices. - Add Intel Xeon E3-1200 v3 Host Bridge, DRAM. - Add Intel DH89xx's SMBus controller. - Add Intel Z68 LPC. - Add yet another Intel 82599 device. - Add Intel X540-AT2. - Add some Intel gigabit Ethernet devices. - PR/48150: Noriyuki Koizumi: Add Intel Centrino Advanced-N 6235 Wi-Fi controller. - Add some Intel Wi-Fi devices. - Add entries for Atheros AR9462 and a new flavour of Intel Haswell Integrated Graphics Device. - Add Radeon HD6320 Graphics. - Add ASMedia ASM1042 xHCI USB3 controller. - Add Realtek RTL8402 PCI-E Card Reader. - Add some Realtek devices. - Add DIGI Neo 8-port serial(PCIe) and OXFORD OXPCIe952 Parallel. - Add ASPEED Graphics family. - Add ASPEED AST1150 PCIe-to-PCI bridge. - Add GeForce 210 High Definition Audio Controller. - Add GeForce GT 640M. - Add some Armada 370 IDs. - Add some new Attansic ethernet devices. - Add ITE IT8213. - Add some HP iLO devices. - Add Matrox MGA G200eH. - Add Samsung Electronics XP941 M.2 SSD.
1.1102.2.14	09-Nov-2014	martin	Pullup the following revisions, requested by msaitoh in ticket #1188: sys/dev/pci/pcidevs 1.1172 sys/dev/pci/if_wm.c 1.263-1.266 via patch sys/dev/pci/if_wmreg.h 1.55-1.56 sys/dev/pci/if_wmvar.h 1.17 share/man/man4/wm.4 1.26-1.27 and 1.29 via patch - Add I354 support. - Insert completion barrier between register write and delay(). - Fix the definition of CTRL_GIO_M_DIS. This bit is not bit 3 but bit 2. - Cleanup
1.1102.2.13	07-Sep-2013	bouyer	sys/dev/pci/if_bge.c 1.203-1.237, 1.239-1.241, 1.243-1.258 via patch sys/dev/pci/if_bgereg.h 1.58-1.74, 1.76-1.83 via patch sys/dev/pci/if_bgevar.h 1.10-1.16 via patch sys/dev/pci/pcidevs 1.1149 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/brgphy.c 1.61-1.63, 1.65, 1.67 via patch sys/dev/mii/miivar.h 1.61 via patch sys/dev/mii/miidevs 1.112-1.113 via patch sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/bge.4 1.13-1.15 via patch Add some device support, fix a lot of bugs and add some enahcements. - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F, BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - brgphy(4): Add BCM5756, BCM5717C, BCM5719C, BCM5720C and BCM57780. - Add some bugfixes and enhancement from FreeBSD: - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292) - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. From Linux tg3 driver. - Sync with BSD. No functional change: - make bge_chipid() and use it. - use switch() instead of a lot of if()s. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. FreeBSD has the same delay(). - Rename PHY related flags for the consistency. It's the same as FreeBSD. - Remove BGE_10_100_ONLY flag because this was not used. For 10/100 devices, when calling mii_attach(), mask BMSR_EXTSTAT flag to not to check Gigabit flags. It's the same as FreeBSD. - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Sync with FreeBSD and OpenBSD. Almost the same as OpenBSD rev. 1.325: - Sync the ring setup code closer to FreeBSD's driver - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support - Wait for the bootcode to complete initialization for 5717 and newer devices. See BCM5718 programmer's guide's "step 13, Device reset Procedure, Section7". 57781's document has the same note. - 5718 and 57785 document say we should wait 100us. - Change the order of register settings a little to match both Broadcom's document and Linux tg3. - Add detach function. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix a bug that the NVRAM lock timeout occured on sparc64 onboard bge. - Check BGE_SGDIG_STS when the chip is NOT a 5717 A0, like freebsd (OpenBSD rev. 1.327). - Fix the setting of Tx Random Backoff Register. - Add BGE_PCIMISCCTL_PCISTATE_RW in BGE_INIT. It's required to set the PCISTATE register correctly. - Update comments from 57XX-PG105-R.pdf and 5718-PG106-R.pdf. - Add missing bus_dmamap_sync() in bge_intr(). The status word is in DMAed area. Same as other BSDs. - Check the hardware config words and print them. This change only read them and print the values. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - bge_init_tx_ring() uses BGE_RSLOTS (==256) but bge_free_tx_ring() uses BGE_TX_RING_CNT (== 512). Delete BGE_RSLOTS and use BGE_TX_RING_CNT. Same as OpenBSD's if_bge.c rev. 1.86. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Call bge_poll_fw() before writing BGE_MODE_CTL register like the latest linux tg3 dirver. - Set DMA watermark depend on the PCI max payload size. - Add BGE_JUMBO_CAPABLE flag to some chips. With this commit, 5714, 5780, 5717, 5718, 5719 (exclude rev. A0), 5720, 57765 and 57766 are added to support jumbo frame. - Fix the setting of sc->bge_flags for 5717 and newer devices. - Fix a link detect bug on non-autopoll systems. Same as OpenBSD (rev.1.329 and 1.336) and FreeBSD (r213710). - 57765 series is not based on 5717 series. 5717 series is based on 57765 series. - Set the TX DMA segment size based on the MTU size. - Change the TX ring size for 5717 series and 57764 series. - For 57766, set BGE_RDMAMODE_JMB_2K_MMRR for non-jumbo frame. Same as Linux tg3. - For 57765 and newer devices, set BGE_MAX_RX_FRAME_LOWAT to 1. This value is recommended by the document. - Change sysctl related functions for consistency. - Style change. - Use macro. Remove duplicated macro. Remove unused variable. - Fix comments. Add comments. - Remove extra semicolon. Remove unused code. [msaitoh, ticket #939]
1.1102.2.12	14-Jul-2013	riz	Apply changes from msaitoh in ticket #906: sys/dev/pci/pcidevs 1.1145, 1.1147, 1.1150-1.1158 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Pull up pcidevs changes: Intel Atom E600 PCI-LPC bridge, adds a watchdog + HPET support. Update some Intel LPC devices. Change from "3400 USB" to "3400 USB EHCI" for EHCI devices. Fix BCM5785F entry. That is not gigabit Ethernet. Add ALTIMA AC1003, BROADCOM BCM57782 and BCM57786. Add Intel I21[0178] Ethernet. Add IDs for Marvell Armada XP. Obtained from Marvell, Semihalf. Add some Intel devices from document (Intel 8 Series / C220 Chipset Family Platform Controller Hub (PCH) Datasheet). Add some Intel devices from datasheets (4th Generation Intel Core Processor, Intel Xeon Processor E3-1200 v3). Add ATI RADEON_HD7340. [msaitoh, ticket #906]
1.1102.2.11	22-Nov-2012	riz	Add some device IDs for ticket 691.
1.1102.2.10	18-Nov-2012	msaitoh	Pull up following revision(s) (requested by matt in ticket #658): sys/dev/pci/pcidevs: revision 1.1143 Add more ASMEDIA 1061 variants.
1.1102.2.9	31-Oct-2012	riz	sys/dev/pci/pucdata.c 1.77-1.82 sys/dev/pci/pcidevs 1.1125, 1.1129 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/puc.4 1.36-1.38 Add entries for puc(4) devices into pcidevs. Add some Intel devices from document (Intel 7 Series / C216 Chipset Family Platform Controller Hub (PCH) Family Specification Update) into pcidevs. Modify some devices in pcidevs to clarify. Add Intel chipset internal serial over lan devices. Closes PR/45567. Use PCI_VENDOR_* and PCI_PRODUCT_*. Fixes PR#46303. Add EXSYS EX-41098-2 UARTs support. The clock frequency is different from other card(s). Fixes PR#46362 reported by Wolfgang Stukenbrock. Add Lava Computers SSERIAL-PCI single port serial PCI card. Update the manual.
1.1102.2.8	24-Oct-2012	riz	Apply patch (requested by msaitoh in ticket #631): This patch doesn't include any changes that rename ID.
1.1102.2.7	24-Oct-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #628): sys/dev/ic/mfi.c: revision 1.40 sys/dev/ic/mfi.c: revision 1.41 sys/dev/ic/mfi.c: revision 1.42 sys/dev/ic/mfi.c: revision 1.43 sys/dev/ic/mfi.c: revision 1.44 sys/dev/ic/mfi.c: revision 1.45 sys/dev/ic/mfi.c: revision 1.46 sys/dev/pci/pcidevs: revision 1.1128 sys/dev/pci/mfi_pci.c: revision 1.15 sys/dev/pci/mfi_pci.c: revision 1.16 sys/dev/ic/mfi.c: revision 1.39 sys/dev/ic/mfivar.h: revision 1.16 sys/dev/ic/mfireg.h: revision 1.6 sys/dev/ic/mfivar.h: revision 1.17 sys/dev/ic/mfireg.h: revision 1.7 sys/dev/ic/mfivar.h: revision 1.18 sys/dev/ic/mfivar.h: revision 1.19 Add some support for 64bit DMA but stick to 32bit DMA for now. From OpenBSD mfi.c rev 1.119. Add LSI MegaRAID SAS2208 MFI_IOP_SKINNY is an enum not a single bit value, so it can't be used as a bit flag. As sc_flags is really used to hold an enum mfi_iop value, change it to enum mfi_iop and rename to sc_ioptype. While there init it in mfi_attach() instead of mfi_pci_attach(). sg64.len is 32bits, so use htole32() Really init sc_ioptype in mfi_attach() Add support newer LSI RAID controllers based on the SAS2208 chip, codenamed "ThunderBolt". Add tagged queuing support for all adapters supported by mfi(4). Tested with a MegaRAID SAS 9265-8i adapter, and an older Dell PERC 5/i. Avoid "unused variable" warning for non-DIAGNOSTIC kernels. Pointed out by Havard Eidnes After discussion on tech-kern@, fix performance issue related to cache flush commands from WAPBL, by skipping the cache flush if the BBU is present and considered good. Users which still want the write back cache with a non-working BBU can set vfs.wapbl.flush_disk_cache to 0. - add commands to monitor the BBU state. Add a boolean BBU sensor to monitor the BBU state via sysmon_envsys(9). - if the BBU is considered good, turn SCSI_SYNCHRONIZE_CACHE_10 and SCSI_SYNCHRONIZE_CACHE_16 commands from upper layer into NOOPs. While there, handle SCSI_SYNCHRONIZE_CACHE_16 in addition to SCSI_SYNCHRONIZE_CACHE_10. - Add a shutdown pmf(9) handler, which flushes the cache and shutdown the firmware - on detach, also flush cache and shutdown firmware. - on attach, print the firmware-provided name, and the BBU state Tested on a LSI MegaRAID SAS 9265-8i and a PERC 5/i Integrated Make MFI_DEBUG build on i386. Report BBU state changes with aprint_normal(), it seems that sysmon_envsys() doens't report changes for ENVSYS_INDICATOR as it does for ENVSYS_DRIVE.
1.1102.2.6	28-Jun-2012	riz	branches: 1.1102.2.6.2; Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/if_wm.c 1.228 sys/dev/pci/if_wmreg.h 1.47 sys/dev/pci/if_wmvar.h 1.13 sys/dev/pci/pcidevs 1.1117 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/wm.4 1.22-1.25 Add support for Intel I350 Ethernet. Update document.
1.1102.2.5	23-Apr-2012	riz	Pull up following revision(s) (requested by bouyer in ticket #191): sys/dev/pci/pcidevs: revision 1.1116 Add various newer LSI RAID/SAS controllers, from OpenBSD.
1.1102.2.4	16-Apr-2012	riz	Pull up following revision(s) (requested by cegger in ticket #180): sys/arch/x86/pci/amdtemp.c: revision 1.14 sys/dev/pci/pcidevs: revision 1.1115 Add AMD Family15h ids - support AMD Family15h - deregister pmf on detach
1.1102.2.3	03-Apr-2012	riz	Pull up following revision(s) (requested by tsutsui in ticket #151): sys/dev/pci/viaide.c: revision 1.78 sys/dev/pci/pcidevs: revision 1.1106 Add VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. Regen from pcidevs rev 1.1106: Add VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. Add UDMA support for VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. This should be pulled up to netbsd-6, since ZOTAC ZBOX nano VD01 http://www.zotacusa.com/zbox-nano-vd01.html exhibited at NetBSD booth in Open Source Conference 2012 Tokyo Spring was favored by many people.
1.1102.2.2	03-Apr-2012	riz	Pull up following revision(s) (requested by pgoyette in ticket #149): sys/dev/pci/pcidevs: revision 1.1108 sys/dev/pci/jmide.c: revision 1.13 Add JMB 362 (as found on ASUS M5A99X motherboard) Add match for JMB362 (with two SATA ports).
1.1102.2.1	22-Mar-2012	riz	Pull up following revision(s) (requested by sborrill in ticket #137): sys/dev/pci/pcidevs: revision 1.1107 Add some Symbios SAS2108 devices. Add new Symbios SAS2008 device (as used by IBM ServeRAID M1015).
1.1102.2.6.2.1	18-Nov-2012	matt	Merge tickets 658 and 659.
1.1132.2.5	03-Dec-2017	jdolecek	update from HEAD
1.1132.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1132.2.3	23-Jun-2013	tls	resync from head
1.1132.2.2	25-Feb-2013	tls	resync with head
1.1132.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.1158.4.1	23-Jul-2013	riastradh	sync with HEAD
1.1158.2.2	18-May-2014	rmind	sync with head
1.1158.2.1	28-Aug-2013	rmind	sync with head
1.1184.2.1	10-Aug-2014	tls	Rebase.
1.1199.2.12	11-Aug-2018	martin	Pull up the following, requested by msaitoh in ticket #1628: share/man/man4/wm.4 1.40 via patch sys/dev/mii/ihphyreg.h 1.2 sys/dev/mii/inbmphyreg.h 1.10 sys/dev/pci/if_wm.c 1.504, 1.506, 1.510-1.535, 1.539-1.540, 1.546, 1.548, 1.551-1.552, 1.558, 1.565-1.573, 1.575, 1.579, 1.582, 1.584 via patch sys/dev/pci/if_wmreg.h 1.99-1.103, 1.106-1.107 via patch sys/dev/pci/if_wmvar.h 1.34-1.39 via patch sys/dev/pci/pcidevs 1.1327 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/pci/pcireg.h patch Sync wm(4) up to 2018/08/08 except MSI/MSI-X and NET_MPSAFE: - remove extra "+" - Fix a bug that non-GMII devices don't send a routing message when the link status is changed. - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes. - If a interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Print sc_flags with snprintb(). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange(). - Add missing "else" in wm_nvm_release(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - If TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function are required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Use new nvm.{acquire,release}() for semaphore. - Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register value from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally on I21[01]". PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Print NVM offset and word count when EERD polling failed. - On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux). - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Fix typo in comment. Reported by Shinichi Doyashiki in PR#52885. - Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki. - CID-1427779: Fix uninitialized variables. - Fix a bug that wm_pll_workaround_i210() is not called when a) Chip is I211 or b) Chip is I210 and it uses iNVM (not FLASH). - Do wm_reset_mdicnfg_82580() on 82580 only. - Fix FLASH access on PCH_SPT and newer. Their FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it must be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. - Make wm_nvm_valid_bank_detect_ich8lan() the same as other OSes. - If the extended configuration size in the EXTCNFSIZE register is 0, don't continue in wm_init_lcd_from_nvm(). - Add PCH_CNP support (I219 with Intel 300 series chipset). - Enable I219 support. - I354 uses an external PHY, so don't use wm_set_eee_i350(). - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. - Fix a bug that 82542 misunderstand fiber's signal detection. - Add debug printf()s. - Update comment. - Rename functions and variables. - Add diagnostic code. - Sort registers. - Lowercase hexadecimal values. - KNF.
1.1199.2.11	03-Jan-2018	snj	Pull up following revision(s) (requested by maya in ticket #1547): sys/dev/pci/pucdata.c: revision 1.99-1.100 sys/dev/pci/pcidevs: revision 1.1278 Add Intel 200 series chipset devices from "Table 2-2. PCH-H Device and Revision ID Table, Intel 200 Series Chipset Family Platform Controller Hub(PCI) Datasheet Volume 1 of 2 (335192-001)". -- Add Intel 200 series devices. -- PR/52868: Petar Bogdanovic: Add support for Manhattan 158220 card
1.1199.2.10	25-Jul-2017	snj	Apply patch (requested by nonaka in ticket #1413): iwm(4): Add a driver for Intel Wireless 726x, 316x, 826x and 416x series. Ported from OpenBSD.
1.1199.2.9	09-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1300): sys/dev/pci/pcidevs: revision 1.1270 sys/dev/pci/if_wm.c: revision 1.426 i82567V-3 is not ICH9 but ICH8. -- 82567V-3 is not ICH9 but ICH8.
1.1199.2.8	09-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1298): sys/dev/pci/piixide.c: revision 1.65 sys/dev/pci/pcidevs: revision 1.1246 - s/82081/82801/ from OpenBSD. - 0x2829 is SATA. 0x282a is RAID. -- Rename macro (sync with pcidevs rev. 1.1246). No functional change.
1.1199.2.7	08-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1291): sys/dev/pci/pcidevs: revisions 1.1221-1.1245, 1.1247-1.1262, 1.1264-1.1269, 1.1271-1.1273 - Add SystemBase SB16C1050 UARTs. - Add D-Link DFE520TX (part of PR#49889 reported by Maxim Tsyplakov). - Add Realtek RTL8188EE and RTL8192CE - Add PLX Technology PEX 8111. - Add some Red Hat devices. - Add Samsung SM951. - Add several NVIDIA geforce and NVS cards. - Add some ATI/AMD devices. - Add Radeon HD 2400 and TSB82AA2 IEEE 1394 Host Controller. Reported by T. M. Pederson (PR#40987). - Add some Edimax, AWT, Fusion-io and Ralink devices of ral(4). - Add MSI RT3090, Edimax RT3591, Ralink RT 536[02] and RT5390 of ral(4). - Add vendor ID of Super Micro. - Add MegaRAID SAS3008. - Fix entries. The following problems were found while sorting entries: - Remove the following entries because the value is duplicated: CMDTECH_240 (CMDTECH_AAR_1210SA has the same value) INTEL_C610_SATA_RAID (INTEL_82801H_SATA_RAID has the same vaule) INTEL_C610_PCIE_1_3 (82801BA_HPB has the same value) GEFORCE_610M3 (GEFORCE_610M2 has the same value) - RADEON_RV350_NQ is not 0x4e41 but 0x4e51 - 82801JD_SATA_RAID is not 0x3a02 but 0x3a05. - The following two entries have the same number. Fix one of them: DH89XXCL_USB_2 to 0x23b5 (was 0x23b4 (== DH89XXCL_USB_1)) 5500_HB to 0x3404 and change the desc (was 0x3403 (== 3400_HB)) 82X58_PCIE_0_1 to 0x3421 (was 0x3420 (== 82X58_PCIE_0_0)) 82801JI_THERMAL to 0x3a32 (was 0x3a30 (== 82801JI_SMB)) - Add the following Intel entries: - PCIe NVMe SSD. - Dual Band Wireless AC 3165. - I219 family Ethernet devices. - Dual Band Wireless AC 3165, 4165 and 8260 devices. - 82597EX_SR (part of PR#47750). - Braswell devices. - Core 6G devices. - 100 series chipset devices. - Xeon D devices. - Xeon 5[56]00 devices. - Xeon E7 v4 devices. - Add "Core i7-6xxxK" to some Intel devices' description. - Fix Intel XE5_V3_IMC1_DDRIO_[34]'s values. - Add another ID for Intel 82Q45 KT. - Add some entries of Intel 965 Express Chipset. - Add some Intel Xeon E5 v3 devices. - Add some Intel 1st gen Core PCI devices. - Add some Intel Core 5G devices. - Add some Intel Core i7-6xxxK devices. - Add some Intel Core i7-800 and i5-700 devices. - Add some Intel 10G devices. - Add ASMedia ASM1042A USB 3.0 Host Controller. - Add ASMEDIA ASM1083/1085/1142. - Add Marvell Armada38x and 88AP510. - Add Qualcomm Atheros QCA986x/988x. - Add Atheros AR9565 Wireless LAN. - Sort entries in device ID's order. - Fix typo.
1.1199.2.6	19-Apr-2015	riz	branches: 1.1199.2.6.4; Apply patch (requested by msaitoh in ticket #699): sys/dev/pci/pcidevs 1.1200, 1.1203, 1.1208-1.1212, 1.1214-1.1220 via patch Sync pcidevs with rev. 1.1220: - Add Samsung Electronics XP941 M.2 SSD - Add Intel Atom Z36xx and Z37xx devices from the datasheet. - Add Intel QuickAssist Virtual Function Device IDs - Add some Intel 10/40G devices. - Add Some Intel E7520 devices - Add Some Intel IOP332 devices - Add Xeon E5 v3 and C61x devices. - Add Mobile 5th Generation Intel Core devices. - Add Intel 63xxESB AC'97 Audio Controller - Fix product ID of INTEL 5000_PCIE_4. - Add some Intel 9 Series devices - Add some Intel devices. - Sort VIATECH's entries in the device ID's order. - Add some VIATECH devices. - Add COMPAQ iLOs. - Add HP IPMI device. - Add another HP vendor ID (was 3PAR) - Add two PLX PCI Bridges. - Add Radeon HD 54xx Audio - Add Dell DARC 4 devices - Add nForce Host Bridges. - Add MegaRAID SAS3108 [msaitoh, ticket #699]
1.1199.2.5	17-Feb-2015	martin	Pull up the following, requested by msaitoh in ticket 528: sys/dev/pci/pcidevs: revision 1.1206-1.1207 - Rename DH89XX_QA to DH89XXCC_IQIA - Modify the description to DH89xxCC PCIe Endpoint and QuickAssist (include typo fix) - Rename DH89xxCC'"'"'s names from DH89XX_ to DH89XXCC_. - Add some DH89xxCC'"'"'s devices. - Add DH89XXCL'"'"'s devices. - Rename PCI_PRODUCT_INTEL_DH89XX_SMB to PCI_PRODUCT_INTEL_DH89XXCC_SMB - Add PCI_PRODUCT_INTEL_DH89XXCL_SMB
1.1199.2.4	04-Feb-2015	snj	Pull up following revision(s) (requested by nonaka in ticket #473): share/man/man4/lii.4: revision 1.3 sys/dev/pci/if_alc.c: revision 1.12 sys/dev/pci/pcidevs: revision 1.1213 sys/dev/pci/if_alcreg.h: revision 1.3 share/man/man4/atphy.4: revision 1.3 share/man/man4/alc.4: revision 1.4 PR/49584: Leonardo Taccari: alc(4): add support for AR816x/AR817x chipsets
1.1199.2.3	28-Jan-2015	martin	Pull up following revision(s) (requested by nonaka in ticket #467): external/intel-fw-public/iwl2030/dist/iwlwifi-2030-6.ucode: revision 1.1 distrib/sets/lists/base/mi: revision 1.1091 share/man/man4/iwn.4: revision 1.13 external/intel-fw-public/iwl100/Makefile: revision 1.1 external/intel-fw-public/iwl2030/Makefile: revision 1.1 external/intel-fw-public/iwl135/Makefile: revision 1.1 sys/dev/pci/pcidevs: revision 1.1205 external/intel-fw-public/Makefile: revision 1.8 external/intel-fw-public/iwl135/dist/iwlwifi-135-6.ucode: revision 1.1 external/intel-fw-public/iwl2030/dist/LICENSE.iwlwifi-2030-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/LICENSE.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/LICENSE.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/iwlwifi-100-5.ucode: revision 1.1 external/intel-fw-public/iwl135/dist/LICENSE.iwlwifi-135-ucode: revision 1.1 external/intel-fw-public/iwl135/dist/README.iwlwifi-135-ucode: revision 1.1 sys/dev/pci/if_iwnvar.h: revision 1.16 external/intel-fw-public/iwl2000/dist/iwlwifi-2000-6.ucode: revision 1.1 external/intel-fw-public/iwl100/dist/README.iwlwifi-100-ucode: revision 1.1 sys/dev/pci/if_iwn.c: revision 1.72 sys/dev/pci/if_iwnreg.h: revision 1.14 external/intel-fw-public/iwl105/dist/README.iwlwifi-105-ucode: revision 1.1 external/intel-fw-public/iwl100/dist/LICENSE.iwlwifi-100-ucode: revision 1.1 external/intel-fw-public/iwl2000/dist/README.iwlwifi-2000-ucode: revision 1.1 external/intel-fw-public/iwl105/dist/iwlwifi-105-6.ucode: revision 1.1 external/intel-fw-public/iwl2000/Makefile: revision 1.1 external/intel-fw-public/iwl105/Makefile: revision 1.1 external/intel-fw-public/iwl2030/dist/README.iwlwifi-2030-ucode: revision 1.1 Add some Intel Wi-Fi devices. Add support for Intel Centrino Wireless-N 100/105/130/135/2200/2230. From OpenBSD.
1.1199.2.2	17-Jan-2015	martin	Pull up revision 1.1204 for ticket #432
1.1199.2.1	07-Nov-2014	snj	Pull up following revision(s) (requested by msaitoh in ticket #190): sys/dev/pci/pcidevs 1.1201-1.1202 via patch sys/dev/pci/if_wm.c 1.291-1.301, 1.304, 1.306-1.307 via patch sys/dev/pci/if_wmreg.h 1.61-1.65 via patch sys/dev/pci/if_wmvar.h 1.21 via patch share/man/man4/wm.4 1.31 via patch - Add Internal SERDES mode support newer than or equal to 82575. - Add new I218 devices. - Add DH89xxCC device. - Add some old devices. - Fix a bug that wm_get_swsm_semaphore() timed out when attaching device on some machines. - Fix a bug that the offset of alt MAC address is wrongly calculated to 0 when alt MAC address function is really used. This bug does not appear as real bug if the same MAC address is written in the default location and alt MAC address's location. - Don't print "failed to detect NVM bank" message. - Delete 82580ER related code. - Fix typo in comment. - Fix debug message. - Cleanup
1.1199.2.6.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.1205.2.11	28-Aug-2017	skrll	Sync with HEAD
1.1205.2.10	05-Feb-2017	skrll	Sync with HEAD
1.1205.2.9	05-Dec-2016	skrll	Sync with HEAD
1.1205.2.8	05-Oct-2016	skrll	Sync with HEAD
1.1205.2.7	09-Jul-2016	skrll	Sync with HEAD
1.1205.2.6	29-May-2016	skrll	Sync with HEAD
1.1205.2.5	19-Mar-2016	skrll	Sync with HEAD
1.1205.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1205.2.3	21-Sep-2015	skrll	Sync with HEAD
1.1205.2.2	06-Jun-2015	skrll	Sync with HEAD
1.1205.2.1	06-Apr-2015	skrll	Sync with HEAD
1.1253.2.6	26-Apr-2017	pgoyette	Somehow this got out of sync on the branch.
1.1253.2.5	26-Apr-2017	pgoyette	Sync with HEAD
1.1253.2.4	20-Mar-2017	pgoyette	Sync with HEAD
1.1253.2.3	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1253.2.2	04-Nov-2016	pgoyette	Sync with HEAD
1.1253.2.1	06-Aug-2016	pgoyette	Sync with HEAD
1.1278.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.1286.2.2	11-May-2017	pgoyette	Sync with HEAD
1.1286.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.1289.2.31	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #1931): sys/dev/pci/pcidevs: revision 1.1477 sys/dev/pci/pcidevs: revision 1.1494 sys/dev/pci/pcidevs: revision 1.1498 sys/dev/pci/pcidevs: revision 1.1499 sys/dev/pci/pcidevs: revision 1.1500 sys/dev/pci/pcidevs: revision 1.1501 sys/dev/pci/pcidevs: revision 1.1502 Add ID for the EXAR XR17V354 PCIe UART. From Hakan Engvall. PR kern/57202 pcidevs: Add missing I225/I226 series devices pcidevs: Add Mellanox ConnectX-6 Lx. add some newer Areca boards in prep for updating driver Add Intel C26[26] eSPI. Update Intel Raptor Lake devices. Add Intel I219-{LM,V}(20,21)
1.1289.2.30	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1915: sys/dev/pci/pcidevs 1.1497 sys/dev/pci/if_wm.c 1.689,1.790-1.791 via patch sys/dev/pci/if_wmreg.h 1.120 via patch sys/dev/pci/if_wmvar.h 1.51 - Use 12K for packet buffer for jumbo frame on PCH2 and newer. - Add new workaround for Tiger Lake and newer to avoid packet loss. - Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.1289.2.29	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1913): sys/dev/pci/pcidevs: revision 1.1493 sys/dev/pci/pcidevs: revision 1.1495 Add Intel 82599 LS. Add Pericom(Diodes) PCIe switches.
1.1289.2.28	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #1911): sys/dev/pci/pcidevs: revision 1.1496 Add entry for ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier
1.1289.2.27	21-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1887): sys/dev/pci/pcidevs: revision 1.1483-1.1492 Add AMD 19h/7xh Root Complex. pcidevs: add ASMedia ASM106x Add Intel Snow Ridge devices. Fix some descriptions of Intel Snow Ridge. Add Intel I226 devices. Add device ids of the VX800 chipset and s3 chrome 500 series GPU. Descriptions mainly aligned with VX800/820 Series System Programming Manual. fix duplicate definition. Add some Intel Xeon D-2100 devices. Fix description of Intel 0x37d0. Simplify descriptions of Snow Ridge internal Ethernet devices. Rename PCI to PCIE for naming consistency.
1.1289.2.26	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #1855: sys/dev/pci/pcidevs 1.1481-1.1482 Add Intel Alder Lake-N, Raptor Lake and 700 series chipset devices.
1.1289.2.25	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1846): sys/dev/pci/pcidevs: revision 1.1478 sys/dev/pci/pcidevs: revision 1.1479 sys/dev/pci/pcidevs: revision 1.1480 Add Samsung SM990. Add devices from PPR for AMD Family 19h Model 61h Revision B1 processors. The SATA device ID for Apollo Lake is not 0x5ae0 but 0x5ae3.
1.1289.2.24	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1783): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1469 sys/dev/pci/pcidevs: revision 1.1470 sys/dev/pci/pcidevs: revision 1.1471 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Add some AMD 17h/9xh devices from OpenBSD. Add AMD F17/Axh devices. Add some Xeon Scalable devices from OpenBSD. Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.1289.2.23	11-Oct-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1771: sys/dev/pci/pcidevs 1.1461-1.1468 via patch - Add several samsung nvme entries. - Add more Alder Lake devices. - Jasper Lake Intel Trace Hub on Compute Die is not 0x4da6 but 0x4e29. - Add Intel Core 8G (8core, H, Halo) Host Bridge, DRAM. - Add AMD 19h/6xh Root Complex. - Add AMD FCH SATA Controller D. - Add NVIDIA GeForce GTX 770. - Sort by number.
1.1289.2.22	02-Aug-2022	snj	Pull up following revision(s) (requested by msaitoh in ticket #1754): sys/dev/pci/pcidevs: 1.1445-1.1460 via patch Update pcidevs: - Add Intel Alder Lake devices and Intel 600 Series PCH devices. - Add some Intel Xeon Scalable / Skylake-E devices. - Fix AMD F16_HB from 0x1568 to 0x1538. - Add some devices for AMD and improve some descriptions to clarify. - Add VMware AHCI and NVMe. - Update Intel 700 series Ethernet devices. - Add some Broadcom devices. - Add some Broadcom / LSI RAID cards. - Fix typos and whitespace.
1.1289.2.21	03-Dec-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1710: sys/dev/pci/pcidevs 1.1441-1.1444 - Add Intel Gemini Lake TXE HECI 1. - Add Intel Elkhart Lake and Rocket Lake devices. - Update Jasper Lake's Processor Transaction Routers.
1.1289.2.20	23-Oct-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1700: sys/dev/pci/pcidevs 1.1420-1.1440 - Add modern QUMRANET/Red Hat VIRTIO range PCI devices. - Add some more product IDs for mcx(4). - Add current generation NVIDIA graphics cards (3050-3090 etc.). - Add RDC R6022 PCI-Host bridge. - Add a whole bunch of radeon devices. - Add Realtek RTL8821CE. - Add Intel 660p SSD, and expand the 760p description. - Add Intel I219's version number. - Add Intel I219V 15-19 and I219LM 16-19. - Add Intel I225V, I225LM and WiFi 6 AX201. - Add newer Intel PCH internal devices. - Add Intel Jasper Lake devices. - PDC20265 is Ultra/100, not 66. - Fix Realtek RTL8125 description. - Sort Cavium devices.
1.1289.2.19	20-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/pcidevs 1.1392-1.1403, 1.1406-1.1419 - Add Intel Xeon D-1500 NTB-secondary and Xeon D NS QuickData DMA channel 0-7. - Add Intel I219 LM10-LM15 and V10-V14. - Add AMD Family14h PCIe. - Add Realtek Killer E3000. - Add Aquantia AQC 10G network adapters. - Add Radeon HD4290 - Add Farallon PN9000SX Ethernet. - NVIDIA 0x036[0-7] are nForce MCP55 LPC Bridge. - Add VIA VX900 chipset. - Add some NVIDIA devices. - Add some Intel UHD Graphics devices. - Add Intel Comet Lake, Whiskey Lake U and Amber Lake Y devices. - Modify description of Intel 0x591e from HD Graphics to UHD Graphics. - Add Intel XMM 7360 LTE Modem. - Add Western Digital WD Blue SN550 NVMe SSD. - Add ATI Radeon R5/R6/R7 Graphics. - Add IDs for Ampere eMAG PCIe Root Ports. - Add RTL8192EE Wireless LAN 802.11n PCI-E NIC. - Add ASIX AX99100 Multi I/O (Serial,Parallel,I2C,SPI,LocalBus,GPIO) Controller. - Add a couple of additional device IDs for the AMD Cryptographic Coprocessor. - Remove duplicated entries.
1.1289.2.18	20-Jul-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1578): sys/dev/pci/pcidevs: revision 1.1404 sys/dev/pci/pcidevs: revision 1.1405 sys/arch/x86/pci/rdcpcib.c: revision 1.3 Add more RDC products (Andrius V) Recognize more rdc devices (Andrius V.) Fix typo
1.1289.2.17	19-Nov-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1448: sys/dev/pci/pcidevs 1.1384-1.1391 - Add micron/crucial SM2263 NVMe. - Add some device found on ASUS X570-P with Ryzen 3200G CPU. - Spell it "PCIe' when using the name. - Add Mellanox ConnectX-4, ConnectX-4 Lx, ConnectX-5, and ConnectX-5 Ex - Modify Attansic Ethernet devices' description to clarify. - Add Killer E2400 and E2500. - Spell "controller" correctly. - Add Realtek RTL8125. - Update Intel's NVMe SSDs. - Add Intel Apollo Lake TXE HECI. - Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.1289.2.16	06-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1326): sys/dev/pci/pcidevs: revision 1.1381 sys/dev/pci/pcidevs: revision 1.1382 sys/dev/pci/pcidevs: revision 1.1383 - Intel C620 Virtual Switch Port for Termal Sensor is not 0x37c6 but 0x37c7. - Add Some AMD Ryzen 3xxx's devices. - Add some AMD devices mainly taken from OpenBSD. - Rename some device names for consistency. - Use _HDA for HD Audio. Add BCM2711 PCIe Host bridge for RPI4
1.1289.2.15	17-Jul-2019	martin	Pull up the foolowing revisions (via patch), requested by msaitoh in ticket #1291: sys/dev/pci/pcidevs 1.1367-1.1380 - Add other two Intel Core 8G host bridges. - Add Intel Xeon E devices. - Add some new Intel C620 LPC devices. - Add some Intel Xeon Scalable processor's devices. - Add some Intel Whisky Lake devices. - Add Xeon D XFI. - Add "(Hyper-V)" to Intel 0x15a9's description. - Add Intel C3000 QSFP devices. - Add Some Qlogic devices from OpenBSD. - Rename Qlogic 0x3022's name from ISP3022 to ISP4022. - Add IDs for SATA controllers VT8237S and VX855 - Modify some descriptions of Virtio devices from OpenBSD to identify the functions. - Add Phison PS5000 and PS5016 NVMe. - Add some ServerWorks HT1100 devices from OpenBSD. - Remove extra white spaces.
1.1289.2.14	07-Mar-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1205): sys/dev/pci/pcidevs: revision 1.1363 sys/dev/pci/pcidevs: revision 1.1364 sys/dev/pci/pcidevs: revision 1.1365 sys/dev/pci/pcidevs: revision 1.1366 Add Tundra (now IDT) TSI381 and PEB383 from OpenBSD. Add some Broadcom Ethernet devices from {Open,Free}BSD. Add SK-NET SK-9Mxx Gigabit Ethernet. Add Intel I219 variations for Cannon Lake.
1.1289.2.13	11-Feb-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1188): sys/dev/pci/pcidevs: revision 1.1360 sys/dev/pci/pcidevs: revision 1.1361 sys/dev/pci/pcidevs: revision 1.1362 sys/dev/pci/pcidevs: revision 1.1357 sys/dev/pci/pcidevs: revision 1.1359 Add yet another Intel Core QuickPath Generic Non-Core Register. Add some RADEON devices. Mainly taken from OpenBSD. Add AMD Family 17h devices from OpenBSD. From OpenBSD: - Add NetXen's vendor ID and devices. - Add Sun Neptune. - Add Some Freescale MPC834x devices from OpenBSD. - Add Intel I210 SGMII Flash-less device from the spec update document.
1.1289.2.12	04-Dec-2018	martin	Pull up following revision(s) (requested by bouyer in ticket #1123): sys/dev/pci/pcidevs: revision 1.1356 Add 2 Symbios products, from OpenBSD
1.1289.2.11	04-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1119): sys/dev/pci/pcidevs: revision 1.1350 sys/dev/pci/pcidevs: revision 1.1351 sys/dev/pci/pcidevs: revision 1.1352 sys/dev/pci/pcidevs: revision 1.1353 sys/dev/pci/pcidevs: revision 1.1354 sys/dev/pci/pcidevs: revision 1.1355 sys/dev/pci/pcidevs: revision 1.1345 sys/dev/pci/pcidevs: revision 1.1346 sys/dev/pci/pcidevs: revision 1.1347 sys/dev/pci/pcidevs: revision 1.1349 Add some Intel devices. Some of them are found in Core i7-8550U. Add Toshiba's XG4 and XG5 NVMe SSD. - Add C62x to the descriptions of Intel C620 series devices. - Add 0xa1ca for Intel C629 LPC or eSPI. Add AMD Seattle Host Bridge and PCIE root port devices Add two otehr I2C devices of Intel. "s/ are are / are /" in comment. No functional change. Add some Intel 100 series and C230 series devices. Add a whole bunch of Cavium ThunderX devices Add Rockchip vendor ID and a device. Requested by jak@ From OpenBSD Add Amazon.com UART, NVMe, and ENA PCI IDs.
1.1289.2.10	17-Oct-2018	martin	Pull up following revision(s) (requested by uwe in ticket #1064): sys/dev/pci/aac_pci.c: revision 1.39 sys/dev/pci/pcidevs: revision 1.1348 Add Adaptec RAID 2445 and 2805. From Dima Veselov.
1.1289.2.9	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #934): sys/dev/pci/pcidevs: 1.1332-1333, 1335-1339, 1341-1344 via patch - Add Symbios/LSI RAID and SAS controllers. - Add some broadcom 802.11ac wifi devices. - Add some Marvell Yukon devices. - Add Intel 82Q45 IDER. - Add some Intel 4 series chipset devices. - Add some Intel 82965 devices. - Change Intel 0x2a06 from IDE to IDER. - Add a bunch of nvidia devices from various generations. - Add two more Yukon variants (from OpenBSD). - Add Mobility Radeon HD 4300(0x9552) and Mobility Radeon HD 4500(0x9553) from OpenBSD. - Add nSYS Technologies. - Add Western Digital (Sandisk) WD Black NVMe SSD as per suggestion from thorpej@
1.1289.2.8	14-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #878): share/man/man4/rtsx.4: revision 1.5 sys/dev/pci/pcidevs: revision 1.1331 sys/dev/ic/rtsxreg.h: revision 1.3 sys/dev/ic/rtsxvar.h: revision 1.3 sys/dev/pci/rtsx_pci.c: revision 1.7 sys/dev/ic/rtsx.c: revision 1.3 add realtek RT525A PCI-E card reader found in my Dell XPS 15 9550. rtsx(4): Add support for RT525A, from openbsd. ok rkujawa Denote RTS525A, bump date
1.1289.2.7	07-Jun-2018	martin	Pull up following revision(s) (requested by jnemeth in ticket #846): sys/dev/pci/pcidevs: revision 1.1334 sys/dev/pci/pucdata.c: revision 1.102 add Oxford Semiconductor OXPCIe954 PCIe 4-port serial card
1.1289.2.6	18-Apr-2018	martin	Pull up following revision(s) (requested by nonaka in ticket #780): sys/dev/pci/pcidevs: revision 1.1328-1.1330 Add some 8th Generation Intel Core Processor devices. Add Intel SSD 760p. Added some NVMe devices.
1.1289.2.5	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #751): sys/dev/pci/pcidevs: revision 1.1308 sys/dev/pci/pcidevs: revision 1.1327 Add Nvidia GeForce 9400M as found in the late 2009 13" MacBook (MacBook6,1) Add new I219 devices (using with 300 series chipset).
1.1289.2.4	09-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #723): sys/dev/pci/pcidevs: revision 1.1314 sys/dev/pci/pcidevs: revision 1.1315 sys/dev/pci/pcidevs: revision 1.1316 sys/dev/pci/pcidevs: revision 1.1317 sys/dev/pci/pcidevs: revision 1.1318 sys/dev/pci/pcidevs: revision 1.1302 sys/dev/pci/pcidevs: revision 1.1303 sys/dev/pci/pcidevs: revision 1.1304 sys/dev/pci/pcidevs: revision 1.1305 sys/dev/pci/pcidevs: revision 1.1306 sys/dev/pci/pcidevs: revision 1.1307 sys/dev/pci/pcidevs: revision 1.1309 sys/dev/pci/pcidevs: revision 1.1320 sys/dev/pci/pcidevs: revision 1.1321 sys/dev/pci/pcidevs: revision 1.1322 sys/dev/pci/pcidevs: revision 1.1324 sys/dev/pci/pcidevs: revision 1.1325 sys/dev/pci/pcidevs: revision 1.1326 sys/dev/pci/pcidevs: revision 1.1310 sys/dev/pci/pcidevs: revision 1.1311 sys/dev/pci/pcidevs: revision 1.1312 sys/dev/pci/pcidevs: revision 1.1313 Add some Chelsio devices. - s/MMC/eMMC/ for Intel 0x0f14 (Bay Trail) - Add Atom E3800 (Bay Trail) eMMC 4.5 Add Intel Apollo Lake devices. Add the GeForce4 MX440 Go, as found on the 17" flat panel G4 iMac (PowerMac4,5). Thanks to Jasper Wallace <jasper AT pointless DOT net> for the loan of the hardware. Add Nvidia GeForce FX Go5200, as found in the PowerBook6,4 Add Marvell Yukon 88E8040T grammar Add some ServerEngines (Emulex) products from OpenBSD. Add some Intel Ethernet devices. Add C620 devices. Fix C62x9 sSATA Controller (RAID) device ID. s/C62x9/C62x/ Add some Xeon E5 (v1) and E5 v2 devices. Add AMD KERNCZ chipset for Zen processors. Add Intel Gemini Lake devices. apply some updates and recommended corrections to pcidevs descriptions PR kern/45733 by Greg A.Woods Intel 0x15c4 is SFI based and 0x15ce is KR based. Add Intel Wireless AC 9260. Add two VIA xHCI devices from OpenBSD. Add Intel X299 and Z370 LPC. Add Intel Coffee Lake internal graphics devices. From ark.intel.com. Add Intel 300 series devices.
1.1289.2.3	08-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #616): sys/dev/pci/if_msk.c: revision 1.55 sys/dev/pci/pcidevs: revision 1.1299 update entry for Marvel Yukon 8058, and fix URL for pci ids (Rocky Hotas)
1.1289.2.2	21-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #364): sys/dev/pci/pcidevs: revision 1.1300 sys/dev/pci/pcidevs: revision 1.1301 sys/dev/pci/pcidevs: revision 1.1290 sys/dev/pci/pcidevs: revision 1.1292 sys/dev/pci/pcidevs: revision 1.1293 sys/dev/pci/pcidevs: revision 1.1294 sys/dev/pci/pcidevs: revision 1.1295 sys/dev/pci/pcidevs: revision 1.1296 sys/dev/pci/pcidevs: revision 1.1297 sys/dev/pci/pcidevs: revision 1.1298 - Change VMware's 0x0770 from "Virtual USB" to "Virtual EHCI" - Add VMware's 0x0774(UHCI) and 0x0778(xHCI) Add Intel 82Q965 HECI. Add Atom Z8000 devices from "Intel Atom Z8000 Processor Series Datasheet (Volume 2 of 2)". - Add some AMD devices. - Add Broadcom BCM5762. Add Intel 5520/X58 QuickPath from OpenBSD. PCI device ID of Atom x5-Z8350's video is 0x22b0. ark.intel.com says 0x22b0 is Intel HD Graphics 400. Add C3000 devices. Add Intel QuickAssist devices. Add Areca ARC1214 & ARC1880. From OpenBSD. Add PEX 8605 4port 4lane PCIe Gen 2 switch.
1.1289.2.1	25-Jun-2017	snj	Pull up following revision(s) (requested by jdolecek in ticket #55): share/man/man4/siisata.4: revision 1.6 sys/dev/pci/pcidevs: revision 1.1291 sys/dev/pci/siisata_pci.c: revision 1.15 add entry for Adaptec 1220SA -- match Adaptec 1220SA; that card actually uses Silicon Image 3132, and works well with the siisata(4) driver -- add Adaptec 1220SA
1.1319.2.15	26-Jan-2019	pgoyette	Sync with HEAD
1.1319.2.14	18-Jan-2019	pgoyette	Synch with HEAD
1.1319.2.13	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1319.2.12	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.1319.2.11	20-Oct-2018	pgoyette	Sync with head
1.1319.2.10	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1319.2.9	28-Jul-2018	pgoyette	Sync with HEAD
1.1319.2.8	25-Jun-2018	pgoyette	Sync with HEAD
1.1319.2.7	21-May-2018	pgoyette	Sync with HEAD
1.1319.2.6	02-May-2018	pgoyette	Synch with HEAD
1.1319.2.5	22-Apr-2018	pgoyette	Sync with HEAD
1.1319.2.4	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.1319.2.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1319.2.2	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.1319.2.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.1343.2.3	21-Apr-2020	martin	Sync with HEAD
1.1343.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1343.2.1	10-Jun-2019	christos	Sync with HEAD
1.1383.2.26	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1978): sys/dev/pci/pcidevs: revision 1.1528 Add VIA VT8233C PCI-ISA Bridge.
1.1383.2.25	08-May-2025	martin	Additionally pull up sys/dev/pci/pcidevs 1.1520 accidently missed in previous for ticket #1944
1.1383.2.24	08-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1944): sys/dev/pci/pcidevs: revision 1.1521 sys/dev/pci/pcidevs: revision 1.1522 sys/dev/pci/pcidevs: revision 1.1518 sys/dev/pci/pcidevs: revision 1.1519 Various VIA IDE controller related additions/changes.
1.1383.2.23	03-Jul-2024	martin	Pull up following revision(s) (requested by nia in ticket #1852): sys/dev/pci/pcidevs: revision 1.1507 Add Brainboxes and devices From Cameron Williams in PR 58358.
1.1383.2.22	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #1790): sys/dev/pci/pcidevs: revision 1.1494 sys/dev/pci/pcidevs: revision 1.1498 sys/dev/pci/pcidevs: revision 1.1499 sys/dev/pci/pcidevs: revision 1.1500 sys/dev/pci/pcidevs: revision 1.1501 sys/dev/pci/pcidevs: revision 1.1502 pcidevs: Add missing I225/I226 series devices pcidevs: Add Mellanox ConnectX-6 Lx. add some newer Areca boards in prep for updating driver Add Intel C26[26] eSPI. Update Intel Raptor Lake devices. Add Intel I219-{LM,V}(20,21)
1.1383.2.21	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1754): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.1383.2.20	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1750): sys/dev/pci/pcidevs: revision 1.1493 sys/dev/pci/pcidevs: revision 1.1495 Add Intel 82599 LS. Add Pericom(Diodes) PCIe switches.
1.1383.2.19	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #1748): sys/dev/pci/pcidevs: revision 1.1496 Add entry for ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier
1.1383.2.18	21-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1719): sys/dev/pci/pcidevs: revision 1.1483-1.1492 Add AMD 19h/7xh Root Complex. pcidevs: add ASMedia ASM106x Add Intel Snow Ridge devices. Fix some descriptions of Intel Snow Ridge. Add Intel I226 devices. Add device ids of the VX800 chipset and s3 chrome 500 series GPU. Descriptions mainly aligned with VX800/820 Series System Programming Manual. fix duplicate definition. Add some Intel Xeon D-2100 devices. Fix description of Intel 0x37d0. Simplify descriptions of Snow Ridge internal Ethernet devices. Rename PCI to PCIE for naming consistency.
1.1383.2.17	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #1671: sys/dev/pci/pcidevs 1.1481-1.1482 Add Intel Alder Lake-N, Raptor Lake and 700 series chipset devices.
1.1383.2.16	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1654): sys/dev/pci/pcidevs: revision 1.1478 sys/dev/pci/pcidevs: revision 1.1479 sys/dev/pci/pcidevs: revision 1.1480 Add Samsung SM990. Add devices from PPR for AMD Family 19h Model 61h Revision B1 processors. The SATA device ID for Apollo Lake is not 0x5ae0 but 0x5ae3.
1.1383.2.15	30-Jan-2023	martin	Pull up following revision(s) (requested by thorpej in ticket #1580): sys/dev/pci/pcidevs: revision 1.1477 Add ID for the EXAR XR17V354 PCIe UART. PR kern/57202
1.1383.2.14	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1566): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1469 sys/dev/pci/pcidevs: revision 1.1470 sys/dev/pci/pcidevs: revision 1.1471 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Add some AMD 17h/9xh devices from OpenBSD. Add AMD F17/Axh devices. Add some Xeon Scalable devices from OpenBSD. Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.1383.2.13	11-Oct-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1537): sys/arch/x86/pci/imcsmb/imc.c: revision 1.5 sys/dev/pci/pcidevs: revision 1.1461-1.1468 add several samsung nvme entries Add more Alder Lake devices. Jasper Lake Intel Trace Hub on Compute Die is not 0x4da6 but 0x4e29. Add Intel Core 8G (8core, H, Halo) Host Bridge, DRAM. Sort by number. No functional change. Add AMD 19h/6xh Root Complex. Add AMD FCH SATA Controller D add NVIDIA GeForce GTX 770 Fix compile error.
1.1383.2.12	27-Jul-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1482: sys/dev/pci/pcidevs 1.1422,1.1445-1.1460 via patch sys/arch/x86/pci/amdsmn.c 1.13 Update pcidevs: - Add Intel Alder Lake devices and Intel 600 Series PCH devices. - Add some Intel Xeon Scalable / Skylake-E devices. - Fix AMD F16_HB from 0x1568 to 0x1538. - Add some devices for AMD and improve some descriptions to clarify. - Add VMware AHCI and NVMe. - Update Intel 700 series Ethernet devices. - Add some Broadcom devices. - Add some Broadcom / LSI RAID cards. - Fix typos and whitespace.
1.1383.2.11	03-Dec-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1380: sys/dev/pci/pcidevs 1.1441-1.1444 - Add Intel Gemini Lake TXE HECI 1. - Add Intel Elkhart Lake and Rocket Lake devices. - Update Jasper Lake's Processor Transaction Routers.
1.1383.2.10	21-Oct-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1362: sys/dev/pci/pcidevs 1.1419-1.1440 - Add newer ciss(4) devices. - Add modern QUMRANET/Red Hat VIRTIO range PCI devices. - Add some more product IDs for mcx(4). - Add current generation NVIDIA graphics cards (3050-3090 etc.). - Add RDC R6022 PCI-Host bridge. - Add a whole bunch of radeon devices. - Add Realtek RTL8821CE. - Add Intel 660p SSD, and expand the 760p description. - Add Intel I219's version number. - Add Intel I219V 15-19 and I219LM 16-19. - Add Intel I225V, I225LM and WiFi 6 AX201. - Add newer Intel PCH internal devices. - Add Intel Jasper Lake devices. - PDC20265 is Ultra/100, not 66. - Fix Realtek RTL8125 description. - Sort Cavium devices.
1.1383.2.9	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #991: sys/dev/pci/pcidevs 1.1403-1.1407, 1.1409, 1.1412-1.1418 via patch - Add NVIDIA Quadro NVS 295. - Add more RDC products from Andrius V. - Add some Intel UHD Graphics devices. Mainly taken from OpenBSD. - Add Intel Comet Lake, Whiskey Lake U and Amber Lake Y devices. - Modify description of Intel 0x591e from HD Graphics to UHD Graphics. - Add Intel XMM 7360 LTE Modem. - Add Western Digital WD Blue SN550 NVMe SSD. - Add ATI Radeon R5/R6/R7 Graphics. - Add IDs for Ampere eMAG PCIe Root Ports. - Add RTL8192EE Wireless LAN 802.11n PCI-E NIC. - Add ASIX AX99100 Multi I/O (Serial,Parallel,I2C,SPI,LocalBus,GPIO) Controller. - Add a couple of additional device IDs for the AMD Cryptographic Coprocessor. - Remove duplicated entries.
1.1383.2.8	07-Jul-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #980): sys/dev/pci/if_aq.c: revision 1.4 sys/dev/pci/if_aq.c: revision 1.5 sys/arch/amd64/conf/GENERIC: revision 1.553 sys/dev/pci/files.pci: revision 1.419 sys/arch/amd64/conf/XEN3_DOM0: revision 1.170 sys/dev/pci/if_aq.c: revision 1.9 share/man/man4/Makefile: revision 1.693 sys/dev/pci/pcidevs: revision 1.1411 share/man/man4/aq.4: revision 1.1 share/man/man4/aq.4: revision 1.3 sys/arch/i386/conf/ALL: revision 1.479 share/man/man4/aq.4: revision 1.4 sys/dev/pci/if_aq.c: revision 1.10 sys/dev/pci/files.pci: revision 1.421 sys/dev/pci/if_aq.c: revision 1.11 sys/dev/pci/if_aq.c: revision 1.12 sys/dev/pci/if_aq.c: revision 1.13 sys/dev/pci/if_aq.c: revision 1.14 sys/dev/pci/if_aq.c: revision 1.15 sys/dev/pci/if_aq.c: revision 1.16 sys/dev/pci/pcidevs: revision 1.1408 sys/arch/amd64/conf/ALL: revision 1.135 sys/net/ethertypes.h: revision 1.19 sys/arch/i386/conf/GENERIC: revision 1.1218 distrib/sets/lists/man/mi: revision 1.1668 sys/dev/pci/if_aq.c: revision 1.1 sys/dev/pci/if_aq.c: revision 1.2 sys/dev/pci/pcidevs: revision 1.1395 sys/dev/pci/if_aq.c: revision 1.3 sys/arch/evbarm/conf/GENERIC64: revision 1.125 Add the ETHERTYPE_QINQ for 802.1ad VLAN stacking add Aquantia AQC 10G network adapters add support Aquantia AQC seriese 10G network adapters. this driver is based on the FreeBSD version https://github.com/Aquantia/aqtion-freebsd , but drastically rewritten for NetBSD. add aq(4) Add Aquantia AQC100, AQC100S and D100. add support VLAN HW filter set/clear IFF_OACTIVE flag only on txring 0 make counters per queue support internal PHY temperature sensor Found by kUBSan: - Use unsigned to avoid undefined behavior in aq_hw_init(). - Cast to unsigned to avoid undefined behavior in aq_set_mac_addr(). fix descriptions of register map in comment return the ifmedia active status correctly even while the link is not up after attach. pointed out by msaitoh@. thanks. On FIBRE devices, there are times when linkstat interrupt doesn't occur? reported from Andrius V. thanks. - use polling instead of linkstat interrupt when FIBRE - add AQ_FORCE_POLL_LINKSTAT options (not by default) sort product table, and tabify add support AQC100S and D100. not tested, but they are probably the same as the AQC100.
1.1383.2.7	28-Apr-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #864): share/man/man4/rtsx.4: revision 1.6 sys/dev/pci/pcidevs: revision 1.1410 sys/dev/pci/rtsx_pci.c: revision 1.9 Add an eclectic mix of GeForce GTX 680, RTS522A PCI-E Card Reader, intel wifi 9560 + whitespace Add support for Realtek RTS522A
1.1383.2.6	19-Mar-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #784): sys/dev/pci/pcidevs: revision 1.1398 sys/dev/pci/pcidevs: revision 1.1399 sys/dev/pci/pcidevs: revision 1.1400 sys/dev/pci/pcidevs: revision 1.1401 sys/dev/pci/pcidevs: revision 1.1402 add a couple of GPUs i have: product ATI RADEON_HD4290 0x9714 Radeon HD4290 Graphics product NVIDIA GF_440 0x0de0 GeForce GT 440 Add Farallon PN9000SX Ethernet. NVIDIA 0x036[0-7] are nForce MCP55 LPC Bridge. Add device IDs of the VIA VX900 chipset. Add some GeForce devices.
1.1383.2.5	31-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #670): sys/dev/pci/pcidevs: revision 1.1396 sys/dev/pci/pcidevs: revision 1.1397 sys/dev/pci/pcidevs: revision 1.1392 sys/dev/pci/pcidevs: revision 1.1393 sys/dev/pci/pcidevs: revision 1.1394 sys/dev/pci/pcidevs: revision 1.1395 - Add Xeon D-1500 NTB-secondary - Add Xeon D NS QuickData DMA channel 0-7 Add AMD Family14h PCIe. Killer E3000 from OpenBSD add Aquantia AQC 10G network adapters Add Intel I219 LM10-LM15 and V10-V14 from OpenBSD. Remove duplicated entries.
1.1383.2.4	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #449): sys/dev/pci/pcidevs: revision 1.1388 sys/dev/pci/pcidevs: revision 1.1389 sys/dev/pci/pcidevs: revision 1.1390 sys/dev/pci/pcidevs: revision 1.1391 Spell controller correctly Add Realtek RTL8125. - Update Intel's NVMe SSDs. - Modify 0x0953's description to "750 or DC P3[567]00 SSD" - Add DC P4[56]00 - Add Apollo Lake TXE HECI. Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.1383.2.3	24-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #373): sys/dev/pci/pcidevs: revision 1.1386 sys/dev/pci/pcidevs: revision 1.1387 Modify Attansic Ethernet devices' description to clarify. Add Killer E2400 and E2500.
1.1383.2.2	24-Sep-2019	martin	Pull up following revision(s) (requested by jmcneill in ticket #237): sys/dev/pci/pcidevs: revision 1.1385 Add Mellanox ConnectX-4, ConnectX-4 Lx, ConnectX-5, and ConnectX-5 Ex
1.1383.2.1	26-Aug-2019	martin	Pull up following revision(s) (requested by mrg in ticket #112): sys/dev/pci/pcidevs: revision 1.1384 add micron/crucial SM2263 nvme. add some device found on asus x570-p with ryzen 3200G cpu. spell it "PCIe' when using the name.
1.1395.2.2	29-Feb-2020	ad	Sync with head.
1.1395.2.1	25-Jan-2020	ad	Sync with head.
1.1404.2.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.1404.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.1419.2.3	03-Apr-2021	thorpej	Sync with HEAD.
1.1419.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.1419.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.1424.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.1426.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.1471.2.17	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1184): sys/dev/pci/pcidevs: revision 1.1528 Add VIA VT8233C PCI-ISA Bridge.
1.1471.2.16	08-May-2025	martin	Additionally pull up sys/dev/pci/pcidevs 1.1520 accidently missed in previous for ticket #1101
1.1471.2.15	08-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1101): sys/dev/pci/pcidevs: revision 1.1521 sys/dev/pci/pcidevs: revision 1.1522 sys/dev/pci/pcidevs: revision 1.1518 sys/dev/pci/pcidevs: revision 1.1519 Various VIA IDE controller related additions/changes.
1.1471.2.14	06-Dec-2024	snj	Pull up following revision(s) (requested by msaitoh in ticket #1024): sys/dev/pci/pcidevs: 1.504-1.506, 1.508-1.513 Update pcidevs: - Add Zhaoxin devices, mainly from my TVS-675 NAS device, and some additional to fill the gaps from pci-ids database. - Add Phison PS5021 and PS5026. - Add some Intel Rapotr Lake graphics. - Add Realtek RTL8126. - Add the intel "arc" a-series of cards. - Add Areca ARC-1224. - Add some AMD devices. - Add Raspberry Pi5 Ethernet. - Add many Brainboxes devices. Reported in PR kern/55824 by Cameron Williams.
1.1471.2.13	24-Aug-2024	martin	Pull up following revision(s) (requested by jakllsch in ticket #805): sys/dev/pci/pcidevs: revision 1.1503 sys/dev/pci/if_re_pci.c: revision 1.53 Add (Realtek/)Killer E2600 Ethernet Add Killer E2600, adapted from PR 57969
1.1471.2.12	03-Jul-2024	martin	Pull up following revision(s) (requested by nia in ticket #730): sys/dev/pci/pcidevs: revision 1.1507 Add Brainboxes and devices From Cameron Williams in PR 58358.
1.1471.2.11	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #560): sys/dev/pci/pcidevs: revision 1.1499 sys/dev/pci/pcidevs: revision 1.1500 sys/dev/pci/pcidevs: revision 1.1501 sys/dev/pci/pcidevs: revision 1.1502 add some newer Areca boards in prep for updating driver Add Intel C26[26] eSPI. Update Intel Raptor Lake devices. Add Intel I219-{LM,V}(20,21) add some newer Areca boards in prep for updating driver Add Intel C26[26] eSPI. Update Intel Raptor Lake devices. Add Intel I219-{LM,V}(20,21)
1.1471.2.10	03-Nov-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #447): sys/dev/pci/pcidevs: revision 1.1498 sys/dev/pci/if_mcx.c: revision 1.26 share/man/man4/mcx.4: revision 1.3 pcidevs: Add Mellanox ConnectX-6 Lx. mcx(4): Add Mellanox ConnectX-6 Lx from OpenBSD.
1.1471.2.9	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #421): sys/dev/pci/pcidevs: revision 1.1497 sys/dev/pci/if_wm.c: revision 1.790 sys/dev/pci/if_wm.c: revision 1.791 sys/dev/pci/if_wmvar.h: revision 1.51 pcidevs: Add Intel I219{V,LM}({22,23}) wm(4): Add a new workaround for Tiger Lake and newer. - Define new WM_T_PCH_TGP and use it for Tiger Lake and newer. Note that we don't define WM_T_PCH_ADP because we have no any Alder Lake specific workaround yet. - Add new workaround for Tiger Lake (and newer) in wm_init_locked() to avoid packet loss. wm(4): Add I219{V,LM}({22,23}) devices (Raptor Lake).
1.1471.2.8	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #406): sys/dev/pci/pcidevs: revision 1.1493 sys/dev/pci/pcidevs: revision 1.1495 Add Intel 82599 LS. Add Pericom(Diodes) PCIe switches.
1.1471.2.7	13-Oct-2023	martin	Pull up following revision(s) (requested by abs in ticket #403): sys/dev/pci/pcidevs: revision 1.1496 Add entry for ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier
1.1471.2.6	08-Oct-2023	martin	Pull up the following, requested by rin in ticket #393: sys/dev/pci/pcidevs 1.1494 (via patch) Add missing I225/I226 series devices
1.1471.2.5	21-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #333): sys/dev/pci/pcidevs: revision 1.1483-1.1492 Add AMD 19h/7xh Root Complex. pcidevs: add ASMedia ASM106x Add Intel Snow Ridge devices. Fix some descriptions of Intel Snow Ridge. Add Intel I226 devices. Add device ids of the VX800 chipset and s3 chrome 500 series GPU. Descriptions mainly aligned with VX800/820 Series System Programming Manual. fix duplicate definition. Add some Intel Xeon D-2100 devices. Fix description of Intel 0x37d0. Simplify descriptions of Snow Ridge internal Ethernet devices. Rename PCI to PCIE for naming consistency.
1.1471.2.4	29-Jul-2023	martin	Pull up the following revisions, via patch, requested by msaitoh in ticket #252: sys/dev/pci/pcidevs 1.1481-1.1482 Add Intel Alder Lake-N, Raptor Lake and 700 series chipset devices.
1.1471.2.3	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #209): sys/dev/pci/pcidevs: revision 1.1478 sys/dev/pci/pcidevs: revision 1.1479 sys/dev/pci/pcidevs: revision 1.1480 Add Samsung SM990. Add devices from PPR for AMD Family 19h Model 61h Revision B1 processors. The SATA device ID for Apollo Lake is not 0x5ae0 but 0x5ae3.
1.1471.2.2	30-Jan-2023	martin	Pull up following revision(s) (requested by thorpej in ticket #61): sys/dev/pci/pcidevs: revision 1.1477 Add ID for the EXAR XR17V354 PCIe UART. PR kern/57202
1.1471.2.1	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #50): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.1507.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1527.2.1	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #71): sys/dev/pci/pcidevs: revision 1.1528 Add VIA VT8233C PCI-ISA Bridge.
1.1510	27-Oct-2025	andvar	regen.
1.1509	21-Oct-2025	pgoyette	regen
1.1508	21-Oct-2025	pgoyette	Regen for vendor MediaTek
1.1507	17-Oct-2025	andvar	regen.
1.1506	03-May-2025	rin	branches: 1.1506.2; pcidevs{,_data}.h: Regen for EXAR XR17V358
1.1505	01-May-2025	andvar	regen.
1.1504	01-May-2025	andvar	regen.
1.1503	28-Apr-2025	andvar	regen.
1.1502	28-Apr-2025	andvar	regen.
1.1501	20-Apr-2025	andvar	regen.
1.1500	05-Apr-2025	andvar	regen.
1.1499	03-Apr-2025	andvar	regen.
1.1498	03-Mar-2025	andvar	regen.
1.1497	23-Jan-2025	andvar	regen.
1.1496	23-Jan-2025	mrg	regen
1.1495	18-Jan-2025	mrg	regen.
1.1494	03-Jan-2025	mrg	regen.
1.1493	12-Dec-2024	msaitoh	regen.
1.1492	11-Nov-2024	msaitoh	Regen.
1.1491	19-Oct-2024	mlelstv	Regen
1.1490	16-Oct-2024	msaitoh	Regen.
1.1489	20-Aug-2024	dogcow	regen
1.1488	17-Jul-2024	mrg	regen after intel arc.
1.1487	11-Jul-2024	msaitoh	Regen.
1.1486	23-Jun-2024	jakllsch	branches: 1.1486.2; Regen. For PR 58358
1.1485	12-May-2024	msaitoh	Regen.
1.1484	06-Apr-2024	andvar	regen.
1.1483	28-Feb-2024	jakllsch	regen
1.1482	18-Jan-2024	msaitoh	regen.
1.1481	28-Dec-2023	msaitoh	pcidevs: Regen.
1.1480	13-Nov-2023	jnemeth	regen for newer Areca boards
1.1479	26-Oct-2023	msaitoh	pcidevs: regen.
1.1478	11-Oct-2023	msaitoh	Regen.
1.1477	10-Oct-2023	abs	Regenerate for ASMEDIA 0x0624
1.1476	08-Oct-2023	msaitoh	Regen.
1.1475	04-Oct-2023	rin	Regen for pcidevs rev 1.1494
1.1474	04-Oct-2023	msaitoh	Regen.
1.1473	10-Aug-2023	andvar	regen.
1.1472	09-Aug-2023	msaitoh	Regen.
1.1471	09-Aug-2023	msaitoh	Regen.
1.1470	05-Aug-2023	andvar	regen.
1.1469	05-Aug-2023	andvar	Regen.
1.1468	04-Aug-2023	msaitoh	Regen.
1.1467	04-Aug-2023	msaitoh	Regen.
1.1466	31-Jul-2023	tnn	pcidevs: regen
1.1465	28-Jul-2023	msaitoh	Regen.
1.1464	23-Jul-2023	msaitoh	Regen.
1.1463	22-Jul-2023	msaitoh	regen.
1.1462	22-Jul-2023	msaitoh	Add Alder Lake-N and Raptor Lake devices.
1.1461	06-Apr-2023	msaitoh	Regen.
1.1460	14-Feb-2023	msaitoh	Regen.
1.1459	03-Feb-2023	msaitoh	Regen.
1.1458	29-Jan-2023	thorpej	Regen for EXAR XR17V354. PR kern/57202
1.1457	18-Jan-2023	msaitoh	Regen.
1.1456	14-Jan-2023	ryo	regen
1.1455	04-Jan-2023	msaitoh	Regen.
1.1454	30-Dec-2022	msaitoh	Regen.
1.1453	28-Dec-2022	msaitoh	Regen.
1.1452	22-Nov-2022	msaitoh	branches: 1.1452.2; Regen.
1.1451	26-Oct-2022	msaitoh	Regen.
1.1450	26-Oct-2022	msaitoh	Regen.
1.1449	07-Oct-2022	snj	regen
1.1448	02-Oct-2022	martin	regen (AMD FCH SATA D added)
1.1447	01-Oct-2022	msaitoh	Regen.
1.1446	28-Sep-2022	msaitoh	Regen.
1.1445	13-Sep-2022	msaitoh	Regen.
1.1444	11-Sep-2022	mrg	regen
1.1443	27-Jun-2022	msaitoh	Regen.
1.1442	17-Jun-2022	msaitoh	Regen.
1.1441	02-Jun-2022	mrg	regen.
1.1440	11-May-2022	markd	Regen
1.1439	04-May-2022	martin	Regen
1.1438	04-May-2022	nia	regen
1.1437	27-Apr-2022	msaitoh	Regen.
1.1436	22-Apr-2022	skrll	Regen
1.1435	24-Mar-2022	yamaguchi	regen
1.1434	16-Mar-2022	yamaguchi	Regen.
1.1433	15-Feb-2022	msaitoh	Regen.
1.1432	02-Feb-2022	msaitoh	Regen.
1.1431	25-Jan-2022	msaitoh	Regen.
1.1430	01-Jan-2022	msaitoh	Regen.
1.1429	30-Nov-2021	msaitoh	Regen.
1.1428	03-Nov-2021	msaitoh	Regen.
1.1427	27-Oct-2021	msaitoh	Regen.
1.1426	27-Oct-2021	msaitoh	Regen.
1.1425	13-Oct-2021	msaitoh	Regen.
1.1424	12-Oct-2021	msaitoh	Regen.
1.1423	11-Oct-2021	msaitoh	Regen.
1.1422	02-Oct-2021	mrg	regen.
1.1421	23-Sep-2021	msaitoh	Regen.
1.1420	06-Sep-2021	mrg	regen.
1.1419	03-Sep-2021	mrg	regen.
1.1418	30-Aug-2021	andvar	regen.
1.1417	30-Aug-2021	mrg	regen.
1.1416	30-Aug-2021	mrg	regen.
1.1415	17-Jul-2021	mrg	regen.
1.1414	12-Jul-2021	msaitoh	Regen.
1.1413	29-Jun-2021	pgoyette	Regen for new sys/dev/devlist2h.awk
1.1412	21-Apr-2021	jakllsch	branches: 1.1412.2; regenerate
1.1411	02-Apr-2021	rin	regen
1.1410	30-Jan-2021	jmcneill	branches: 1.1410.2; regen
1.1409	29-Dec-2020	skrll	regen
1.1408	20-Nov-2020	msaitoh	Regen.
1.1407	19-Nov-2020	reinoud	Regenated from pcidevs
1.1406	14-Jul-2020	jdolecek	branches: 1.1406.2; regen
1.1405	09-Jul-2020	msaitoh	Regen.
1.1404	24-Jun-2020	thorpej	Regen.
1.1403	13-Jun-2020	ryo	regen
1.1402	11-Jun-2020	jdolecek	regen
1.1401	26-May-2020	martin	regen
1.1400	01-May-2020	jmcneill	regen
1.1399	23-Apr-2020	msaitoh	Regen.
1.1398	22-Apr-2020	christos	regen
1.1397	22-Apr-2020	prlw1	regen
1.1396	22-Apr-2020	msaitoh	Regen.
1.1395	18-Apr-2020	simonb	Regen.
1.1394	16-Apr-2020	msaitoh	Regen.
1.1393	10-Apr-2020	christos	regen
1.1392	07-Apr-2020	christos	branches: 1.1392.2; regen
1.1391	01-Apr-2020	msaitoh	Regen.
1.1390	11-Mar-2020	msaitoh	Regen.
1.1389	11-Mar-2020	fcambus	Regen.
1.1388	08-Mar-2020	msaitoh	Regen.
1.1387	05-Mar-2020	msaitoh	Regen.
1.1386	19-Feb-2020	mrg	regen.
1.1385	28-Jan-2020	msaitoh	Regen.
1.1384	21-Jan-2020	msaitoh	Regen.
1.1383	01-Jan-2020	ryo	branches: 1.1383.2; regen
1.1382	29-Dec-2019	sevan	regen
1.1381	18-Dec-2019	msaitoh	Regen.
1.1380	27-Nov-2019	msaitoh	Regen.
1.1379	14-Nov-2019	msaitoh	Regen.
1.1378	08-Nov-2019	msaitoh	Regen.
1.1377	29-Oct-2019	msaitoh	Regen.
1.1376	12-Oct-2019	skrll	regen
1.1375	21-Sep-2019	msaitoh	Regen.
1.1374	21-Sep-2019	msaitoh	Regen.
1.1373	24-Aug-2019	jmcneill	regen
1.1372	21-Aug-2019	mrg	regen for new micron/crucial/amd entries, and PCIe spelling.
1.1371	28-Jul-2019	mlelstv	branches: 1.1371.2; regen
1.1370	24-Jul-2019	msaitoh	Regen.
1.1369	18-Jul-2019	msaitoh	Regen.
1.1368	12-Jul-2019	msaitoh	Regen.
1.1367	12-Jul-2019	msaitoh	Regen.
1.1366	09-Jul-2019	msaitoh	Regen.
1.1365	05-Jul-2019	msaitoh	Regen.
1.1364	02-Jul-2019	msaitoh	Regen.
1.1363	04-Jun-2019	msaitoh	Regen.
1.1362	01-Jun-2019	jdolecek	regen
1.1361	30-May-2019	mlelstv	regen
1.1360	15-Apr-2019	msaitoh	Regen.
1.1359	15-Apr-2019	msaitoh	Regen.
1.1358	12-Apr-2019	msaitoh	Regen.
1.1357	11-Apr-2019	mrg	regen.
1.1356	08-Mar-2019	msaitoh	Regen.
1.1355	05-Mar-2019	msaitoh	Regen.
1.1354	20-Feb-2019	msaitoh	Regen.
1.1353	20-Feb-2019	msaitoh	Regen.
1.1352	14-Feb-2019	msaitoh	Regen.
1.1351	08-Feb-2019	msaitoh	Regen.
1.1350	06-Feb-2019	msaitoh	Regen.
1.1349	28-Jan-2019	msaitoh	Regen.
1.1348	26-Jan-2019	msaitoh	Regen.
1.1347	21-Jan-2019	msaitoh	Regen.
1.1346	26-Dec-2018	msaitoh	Regen.
1.1345	03-Dec-2018	bouyer	Regen for new Symbios devices
1.1344	29-Nov-2018	jmcneill	regen
1.1343	24-Nov-2018	skrll	Regen
1.1342	12-Nov-2018	msaitoh	Regen.
1.1341	08-Nov-2018	msaitoh	Regen.
1.1340	16-Oct-2018	jmcneill	regen
1.1339	15-Oct-2018	uwe	Adaptec RAID 2445 and 2805
1.1338	23-Aug-2018	msaitoh	Regen.
1.1337	30-Jul-2018	msaitoh	Regen.
1.1336	30-Jul-2018	msaitoh	Regen.
1.1335	06-Jul-2018	prlw1	regen
1.1334	20-Jun-2018	msaitoh	branches: 1.1334.2; Regen.
1.1333	14-Jun-2018	msaitoh	Regen.
1.1332	12-Jun-2018	jdolecek	regen
1.1331	09-Jun-2018	jdolecek	regen
1.1330	31-May-2018	mrg	regen.
1.1329	31-May-2018	msaitoh	Regen.
1.1328	28-May-2018	msaitoh	Regen.
1.1327	26-May-2018	jakllsch	regen
1.1326	24-May-2018	msaitoh	Regen.
1.1325	21-May-2018	jnemeth	regen
1.1324	03-May-2018	maya	regen
1.1323	28-Apr-2018	mlelstv	regen
1.1322	24-Apr-2018	maya	regen
1.1321	17-Apr-2018	nonaka	regen
1.1320	16-Apr-2018	msaitoh	Regen.
1.1319	09-Apr-2018	msaitoh	Regen.
1.1318	09-Apr-2018	msaitoh	Regen.
1.1317	04-Apr-2018	msaitoh	Regen.
1.1316	02-Apr-2018	msaitoh	Regen.
1.1315	31-Mar-2018	sevan	regen
1.1314	26-Mar-2018	msaitoh	Regen.
1.1313	20-Mar-2018	nonaka	regen
1.1312	16-Mar-2018	msaitoh	Regen.
1.1311	04-Mar-2018	jdolecek	branches: 1.1311.2; regen
1.1310	04-Mar-2018	jdolecek	regen
1.1309	02-Mar-2018	msaitoh	Regen.
1.1308	24-Feb-2018	mlelstv	Regen.
1.1307	22-Feb-2018	msaitoh	Regen.
1.1306	22-Feb-2018	msaitoh	Regen.
1.1305	22-Feb-2018	msaitoh	Regen.
1.1304	22-Feb-2018	msaitoh	Regen.
1.1303	21-Feb-2018	msaitoh	Regen.
1.1302	17-Feb-2018	sevan	regen
1.1301	16-Feb-2018	sevan	regen
1.1300	25-Jan-2018	sborrill	regen
1.1299	24-Jan-2018	sevan	regen
1.1298	21-Jan-2018	sevan	regen
1.1297	09-Jan-2018	msaitoh	Regen.
1.1296	09-Jan-2018	msaitoh	Regen.
1.1295	18-Dec-2017	msaitoh	Regen.
1.1294	09-Nov-2017	msaitoh	Regen.
1.1293	30-Oct-2017	sevan	regen
1.1292	20-Oct-2017	christos	regen
1.1291	15-Sep-2017	hikaru	Regen.
1.1290	17-Aug-2017	msaitoh	Regen.
1.1289	02-Aug-2017	msaitoh	regen.
1.1288	26-Jul-2017	msaitoh	Regen.
1.1287	22-Jul-2017	msaitoh	Regen.
1.1286	07-Jul-2017	msaitoh	Regen.
1.1285	04-Jul-2017	msaitoh	Regen.
1.1284	27-Jun-2017	wiz	regen (comment change only)
1.1283	17-Jun-2017	jdolecek	regen
1.1282	06-Jun-2017	msaitoh	Regen.
1.1281	23-May-2017	nonaka	branches: 1.1281.2; regen
1.1280	08-May-2017	msaitoh	- Add RICOH R5C576 SD Card Controller. - Add some SiS devices. Taken from OpenBSD. - Change SiS 0x6325 from 6325 AGP VGA to 650 AGP VGA.
1.1279	27-Apr-2017	msaitoh	regen.
1.1278	19-Apr-2017	msaitoh	branches: 1.1278.2; regen.
1.1277	13-Apr-2017	msaitoh	regen.
1.1276	31-Mar-2017	skrll	regen
1.1275	31-Mar-2017	msaitoh	regen.
1.1274	23-Mar-2017	msaitoh	regen.
1.1273	14-Mar-2017	nonaka	regen
1.1272	27-Feb-2017	msaitoh	regen.
1.1271	16-Jan-2017	msaitoh	regen.
1.1270	15-Jan-2017	msaitoh	branches: 1.1270.2; regen.
1.1269	15-Jan-2017	msaitoh	regen.
1.1268	10-Jan-2017	nonaka	regen
1.1267	27-Dec-2016	msaitoh	Regen.
1.1266	26-Dec-2016	msaitoh	Regen.
1.1265	01-Dec-2016	msaitoh	Regen.
1.1264	23-Nov-2016	christos	regen
1.1263	28-Oct-2016	msaitoh	regen.
1.1262	26-Oct-2016	msaitoh	regen.
1.1261	24-Oct-2016	msaitoh	Regen.
1.1260	20-Oct-2016	msaitoh	regen.
1.1259	19-Oct-2016	msaitoh	regen.
1.1258	04-Oct-2016	kiyohara	Regen.
1.1257	16-Sep-2016	msaitoh	regen.
1.1256	01-Sep-2016	msaitoh	regen.
1.1255	23-Aug-2016	msaitoh	regen.
1.1254	23-Aug-2016	msaitoh	regen.
1.1253	23-Aug-2016	msaitoh	regen.
1.1252	19-Aug-2016	msaitoh	regen.
1.1251	17-Aug-2016	msaitoh	regen.
1.1250	17-Aug-2016	msaitoh	regen.
1.1249	08-Aug-2016	msaitoh	regen.
1.1248	29-Jul-2016	nonaka	regen
1.1247	28-Jul-2016	nonaka	regen
1.1246	06-Jul-2016	christos	branches: 1.1246.2; regen
1.1245	27-Jun-2016	pgoyette	Regen
1.1244	01-Jun-2016	pgoyette	Regen
1.1243	30-May-2016	nonaka	regen
1.1242	27-Apr-2016	christos	regen
1.1241	26-Apr-2016	msaitoh	regen.
1.1240	12-Feb-2016	mrg	regen.
1.1239	01-Feb-2016	msaitoh	Regen.
1.1238	22-Jan-2016	msaitoh	Regen.
1.1237	01-Dec-2015	msaitoh	regen.
1.1236	01-Dec-2015	msaitoh	regen.
1.1235	16-Nov-2015	msaitoh	Regen.
1.1234	16-Nov-2015	msaitoh	Regen.
1.1233	06-Nov-2015	nonaka	regen
1.1232	22-Oct-2015	msaitoh	Regen.
1.1231	21-Oct-2015	nonaka	regen
1.1230	19-Oct-2015	tnn	regen
1.1229	28-Aug-2015	nonaka	regen
1.1228	27-Aug-2015	nonaka	regen
1.1227	26-Aug-2015	msaitoh	regen.
1.1226	23-Aug-2015	jakllsch	pcidevs .h regen
1.1225	14-Aug-2015	nonaka	regen
1.1224	14-Aug-2015	knakahara	regen
1.1223	13-Aug-2015	msaitoh	regen.
1.1222	26-Jun-2015	msaitoh	regen.
1.1221	02-Jun-2015	nonaka	regen
1.1220	16-May-2015	msaitoh	regen.
1.1219	13-May-2015	nonaka	regen
1.1218	11-May-2015	msaitoh	regen.
1.1217	09-May-2015	christos	regen
1.1216	04-May-2015	ryo	regen
1.1215	03-May-2015	msaitoh	regen.
1.1214	02-Apr-2015	tnn	regen
1.1213	20-Mar-2015	msaitoh	regen.
1.1212	20-Mar-2015	msaitoh	regen.
1.1211	18-Mar-2015	msaitoh	regen.
1.1210	04-Mar-2015	knakahara	regen.
1.1209	15-Feb-2015	msaitoh	regen.
1.1208	25-Jan-2015	msaitoh	Regen.
1.1207	18-Jan-2015	christos	regen
1.1206	16-Jan-2015	msaitoh	regen.
1.1205	14-Jan-2015	msaitoh	regen.
1.1204	31-Dec-2014	veego	Regen
1.1203	23-Dec-2014	msaitoh	Regen.
1.1202	17-Dec-2014	hikaru	Regen.
1.1201	15-Dec-2014	msaitoh	Regen.
1.1200	15-Dec-2014	msaitoh	regen.
1.1199	30-Oct-2014	nonaka	branches: 1.1199.2; regen
1.1198	29-Oct-2014	nonaka	regen
1.1197	02-Oct-2014	msaitoh	Regen.
1.1196	25-Sep-2014	msaitoh	Regen.
1.1195	21-Sep-2014	christos	regen
1.1194	16-Sep-2014	msaitoh	Regen.
1.1193	01-Sep-2014	msaitoh	Regen.
1.1192	30-Jul-2014	msaitoh	branches: 1.1192.2; regen.
1.1191	25-Jul-2014	msaitoh	Regen.
1.1190	08-Jul-2014	msaitoh	Regen.
1.1189	03-Jul-2014	msaitoh	Regen.
1.1188	02-Jul-2014	msaitoh	Regen.
1.1187	23-Jun-2014	msaitoh	Regen.
1.1186	09-Jun-2014	msaitoh	regen.
1.1185	30-May-2014	msaitoh	Regen.
1.1184	28-May-2014	msaitoh	Regen.
1.1183	28-May-2014	msaitoh	regen.
1.1182	25-May-2014	njoly	Regen for GeForce GT 640M.
1.1181	26-Apr-2014	mrg	regen.
1.1180	09-Apr-2014	msaitoh	regen
1.1179	09-Apr-2014	msaitoh	regen.
1.1178	07-Apr-2014	njoly	Regen for Intel Ivy Bridge PCI Express Root Port.
1.1177	15-Mar-2014	kiyohara	branches: 1.1177.2; Regen.
1.1176	14-Feb-2014	njoly	Regen for GeForce 210 High Definition Audio Controller.
1.1175	26-Jan-2014	msaitoh	Regen.
1.1174	21-Jan-2014	msaitoh	Regen.
1.1173	10-Jan-2014	msaitoh	regen.
1.1172	09-Jan-2014	christos	regen
1.1171	09-Jan-2014	msaitoh	regen.
1.1170	04-Jan-2014	msaitoh	regen.
1.1169	01-Jan-2014	msaitoh	regen.
1.1168	30-Dec-2013	msaitoh	regen.
1.1167	27-Dec-2013	msaitoh	regen.
1.1166	25-Dec-2013	msaitoh	regen.
1.1165	23-Dec-2013	msaitoh	regen.
1.1164	22-Dec-2013	nonaka	regen
1.1163	21-Dec-2013	msaitoh	regen.
1.1162	15-Dec-2013	pgoyette	Regen
1.1161	13-Dec-2013	pedro	regen
1.1160	12-Dec-2013	nonaka	regen
1.1159	08-Dec-2013	msaitoh	regen.
1.1158	13-Nov-2013	msaitoh	regen.
1.1157	07-Nov-2013	msaitoh	Regen.
1.1156	07-Nov-2013	msaitoh	Add Quark X1000 devices.
1.1155	07-Nov-2013	msaitoh	Regen.
1.1154	14-Sep-2013	soren	Regen.
1.1153	04-Sep-2013	christos	regen
1.1152	23-Aug-2013	christos	regen
1.1151	17-Jul-2013	soren	Regen.
1.1150	13-Jun-2013	matt	branches: 1.1150.2; 1.1150.4; Regen.
1.1149	05-Jun-2013	msaitoh	regen.
1.1148	04-Jun-2013	msaitoh	regen
1.1147	03-Jun-2013	msaitoh	Regen.
1.1146	20-Apr-2013	rkujawa	Regen PCI devs database.
1.1145	18-Apr-2013	msaitoh	regen.
1.1144	18-Apr-2013	msaitoh	regen.
1.1143	06-Mar-2013	msaitoh	regen.
1.1142	03-Mar-2013	msaitoh	regen.
1.1141	27-Feb-2013	msaitoh	regen.
1.1140	20-Feb-2013	msaitoh	regen.
1.1139	30-Jan-2013	msaitoh	regen.
1.1138	06-Dec-2012	msaitoh	Regen.
1.1137	05-Dec-2012	christos	Regen
1.1136	29-Nov-2012	msaitoh	regen.
1.1135	21-Oct-2012	matt	Regen.
1.1134	20-Oct-2012	matt	Regen.
1.1133	19-Oct-2012	prlw1	regen
1.1132	29-Sep-2012	gsutre	regen
1.1131	25-Sep-2012	gson	regen
1.1130	23-Sep-2012	chs	regen
1.1129	18-Sep-2012	matt	Regen.
1.1128	17-Sep-2012	matt	Regen.
1.1127	17-Sep-2012	tsutsui	Regen from pcidevs rev 1.1134: > Add Broadcom BCM57762 Gigabit Ethernet, per PR kern/46961.
1.1126	13-Sep-2012	nonaka	regen.
1.1125	10-Aug-2012	bouyer	branches: 1.1125.2; Regen: add some PCI AHCI controllers.
1.1124	07-Aug-2012	msaitoh	regen.
1.1123	06-Aug-2012	msaitoh	regen.
1.1122	05-Aug-2012	msaitoh	regen.
1.1121	05-Aug-2012	bouyer	Regen: add MegaRAID SAS2208
1.1120	27-Jul-2012	matt	Regen.
1.1119	26-Jul-2012	uwe	Regen for VirtualBox ids.
1.1118	25-Jul-2012	msaitoh	regen.
1.1117	23-Jul-2012	msaitoh	regen.
1.1116	22-Jul-2012	matt	Regen.
1.1115	20-Jul-2012	msaitoh	regen.
1.1114	19-Jul-2012	msaitoh	regen.
1.1113	12-Jul-2012	kiyohara	Regen.
1.1112	26-Jun-2012	drochner	regen
1.1111	25-May-2012	msaitoh	regen for Intel I350 (wm)
1.1110	18-Apr-2012	bouyer	Regen: Add various newer LSI RAID/SAS controllers
1.1109	13-Apr-2012	cegger	regen.
1.1108	09-Apr-2012	mlelstv	regen
1.1107	08-Apr-2012	jakllsch	regen.
1.1106	08-Apr-2012	jakllsch	regen.
1.1105	07-Apr-2012	christos	regen
1.1104	04-Apr-2012	uwe	Regen for Marvell Yukon 88e8040 ethernet.
1.1103	28-Mar-2012	pgoyette	Regen for JMB362
1.1102	21-Mar-2012	sborrill	Regen
1.1101	18-Mar-2012	tsutsui	Regen from pcidevs rev 1.1106: > Add VIA VX900 media system processor IDE controller. > PR/46166 from Hiroshi Tokuda.
1.1100	03-Mar-2012	nonaka	regen.
1.1099	02-Mar-2012	nonaka	regen.
1.1098	02-Mar-2012	nonaka	regen.
1.1097	15-Feb-2012	tsutsui	branches: 1.1097.2; Regen from pcidevs rev 1.1102: > Add VIA VX900 Host Bridge, to handle its quirk. PR/46018 > Ok releng@
1.1096	18-Oct-2011	jmcneill	branches: 1.1096.2; 1.1096.6; regen
1.1095	06-Oct-2011	jmcneill	regen
1.1094	04-Oct-2011	njoly	Regen for 3WARE 9750 Series RAID.
1.1093	23-Sep-2011	njoly	Regen for NVIDIA GeForce 210.
1.1092	05-Sep-2011	msaitoh	regen.
1.1091	02-Sep-2011	msaitoh	regen.
1.1090	15-Aug-2011	msaitoh	regen.
1.1089	11-Aug-2011	bouyer	Regen: add SiS 315 Pro VGA
1.1088	09-Aug-2011	mbalmer	re-gen after PWDOG1 addition.
1.1087	06-Aug-2011	jakllsch	regen
1.1086	06-Aug-2011	mbalmer	re-gen.
1.1085	04-Aug-2011	jakllsch	regen
1.1084	23-Jul-2011	matt	Regen.
1.1083	28-Jun-2011	sborrill	Regen
1.1082	27-Jun-2011	nonaka	regen.
1.1081	07-Jun-2011	cegger	regen.
1.1080	06-Jun-2011	msaitoh	regen.
1.1079	28-May-2011	phx	branches: 1.1079.2; regen
1.1078	28-May-2011	phx	regen
1.1077	28-May-2011	ryo	Regen.
1.1076	27-May-2011	matt	Regen.
1.1075	22-May-2011	mrg	regenerate.
1.1074	21-May-2011	msaitoh	regen
1.1073	20-May-2011	msaitoh	regen.
1.1072	15-May-2011	msaitoh	regen.
1.1071	14-May-2011	martin	Regen
1.1070	13-May-2011	msaitoh	regen.
1.1069	02-May-2011	manu	Add support for new SUNIX PCI seriel and/or parallel adapters, with new SUNIX vendor Id 0x1fd4 and product Id 0x1999. Constants were taken from Linux open source driver distributed by SUNIX. The following models should be supported now: 5008 1 parallel port 5016 16 serial ports, only 8 supported right now (PUC_MAX_PORTS limit) 5027 1 serial port 5037 2 serial ports 5056 4 serial ports 5066 8 serial ports, this model is the only one that was actually tested. 5069 1 serial port and 1 parallel port 5079 2 serial ports and 1 parallel port 5099 4 serial ports and 1 parallel port
1.1068	13-Apr-2011	wiz	regen (+ ATI Radeon HD4650)
1.1067	04-Apr-2011	bouyer	Regen: Add RDC PCI-ISA bridge
1.1066	04-Apr-2011	dyoung	Regen.
1.1065	16-Mar-2011	matt	regen.
1.1064	05-Mar-2011	matt	Regen.
1.1063	25-Feb-2011	jmcneill	regen
1.1062	25-Feb-2011	jmcneill	regen
1.1061	24-Feb-2011	matt	Regen.
1.1060	24-Feb-2011	matt	Regen.
1.1059	23-Feb-2011	jmcneill	regen
1.1058	22-Feb-2011	jmcneill	regen
1.1057	21-Feb-2011	njoly	Regen: Add two more Intel 82Q45 devices.
1.1056	20-Feb-2011	msaitoh	regen.
1.1055	18-Feb-2011	jmcneill	regen
1.1054	03-Feb-2011	njoly	Regen for Nvidia devices addition.
1.1053	26-Jan-2011	bouyer	Regen: Add RDC Semiconductor R6040 10/100 Ethernet
1.1052	11-Jan-2011	dyoung	branches: 1.1052.2; 1.1052.4; Regen.
1.1051	16-Dec-2010	wiz	regen.
1.1050	11-Dec-2010	matt	Regen
1.1049	11-Nov-2010	cegger	regen.
1.1048	06-Nov-2010	jakllsch	regen.
1.1047	14-Oct-2010	taca	Regen.
1.1046	26-Aug-2010	dogcow	Regen for BCM43xx update.
1.1045	01-Aug-2010	msaitoh	regen.
1.1044	30-Jul-2010	njoly	Regen for ICH10 devices update.
1.1043	29-Jul-2010	njoly	Regen for Intel LPC updates.
1.1042	29-Jul-2010	njoly	Regen for Matrox MGA G200eW.
1.1041	28-Jul-2010	msaitoh	regen.
1.1040	13-Jul-2010	msaitoh	regen.
1.1039	13-Jul-2010	kiyohara	Regen.
1.1038	08-Jul-2010	msaitoh	regen.
1.1037	08-Jul-2010	msaitoh	regen
1.1036	07-Jul-2010	msaitoh	regen.
1.1035	06-Jul-2010	msaitoh	regen.
1.1034	06-Jul-2010	msaitoh	regen.
1.1033	02-Jul-2010	mrg	regen.
1.1032	16-Jun-2010	riz	Regen for added Intel IDs.
1.1031	08-Jun-2010	pgoyette	Regen
1.1030	08-Jun-2010	mrg	regenerate
1.1029	15-May-2010	jakllsch	Regen.
1.1028	12-May-2010	christos	regen
1.1027	10-May-2010	matt	Regen.
1.1026	08-May-2010	pgoyette	Regen
1.1025	07-May-2010	kiyohara	Regen.
1.1024	17-Apr-2010	christos	regen
1.1023	02-Apr-2010	jakllsch	regen
1.1022	14-Mar-2010	mrg	branches: 1.1022.2; regenerate
1.1021	25-Feb-2010	macallan	regen
1.1020	16-Feb-2010	msaitoh	regen.
1.1019	09-Feb-2010	hubertf	regen
1.1018	09-Feb-2010	msaitoh	regen.
1.1017	08-Feb-2010	msaitoh	regen.
1.1016	03-Feb-2010	macallan	branches: 1.1016.2; regen
1.1015	30-Jan-2010	jakllsch	regen.
1.1014	24-Jan-2010	msaitoh	regen.
1.1013	22-Jan-2010	hubertf	regen - thanks for the reminder @ veego
1.1012	19-Jan-2010	mbalmer	regenerate
1.1011	14-Jan-2010	sborrill	Regen
1.1010	12-Jan-2010	msaitoh	regen
1.1009	12-Jan-2010	nonaka	regen (Added RENESAS SH7780/SH7785 PCI Controller.)
1.1008	11-Jan-2010	msaitoh	regen
1.1007	07-Jan-2010	msaitoh	regen
1.1006	07-Jan-2010	jdc	Regenerate for: Add Sun Cassini/NS Saturn devices.
1.1005	03-Jan-2010	jnemeth	regen
1.1004	29-Dec-2009	msaitoh	regen
1.1003	30-Nov-2009	sborrill	Regen
1.1002	21-Nov-2009	jakllsch	regen.
1.1001	18-Nov-2009	bouyer	Regen: Complete BROADCOM BCM5709 family, and add BCM5716 family.
1.1000	18-Nov-2009	jakllsch	regen.
1.999	13-Oct-2009	cegger	regen.
1.998	13-Oct-2009	cegger	regen.
1.997	05-Oct-2009	cegger	regen.
1.996	20-Sep-2009	christos	Regen
1.995	01-Aug-2009	matt	Regen.
1.994	07-Jul-2009	cegger	regen.
1.993	29-Jun-2009	hubertf	regen
1.992	12-Jun-2009	rjs	Regen.
1.991	11-Jun-2009	cegger	regen.
1.990	29-May-2009	skrll	Regen.
1.989	18-May-2009	skrll	Regen
1.988	13-May-2009	cegger	regen.
1.987	09-May-2009	sborrill	regen
1.986	30-Apr-2009	skrll	Regen (again)
1.985	30-Apr-2009	skrll	Regen.
1.984	28-Apr-2009	ginsbach	Regen for Mellanox additions.
1.983	25-Apr-2009	tsutsui	Regen from pcidevs rev 1.983: > Tweak some Realtek re(4) device names.
1.982	07-Apr-2009	msaitoh	regen
1.981	10-Mar-2009	msaitoh	regen
1.980	26-Feb-2009	mjf	Regen.
1.979	19-Feb-2009	sborrill	Regen
1.978	11-Feb-2009	markd	regen
1.977	06-Feb-2009	jakllsch	branches: 1.977.2; regen.
1.976	27-Jan-2009	markd	regen
1.975	19-Jan-2009	jmcneill	regen for Attansic L1E Gigabit Ethernet Adapter
1.974	07-Jan-2009	jnemeth	regen
1.973	27-Dec-2008	tsutsui	Regen from pcidevs rev 1.971: > Fix tyop in previous.
1.972	27-Dec-2008	jnemeth	regen for et (Agere) driver
1.971	21-Dec-2008	nonaka	Regen.
1.970	16-Dec-2008	christos	regen
1.969	29-Nov-2008	christos	regen
1.968	18-Nov-2008	matt	Regen.
1.967	17-Nov-2008	matt	Regen.
1.966	08-Nov-2008	christos	regen.
1.965	04-Nov-2008	cegger	regen.
1.964	04-Nov-2008	cegger	regen.
1.963	15-Oct-2008	bouyer	branches: 1.963.2; 1.963.4; Regen: fix entry description.
1.962	15-Oct-2008	bouyer	Regen: add more ICH9 ethernet devices.
1.961	01-Oct-2008	christos	regen
1.960	01-Oct-2008	christos	ICH10 support from Brad du Plessis
1.959	28-Sep-2008	bouyer	Regen for JMicron Ethernet Controllers.
1.958	18-Sep-2008	cegger	regen.
1.957	18-Sep-2008	cegger	regen.
1.956	13-Sep-2008	njoly	Regen for Nvidia GeForce 6150 devices addition.
1.955	08-Sep-2008	cegger	regen.
1.954	08-Sep-2008	cegger	regen.
1.953	06-Sep-2008	rmind	Regen.
1.952	05-Sep-2008	cegger	regen.
1.951	27-Aug-2008	christos	regen
1.950	23-Aug-2008	tsutsui	Regen from pcidevs rev 1.949: > - add a Realtek 8139D variant (from FreeBSD) > - remove a duplicated entry of Realtek 0x8138
1.949	22-Aug-2008	tnn	regen (Intel 945GME)
1.948	19-Aug-2008	matthias	Regen.
1.947	19-Aug-2008	matthias	Regen.
1.946	06-Aug-2008	dyoung	Regen.
1.945	29-Jul-2008	tron	Regen.
1.944	15-Jul-2008	tonnerre	Regen.
1.943	25-May-2008	chs	branches: 1.943.2; 1.943.4; regen
1.942	20-May-2008	cegger	regen.
1.941	13-May-2008	rjs	Regen.
1.940	08-May-2008	sborrill	Regen
1.939	30-Apr-2008	joerg	branches: 1.939.2; Regen.
1.938	17-Apr-2008	christos	branches: 1.938.2; 1.938.4; regen
1.937	15-Apr-2008	cegger	regen.
1.936	09-Apr-2008	joerg	Regen.
1.935	03-Apr-2008	riz	Regen.
1.934	01-Apr-2008	cegger	regen.
1.933	26-Mar-2008	jmcneill	Regen for PR#38301
1.932	24-Feb-2008	xtraeme	regen
1.931	22-Feb-2008	tsutsui	Regen from pcidevs rev 1.930: > Add I-O DATA RSA-PCI 2 port serial board.
1.930	13-Feb-2008	tron	branches: 1.930.2; 1.930.6; Regen.
1.929	10-Feb-2008	ichiro	regen
1.928	05-Feb-2008	ichiro	regen
1.927	11-Jan-2008	tnn	regen (add ServeRAID 6/7)
1.926	11-Jan-2008	ragge	Regen
1.925	03-Jan-2008	jnemeth	regen -- Symbios Logic 0x0030 is used for both 53c1020/53c1030 from Dave Huang
1.924	30-Dec-2007	jdc	Regenerate for "Add Apple Shasta GMAC (From FreeBSD)".
1.923	28-Dec-2007	dholland	Regen.
1.922	24-Dec-2007	hamajima	regen.
1.921	09-Dec-2007	jmcneill	branches: 1.921.2; Merge jmcneill-pm branch.
1.920	26-Nov-2007	bjs	branches: 1.920.2; 1.920.4; Add PCI IDs for the ATI Radeon X800 GTO (R430). Most sources I have consulted indicated that this ID is an R430 (including xf86-video-ati). PCI_PRODUCT_ATI_RADEON_R430_554F (primary) PCI_PRODUCT_ATI_RADEON_R430_556F (secondary) vga0 at pci1 dev 0 function 0: ATI Technologies product 0x554f (rev. 0x00) ATI Technologies product 0x556f (miscellaneous display) at pci1 dev 0 function 1 not configured
1.919	24-Nov-2007	markd	regen.
1.918	23-Nov-2007	xtraeme	regen
1.917	19-Nov-2007	ad	Regen.
1.916	15-Nov-2007	manu	Regen
1.915	14-Nov-2007	xtraeme	regen
1.914	12-Nov-2007	joerg	Regen.
1.913	30-Oct-2007	xtraeme	branches: 1.913.2; regen
1.912	30-Oct-2007	xtraeme	regen
1.911	30-Oct-2007	xtraeme	regen
1.910	30-Oct-2007	jnemeth	regen
1.909	24-Oct-2007	xtraeme	regen
1.908	22-Oct-2007	manu	Regen
1.907	22-Oct-2007	manu	Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.906	22-Oct-2007	nisimura	regen
1.905	14-Oct-2007	xtraeme	branches: 1.905.2; regen
1.904	09-Oct-2007	garbled	Add a bunch of IBM PCI dev id's for graphics cards and PCI bridges
1.903	20-Sep-2007	xtraeme	branches: 1.903.2; regen
1.902	17-Sep-2007	msaitoh	regen
1.901	11-Sep-2007	jnemeth	regen
1.900	09-Sep-2007	xtraeme	regen
1.899	08-Sep-2007	ober	regen. added Intel Pro/Wifi 4965AGN
1.898	07-Sep-2007	cube	Regen.
1.897	02-Sep-2007	ober	regen. added Intel LPC 82801HEM and 82801IO
1.896	01-Sep-2007	ober	* empty log message *
1.895	31-Aug-2007	xtraeme	regen
1.894	30-Aug-2007	xtraeme	regen
1.893	28-Aug-2007	msaitoh	regen
1.892	26-Aug-2007	xtraeme	branches: 1.892.2; regen
1.891	19-Aug-2007	riz	Regen.
1.890	14-Aug-2007	jnemeth	regen
1.889	11-Aug-2007	jnemeth	regem
1.888	08-Aug-2007	tron	Regen.
1.887	06-Aug-2007	martin	branches: 1.887.2; regen
1.886	04-Aug-2007	kiyohara	Regen.
1.885	07-Jul-2007	jklos	branches: 1.885.2; 1.885.6; Added Chelsio vendor and various Chelsio hardware.
1.884	06-Jul-2007	hpeyerl	regen
1.883	27-Jun-2007	xtraeme	regen
1.882	26-Jun-2007	xtraeme	regen
1.881	28-May-2007	mrg	regenerate.
1.880	19-May-2007	tsutsui	Regen from pcidevs rev 1.880: > Add IBM GXT250P video.
1.879	16-May-2007	mrg	regenerate
1.878	02-May-2007	bouyer	Regen: added JMicron Technology products
1.877	29-Apr-2007	bouyer	Regen: remove extra "Intel" from previous commit.
1.876	29-Apr-2007	bouyer	Regen: add Intel i82801H IFE G/GT LAN Controller
1.875	27-Mar-2007	pooka	regen
1.874	25-Mar-2007	isaki	Regen; ATI ES1000.
1.873	17-Mar-2007	dyoung	Regen.
1.872	10-Mar-2007	christos	branches: 1.872.2; 1.872.4; regen
1.871	09-Mar-2007	xtraeme	branches: 1.871.2; regen
1.870	03-Mar-2007	jmcneill	Regen.
1.869	01-Mar-2007	jmcneill	Regen.
1.868	16-Feb-2007	markd	branches: 1.868.2; regen for i965Q
1.867	10-Feb-2007	mlelstv	Get the correct RCS-Id pulled in
1.866	10-Feb-2007	mlelstv	Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.865	30-Jan-2007	msaitoh	regen
1.864	28-Jan-2007	chs	regen
1.863	27-Jan-2007	isaki	Regen.
1.862	21-Jan-2007	isaki	Regen; AMD Geode LX and CS5536.
1.861	14-Jan-2007	rpaulo	regen
1.860	06-Jan-2007	jmcneill	Regen.
1.859	04-Jan-2007	jmcneill	Regen.
1.858	04-Jan-2007	jmcneill	Regen.
1.857	31-Dec-2006	ichiro	regen
1.856	10-Dec-2006	bouyer	Regen: another PCI id for Dell PERC 5
1.855	09-Dec-2006	manu	Add support for 4 and 8 ports modem boards from Perle systems
1.854	26-Nov-2006	tsutsui	branches: 1.854.2; Regen from pcidevs rev 1.855: > - add more Broadcom product IDs, and fix two alternative IDs > (from FreeBSD and OpenBSD) > - sort by product ID
1.853	26-Nov-2006	enami	Regen.
1.852	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.853: > Add an ID of Broadcom BCM5752M. From Matt Kolb on tech-kern.
1.851	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.852: > - add Realtek RTL8100 (from FreeBSD) > - fix entry of RTL8169SC > - remove vague comment in Realtek entry
1.850	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.851: > Add and note more Realtek Ethernet chips.
1.849	13-Nov-2006	xtraeme	regen
1.848	13-Nov-2006	bouyer	Regen: add NetMos 9855 Quad UART and 1284 Printer port
1.847	08-Nov-2006	rpaulo	regen
1.846	08-Nov-2006	martti	Regenerated.
1.845	07-Nov-2006	martti	Regenerated.
1.844	03-Nov-2006	abs	Regenerate after adding Broadcom BCM5754 (0x167a)
1.843	31-Oct-2006	manu	regen
1.842	25-Oct-2006	bouyer	Regen: Add VIA VT8237A Integrated SATA Controller
1.841	23-Oct-2006	rjs	Regen.
1.840	21-Oct-2006	bouyer	Regen: added some Intel gigabit ethernet products
1.839	09-Oct-2006	cube	Regen.
1.838	02-Oct-2006	cube	Regen (fix pasto).
1.837	02-Oct-2006	cube	Regen (MCP55 ISA brige).
1.836	29-Sep-2006	cube	Regen.
1.835	27-Sep-2006	cube	Regen (Intel/0x1094).
1.834	27-Sep-2006	sketch	Regenerate for TI ACX changes.
1.833	26-Sep-2006	heas	Regen for pcidevs 1.833
1.832	24-Sep-2006	briggs	Regen with new IDs for devices in the last-generation Apple PowerBook G4.
1.831	07-Sep-2006	riz	branches: 1.831.2; Regen.
1.830	07-Sep-2006	itohy	branches: 1.830.2; Regen from pcidevs rev 1.830.
1.829	03-Sep-2006	xtraeme	regen
1.828	03-Sep-2006	xtraeme	regen
1.827	03-Sep-2006	xtraeme	regen
1.826	03-Sep-2006	xtraeme	regen
1.825	03-Sep-2006	xtraeme	regen
1.824	03-Sep-2006	xtraeme	regen
1.823	27-Aug-2006	christos	Regen
1.822	24-Aug-2006	riz	Regen (Marvell product IDs)
1.821	23-Aug-2006	cube	Regen (Broadcom NetXtreme II).
1.820	22-Aug-2006	riz	Regen for GALILEO -> MARVELL
1.819	13-Aug-2006	dsainty	Regen: Add the Promise PDC20771, present on the Promise FastTrak TX2300
1.818	10-Aug-2006	simonb	Regen.
1.817	09-Aug-2006	jnemeth	regen for Token Ring additions
1.816	08-Aug-2006	cube	Regen (Intel 63xxESB devices).
1.815	06-Aug-2006	jmcneill	branches: 1.815.2; Regen.
1.814	03-Aug-2006	bsh	regen
1.813	01-Aug-2006	xtraeme	regen
1.812	01-Aug-2006	bsh	regen
1.811	30-Jul-2006	jmcneill	regen
1.810	30-Jul-2006	oster	Regen
1.809	30-Jul-2006	simonb	Regen.
1.808	21-Jul-2006	gdamore	Regen.
1.807	21-Jul-2006	gdamore	Update ADMtek pci device ids to match data sheets. While here, add the 9513 device (although I've never seen one, I have a datasheet for it. :-)
1.806	18-Jul-2006	itohy	Regen from pcidevs rev 1.807.
1.805	13-Jul-2006	kent	regen for VT82xx HDA
1.804	07-Jul-2006	simonb	Regen.
1.803	30-Jun-2006	nisimura	Regen after Micrel KSZ8842/8841 product addtion.
1.802	30-Jun-2006	xtraeme	regen
1.801	30-Jun-2006	xtraeme	regen
1.800	27-Jun-2006	kent	regen for NVIDA MCP55_HDA
1.799	26-Jun-2006	xtraeme	regen
1.798	26-Jun-2006	xtraeme	regen
1.797	25-Jun-2006	soren	Regen.
1.796	18-Jun-2006	christos	Regen
1.795	18-Jun-2006	xtraeme	regen
1.794	15-Jun-2006	heinz	Regen (LSI SAS, parallel SCSI and FC devices).
1.793	14-Jun-2006	grant	branches: 1.793.2; regen.
1.792	06-Jun-2006	rpaulo	regen
1.791	05-Jun-2006	martin	Regen (DWL-610 added)
1.790	05-Jun-2006	msaitoh	regen
1.789	31-May-2006	riz	Regen.
1.788	19-May-2006	markd	branches: 1.788.2; regen
1.787	30-Apr-2006	xtraeme	regen
1.786	30-Apr-2006	xtraeme	regen
1.785	29-Apr-2006	gdamore	Regen.
1.784	25-Apr-2006	wrstuden	Regen to catch change for 3ware cards.
1.783	20-Apr-2006	gdamore	Regen.
1.782	15-Apr-2006	jmcneill	Regen.
1.781	06-Apr-2006	simonb	Regen.
1.780	05-Apr-2006	nonaka	Regen.
1.779	26-Mar-2006	jonathan	Regen pcidevs{,_data}.h from pcidevs rev 1.778 (fixes cut-and-paste typo in description of bcm5780S).
1.778	25-Mar-2006	rpaulo	regen
1.777	21-Mar-2006	he	Regenerated files after adding IDs for HP/Compaq Smart Array controllers.
1.776	12-Mar-2006	chs	branches: 1.776.2; regen
1.775	09-Mar-2006	he	Commit regenerated files after adding Nortel Baystack 21 info.
1.774	09-Mar-2006	hamajima	Regen: Add EXAR devices, and fix duplicated NVIDIA product id 0x0057.
1.773	07-Mar-2006	bouyer	Regen: added more serverwork SATA controllers.
1.772	07-Mar-2006	bouyer	branches: 1.772.2; Regen for serverworks K2 SATA addition.
1.771	06-Mar-2006	martin	Regen (MOXA devices added)
1.770	01-Mar-2006	dogcow	branches: 1.770.2; add a bunch of Nvidia nForce ethernet IDs (via OpenBSD.)
1.769	28-Feb-2006	gdamore	Regen.
1.768	21-Feb-2006	chuck	regen.
1.767	16-Feb-2006	kochi	regen.
1.766	12-Feb-2006	tron	Regen.
1.765	12-Feb-2006	tron	Regen.
1.764	12-Feb-2006	tron	Regen.
1.763	07-Feb-2006	cube	Regen, correctly this time. Sorry for the mess.
1.762	07-Feb-2006	xtraeme	regen
1.761	06-Feb-2006	abs	regen to add SIS 190 ethernet, as found in ASUS T2-AE1
1.760	02-Feb-2006	christos	branches: 1.760.2; regen
1.759	26-Jan-2006	bouyer	branches: 1.759.2; Regen: Add Promise PDC2057x
1.758	16-Jan-2006	christos	regen
1.757	10-Jan-2006	christos	Regen
1.756	02-Jan-2006	manu	regen
1.755	02-Jan-2006	tron	Regen.
1.754	02-Jan-2006	tron	Regen.
1.753	01-Jan-2006	xtraeme	regen
1.752	29-Dec-2005	silke	branches: 1.752.2; Regen from revision 1.751 of pcidevs.
1.751	19-Dec-2005	jonathan	Regem from src/sys/dev/pci/pcidevs rev 1.750, adding an entry for SK PCi-Express gigabit ethernet boards based on Marvell's Yukon-2 chipset.
1.750	19-Dec-2005	jonathan	regen after sys/dev/pci/pcidevs 1.749 (addition of SiI 3124)
1.749	18-Dec-2005	sekiya	Regen.
1.748	11-Dec-2005	jonathan	Regen after pcidevs rev 1.746, 1.747 (Add device IDs for LSI Megaraid, and Intel IDs for software-compatible Intel RAID products).
1.747	10-Dec-2005	jonathan	regen from pcidevs rev.1.745 with IDs for broadcom bcm5714, bcm5715, bcm5780.
1.746	09-Dec-2005	xtraeme	regen
1.745	08-Dec-2005	xtraeme	regen
1.744	07-Dec-2005	jonathan	regen from sys/dev/pci/pcidevs rev 1.742 (add bcm5714 et al.).
1.743	06-Dec-2005	fvdl	Regen for BCM5789.
1.742	04-Dec-2005	christos	Regen.
1.741	04-Dec-2005	christos	Regen.
1.740	27-Nov-2005	jdolecek	regen - 3ware ATA RAID description/ID fixes
1.739	22-Nov-2005	rpaulo	regen
1.738	18-Nov-2005	rpaulo	regen
1.737	18-Nov-2005	rpaulo	regen
1.736	01-Nov-2005	abs	branches: 1.736.2; Regen after adding Radeon 9000 PCI (0x4966)
1.735	05-Oct-2005	jonathan	branches: 1.735.2; Regen with PCI ids for Intel pro/1000 PT pci-e gigabit ethernet.
1.734	04-Sep-2005	tron	Regen.
1.733	21-Aug-2005	chs	regen
1.732	21-Aug-2005	gavan	regen
1.731	30-Jul-2005	grant	regen
1.730	30-Jul-2005	grant	regen
1.729	29-Jul-2005	cube	Regen.
1.728	25-Jul-2005	christos	regen.
1.727	01-Jul-2005	drochner	branches: 1.727.2; regen
1.726	01-Jul-2005	augustss	Regen.
1.725	28-Jun-2005	thorpej	Regen -- static'ify pci_vendors[] and pci_products[]
1.724	26-Jun-2005	seb	Regen.
1.723	24-Jun-2005	seb	Regen.
1.722	20-Jun-2005	briggs	Regen.
1.721	19-Jun-2005	chs	regen
1.720	15-Jun-2005	bouyer	Regen: add Intel 82801FBM Serial ATA Controller.
1.719	14-Jun-2005	bouyer	Regen: add SiS 965.
1.718	13-Jun-2005	tron	Regen.
1.717	13-Jun-2005	nonaka	Regen.
1.716	13-Jun-2005	nonaka	Regen.
1.715	07-Jun-2005	skrll	Regen.
1.714	28-May-2005	kent	regen
1.713	27-May-2005	augustss	Regen.
1.712	18-May-2005	riz	Regen.
1.711	12-May-2005	cube	Regen
1.710	09-May-2005	martin	Regen
1.709	04-May-2005	cube	Regen.
1.708	02-May-2005	rh	regen
1.707	20-Apr-2005	bouyer	Regen for "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller".
1.706	06-Apr-2005	martti	Regen for Dell CERC SATA RAID 1.5/6ch
1.705	24-Mar-2005	tron	Regen.
1.704	19-Mar-2005	tron	Regen.
1.703	19-Mar-2005	tron	Regen.
1.702	16-Mar-2005	xtraeme	branches: 1.702.2; regen
1.701	16-Mar-2005	uwe	Regen for ATI Radeon 9100 IGP chipset.
1.700	15-Mar-2005	xtraeme	regen
1.699	15-Mar-2005	xtraeme	regen
1.698	15-Mar-2005	uwe	Regen for ENE CB710, CB720 CardBus controllers.
1.697	07-Mar-2005	christos	Regen
1.696	27-Feb-2005	perry	regen
1.695	20-Feb-2005	jdolecek	regen: change the VIATECH VT61XX to VT612X
1.694	02-Feb-2005	christos	Regen
1.693	02-Feb-2005	abs	Regen to pick up corrected code for VIA VT6421
1.692	01-Feb-2005	abs	regen to pick up the 'VT642 SATA RAID Controller'.
1.691	27-Jan-2005	jdolecek	regen: add id for nForce2 Ultra 400 audio
1.690	27-Jan-2005	kent	regenerate for nForce 4 devices
1.689	25-Jan-2005	matt	branches: 1.689.2; Regen.
1.688	23-Jan-2005	fredb	Regen.
1.687	18-Jan-2005	tsubai	Regen.
1.686	15-Jan-2005	skrll	branches: 1.686.2; Regen.
1.685	15-Jan-2005	jdolecek	regen: Fix Geforce 6800GT entry
1.684	15-Jan-2005	jdolecek	regen: add bunch of NVIDIA GeForce 6800 devices
1.683	15-Jan-2005	jdolecek	regen: use full 'nForce2 Ultra 400' tag for the NFORCE2_400* enties
1.682	13-Jan-2005	jdolecek	regen: add bunch of NVIDIA nForce2 Ultra 400 chipset devices
1.681	13-Jan-2005	jdolecek	regen: VIA Velocity
1.680	08-Jan-2005	fredb	Regen.
1.679	08-Jan-2005	ginsbach	regen
1.678	06-Jan-2005	he	Regen after adding entry for LSI Logic FC919X.
1.677	27-Dec-2004	reinoud	Update generated files from pcidevs (ENE update)
1.676	18-Dec-2004	xtraeme	regen
1.675	16-Dec-2004	heas	Add entry for the Radeon Mobility T2.
1.674	11-Dec-2004	reinoud	Add ATI IXP audio controllers
1.673	07-Dec-2004	jdarrow	Regen from pcidevs 1.672: Add a new vendor and product ID for a cardbus NIC I've got Belkin 0x1799 F5D6020v3 0x6020
1.672	06-Dec-2004	cube	Regen (ATI IXP IDE).
1.671	05-Dec-2004	bouyer	Regen: Cleanup in VIA chipset names. From Lloyd Parkes in kern/15466.
1.670	30-Nov-2004	grant	regen
1.669	22-Nov-2004	rumble	regen
1.668	19-Nov-2004	tacha	regen.
1.667	17-Nov-2004	jdarrow	Regen from pcidevs 1.666 (add two wireless cards)
1.666	17-Nov-2004	junyoung	Regen.
1.665	16-Nov-2004	skrll	Regen
1.664	14-Nov-2004	soren	Regen.
1.663	06-Nov-2004	xtraeme	Regen for: Add Nvidia nForce3 250 chipset products, from Kouichirou Hiratsuka in PR 27843.
1.662	28-Oct-2004	martti	Regen for Dell PERC 4/SC
1.661	26-Oct-2004	augustss	Regen.
1.660	22-Oct-2004	cube	Regen.
1.659	20-Oct-2004	martti	Regen for DELL PERC 4e/Si
1.658	13-Oct-2004	augustss	Regen.
1.657	13-Oct-2004	drochner	regen
1.656	11-Oct-2004	jmmv	Regen after addition of ATI Technologies Radeon 9200SE Secondary identifier.
1.655	05-Oct-2004	augustss	Regen.
1.654	30-Sep-2004	cube	Regen.
1.653	18-Sep-2004	martin	Regen (FC929X added)
1.652	17-Sep-2004	xtraeme	Regen for GeForce FX 5700 LE.
1.651	17-Sep-2004	xtraeme	Add GeForceFX 5700 LE and regen.
1.650	16-Sep-2004	ragge	Regen.
1.649	10-Sep-2004	darrenr	Regen'd to add Sierra Wireless AirCard 580
1.648	26-Aug-2004	itohy	Regen from pcidevs rev 1.647.
1.647	05-Aug-2004	mycroft	Regen.
1.646	05-Aug-2004	cube	Regen.
1.645	04-Aug-2004	mycroft	Regen.
1.644	04-Aug-2004	mycroft	Regen.
1.643	04-Aug-2004	mycroft	Regen.
1.642	02-Aug-2004	mycroft	Regen.
1.641	02-Aug-2004	mycroft	Regen.
1.640	02-Aug-2004	mycroft	Regen.
1.639	02-Aug-2004	mycroft	Regen.
1.638	29-Jul-2004	cube	Regen.
1.637	27-Jul-2004	christos	regen
1.636	27-Jul-2004	rumble	regen.
1.635	25-Jul-2004	christos	regen
1.634	21-Jul-2004	mycroft	Regen.
1.633	17-Jul-2004	mycroft	Regen.
1.632	15-Jul-2004	junyoung	Regen for 3c920B-EMB-WNM Integrated Fast Ethernet.
1.631	05-Jul-2004	mycroft	Regen.
1.630	04-Jul-2004	wiz	regen from pcidevs, so that this file has the correct CVS Id.
1.629	04-Jul-2004	mycroft	Update Serverworks chip IDs, per: http://www.plasma-online.de/english/identify/picture/serverworks.html This seems remarkably more sane.
1.628	15-Jun-2004	bouyer	Regen for pcidevs 1.628: Add the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.627	27-May-2004	christos	Regen
1.626	27-May-2004	abs	regen from pcidevs 1.626: Add ALI 'M5257 PCI Software Modem'
1.625	12-May-2004	tron	Regen for PR kern/25166.
1.624	11-May-2004	thorpej	Regen: Add i6300ESB IDE / SATA controller PCI IDs.
1.623	10-May-2004	lukem	regen for: add Nvidia GeForce4 Ti 4200 Go AGP 8x (0x286)
1.622	10-May-2004	gendalia	regen (Add support for Adaptec AAR 2810SA raid controller.)
1.621	07-May-2004	sekiya	regen.
1.620	22-Apr-2004	skd	regen.
1.619	17-Apr-2004	matt	Regen.
1.618	03-Apr-2004	martti	Added another SB Live! 5.1 soundcard and joystick version
1.617	03-Apr-2004	mycroft	Add RealTek RT8180.
1.616	01-Apr-2004	tron	Regen from "pcidevs".
1.615	26-Mar-2004	gendalia	branches: 1.615.2; regen for Adaptect 2200S (Vulcan 2M) / Dell PERC 320/DC
1.614	25-Mar-2004	drochner	regen
1.613	12-Mar-2004	drochner	regen
1.612	03-Mar-2004	gavan	Add Creative Labs SB Audigy LS
1.611	26-Feb-2004	drochner	regen
1.610	28-Jan-2004	chs	regen
1.609	26-Jan-2004	blymn	Added note for updating these files.
1.608	26-Jan-2004	blymn	Regen with correct rcs id.
1.607	26-Jan-2004	blymn	Add amd8151 agp bridge id.
1.606	19-Jan-2004	ragge	Err, get correct RCS IDs. As Love said.
1.605	19-Jan-2004	ragge	Add "Intel PRO/10GbE LR Server Adapter". Regen.
1.604	06-Jan-2004	matt	Regen.
1.603	05-Jan-2004	matt	Regen.
1.602	20-Dec-2003	thorpej	Regen: Add product ID for the SiI 3124 SATALink.
1.601	19-Dec-2003	thorpej	Regen: Add product ID for the Silicon Image 3114 SATALink.
1.600	17-Dec-2003	tls	Add support for some more Marvell-based adapters to the sk driver: D-Link DGE-530T Linksys EG1032v2 Linksys EG1064v2 The Linux driver treats all of these the same. Tested with a DGE-530T; it works, though performance on at least one of my systems is atrocious. Fix setting of ethernet capabilities into interface capabilites word in attach routine. Manual page: note that jumbo support doesn't work and checksum support doesn't exist.
1.599	14-Dec-2003	tsutsui	Regen: added ATI Radeon 9200SE.
1.598	09-Dec-2003	ad	Regen.
1.597	08-Dec-2003	pooka	regen for broadcom 5703a3
1.596	08-Dec-2003	scw	Regen
1.595	06-Dec-2003	jdolecek	regen: add NetMos NM9805 and NM9815
1.594	05-Dec-2003	grant	regen
1.593	28-Nov-2003	chs	regen
1.592	28-Nov-2003	cube	Regen.
1.591	26-Nov-2003	matt	Regen.
1.590	22-Nov-2003	fvdl	Regenerate after rearranging the Promise ids.
1.589	03-Nov-2003	cube	Regen.
1.588	25-Oct-2003	chs	regen
1.587	23-Oct-2003	fvdl	Regen.
1.586	23-Oct-2003	briggs	Regenerate for ENE Cardbus product IDs.
1.585	22-Oct-2003	fvdl	Regen.
1.584	18-Oct-2003	matt	Regen.
1.583	18-Oct-2003	lukem	regen for more ServerWorks devices
1.582	15-Oct-2003	itojun	regen
1.581	14-Oct-2003	martti	Regen: Dell PERC 4/Di
1.580	10-Oct-2003	thorpej	Regen: Add several more Intel Gig-E product IDs.
1.579	08-Oct-2003	jonathan	Regen from pcidevs rev 1.581 with Hifn 7954/5 and 7956.
1.578	26-Sep-2003	matt	Regen.
1.577	25-Sep-2003	joda	regen
1.576	21-Sep-2003	nisimura	Regen.
1.575	21-Sep-2003	nisimura	Regen to fix the company name.
1.574	15-Sep-2003	bouyer	Regen: rename VT8237_RAID to VT8237_SATA
1.573	13-Sep-2003	jdolecek	regen - fix description of one Intel PRO/100 VM variant, as reported by Lars Heidieker on current-users@
1.572	28-Aug-2003	thorpej	Regen; spell Hifn correctly, add more Hifn products.
1.571	27-Aug-2003	kleink	Regen.
1.570	26-Aug-2003	jdolecek	regen: add SysKonnect 9821
1.569	26-Aug-2003	tron	Regen.
1.568	23-Aug-2003	dogcow	regen (add VIA VT6105)
1.567	20-Aug-2003	jonathan	regen from pcidevs 1.567 with fix for typo in bcm5823 product-id.
1.566	17-Aug-2003	chs	regen
1.565	17-Aug-2003	bouyer	Regen: Add 3com 3c940 Gigabit Ethernet, and VIA KT600 PCI devices.
1.564	14-Aug-2003	itojun	regen
1.563	07-Aug-2003	wiz	regen (fix ID of Radeon 9100 Secondary)
1.562	01-Aug-2003	wiz	regen (avoid duplicate definition, PR 22331)
1.561	01-Aug-2003	jonathan	regen from pcidevs rev 560 (adding Sun's bcm582x-compatible crypto accelerators)
1.560	29-Jul-2003	thorpej	Regen: Add the product ID used for the i82546EB chips present on the Intel PRO/1000 MT Quad Port Server Adapter.
1.559	29-Jul-2003	martin	Regen.
1.558	27-Jul-2003	jonathan	regen from pcidevs rev 1.556 with vendor Cavium and product Nitrox.
1.557	27-Jul-2003	jonathan	Regen from pcidevs 1.555 (Bluesteel,Broadcom security coprocessors)
1.556	23-Jul-2003	kleink	Regen from 1.554.
1.555	17-Jul-2003	hannken	Regen.
1.554	17-Jul-2003	hannken	Add support for the BCM5705: - Only BCM5705M asic rev A1 was tested. Thanks to Bill Paul (wpaul@freebsd.org) for help and support. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.553	14-Jul-2003	jmmv	Regen (+ ATI Radeon 9200)
1.552	14-Jul-2003	wiz	regen (+MGA G550).
1.551	12-Jul-2003	hubertf	Regen for Toshiba SD and SM cards
1.550	06-Jul-2003	dyoung	Add driver atw for PCI/Mini-PCI/Cardbus 802.11b NICs based on the ADMtek ADM8211. Read the man page for bugs and other outstanding issues.
1.549	06-Jul-2003	simonb	Regen: Trailing white-space and TABs nits.
1.548	06-Jul-2003	simonb	Regen; E7500 and E7505 updates.
1.547	06-Jul-2003	tron	Regen.
1.546	06-Jul-2003	simonb	Regen: Add an entry for the Intel E7505 Chipset Memory Controller Hub.
1.545	06-Jul-2003	tron	Regen.
1.544	06-Jul-2003	tron	Regen.
1.543	28-Jun-2003	bouyer	branches: 1.543.2; Regen: some more Intel devices.
1.542	27-Jun-2003	christos	Regen
1.541	27-Jun-2003	drochner	regen
1.540	26-Jun-2003	ichiro	regen.
1.539	14-Jun-2003	ichiro	regen.
1.538	13-Jun-2003	kent	Regen
1.537	08-Jun-2003	ichiro	regen.
1.536	07-Jun-2003	wiz	regen (Add some Brooktree device IDs).
1.535	07-Jun-2003	wiz	regen (+ATI 9100, 9600TX, 9500/9700)
1.534	07-Jun-2003	ichiro	regen.
1.533	07-Jun-2003	ichiro	regen.
1.532	07-Jun-2003	ichiro	regen.
1.531	28-Apr-2003	haya	Regenerate for pcidevs 1.529 and 1.530 changes.
1.530	24-Apr-2003	nathanw	Regen.
1.529	22-Apr-2003	fvdl	Regenerated after AMD 8131 additions.
1.528	15-Apr-2003	augustss	Regen.
1.527	15-Apr-2003	rjs	Regen.
1.526	13-Apr-2003	explorer	typo, s/TX4/TX2/
1.525	12-Apr-2003	explorer	add an older pci id for SATA150TX2PLUS, too
1.524	12-Apr-2003	christos	regen
1.523	11-Apr-2003	explorer	add Promise ATA/150 TX4 id
1.522	05-Apr-2003	lha	Regenerate. (add Radeon Mobility M7 LW (AGP))
1.521	03-Apr-2003	kent	regen
1.520	29-Mar-2003	jdc	Regenerate. (Add support for USRobotics Wireless Access PCI adapter (model 132415).)
1.519	25-Mar-2003	tsubai	Regen.
1.518	20-Mar-2003	thorpej	Regen; add more LSI MPT IDs.
1.517	18-Mar-2003	thorpej	Regen; added Silicon Image SATALink 3112.
1.516	13-Mar-2003	joda	regen: more nvidia devices
1.515	06-Mar-2003	bouyer	Regen: Remove duplicate in previous.
1.514	06-Mar-2003	bouyer	Regen: Add several SiS host bridges, from FreeBSD.
1.513	03-Mar-2003	soren	Regen.
1.512	02-Mar-2003	salo	Regen.
1.511	28-Feb-2003	enami	Regen.
1.510	26-Feb-2003	salo	Regen.
1.509	18-Feb-2003	grant	regen
1.508	03-Feb-2003	pooka	regen
1.507	03-Feb-2003	joda	regen
1.506	02-Feb-2003	bouyer	Regen: add SiS 651 host bridge.
1.505	01-Feb-2003	matt	Regen.
1.504	24-Jan-2003	thorpej	Regen; Tweak a Promise description.
1.503	24-Jan-2003	thorpej	Regen; add some more Promise products.
1.502	24-Jan-2003	thorpej	Regen; add Nvidia nForce2 ATA133 IDE.
1.501	17-Jan-2003	jonathan	regen from pcidevs rev 1.497: Add pcidevs entries for more Broadcom 570x family variants.
1.500	16-Jan-2003	gendalia	regen after pcidevs is committed so the correct version number is embedded. Thanks to wiz.
1.499	16-Jan-2003	gendalia	Add 3com 3c905CX-TX and Nvidia nForce2 MCP-T ac-97, on the ASUS A7N8X Deluxe motherboard, Chipset: NVIDIA nForce2 MCP-T
1.498	10-Jan-2003	augustss	Regen.
1.497	30-Dec-2002	explorer	update
1.496	26-Dec-2002	matt	Regen.
1.495	26-Dec-2002	matt	Regen.
1.494	11-Dec-2002	jdolecek	regen: add Madge Networks Smart 16/4 PCI Ringnode Mk2 entry
1.493	03-Dec-2002	explorer	add Ricoh 5C552 Cardbus + Firewire product ID
1.492	26-Nov-2002	thorpej	Regen: Add IBM 133 PCI-X Bridge.
1.491	26-Nov-2002	thorpej	Regen: Add Intel i80321 I/O Processor.
1.490	19-Nov-2002	bsh	Regen. for additional vendor IDs for Samsung Electronics Co. Ltd.
1.489	15-Nov-2002	tsutsui	Regen from pcidevs rev 1.486: > Remove an entry for AMD PCNETS_PCI. > > AMD Am79C974 PCnet-SCSI Ethernet and SCSI conrtoller is > a multi-function PCI device which has two device IDs > of Am53C974(PCscsi-PCI) and Am79C970(PCnet-PCI), > and there is no its own ID.
1.488	15-Nov-2002	ad	Regen.
1.487	09-Nov-2002	grant	regen for ATI Rage Mobility M1.
1.486	08-Nov-2002	msaitoh	regen
1.485	07-Nov-2002	thorpej	Regen; add Intel i31244 Serial ATA controller.
1.484	07-Nov-2002	martin	Regen (Myson MTD804 added)
1.483	04-Nov-2002	wiz	regen (typo fix from PR 18921)
1.482	25-Oct-2002	kent	regen
1.481	24-Oct-2002	grant	regen
1.480	21-Oct-2002	soren	Regen.
1.479	21-Oct-2002	grant	regen
1.478	18-Oct-2002	yamt	regen; fix simple typos
1.477	16-Oct-2002	leo	Regen.
1.476	05-Oct-2002	kent	regen
1.475	24-Sep-2002	pooka	regen
1.474	20-Sep-2002	pooka	regen
1.473	19-Sep-2002	nathanw	Regen.
1.472	11-Sep-2002	bouyer	Regen: add 2 acard pciide controllers.
1.471	06-Sep-2002	uwe	Regen for Integraphics Systems CyberPro 2000 and 2010.
1.470	23-Aug-2002	bouyer	Regen: Add Promise Ultra133TX2 and Ultra133TX2v2 PCI ID
1.469	16-Aug-2002	heinz	regen
1.468	10-Aug-2002	toshii	Regen; added SiS745 host bridge.
1.467	08-Aug-2002	drochner	regen
1.466	04-Aug-2002	thorpej	Regen; added ADI BECC product ID.
1.465	03-Aug-2002	kent	regen
1.464	02-Aug-2002	soren	Regen.
1.463	02-Aug-2002	ad	Regen.
1.462	31-Jul-2002	bouyer	Regen: remove duplicate entry for 82801DB_LPC.
1.461	30-Jul-2002	bouyer	Regen: Add Intel 845G PCI IDs
1.460	29-Jul-2002	augustss	Regen.
1.459	26-Jul-2002	onoe	regen: Add Silicon Image (CMDTECH) 0680
1.458	22-Jul-2002	msaitoh	regen
1.457	12-Jul-2002	enami	Regen.
1.456	09-Jul-2002	matt	Regen.
1.455	09-Jul-2002	thorpej	Regen: Clean up Intel i8254x product names, and add i82540, i82545, and i82546 product IDs.
1.454	07-Jul-2002	augustss	Regen.
1.453	05-Jul-2002	enami	Regen.
1.452	04-Jul-2002	rjs	Regenerate.
1.451	03-Jul-2002	msaitoh	regen
1.450	27-Jun-2002	pooka	regen
1.449	26-Jun-2002	cjs	Regenerate.
1.448	26-Jun-2002	cjs	Regenerate.
1.447	22-Jun-2002	fvdl	Regenerate.
1.446	22-Jun-2002	cjs	Regenerate.
1.445	21-Jun-2002	thorpej	Regenerate: Add product ID for Intel 82452KX/GX PCI-Host bridge.
1.444	17-Jun-2002	jdolecek	regen: add VIA Technologies VT8653 and VT8633 entries move VT82C686A_MC97 entry to where it numerically belongs
1.443	10-Jun-2002	cjs	Regenerate.
1.442	09-Jun-2002	taca	Regen.
1.441	04-Jun-2002	fvdl	Regen.
1.440	03-Jun-2002	briggs	Regen. for MPC8245 host bridge.
1.439	02-Jun-2002	seb	regen for Titan PCI-800L.
1.438	28-May-2002	christos	regen
1.437	28-May-2002	tron	Regen from "pcidevs".
1.436	27-May-2002	drochner	regen
1.435	25-May-2002	ichiro	regen
1.434	24-May-2002	lukem	regen for NetMos Quad UART and 1284 Printer port
1.433	19-May-2002	bouyer	branches: 1.433.2; Regen: add HPT372 and 74 PCI IDE controllers.
1.432	19-May-2002	sommerfeld	Regen
1.431	19-May-2002	sommerfeld	Regen
1.430	18-May-2002	matt	Regen.
1.429	15-May-2002	augustss	branches: 1.429.2; Regen.
1.428	14-May-2002	augustss	Regen.
1.427	03-May-2002	augustss	Regen.
1.426	29-Apr-2002	nonaka	Regen. (Added two IBM devices.)
1.425	26-Apr-2002	ad	Regen.
1.424	18-Apr-2002	thorpej	Regen: Added Symbios 53c875A product ID.
1.423	18-Apr-2002	tshiozak	add PCI/USB vendor ID for Kurusugawa Electronics, Inc. PCI-ID: 0x6809, USB-ID: 0x0b7e.
1.422	16-Apr-2002	drochner	regen
1.421	15-Apr-2002	mycroft	Regen.
1.420	15-Apr-2002	ad	Regen.
1.419	05-Apr-2002	mycroft	Regen.
1.418	27-Mar-2002	abs	Regen for 'CONEXANT 56KFAXMODEM 0x1085 HW 56K Fax Modem'
1.417	19-Mar-2002	bouyer	Regen: add AMD768 devices
1.416	11-Mar-2002	joda	regen
1.415	01-Mar-2002	kleink	Regen: add Motorola Raven, Falcon and Hawk product IDs.
1.414	28-Feb-2002	junyoung	Regen.
1.413	28-Feb-2002	junyoung	Regen; fix for duplicate ATI Rage Pro.
1.412	27-Feb-2002	martin	regen
1.411	27-Feb-2002	junyoung	Regen.
1.410	24-Feb-2002	kleink	Regen: add IBM 82660 PowerPC to PCI Bridge and Memory Controller.
1.409	21-Feb-2002	simonb	regen: Use BCM prefix instead of SB prefix in previous.
1.408	20-Feb-2002	simonb	regen: Add the SiByte/Broadcom SB1250 PCI and LDT bridge IDs.
1.407	20-Feb-2002	simonb	regen: Rename the SIBYTE entry to "Broadcom Corp. (SiByte)".
1.406	16-Feb-2002	thorpej	Regen: Add the Elan SC520 System Controller.
1.405	04-Feb-2002	junyoung	Regen.
1.404	02-Feb-2002	tsutsui	Regen; add Initio INIC-850/1060 SCSI and variants.
1.403	02-Feb-2002	tsutsui	Regen; add HiNT Corp. HB1 PCI-PCI bridge.
1.402	30-Jan-2002	tron	Regen.
1.401	29-Jan-2002	junyoung	Regen.
1.400	28-Jan-2002	thorpej	Regen: Add product ID for the Intel i82544GC 64-bit variant.
1.399	26-Jan-2002	tron	Regen.
1.398	26-Jan-2002	ichiro	regen
1.397	22-Jan-2002	junyoung	Regen.
1.396	22-Jan-2002	junyoung	Regen. Add Mach64 GS.
1.395	21-Jan-2002	ichiro	regen
1.394	21-Jan-2002	eeh	Regen.
1.393	21-Jan-2002	thorpej	Regen; Broadcom BCM570x-related IDs.
1.392	15-Jan-2002	thorpej	Regen; add Intel i82544GC Gig-E.
1.391	14-Jan-2002	augustss	Regen.
1.390	14-Jan-2002	augustss	Regen.
1.389	13-Jan-2002	augustss	Regen.
1.388	09-Jan-2002	augustss	Regen.
1.387	05-Jan-2002	junyoung	Regen.
1.386	05-Jan-2002	augustss	Regen.
1.385	03-Jan-2002	rafal	Regen (added Rainbow Technologies & Rainbow's CryptoSwift 200).
1.384	01-Jan-2002	augustss	Regen.
1.383	30-Dec-2001	ichiro	regen
1.382	29-Dec-2001	ichiro	regen
1.381	20-Dec-2001	thorpej	Regen; added SiS 630 PCI-Host bridge.
1.380	16-Dec-2001	thorpej	Regen: Add Middle Digital, Inc. vendor ID. Add product IDs for the Middle Digital PCI-Weasel Virtual VGA, Serial Port, and Control registers.
1.379	15-Dec-2001	soren	Regen.
1.378	13-Dec-2001	bouyer	Regen: Add some promise controllers, from FreeBSD
1.377	11-Dec-2001	ad	Regen.
1.376	11-Dec-2001	kleink	Regen: add product ID for the Motorola MPC107.
1.375	07-Dec-2001	onoe	Regen: Add HPB and LPC functions of 82801BAM.
1.374	07-Nov-2001	thorpej	Regen; correct i80312 description.
1.373	07-Nov-2001	thorpej	Regen; add Cyclone PCI-700 (IQ80310).
1.372	25-Oct-2001	augustss	Regen.
1.371	24-Oct-2001	thorpej	Regen; Add product IDs for the Powerhouse Systems PowerPro and PowerTop PowerPC system controllers.
1.370	24-Oct-2001	thorpej	Regen; more Intel Gig-E fixup, and add Intel i82544 IDs.
1.369	16-Oct-2001	tron	branches: 1.369.2; Regen.
1.368	13-Oct-2001	ichiro	regen.
1.367	11-Oct-2001	tron	Regen.
1.366	03-Oct-2001	simonb	Regen: Minor correction to the name of the 405GP PCI bridge.
1.365	01-Oct-2001	thorpej	Regen; added product ID for RealTek 8139[BC].
1.364	28-Sep-2001	thorpej	Regen; more corrections to Intel Gig-e product IDs.
1.363	27-Sep-2001	itojun	regen
1.362	15-Sep-2001	chris	Regen
1.361	10-Sep-2001	eeh	Regen.
1.360	05-Sep-2001	drochner	branches: 1.360.2; regen
1.359	04-Sep-2001	thorpej	Regen; added Intel IXP1200 Network Processor.
1.358	04-Sep-2001	pooka	regen.. again.. sigh
1.357	04-Sep-2001	pooka	regen
1.356	30-Aug-2001	soren	Regen.
1.355	28-Aug-2001	tsubai	Regen.
1.354	27-Aug-2001	enami	Regen.
1.353	27-Aug-2001	ad	Regen.
1.352	23-Aug-2001	thorpej	Regen; take part numbers out of PCnet-PCI product descriptions.
1.351	23-Aug-2001	thorpej	Regen; Add Promise ATA/100 TX2.
1.350	20-Aug-2001	fvdl	Regen.
1.349	20-Aug-2001	gehenna	Regen
1.348	18-Aug-2001	thorpej	Regen; added Avance Logic ALS4000.
1.347	17-Aug-2001	matt	regen.
1.346	31-Jul-2001	matt	regen
1.345	28-Jul-2001	pooka	regen
1.344	25-Jul-2001	thorpej	Regen; Added Antares Gigabit Ethernet.
1.343	23-Jul-2001	bouyer	Regen for TITAN PCI100H.
1.342	30-Jun-2001	thorpej	branches: 1.342.2; Regen.
1.341	28-Jun-2001	thorpej	Add Tamarack Microelectronics, Inc. vendor ID, and product ID for Tamarack TC9021 Gigabit Ethernet.
1.340	27-Jun-2001	mcr	added Solidum PAX.ware 1100 and Lucent ORCA PCI targets
1.339	20-Jun-2001	thorpej	Regen; add/update 3Com 3CR990 products.
1.338	20-Jun-2001	thorpej	Regen; add SysKonnect SKNET-GE.
1.337	20-Jun-2001	imp	Regen.
1.336	18-Jun-2001	onoe	regen: Add Cisco/Aironet 350
1.335	17-Jun-2001	tsubai	Regen.
1.334	13-Jun-2001	simonb	Regen; add IBM 405GP PCI Bridge & ROB-in i960RP Microprocessor.
1.333	11-Jun-2001	thorpej	Regen; Add Level One vendor ID, and LXT-1001 10/100/1000 Ethernet product ID.
1.332	11-Jun-2001	tron	Regen from "pcidevs".
1.331	31-May-2001	nisimura	Regen; product IDs for IDT RC32334 and RC32332 systeml controllers.
1.330	21-May-2001	thorpej	Regen; add ICH3 network controller IDs.
1.329	18-May-2001	enami	Regen.
1.328	17-May-2001	drochner	regen
1.327	14-May-2001	thorpej	Regen; added DP83820 Gigabit Ethernet chip.
1.326	06-May-2001	fvdl	Regen.
1.325	25-Apr-2001	simonb	Regen.
1.324	21-Apr-2001	tsutsui	Regen from pcidevs rev. 1.318: > Correct an entry for Acard UDMA IDE controller. > The chip is actually marked as ATP860-A, not ATP860R.
1.323	13-Apr-2001	hubertf	Regen for moved Trident CyberBlade entry
1.322	12-Apr-2001	itojun	regen.
1.321	12-Apr-2001	matt	Regen.
1.320	09-Apr-2001	ad	Regen.
1.319	31-Mar-2001	minoura	Regen.
1.318	21-Mar-2001	soren	Regen.
1.317	21-Mar-2001	hubertf	regen for trident cyberclade
1.316	20-Mar-2001	mrg	regen
1.315	17-Mar-2001	thorpej	Regen.
1.314	13-Mar-2001	bouyer	Regen.
1.313	07-Mar-2001	matt	Regen.
1.312	05-Mar-2001	matt	branches: 1.312.2; Regen
1.311	04-Mar-2001	msaitoh	regen
1.310	02-Mar-2001	thorpej	Regen -- Add Mylex AcceleRAID and eXtremeRAID product and subsystem IDs.
1.309	02-Mar-2001	thorpej	Regen; added Xircom WinGlobal modem product ID.
1.308	18-Feb-2001	bouyer	Regen.
1.307	18-Feb-2001	matt	regen
1.306	12-Feb-2001	tshiozak	add CMI8738B.
1.305	11-Feb-2001	ad	Regen.
1.304	04-Feb-2001	briggs	Regen.
1.303	02-Feb-2001	hubertf	Regen: Add M5243 PCI-AGP Bridge Fixes PR 11466 by Jared D. McNeill <jmcneill@invisible.yi.org>
1.302	31-Jan-2001	thorpej	Regen; add SEGA Dreamcast Broadband Adapter.
1.301	26-Jan-2001	rh	regen
1.300	26-Jan-2001	rh	regen
1.299	22-Jan-2001	jdolecek	regen: US Robotics (3Com) 56K Voice Internal PCI Modem (WinModem) entry added
1.298	22-Jan-2001	mjl	Regen for stallion.
1.297	17-Jan-2001	thorpej	Regen; Intel Gigabit Ethernet update.
1.296	15-Jan-2001	thorpej	Regen.
1.295	12-Jan-2001	itojun	regen
1.294	09-Jan-2001	rh	regen
1.293	09-Jan-2001	thorpej	Regen; added Symbios 53c1510D.
1.292	08-Jan-2001	thorpej	Regen; added Accton EN2242.
1.291	08-Jan-2001	thorpej	Regen; added DEC SWXCR RAID.
1.290	08-Jan-2001	rh	regen
1.289	03-Jan-2001	mycroft	Regen.
1.288	31-Dec-2000	ad	Regen.
1.287	30-Dec-2000	eeh	Regen.
1.286	29-Dec-2000	eeh	Add a bus type for psycho controllers.
1.285	27-Dec-2000	tsutsui	Regen. (Fix Acard UDMA IDE)
1.284	27-Dec-2000	augustss	Regen.
1.283	23-Dec-2000	tsutsui	Regen. (Adding Acard UDMA IDE controllers)
1.282	14-Dec-2000	onoe	regen
1.281	05-Dec-2000	drochner	regen
1.280	20-Nov-2000	onoe	regen. CXD32222 -> CXD3222 (Sony OHCI Controller)
1.279	18-Nov-2000	thorpej	Regen.
1.278	14-Nov-2000	bouyer	Regen.
1.277	04-Nov-2000	thorpej	Regen.
1.276	03-Nov-2000	ad	regen.
1.275	03-Nov-2000	ad	Regen.
1.274	27-Oct-2000	thorpej	Regen.
1.273	27-Oct-2000	thorpej	Regen.
1.272	19-Oct-2000	thorpej	Regen.
1.271	14-Oct-2000	simonb	Regen
1.270	12-Oct-2000	itojun	regen
1.269	03-Oct-2000	thorpej	Regen: ADMtek AN985
1.268	27-Sep-2000	augustss	Regen.
1.267	20-Sep-2000	thorpej	Regen.
1.266	19-Sep-2000	fvdl	Regen.
1.265	13-Sep-2000	ad	Regen.
1.264	13-Sep-2000	tron	Regen.
1.263	13-Sep-2000	tron	Regen.
1.262	13-Sep-2000	tron	Regen.
1.261	07-Sep-2000	soren	Regen.
1.260	04-Sep-2000	soren	Regen.
1.259	04-Sep-2000	itojun	regen
1.258	02-Sep-2000	cgd	regen for changed pcidevs
1.257	29-Aug-2000	wiz	regen
1.256	21-Aug-2000	enami	Regen.
1.255	21-Aug-2000	haya	Regen.
1.254	07-Aug-2000	billc	Manually edit the headers to show which version of pcidevs this was generated from. XXX - this really should be automated
1.253	07-Aug-2000	billc	Added PCI IDs for 3Com's Home Connect (3c450) and Office Connect (3cSOHO100). They both are register compat with later versions of the 3c905 (tornado/ vortex) series.
1.252	04-Aug-2000	ad	Regen.
1.251	23-Jul-2000	jonathan	* Add entries for older Yamaha YMF chips (724, 740) plus the newer steppings (740C, 724F) which use the same microcode as the DS-1S/DS-1E. Regen.
1.250	20-Jul-2000	ad	Regen.
1.249	10-Jul-2000	ad	Regen.
1.248	09-Jul-2000	mycroft	Regen.
1.247	08-Jul-2000	mrg	regen
1.246	04-Jul-2000	enami	Regen.
1.245	03-Jul-2000	gmcgarry	Regen.
1.244	24-Jun-2000	joda	regen
1.243	21-Jun-2000	tron	Regen.
1.242	16-Jun-2000	soren	branches: 1.242.2; Regen.
1.241	13-Jun-2000	tron	Regen.
1.240	12-Jun-2000	nathanw	Regen.
1.239	12-Jun-2000	bouyer	Regen.
1.238	12-Jun-2000	bouyer	Regen.
1.237	09-Jun-2000	augustss	Regen.
1.236	07-Jun-2000	ad	Regen.
1.235	07-Jun-2000	mrg	regen.
1.234	07-Jun-2000	soren	Regen.
1.233	06-Jun-2000	kleink	Regen.
1.232	05-Jun-2000	kleink	Regen.
1.231	05-Jun-2000	kleink	Regen.
1.230	30-May-2000	matt	Regen
1.229	28-May-2000	drochner	regen
1.228	27-May-2000	scw	Regen.
1.227	27-May-2000	soren	branches: 1.227.2; Regen.
1.226	26-May-2000	matt	Regen.
1.225	18-May-2000	drochner	regen
1.224	18-May-2000	garbled	regen
1.223	14-May-2000	soren	Regen.
1.222	12-May-2000	thorpej	Regen.
1.221	07-May-2000	jonathan	PCI product codes for Yamaha YMF744B (DS-1S) YMF754 (DS-1E) audio chips.
1.220	30-Apr-2000	dante	Add ASB-3940U3W host adapters (AdvanSys)
1.219	28-Apr-2000	augustss	Regen.
1.218	26-Apr-2000	thorpej	Regen.
1.217	22-Apr-2000	augustss	Regen.
1.216	19-Apr-2000	bouyer	Regen
1.215	18-Apr-2000	cgd	regen for changed pcidevs
1.214	17-Apr-2000	thorpej	Regen.
1.213	03-Apr-2000	joda	regen
1.212	31-Mar-2000	tsarna	Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.211	30-Mar-2000	joda	regen
1.210	27-Mar-2000	ad	Regen.
1.209	23-Mar-2000	joda	regen
1.208	23-Mar-2000	ad	Regen.
1.207	21-Mar-2000	joda	regen
1.206	16-Mar-2000	thorpej	Regen.
1.205	14-Mar-2000	tsarna	Regen
1.204	13-Mar-2000	ad	Regen.
1.203	11-Mar-2000	jdolecek	regen
1.202	09-Mar-2000	jdolecek	regen
1.201	06-Mar-2000	mark	Regenerate from pcidevs.
1.200	04-Mar-2000	tron	Regen from "pcidevs".
1.199	16-Feb-2000	soren	Regen.
1.198	08-Feb-2000	bouyer	regen
1.197	03-Feb-2000	dante	Add AdvanSys U2W host adapters
1.196	03-Feb-2000	joda	regen
1.195	31-Jan-2000	thorpej	Regen.
1.194	25-Jan-2000	haya	Regen.
1.193	24-Jan-2000	joda	regen
1.192	23-Jan-2000	tsubai	Regen.
1.191	12-Jan-2000	haya	Regen.
1.190	11-Jan-2000	matt	Regen.
1.189	09-Jan-2000	soren	Regen.
1.188	05-Jan-2000	drochner	regen
1.187	31-Dec-1999	hubertf	Add an entry for the Trident Microsystems CYBER 9525 (as found in an Toshiba Satellite 2650 XDVD).
1.186	17-Dec-1999	drochner	regen
1.185	17-Dec-1999	augustss	Regen.
1.184	17-Dec-1999	soren	Regen.
1.183	17-Dec-1999	soren	Regen.
1.182	15-Dec-1999	kleink	Regen.
1.181	04-Dec-1999	castor	Regen
1.180	04-Dec-1999	thorpej	Regen.
1.179	01-Dec-1999	drochner	regen
1.178	30-Nov-1999	drochner	regen
1.177	17-Nov-1999	augustss	Regen.
1.176	12-Nov-1999	is	Identification of DOMEX PCI-SCSI board, by Thilo Manske, PR 8772.
1.175	07-Nov-1999	cgd	regen
1.174	06-Nov-1999	rh	regen.
1.173	06-Nov-1999	augustss	Regen.
1.172	04-Nov-1999	tsarna	regen
1.171	04-Nov-1999	augustss	Regen.
1.170	04-Nov-1999	joda	regen
1.169	01-Nov-1999	augustss	Regen.
1.168	29-Oct-1999	matt	Regen.
1.167	29-Oct-1999	haya	Maintainance the name of Ricoh PCI-CardBus bridges.
1.166	11-Oct-1999	hannken	branches: 1.166.2; 1.166.4; Regen.
1.165	29-Sep-1999	thorpej	branches: 1.165.2; Regen.
1.164	29-Sep-1999	thorpej	Regen.
1.163	27-Sep-1999	ad	Regen.
1.162	11-Sep-1999	soren	Regen.
1.161	10-Sep-1999	augustss	Regen.
1.160	08-Sep-1999	thorpej	Regen.
1.159	08-Sep-1999	thorpej	Regen.
1.158	08-Sep-1999	thorpej	Regen.
1.157	01-Sep-1999	drochner	regen
1.156	28-Aug-1999	augustss	Regen.
1.155	27-Aug-1999	thorpej	Regen.
1.154	24-Aug-1999	soren	Regenerate.
1.153	17-Aug-1999	drochner	regen
1.152	16-Aug-1999	mcr	Added Solidum Systems PCI vendor ID and Device IDs
1.151	12-Aug-1999	bouyer	Regen.
1.150	01-Aug-1999	veego	Regen.
1.149	29-Jul-1999	thorpej	Regen.
1.148	28-Jul-1999	soren	Regen.
1.147	27-Jul-1999	castor	regen
1.146	27-Jul-1999	thorpej	Regen.
1.145	20-Jul-1999	drochner	regen
1.144	15-Jul-1999	ross	Regen.
1.143	15-Jul-1999	augustss	Regen.
1.142	14-Jul-1999	fvdl	Regen.
1.141	12-Jul-1999	kleink	Regen.
1.140	01-Jul-1999	itojun	IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628. (Sorry for a big commit, I can't separate this into several pieces...) Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details. - sys/kern: do not assume single mbuf, accept chained mbuf on passing data from userland to kernel (or other way round). - "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ package (ftp://ftp.csl.sony.co.jp/pub/kjc/). - sys/netinet/tcp: IPv4/v6 dual stack tcp support. - sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those file to be there so we patch it up. - sys/netinet: IPsec additions are here and there. - sys/netinet6/: most of IPv6 code sits here. - sys/netkey: IPsec key management code - dev/pci/pcidevs: regen In my understanding no code here is subject to export control so it should be safe.
1.139	27-Jun-1999	drochner	regen
1.138	19-Jun-1999	augustss	Regen.
1.137	18-Jun-1999	tsubai	Regen.
1.136	05-Jun-1999	tron	Regen.
1.135	05-Jun-1999	tron	Regen.
1.134	02-Jun-1999	drochner	regen
1.133	29-May-1999	mrg	regen.
1.132	29-May-1999	mrg	regen.
1.131	29-May-1999	fvdl	Regen.
1.130	27-May-1999	mrg	regen.
1.129	23-May-1999	veego	Regen.
1.128	20-May-1999	augustss	Regen.
1.127	17-May-1999	drochner	regen
1.126	07-May-1999	hwr	Regen.
1.125	07-May-1999	hwr	Regen.
1.124	07-May-1999	drochner	regen
1.123	07-May-1999	matt	Regen
1.122	06-May-1999	mrg	regen
1.121	12-Apr-1999	jonathan	Add entries for TI, Ricoh, Toshiba PCI-to-cardbus bridge chips, so we can identify them as cardbus chips supported by the cardbus patches. Add entry for OPTI chipsets whose interrupts arent properly set up by some BIOSes. From cardbus patches<ftp://nandra.iri.co.jp/pub/NetBSD/CardBus by HAYAKAWA Koichi <haya@tcad.ulsi.sony.co.jp>.
1.120	25-Mar-1999	explorer	branches: 1.120.2; Add LAN Media Corporation info
1.119	19-Mar-1999	mycroft	Regen.
1.118	17-Mar-1999	thorpej	Regen.
1.117	16-Mar-1999	tron	Regen from "pcidevs".
1.116	16-Mar-1999	mjacob	add Qlogic 1080 && 1240 product IDs
1.115	11-Mar-1999	sakamoto	Regen.
1.114	09-Mar-1999	drochner	regen
1.113	19-Feb-1999	thorpej	Regen.
1.112	19-Feb-1999	thorpej	Regen.
1.111	19-Feb-1999	thorpej	Regen.
1.110	14-Feb-1999	tron	Regen from pcidevs.
1.109	05-Feb-1999	thorpej	Regen.
1.108	28-Jan-1999	augustss	Regen.
1.107	16-Jan-1999	drochner	regen
1.106	30-Dec-1998	hwr	Regen
1.105	21-Dec-1998	drochner	regen
1.104	20-Dec-1998	drochner	regen
1.103	18-Dec-1998	thorpej	Regen.
1.102	09-Dec-1998	thorpej	Regen.
1.101	05-Dec-1998	matt	Add Sigma Designs REALmagic Hollywood Plus MPEG-2 decoder
1.100	30-Nov-1998	jtk	regen
1.99	29-Nov-1998	thorpej	Regen.
1.98	29-Nov-1998	thorpej	Regen.
1.97	27-Nov-1998	drochner	regen
1.96	26-Nov-1998	mcr	regenerated pcidevs.h and pcidevs_data.h due to kern/6280 (not kern/5847 as originally noted)
1.95	20-Oct-1998	thorpej	Regen.
1.94	19-Oct-1998	bouyer	Regen.
1.93	05-Oct-1998	tron	Regen.
1.92	04-Oct-1998	tron	Regen.
1.91	01-Oct-1998	fvdl	Regen.
1.90	26-Sep-1998	dante	Change product ADVSYS 2300 to ADVSYS WIDE
1.89	22-Sep-1998	mark	Regenerate from pcidevs.
1.88	18-Sep-1998	drochner	regen
1.87	18-Sep-1998	mark	Regen.
1.86	06-Sep-1998	veego	Regen.
1.85	26-Aug-1998	drochner	regen
1.84	26-Aug-1998	dante	Changed ADVSYSPROD to ADVSYS vendor ID Added 1200A, 1200B, ULTRA, 2300 AdvanSys products ID
1.83	07-Aug-1998	thorpej	Regen.
1.82	28-Jul-1998	thorpej	Regen.
1.81	27-Jul-1998	tron	Regen from pcidevs.
1.80	27-Jul-1998	tron	Regen from pcidevs.
1.79	27-Jul-1998	augustss	Regen.
1.78	25-Jul-1998	explorer	commit this so it gets the right generated from tag
1.77	25-Jul-1998	explorer	define RISCom and RISCom N2 product
1.76	22-Jul-1998	drochner	regen
1.75	19-Jul-1998	tron	Regen from pcidevs.
1.74	13-Jul-1998	mark	Regen.
1.73	10-Jul-1998	tron	Regen from pcidevs.
1.72	30-Jun-1998	augustss	Regen.
1.71	15-Jun-1998	bouyer	Regen.
1.70	15-Jun-1998	bouyer	Regen.
1.69	15-Jun-1998	bouyer	Regen.
1.68	25-May-1998	mark	Regen from pcidevs.
1.67	14-May-1998	augustss	Regen.
1.66	13-May-1998	augustss	Regen.
1.65	05-May-1998	thorpej	Regen.
1.64	04-May-1998	thorpej	Regen.
1.63	30-Apr-1998	mycroft	Regen.
1.62	16-Apr-1998	drochner	regen
1.61	28-Mar-1998	thorpej	Regen.
1.60	28-Mar-1998	cgd	regen
1.59	07-Mar-1998	augustss	Regen.
1.58	04-Mar-1998	cgd	regen
1.57	04-Mar-1998	cgd	regen
1.56	27-Feb-1998	mycroft	Regen.
1.55	07-Feb-1998	augustss	Newly generated version.
1.54	07-Feb-1998	augustss	Add the AGP device in the 82443LX.
1.53	27-Jan-1998	thorpej	Regen: add Packet Engines G-NIC gigabit Ethernet.
1.52	22-Jan-1998	mycroft	Regen.
1.51	09-Jan-1998	thorpej	Regen.
1.50	08-Jan-1998	thorpej	Regen: back out RCS ID related changes.
1.49	05-Jan-1998	perry	regened with RCSIDs in place
1.48	16-Nov-1997	christos	Regen.
1.47	27-Oct-1997	thorpej	Regen.
1.46	14-Oct-1997	thorpej	branches: 1.46.2; Regen.
1.45	08-Oct-1997	mycroft	Regen.
1.44	06-Oct-1997	lonhyn	pcidevs change regen.
1.43	01-Oct-1997	drochner	regen
1.42	23-Sep-1997	mycroft	Regen.
1.41	15-Sep-1997	bouyer	Regen.
1.40	12-Sep-1997	drochner	regen
1.39	28-Aug-1997	thorpej	Regen: added SMC83C170 entry.
1.38	23-Aug-1997	drochner	regen
1.37	16-Aug-1997	mjacob	Add in PCI_PRODUCT_QLOGIC_ISP2100.
1.36	14-Aug-1997	drochner	Regen with new RCS ID. (thanks, Bernd)
1.35	14-Aug-1997	drochner	Include some more stuff.
1.34	03-Aug-1997	fvdl	Regen.
1.33	19-Jul-1997	cgd	branches: 1.33.2; regen for new pcidevs
1.32	24-Jun-1997	thorpej	Correct Intel Triton MX PCI IDs, from Tom Yu <tlyu@mit.edu>, PR #3478.
1.31	24-Jun-1997	kml	Added definitions for Essential Communications RoadRunner HIPPI and Gigabit Ethernet interfaces.
1.30	13-May-1997	jtk	regen
1.29	04-Apr-1997	augustss	Regen.
1.28	20-Mar-1997	chuck	regen
1.27	16-Feb-1997	veego	regen.
1.26	01-Jan-1997	veego	branches: 1.26.4; Add the Matrox Millenium and regen the header files.
1.25	30-Nov-1996	jtk	forgot to regenerate these after committing pcidevs for its NetBSD header tag
1.24	28-Nov-1996	jtk	add Cirrus CL-7541, 7543 and ACC 2188; regen .h files
1.23	28-Nov-1996	thorpej	Regen.
1.22	19-Oct-1996	jonathan	regen from pcidevs 1.20
1.21	15-Oct-1996	christos	regen
1.20	13-Oct-1996	christos	regen
1.19	08-Oct-1996	gibbs	dev/microcode/aic7xxx/aic7xxx.seq, dev/microcode/aic7xxx_seq.h, dev/ic/aic7xxxreg.h: Remove intrinsic knowledge about SDTR and WDTR messages and replace it with a generic message system that allows the kernel driver to handle SDTR, WDTR and any other type of extended message it chooses too. This makes the sequencer code much simpler, makes extended message handling debuggable since the bulk of the work is in the kernel driver, and saves lots of instruction space. Regen microcode header file. dev/ic/aic7xxx.c, dev/ic/aic7xxxvar.h: Add code to handle WDTR and SDTR negotiation in light of the changes in the message interface to the sequencer. Don't reject targets that negotiate async by sending an SDTR with a 0 offset. Use an sdtr message with 0,0 to negotiate async when a target suggests a period that is too long for us to handle. Some tape and cdrom drives don't like us doing the message reject that we did in the past. Fix a problem with handing the QUEUE FULL condition. Fix a race condition (most likely the cause of the SCB paging problems) that might allow the sequencer to get unpaused before the condition that caused it to be paused (a SEQINT) was handled. Race condition pointed out by Doug Ledford <dledford@dialnet.net> and by "Dan Willis" <dan@plutotech.com>. dev/pci/ahc_pci.c: Add support for the 2940AU, an aic7860 based controller. dev/pci/pcidevs.h, dev/pci/pcidevs_data.h: Add product IDs for the 2940AU, aic7860 and aic7855. Regen data file. scsi/scsi_message.h: Add MSG_EXT_SDTR_LEN and MSG_EXT_WDTR_LEN - the length of bytes in these extended messages. Thanks to Chuck Cranor <chuck@maria.wustl.edu> for testing these changes out for me.
1.18	24-Sep-1996	christos	New cyclades driver from Timo Rossi.
1.17	29-Aug-1996	cgd	regen for new pcidevs
1.16	26-Aug-1996	cgd	regen for new pcidevs
1.15	23-Aug-1996	cgd	regen for new pcidevs
1.14	15-Jul-1996	cgd	regen for updated pcidevs
1.13	22-Jun-1996	chuck	regenerate to fix "generated from" comment (next commit pcidevs first, then regenerate)
1.12	22-Jun-1996	chuck	add product code for efficient networks eni155p atm card and regenerate
1.11	16-May-1996	mycroft	branches: 1.11.4; Regen.
1.10	13-May-1996	thorpej	Regenerate.
1.9	07-May-1996	thorpej	Regenerate; pcidevs changed.
1.8	04-Apr-1996	mycroft	Regen.
1.7	19-Feb-1996	christos	pcidevs was changed.
1.6	31-Jan-1996	cgd	update for new version of 'pcidevs'
1.5	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.4	10-Nov-1995	christos	pcidevs was changed
1.3	10-Nov-1995	christos	- pcidevs: add 3c590 and adaptec ultra - aic7xxx.c: fix the probe routines to use pcidevs.h
1.2	20-Jun-1995	cgd	vendor IDs for Acer and SIS
1.1	18-Jun-1995	cgd	built from pcidevs list
1.11.4.1	04-Mar-1997	mycroft	Regen.
1.26.4.1	12-Mar-1997	is	Merge in changes from The Trunk
1.33.2.9	14-Oct-1997	thorpej	Sync w/ trunk.
1.33.2.8	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.7	14-Oct-1997	thorpej	Regen.
1.33.2.6	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.5	22-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.4	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.3	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.2	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.46.2.2	17-Nov-1997	thorpej	Regen.
1.46.2.1	27-Oct-1997	thorpej	Regen.
1.120.2.13	11-Sep-2000	he	Pull up regenerated files (requested by billc): Recognize 3Com's Home Connect (3c450) and Office Connect (3cSOHO100) Ethernet cards.
1.120.2.12	23-Jul-2000	he	Pull up regenerated files (requested by jonathan): Add product codes for various Yamaha audio chips.
1.120.2.11	07-Jul-2000	he	Pull up regenerated files (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0)
1.120.2.10	03-Jun-2000	he	Pull up regenerated files (requested by tsubai): Add support for PCI variants of Cyclom-4Y and -8Y cards. Also fixes PR#7175 (new serial controller).
1.120.2.9	22-May-2000	he	Pull up regenerated files (requested by bouyer): Add product IDs for some VIA Technologies devices.
1.120.2.8	10-Feb-2000	he	Pull up regenerated file (requeste by bouyer): Add support for US Robotics 3CP5609 PCI 16550 Modem.
1.120.2.7	15-Jan-2000	he	Regenerate after fixing DPT SCSI boards description.
1.120.2.6	20-Oct-1999	he	Regenerate these after adding Realtek to pcidevs.
1.120.2.5	05-Oct-1999	he	Regenerate these from recent pcidevs pullup.
1.120.2.4	24-Sep-1999	cgd	regenerate _after_ pcidevs is checked in
1.120.2.3	29-Aug-1999	he	Regenerated. Recognize 3c905C-TX card. (bouyer)
1.120.2.2	22-Jun-1999	perry	regen (fvdl): recognize another 390xB card
1.120.2.1	13-Apr-1999	jonathan	branches: 1.120.2.1.2; 1.120.2.1.4; Pull up sys/dev/pci/pcidevs 1.113->1.114 (CardBus bridge chips) to 1.4. pcidevs.h, pcidevs_data.h: regen.
1.120.2.1.4.2	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.120.2.1.4.1	28-Jun-1999	itojun	KAME/NetBSD 1.4 SNAP kit, dated 19990628. NOTE: this branch (kame) is used just for refernce. this may not compile due to multiple reasons.
1.120.2.1.2.3	02-Aug-1999	thorpej	Regen.
1.120.2.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.120.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.165.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.166.4.1	15-Nov-1999	fvdl	Sync with -current
1.166.2.10	23-Apr-2001	bouyer	Sync with HEAD.
1.166.2.9	21-Apr-2001	bouyer	Sync with HEAD
1.166.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.166.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.166.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.166.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.166.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.166.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.166.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.166.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.227.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.242.2.27	07-Apr-2004	jmc	Regen (requested by itojun in ticket #71) Add National Semiconductors SC1100 (GEODE) PCI devices
1.242.2.26	06-Jun-2002	he	Pull up regenerated files (requested by mycroft): Add Intel ICH3 network controller IDs. Part of fix to fxp driver to work around a bug which can cause some machines to randomly hang, and to drag along performance improvements and improved hardware support.
1.242.2.25	25-Mar-2002	he	Pull up regenerated files (requested by bouyer): Add support for the pciide part of the AMD-768 chipset.
1.242.2.24	06-Feb-2002	he	Pull up regenerated files (requested by tsutsui): Add support for Acard ATP-850/860 pciide controllers.
1.242.2.23	29-Jan-2002	he	Pull up regenerated files (requested by bouyer): Add support for the 53c1010-33, and bring documentation up to date.
1.242.2.22	24-Jan-2002	he	Pull up regenerated files (requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.242.2.21	25-Oct-2001	he	Pull up regenerated files (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.242.2.20	18-Jun-2001	jhawk	Pull up regenerated files (requested by tsubai): Support the new dual USB iBook.
1.242.2.19	11-Jun-2001	he	Pull up regenerated files (requested by tron): Support the D-Link DFE-530TX+. Fixes PR#13161.
1.242.2.18	15-May-2001	he	Pull up regenerated files (requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.242.2.17	06-May-2001	he	Pull up regenerated files (requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.242.2.16	03-May-2001	he	Pull up regenerated files (requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.242.2.15	23-Apr-2001	he	Pull up regenerated files (requested by hubertf): Support Accton EN2242 and other AmdTek AN985 compatible cards with the tlp(4) driver.
1.242.2.14	22-Mar-2001	he	Pull up regenerated files (requested by hubertf): Make the Intel Pro/100 as e.g. found on Toshiba Satellite Pro 4600s work.
1.242.2.13	20-Mar-2001	he	Pull up regenerated files (requested by chuck): Add support for the 3Com 3c555, 3c556 and 3c556B MiniPCI Ethernet cards.
1.242.2.12	26-Feb-2001	he	Pull up regenerated files (requested by bouyer): Add support for the Intel 82801BAM IDE controller.
1.242.2.11	31-Dec-2000	jhawk	Regen.
1.242.2.10	30-Oct-2000	tv	Regen.
1.242.2.9	19-Oct-2000	he	Pull up regenerated files (requested by billc): Add PCI IDs for 3Com Home Connect (3c450) and Office Connect (3cSOHO100), register compatible with later versions of 3c905.
1.242.2.8	21-Sep-2000	thorpej	Regen: Add National Semiconductor DP83815 10/100 Ethernet.
1.242.2.7	13-Sep-2000	tron	Regen.
1.242.2.6	22-Aug-2000	enami	Regen.
1.242.2.5	23-Jul-2000	jonathan	Pull up pcidevs.h rev 1.251 and pcidevs_data.h rev 249, by pulling up pcidevs rev 1.245 (YMF 7xx audio product codes), then regen.
1.242.2.4	16-Jul-2000	jhawk	Pullup pcidevs rev 1.242 and regen pcidevs_data.h, pcidevs.h, approved by jhawk: Add ID for i82559ER.
1.242.2.3	06-Jul-2000	enami	Regen.
1.242.2.2	27-Jun-2000	thorpej	Regen.
1.242.2.1	26-Jun-2000	bouyer	Regen.
1.312.2.29	17-Jan-2003	thorpej	Sync with HEAD.
1.312.2.28	15-Jan-2003	thorpej	Sync with HEAD.
1.312.2.27	03-Jan-2003	thorpej	Sync with HEAD.
1.312.2.26	29-Dec-2002	thorpej	Sync with HEAD.
1.312.2.25	19-Dec-2002	thorpej	Sync with HEAD.
1.312.2.24	11-Dec-2002	thorpej	Sync with HEAD.
1.312.2.23	26-Nov-2002	thorpej	Sync with HEAD.
1.312.2.22	11-Nov-2002	nathanw	Catch up to -current
1.312.2.21	18-Oct-2002	nathanw	Catch up to -current.
1.312.2.20	17-Sep-2002	nathanw	Catch up to -current.
1.312.2.19	27-Aug-2002	nathanw	Catch up to -current.
1.312.2.18	13-Aug-2002	nathanw	Catch up to -current.
1.312.2.17	01-Aug-2002	nathanw	Catch up to -current.
1.312.2.16	20-Jun-2002	nathanw	Catch up to -current.
1.312.2.15	17-Apr-2002	nathanw	Catch up to -current.
1.312.2.14	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.312.2.13	28-Feb-2002	nathanw	Regen.
1.312.2.12	28-Feb-2002	nathanw	Regen.
1.312.2.11	28-Feb-2002	nathanw	Catch up to -current.
1.312.2.10	11-Jan-2002	nathanw	More catchup.
1.312.2.9	08-Jan-2002	nathanw	Catch up to -current.
1.312.2.8	14-Nov-2001	nathanw	Catch up to -current.
1.312.2.7	22-Oct-2001	nathanw	Catch up to -current.
1.312.2.6	08-Oct-2001	nathanw	Catch up to -current.
1.312.2.5	21-Sep-2001	nathanw	Catch up to -current.
1.312.2.4	24-Aug-2001	nathanw	Catch up with -current.
1.312.2.3	21-Jun-2001	nathanw	Catch up to -current.
1.312.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.312.2.1	05-Mar-2001	nathanw	file pcidevs.h was added on branch nathanw_sa on 2001-04-09 01:57:05 +0000
1.342.2.9	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.342.2.8	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.342.2.7	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.342.2.6	16-Mar-2002	jdolecek	Catch up with -current.
1.342.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.342.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.342.2.3	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.342.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.342.2.1	03-Aug-2001	lukem	update to -current
1.360.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.360.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.369.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.429.2.4	29-Aug-2002	gehenna	catch up with -current.
1.429.2.3	15-Jul-2002	gehenna	catch up with -current.
1.429.2.2	20-Jun-2002	gehenna	catch up with -current.
1.429.2.1	30-May-2002	gehenna	Catch up with -current.
1.433.2.36	26-Oct-2005	jmc	regen
1.433.2.35	01-Jun-2005	riz	Regen for ticket 5721.
1.433.2.34	01-May-2005	tron	Regen for ticket 5756.
1.433.2.33	19-Mar-2005	tron	Regen for ticket 1755.
1.433.2.32	11-Nov-2004	he	Pull up regenerated files (requested by itohy in ticket #1741): Add support for the Workbit NinjaSCSI-32 PCI/Cardbus SCSI driver, njs(4).
1.433.2.31	12-Jul-2004	he	Pull up regenerated files (requested by grant in ticket #1720): Add support for SATALink 3112.
1.433.2.30	22-Jun-2004	tron	Regen for ticket #1718.
1.433.2.29	06-Apr-2004	jmc	Regen (requested by briggs in ticket #1648) Pull in changes to support the Intel 82801EB/ER (ICH5) Controller. PR#25002
1.433.2.28	28-Mar-2004	jmc	Regen for ticket #1631
1.433.2.27	15-Mar-2004	jmc	Regen for ticket #1612
1.433.2.26	20-Oct-2003	cyber	Pull up revision 1.569 (requested by tron in ticket #1440): Add entry for Intel i865G Integrated Graphics Device.
1.433.2.25	10-Sep-2003	grant	Regenerate (requested by tron in ticket #1417): Correct names and descriptions for i875 entries to be consistent with other Intel chipsets.
1.433.2.24	15-Aug-2003	tron	Regen.
1.433.2.23	10-Jul-2003	tron	Regen for ticket #1365.
1.433.2.22	30-Jun-2003	grant	Regen (requested by drochner in ticket #1343): add Intel PRO/100M.
1.433.2.21	19-Jun-2003	grant	Regen (requested by hamajima in ticket #1323): Support the PowerBook G4 12-inch (and maybe 17-inch). Kauai ATA, new obio chip, and Japanese keymap for WSDISPLAY_COMPAT_RAWKBD
1.433.2.20	16-Jun-2003	grant	Regenerate (requested by bouyer in ticket #1234): Add several SiS host bridges, from FreeBSD. Remove duplicate in previous.
1.433.2.19	02-Jun-2003	tron	Regen for ticket 1143.
1.433.2.18	28-Jan-2003	jmc	Regen (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.433.2.17	27-Jan-2003	jmc	Regen from pcidevs commits to Add Intel 82801E Communications I/O Controller Hub (C-ICH)
1.433.2.16	12-Dec-2002	he	Pull up regenerated file (requested by ad in ticket #1030): Add support for Adaptec 2005S. Handle DPT ioctls() from native binaries. Fix SCSI probe problem.
1.433.2.15	28-Nov-2002	tron	Regen for ticket #1019.
1.433.2.14	26-Nov-2002	tron	Regen for tickets #1017 and #1018.
1.433.2.13	21-Nov-2002	tron	Regen for ticket #925.
1.433.2.12	16-Nov-2002	he	Pull up regenerated files (requested by ad in ticket #646): Add additional Adaptec RAID controller types. Fixes PR#17719.
1.433.2.11	07-Nov-2002	tron	Regen for ticket #452.
1.433.2.10	01-Nov-2002	tron	Regen for ticket #724.
1.433.2.9	02-Aug-2002	lukem	regen (for ticket #603)
1.433.2.8	01-Aug-2002	lukem	regen (for ticket #589)
1.433.2.7	12-Jul-2002	thorpej	Regen. pullup-1-6 ticket #488 (matt).
1.433.2.6	25-Jun-2002	lukem	regen for ticket #362 from cjs
1.433.2.5	21-Jun-2002	lukem	regen for tickes #344 / #345
1.433.2.4	11-Jun-2002	lukem	regen for ticket #249
1.433.2.3	11-Jun-2002	lukem	regen for ticket #166
1.433.2.2	11-Jun-2002	lukem	regen for ticket #139
1.433.2.1	10-Jun-2002	tv	Regen.
1.543.2.18	11-Dec-2005	christos	Sync with head.
1.543.2.17	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.543.2.16	01-Apr-2005	skrll	Sync with HEAD.
1.543.2.15	08-Mar-2005	skrll	Sync with HEAD.
1.543.2.14	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.543.2.13	04-Feb-2005	skrll	Sync with HEAD.
1.543.2.12	24-Jan-2005	skrll	Sync with HEAD.
1.543.2.11	17-Jan-2005	skrll	Sync with HEAD.
1.543.2.10	18-Dec-2004	skrll	Sync with HEAD.
1.543.2.9	29-Nov-2004	skrll	Sync with HEAD.
1.543.2.8	14-Nov-2004	skrll	Sync with HEAD.
1.543.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.543.2.6	19-Oct-2004	skrll	Sync with HEAD
1.543.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.543.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.543.2.3	03-Sep-2004	skrll	Sync with HEAD
1.543.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.543.2.1	03-Aug-2004	skrll	Sync with HEAD
1.615.2.17	12-Nov-2004	jmc	branches: 1.615.2.17.2; Regen (requested by martti in ticket #953) Regen
1.615.2.16	12-Nov-2004	jmc	Regen (requested by martti in ticket #947) Regen
1.615.2.15	08-Oct-2004	jmc	Regen for ticket #896
1.615.2.14	18-Sep-2004	he	Pull up regenerated files (requested by martin in ticket #856): Add support for PCI-X version of LSI FC929, from PR#26836.
1.615.2.13	30-Aug-2004	tron	Regen for ticket 800.
1.615.2.12	30-Aug-2004	tron	Regen for ticket 800.
1.615.2.11	22-Aug-2004	tron	Regen for ticket 781.
1.615.2.10	28-Jul-2004	tron	Regen for ticket #721.
1.615.2.9	28-Jul-2004	tron	Regen for ticket #718.
1.615.2.8	23-Jul-2004	he	Pull up regenerated files (requested by mycroft in ticket #691): Add support for SiS 741. Fixes PR#26394.
1.615.2.7	05-Jul-2004	he	Pull up regenerated files (requested by mycroft in ticket #606): Update ServerWorks chip IDs, per http://www.plasma-online.de/english/identify/picture/serverworks.html Seems remarkably more sane.
1.615.2.6	22-Jun-2004	tron	Regen for ticket #537.
1.615.2.5	13-May-2004	jdc	Regnerate (requested by tron in ticket #316)
1.615.2.4	11-May-2004	tron	Regen for ticket 284, this time with correct revision id for "pcidevs".
1.615.2.3	11-May-2004	tron	Regen for ticket 284.
1.615.2.2	05-Apr-2004	tron	Regen for ticket 54.
1.615.2.1	01-Apr-2004	jmc	Regen (requested by tron in ticket #38) Correct entry for PCI vendor id 0x11af. It is assigned to Avid Technology Inc. and not to Pro-Log Corporation. PR#25004
1.615.2.17.2.19	21-Nov-2006	ghen	Regen for ticket #10749.
1.615.2.17.2.18	11-Sep-2006	tron	Regen for ticket #10693.
1.615.2.17.2.17	04-Jun-2006	tron	Regen for ticket #10635.
1.615.2.17.2.16	11-Apr-2006	riz	Regen for ticket 10409.
1.615.2.17.2.15	17-Mar-2006	riz	Regen for ticket 10364.
1.615.2.17.2.14	01-Dec-2005	tron	Regen for ticket #10158.
1.615.2.17.2.13	01-Dec-2005	riz	Regen for ticket 5990.
1.615.2.17.2.12	21-Jul-2005	snj	Regen for ticket 5507.
1.615.2.17.2.11	05-Jul-2005	riz	Regen for ticket 2029.
1.615.2.17.2.10	21-Jun-2005	riz	Regen for ticket 2007.
1.615.2.17.2.9	19-May-2005	tron	Regen for ticket 1532.
1.615.2.17.2.8	19-May-2005	tron	Regen for ticket 1531.
1.615.2.17.2.7	19-May-2005	tron	Regen for ticket 1531.
1.615.2.17.2.6	11-May-2005	riz	Regen for ticket 1402.
1.615.2.17.2.5	10-May-2005	riz	Regen for ticket 1341.
1.615.2.17.2.4	05-May-2005	riz	Regen for ticket 1153.
1.615.2.17.2.3	01-May-2005	tron	Regen for ticket 1438.
1.615.2.17.2.2	17-Apr-2005	tron	Regen for ticket 1120.
1.615.2.17.2.1	07-Apr-2005	tron	Regen for ticket 1103.
1.686.2.1	29-Apr-2005	kent	sync with -current
1.689.2.4	26-Mar-2005	yamt	sync with head.
1.689.2.3	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.689.2.2	12-Feb-2005	yamt	sync with head.
1.689.2.1	25-Jan-2005	yamt	file pcidevs.h was added on branch yamt-km on 2005-02-12 18:17:48 +0000
1.702.2.48	10-Nov-2008	snj	Regen for ticket 1976.
1.702.2.47	08-Apr-2008	jdc	Regenerate for pcidevs 1.701.2.49. (requested by bouyer in ticket #1909).
1.702.2.46	19-Nov-2007	bouyer	Regen for ticket #1863
1.702.2.45	19-Nov-2007	bouyer	Regen for ticket 1862
1.702.2.44	14-Oct-2007	riz	Regen for ticket 1770.
1.702.2.43	14-Oct-2007	riz	Regen for ticket 1769.
1.702.2.42	13-Sep-2007	bouyer	Regen for ticket #1822
1.702.2.41	31-Mar-2007	bouyer	regen for ticket 1681
1.702.2.40	03-Mar-2007	bouyer	regen for ticket #1673
1.702.2.39	23-Dec-2006	ghen	Regen for ticket #1609.
1.702.2.38	23-Dec-2006	ghen	Regen for ticket #1608.
1.702.2.37	20-Nov-2006	ghen	Regen for ticket #1584.
1.702.2.36	27-Aug-2006	ghen	branches: 1.702.2.36.2; Regen for ticket #1479.
1.702.2.35	15-Aug-2006	ghen	Regen for ticket #1466.
1.702.2.34	03-Aug-2006	tron	Regenerate for ticket #1446.
1.702.2.33	31-Jul-2006	tron	Regenerate for ticket #1441.
1.702.2.32	30-Jul-2006	tron	Regen for ticket #1432.
1.702.2.31	04-Jun-2006	tron	Regen for ticket #1351.
1.702.2.30	04-Jun-2006	tron	Regen for ticket #1349.
1.702.2.29	04-May-2006	tron	Regen for ticket #1298.
1.702.2.28	30-Apr-2006	riz	Regen for ticket 1295.
1.702.2.27	20-Apr-2006	snj	Regen for ticket 1266.
1.702.2.26	08-Apr-2006	tron	Regen for ticket #1247.
1.702.2.25	17-Mar-2006	tron	Regen for ticket #1203.
1.702.2.24	15-Feb-2006	riz	Regen for ticket 1169.
1.702.2.23	14-Feb-2006	riz	Sigh. Regen _again_ for ticket 1168.
1.702.2.22	14-Feb-2006	riz	Regen for ticket 1168.
1.702.2.21	05-Feb-2006	riz	Regen for ticket 1156.
1.702.2.20	21-Jan-2006	snj	Regen for ticket 1126.
1.702.2.19	10-Jan-2006	snj	Regen for ticket 1112.
1.702.2.18	05-Jan-2006	riz	Regen for ticket 1086.
1.702.2.17	15-Dec-2005	tron	branches: 1.702.2.17.2; Regen for ticket #1047.
1.702.2.16	24-Nov-2005	tron	Regen for ticket #994.
1.702.2.15	21-Nov-2005	tron	Regen for ticket #957.
1.702.2.14	08-Sep-2005	tron	Regen for ticket #752.
1.702.2.13	06-Sep-2005	riz	Regen for ticket 748.
1.702.2.12	24-Jul-2005	riz	Regen for ticket 617.
1.702.2.11	28-Jun-2005	tron	Regen for ticket 485.
1.702.2.10	22-Jun-2005	tron	Regen for ticket 478.
1.702.2.9	21-Jun-2005	tron	Regen for ticket 476.
1.702.2.8	18-Jun-2005	riz	Regen for ticket 450.
1.702.2.7	28-May-2005	snj	Regen for ticket 323.
1.702.2.6	10-May-2005	tron	Regen for ticket 272.
1.702.2.5	21-Apr-2005	tron	Regen for ticket 180.
1.702.2.4	07-Apr-2005	jmc	Regen for ticket #59
1.702.2.3	07-Apr-2005	jmc	Regen for ticket #16
1.702.2.2	07-Apr-2005	jmc	Regen for ticket #15
1.702.2.1	07-Apr-2005	tron	Regen for ticket 116.
1.702.2.36.2.1	04-Oct-2007	bouyer	regen for ticket #1822
1.702.2.17.2.1	04-Oct-2007	bouyer	regen for ticket #1822
1.727.2.10	27-Feb-2008	yamt	sync with head.
1.727.2.9	11-Feb-2008	yamt	sync with head.
1.727.2.8	21-Jan-2008	yamt	sync with head
1.727.2.7	07-Dec-2007	yamt	sync with head
1.727.2.6	15-Nov-2007	yamt	sync with head.
1.727.2.5	27-Oct-2007	yamt	sync with head.
1.727.2.4	03-Sep-2007	yamt	sync with head.
1.727.2.3	26-Feb-2007	yamt	sync with head.
1.727.2.2	30-Dec-2006	yamt	sync with head.
1.727.2.1	21-Jun-2006	yamt	sync with head.
1.735.2.1	02-Nov-2005	yamt	sync with head.
1.736.2.2	29-Nov-2005	yamt	sync with head.
1.736.2.1	22-Nov-2005	yamt	sync with head.
1.752.2.4	01-Mar-2006	yamt	sync with head.
1.752.2.3	18-Feb-2006	yamt	sync with head.
1.752.2.2	01-Feb-2006	yamt	sync with head.
1.752.2.1	15-Jan-2006	yamt	sync with head.
1.759.2.1	09-Sep-2006	rpaulo	sync with head
1.760.2.3	07-Jun-2006	kardel	Sync with head.
1.760.2.2	01-Jun-2006	kardel	Sync with head.
1.760.2.1	22-Apr-2006	simonb	Sync with head.
1.770.2.8	14-Sep-2006	yamt	sync with head.
1.770.2.7	03-Sep-2006	yamt	sync with head.
1.770.2.6	11-Aug-2006	yamt	sync with head
1.770.2.5	26-Jun-2006	yamt	sync with head.
1.770.2.4	24-May-2006	yamt	sync with head.
1.770.2.3	11-Apr-2006	yamt	sync with head
1.770.2.2	01-Apr-2006	yamt	sync with head.
1.770.2.1	13-Mar-2006	yamt	sync with head.
1.772.2.3	11-May-2006	elad	sync with head
1.772.2.2	19-Apr-2006	elad	sync with head.
1.772.2.1	07-Mar-2006	elad	file pcidevs.h was added on branch elad-kernelauth on 2006-04-19 03:25:36 +0000
1.776.2.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.776.2.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.788.2.3	22-Jun-2006	chap	Complete a sync sys/ with head.
1.788.2.2	19-Jun-2006	chap	Sync with head.
1.788.2.1	19-May-2006	chap	file pcidevs.h was added on branch chap-midi on 2006-06-19 04:01:36 +0000
1.793.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.815.2.7	15-Sep-2006	tron	Regen for ticket #153.
1.815.2.6	09-Sep-2006	rpaulo	Regen for ticket #139.
1.815.2.5	07-Sep-2006	riz	Regen for ticket #115.
1.815.2.4	07-Sep-2006	riz	Regen for ticket #113
1.815.2.3	07-Sep-2006	riz	Regen for ticket #112.
1.815.2.2	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #28): sys/dev/pci/pcidevs: revision 1.819 sys/dev/pci/pcidevs.h: revision 1.819 sys/dev/pci/pcidevs_data.h: revision 1.818 (via patch) sys/dev/pci/pdcsata.c: revision 1.10 Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.815.2.1	11-Aug-2006	riz	Regen for ticket #8.
1.830.2.3	01-Feb-2007	ad	Sync with head.
1.830.2.2	12-Jan-2007	ad	Sync with head.
1.830.2.1	18-Nov-2006	ad	Sync with head.
1.831.2.3	18-Dec-2006	yamt	sync with head.
1.831.2.2	10-Dec-2006	yamt	sync with head.
1.831.2.1	22-Oct-2006	yamt	sync with head
1.854.2.25	31-Oct-2008	snj	Regen for ticket 1215.
1.854.2.24	29-Sep-2008	jdc	Regenerate for pullup ticket #1192. (addd SYMBIOS SAS1068_2, SAS1064E_2 and SAS1068E_2)
1.854.2.23	17-May-2008	bouyer	Regen for ticket 1145
1.854.2.22	17-May-2008	bouyer	Regen for ticket 1143
1.854.2.21	19-Apr-2008	bouyer	Regen for ticket 1128
1.854.2.20	09-Mar-2008	bouyer	Regen for ticket #1097
1.854.2.19	23-Feb-2008	bouyer	Regen for ticket #1061
1.854.2.18	22-Feb-2008	bouyer	Regen for ticket #1074
1.854.2.17	11-Jan-2008	xtraeme	Regen for ticket #1037.
1.854.2.16	15-Nov-2007	pavel	Backout accidental commit to the release branch.
1.854.2.15	15-Nov-2007	manu	regen
1.854.2.14	04-Nov-2007	xtraeme	Regen for ticket #954.
1.854.2.13	24-Oct-2007	pavel	regen for ticket #935.
1.854.2.12	22-Oct-2007	pavel	regen for ticket #931.
1.854.2.11	07-Oct-2007	xtraeme	Regen for ticket #923.
1.854.2.10	31-Aug-2007	pavel	regen for ticket #678
1.854.2.9	29-Aug-2007	liamjfoy	Regen for ticket #846
1.854.2.8	29-Aug-2007	liamjfoy	Pull up following revision(s) (requested by masanobu in ticket #850): sys/dev/pci/if_wm.c 1.137 sys/dev/pci/if_wm.c 1.138 sys/dev/pci/if_wm.c 1.141 sys/dev/pci/if_wm.c 1.144 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphy.c 1.12 sys/dev/mii/miidevs 1.72 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/wm.4 1.18 Fix for 82541 and 82547's reset bug. Modify wm_reset() to make some device stable. Add ICH9 support to wm. Add I82566 support to igphy. Pulled done via patch.
1.854.2.7	26-Aug-2007	liamjfoy	Regen pcidevs.h pcidevs_data.h
1.854.2.6	10-Jun-2007	bouyer	Regen for ticket 713
1.854.2.5	02-May-2007	snj	branches: 1.854.2.5.2; Regen for ticket 597.
1.854.2.4	21-Feb-2007	tron	Regenerate for ticket #446.
1.854.2.3	17-Feb-2007	tron	Regen for ticket #449.
1.854.2.2	17-Feb-2007	tron	Regen for ticket #423.
1.854.2.1	12-Dec-2006	tron	Regen for ticket for #267.
1.854.2.5.2.4	03-Jun-2008	skrll	Sync with netbsd-4.
1.854.2.5.2.3	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.854.2.5.2.2	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.854.2.5.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.868.2.5	17-May-2007	yamt	sync with head.
1.868.2.4	07-May-2007	yamt	sync with head.
1.868.2.3	15-Apr-2007	yamt	sync with head.
1.868.2.2	24-Mar-2007	yamt	sync with head.
1.868.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.871.2.9	23-Oct-2007	ad	Sync with head.
1.871.2.8	12-Oct-2007	ad	Sync with head.
1.871.2.7	09-Oct-2007	ad	Sync with head.
1.871.2.6	20-Aug-2007	ad	Sync with HEAD.
1.871.2.5	15-Jul-2007	ad	Sync with head.
1.871.2.4	09-Jun-2007	ad	Sync with head.
1.871.2.3	27-May-2007	ad	Sync with head.
1.871.2.2	10-Apr-2007	ad	Sync with head.
1.871.2.1	13-Mar-2007	ad	Sync with head.
1.872.4.2	29-Mar-2007	reinoud	Pullup to -current
1.872.4.1	18-Mar-2007	reinoud	First attempt to bring branch in sync with HEAD
1.872.2.1	11-Jul-2007	mjf	Sync with head.
1.885.6.13	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.885.6.12	21-Nov-2007	joerg	Sync with HEAD.
1.885.6.11	14-Nov-2007	joerg	Sync with HEAD.
1.885.6.10	31-Oct-2007	joerg	Sync with HEAD.
1.885.6.9	28-Oct-2007	joerg	Sync with HEAD.
1.885.6.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.885.6.7	02-Oct-2007	joerg	Sync with HEAD.
1.885.6.6	06-Sep-2007	jmcneill	Regen.
1.885.6.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.885.6.4	27-Aug-2007	jmcneill	Regen.
1.885.6.3	16-Aug-2007	jmcneill	Sync with HEAD.
1.885.6.2	09-Aug-2007	jmcneill	Sync with HEAD.
1.885.6.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.885.2.3	10-Sep-2007	skrll	Sync with HEAD.
1.885.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.885.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.887.2.2	06-Aug-2007	martin	regen
1.887.2.1	06-Aug-2007	martin	file pcidevs.h was added on branch matt-mips64 on 2007-08-06 22:40:08 +0000
1.892.2.3	23-Mar-2008	matt	sync with HEAD
1.892.2.2	09-Jan-2008	matt	sync with HEAD
1.892.2.1	06-Nov-2007	matt	sync with HEAD
1.903.2.1	14-Oct-2007	yamt	sync with head.
1.905.2.4	21-Nov-2007	bouyer	Sync with HEAD
1.905.2.3	18-Nov-2007	bouyer	Sync with HEAD
1.905.2.2	13-Nov-2007	bouyer	Sync with HEAD
1.905.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.913.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.913.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.913.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.913.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.920.4.1	11-Dec-2007	yamt	sync with head.
1.920.2.1	26-Dec-2007	ad	Sync with head.
1.921.2.3	11-Jan-2008	bouyer	Sync with HEAD
1.921.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.921.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.930.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.930.6.4	05-Oct-2008	mjf	Sync with HEAD.
1.930.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.930.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.930.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.930.2.1	24-Mar-2008	keiichi	sync with head.
1.938.4.9	09-Oct-2010	yamt	sync with head
1.938.4.8	11-Aug-2010	yamt	sync with head.
1.938.4.7	11-Mar-2010	yamt	sync with head
1.938.4.6	19-Aug-2009	yamt	sync with head.
1.938.4.5	18-Jul-2009	yamt	sync with head.
1.938.4.4	20-Jun-2009	yamt	sync with head
1.938.4.3	16-May-2009	yamt	sync with head
1.938.4.2	04-May-2009	yamt	sync with head.
1.938.4.1	16-May-2008	yamt	sync with head.
1.938.2.2	04-Jun-2008	yamt	sync with head
1.938.2.1	18-May-2008	yamt	sync with head.
1.939.2.4	10-Oct-2008	skrll	Sync with HEAD.
1.939.2.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.939.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.939.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.943.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.943.4.1	19-Oct-2008	haad	Sync with HEAD.
1.943.2.2	31-Jul-2008	simonb	Sync with head.
1.943.2.1	18-Jul-2008	simonb	Sync with head.
1.963.4.21	09-Nov-2015	sborrill	Regen
1.963.4.20	30-Apr-2015	snj	regen for ticket 1962
1.963.4.19	07-Sep-2013	bouyer	Regen for ticket 1867
1.963.4.18	19-Jun-2013	bouyer	regen for ticket #1850
1.963.4.17	20-Jan-2013	bouyer	regen for ticket #1842
1.963.4.16	17-Mar-2012	bouyer	Regen for ticket 1739
1.963.4.15	22-Nov-2010	riz	branches: 1.963.4.15.2; Regen for ticket 1470.
1.963.4.14	20-Nov-2010	riz	Regen for ticket 1362.
1.963.4.13	19-Nov-2010	riz	Regen for ticket 1359.
1.963.4.12	19-Nov-2010	riz	Regen for ticket #1358.
1.963.4.11	28-Mar-2010	snj	Regen for ticket 1326.
1.963.4.10	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.963.4.9	27-Jan-2010	sborrill	Regen
1.963.4.8	23-Jan-2010	bouyer	Regen for ticket 1254
1.963.4.7	18-Dec-2009	snj	Regen for ticket 1175.
1.963.4.6	28-Oct-2009	bouyer	Regen for ticket 1114
1.963.4.5	26-Sep-2009	snj	Regen for ticket 1021.
1.963.4.4	21-May-2009	snj	Regen for ticket 771.
1.963.4.3	05-May-2009	bouyer	Regen for ticket 737
1.963.4.2	03-May-2009	snj	Regen for ticket 722.
1.963.4.1	20-Mar-2009	msaitoh	branches: 1.963.4.1.4; Regen for ticket 587.
1.963.4.15.2.2	14-Oct-2011	matt	Regen.
1.963.4.15.2.1	07-Jan-2011	matt	Regen.
1.963.4.1.4.9	05-Nov-2013	matt	Regen.
1.963.4.1.4.8	05-Nov-2013	matt	Pull down xhci support from HEAD
1.963.4.1.4.7	03-Jan-2012	matt	Regen.
1.963.4.1.4.6	31-Dec-2011	matt	Regen.
1.963.4.1.4.5	24-Dec-2011	matt	Regen.
1.963.4.1.4.4	10-May-2010	matt	Regen.
1.963.4.1.4.3	21-Apr-2010	matt	sync to netbsd-5
1.963.4.1.4.2	07-Apr-2010	cliff	add RMI XLR PCI-X bridge
1.963.4.1.4.1	09-Nov-2009	cliff	in pcidevs: - add vendor entry for RMI - add product entry for RMI XLS integrated PCIe Interface regenerate (make) pcidevs.h and pcidevs_data.h
1.963.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.963.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.963.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.977.2.2	23-Jul-2009	jym	Sync with HEAD.
1.977.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1016.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.1016.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.1016.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1022.2.7	12-Jun-2011	rmind	sync with head
1.1022.2.6	31-May-2011	rmind	sync with head
1.1022.2.5	21-Apr-2011	rmind	sync with head
1.1022.2.4	06-Mar-2011	rmind	sync with head (and fix few botches with this)
1.1022.2.3	05-Mar-2011	rmind	sync with head
1.1022.2.2	03-Jul-2010	rmind	sync with head
1.1022.2.1	30-May-2010	rmind	sync with head
1.1052.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.1052.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.1052.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1079.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.1096.6.7	02-Jun-2012	mrg	sync to latest -current.
1.1096.6.6	29-Apr-2012	mrg	sync to latest -current.
1.1096.6.5	05-Apr-2012	mrg	sync to latest -current.
1.1096.6.4	06-Mar-2012	mrg	sync to -current
1.1096.6.3	06-Mar-2012	mrg	sync to -current
1.1096.6.2	04-Mar-2012	mrg	sync to latest -current.
1.1096.6.1	18-Feb-2012	mrg	merge to -current.
1.1096.2.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1096.2.4	16-Jan-2013	yamt	sync with (a bit old) head
1.1096.2.3	30-Oct-2012	yamt	sync with head
1.1096.2.2	23-May-2012	yamt	sync with head.
1.1096.2.1	17-Apr-2012	yamt	sync with head
1.1097.2.18	30-Apr-2015	snj	regen for ticket 1296
1.1097.2.17	19-Apr-2015	riz	Regen for ticket 1293.
1.1097.2.16	07-Dec-2014	martin	regen
1.1097.2.15	04-Dec-2014	snj	regen for ticket 1202
1.1097.2.14	09-Nov-2014	martin	Regen for ticket #1188
1.1097.2.13	07-Sep-2013	bouyer	regen for ticket 939
1.1097.2.12	14-Jul-2013	riz	Regen for ticket #906.
1.1097.2.11	22-Nov-2012	riz	Regen for ticket 691.
1.1097.2.10	18-Nov-2012	msaitoh	Regen for ticket #658
1.1097.2.9	31-Oct-2012	riz	Regen for ticket 643.
1.1097.2.8	24-Oct-2012	riz	Regen for ticket #631
1.1097.2.7	24-Oct-2012	riz	Regen for ticket 628.
1.1097.2.6	28-Jun-2012	riz	branches: 1.1097.2.6.2; Regen for ticket #362.
1.1097.2.5	23-Apr-2012	riz	Regen for ticket #191.
1.1097.2.4	16-Apr-2012	riz	Regen for ticket 180.
1.1097.2.3	03-Apr-2012	riz	Regen for ticket #151.
1.1097.2.2	03-Apr-2012	riz	Regen for ticket #149.
1.1097.2.1	22-Mar-2012	riz	Regen for ticket #137.
1.1097.2.6.2.1	18-Nov-2012	matt	Merge tickets 658 and 659.
1.1125.2.5	03-Dec-2017	jdolecek	update from HEAD
1.1125.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1125.2.3	23-Jun-2013	tls	resync from head
1.1125.2.2	25-Feb-2013	tls	resync with head
1.1125.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.1150.4.1	23-Jul-2013	riastradh	sync with HEAD
1.1150.2.2	18-May-2014	rmind	sync with head
1.1150.2.1	28-Aug-2013	rmind	sync with head
1.1177.2.1	10-Aug-2014	tls	Rebase.
1.1192.2.12	14-Aug-2018	martin	Regen for ticket #1628
1.1192.2.11	03-Jan-2018	snj	regen for ticket 1547
1.1192.2.10	25-Jul-2017	snj	regen for ticket 1413
1.1192.2.9	09-Dec-2016	snj	regen for ticket 1300
1.1192.2.8	09-Dec-2016	snj	regen for ticket 1298
1.1192.2.7	08-Dec-2016	snj	regen for ticket 1291
1.1192.2.6	19-Apr-2015	riz	branches: 1.1192.2.6.4; Regen for ticket #699
1.1192.2.5	17-Feb-2015	martin	Regen (ticket #528)
1.1192.2.4	04-Feb-2015	snj	regen for ticket 473
1.1192.2.3	28-Jan-2015	snj	regen for ticket 467
1.1192.2.2	17-Jan-2015	martin	regen (ticket #432)
1.1192.2.1	07-Nov-2014	snj	regen for ticket 190
1.1192.2.6.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.1199.2.11	28-Aug-2017	skrll	Sync with HEAD
1.1199.2.10	05-Feb-2017	skrll	Sync with HEAD
1.1199.2.9	05-Dec-2016	skrll	Sync with HEAD
1.1199.2.8	05-Oct-2016	skrll	Sync with HEAD
1.1199.2.7	09-Jul-2016	skrll	Sync with HEAD
1.1199.2.6	29-May-2016	skrll	Sync with HEAD
1.1199.2.5	19-Mar-2016	skrll	Sync with HEAD
1.1199.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1199.2.3	21-Sep-2015	skrll	Regen
1.1199.2.2	06-Jun-2015	skrll	Sync with HEAD
1.1199.2.1	06-Apr-2015	skrll	Sync with HEAD
1.1246.2.6	26-Apr-2017	pgoyette	Regen
1.1246.2.5	26-Apr-2017	pgoyette	Sync with HEAD
1.1246.2.4	20-Mar-2017	pgoyette	Sync with HEAD
1.1246.2.3	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1246.2.2	04-Nov-2016	pgoyette	Sync with HEAD
1.1246.2.1	06-Aug-2016	pgoyette	Sync with HEAD
1.1270.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.1278.2.2	11-May-2017	pgoyette	Sync with HEAD
1.1278.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.1281.2.31	03-Feb-2024	martin	regen (ticket #1931)
1.1281.2.30	18-Oct-2023	martin	regen (for ticket #1915)
1.1281.2.29	13-Oct-2023	martin	Regen for ticket #1913 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1281.2.28	13-Oct-2023	martin	Regen for ticket #1911 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1281.2.27	21-Aug-2023	martin	regen (for ticket #1887)
1.1281.2.26	29-Jul-2023	martin	regen (ticket #1855)
1.1281.2.25	21-Jun-2023	martin	Regen for ticket #1846
1.1281.2.24	18-Jan-2023	martin	Regen for ticket #1783
1.1281.2.23	11-Oct-2022	martin	regen (Ticket #1771)
1.1281.2.22	02-Aug-2022	snj	regen for ticket 1754
1.1281.2.21	03-Dec-2021	martin	regen (ticket #1710)
1.1281.2.20	23-Oct-2021	martin	Regen for ticket #1700: - Add modern QUMRANET/Red Hat VIRTIO range PCI devices. - Add some more product IDs for mcx(4). - Add current generation NVIDIA graphics cards (3050-3090 etc.). - Add RDC R6022 PCI-Host bridge. - Add a whole bunch of radeon devices. - Add Realtek RTL8821CE. - Add Intel 660p SSD, and expand the 760p description. - Add Intel I219's version number. - Add Intel I219V 15-19 and I219LM 16-19. - Add Intel I225V, I225LM and WiFi 6 AX201. - Add newer Intel PCH internal devices. - Add Intel Jasper Lake devices. - PDC20265 is Ultra/100, not 66. - Fix Realtek RTL8125 description. - Sort Cavium devices.
1.1281.2.19	20-Jul-2020	martin	Regen for ticket #1579
1.1281.2.18	20-Jul-2020	martin	regen for ticket #1578
1.1281.2.17	19-Nov-2019	martin	regen for ticket #1448
1.1281.2.16	06-Aug-2019	martin	regen (ticket #1326)
1.1281.2.15	17-Jul-2019	martin	regen
1.1281.2.14	07-Mar-2019	martin	Regen for ticket #1205
1.1281.2.13	11-Feb-2019	martin	Regen for ticket #1188
1.1281.2.12	04-Dec-2018	martin	Regen for ticket #1123
1.1281.2.11	04-Dec-2018	martin	Regen for #1119
1.1281.2.10	17-Oct-2018	martin	Regen for ticket #1064
1.1281.2.9	26-Jul-2018	snj	regen for ticket 934
1.1281.2.8	15-Jun-2018	msaitoh	Regen for ticket #878.
1.1281.2.7	07-Jun-2018	martin	Regen (ticket #846)
1.1281.2.6	18-Apr-2018	martin	regen for ticket #780
1.1281.2.5	14-Apr-2018	martin	regen (for ticket #751): Add Nvidia GeForce 9400M as found in the late 2009 13" MacBook (MacBook6,1) Add new I219 devices (using with 300 series chipset).
1.1281.2.4	09-Apr-2018	martin	Regen (ticket #723)
1.1281.2.3	08-Mar-2018	martin	Regen for ticket #616
1.1281.2.2	21-Nov-2017	martin	regen
1.1281.2.1	25-Jun-2017	snj	regen for ticket 55
1.1311.2.15	26-Jan-2019	pgoyette	Sync with HEAD
1.1311.2.14	18-Jan-2019	pgoyette	Synch with HEAD
1.1311.2.13	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1311.2.12	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.1311.2.11	20-Oct-2018	pgoyette	Sync with head
1.1311.2.10	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1311.2.9	28-Jul-2018	pgoyette	Sync with HEAD
1.1311.2.8	25-Jun-2018	pgoyette	Sync with HEAD
1.1311.2.7	21-May-2018	pgoyette	Sync with HEAD
1.1311.2.6	02-May-2018	pgoyette	Synch with HEAD
1.1311.2.5	22-Apr-2018	pgoyette	Sync with HEAD
1.1311.2.4	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.1311.2.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1311.2.2	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.1311.2.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.1334.2.3	21-Apr-2020	martin	Sync with HEAD
1.1334.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1334.2.1	10-Jun-2019	christos	Sync with HEAD
1.1371.2.26	26-Oct-2025	martin	Regen for ticket #1978
1.1371.2.25	08-May-2025	martin	regen for ticket #1944
1.1371.2.24	08-May-2025	martin	Regen for ticket #1944: Various VIA IDE controller related additions/changes.
1.1371.2.23	03-Jul-2024	martin	Regen for ticket #1852 (Brainboxes devices)
1.1371.2.22	03-Feb-2024	martin	regen (ticket #1790)
1.1371.2.21	18-Oct-2023	martin	Regen for ticket #1915
1.1371.2.20	13-Oct-2023	martin	Regen for ticket #1750 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1371.2.19	13-Oct-2023	martin	Regen for ticket #1748 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1371.2.18	21-Aug-2023	martin	regen (for ticket #1719)
1.1371.2.17	29-Jul-2023	martin	regen (ticket #1671)
1.1371.2.16	21-Jun-2023	martin	Regen for ticket #1654
1.1371.2.15	30-Jan-2023	martin	Regen for ticket #1580
1.1371.2.14	18-Jan-2023	martin	Regen for ticket #1566
1.1371.2.13	11-Oct-2022	martin	regen (Ticket #1537)
1.1371.2.12	27-Jul-2022	martin	regen for ticket #1482
1.1371.2.11	03-Dec-2021	martin	regen (ticket #1380)
1.1371.2.10	21-Oct-2021	martin	Regen (for ticket #1362)
1.1371.2.9	10-Jul-2020	martin	Regen for ticket #991
1.1371.2.8	07-Jul-2020	martin	regen
1.1371.2.7	28-Apr-2020	martin	Regen for ticket #864
1.1371.2.6	19-Mar-2020	martin	Regen (for ticket #784)
1.1371.2.5	31-Jan-2020	martin	regen (for ticket #670)
1.1371.2.4	19-Nov-2019	martin	Regen for ticket #449
1.1371.2.3	24-Oct-2019	martin	Regen for ticket #373
1.1371.2.2	24-Sep-2019	martin	regen for ticket #237 (Mellanox device IDs added)
1.1371.2.1	26-Aug-2019	martin	Regen for ticket #112
1.1383.2.2	29-Feb-2020	ad	Sync with head.
1.1383.2.1	25-Jan-2020	ad	Sync with head.
1.1392.2.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.1392.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.1406.2.3	03-Apr-2021	thorpej	Sync with HEAD.
1.1406.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.1406.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.1410.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.1412.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.1452.2.18	26-Oct-2025	martin	Regen for ticket #1184
1.1452.2.17	08-May-2025	martin	regen for ticket #1101
1.1452.2.16	08-May-2025	martin	Regen for ticket #1101: Various VIA IDE controller related additions/changes.
1.1452.2.15	06-Dec-2024	snj	regen for ticket 1024
1.1452.2.14	24-Aug-2024	martin	Regen for ticket #805
1.1452.2.13	03-Jul-2024	martin	Regen for ticket #730 (Brainboxes devices)
1.1452.2.12	03-Feb-2024	martin	regen (ticket #560)
1.1452.2.11	03-Nov-2023	martin	Regen for ticket #447 (Add Mellanox ConnectX-6 Lx)
1.1452.2.10	18-Oct-2023	martin	Regen for ticket #421
1.1452.2.9	13-Oct-2023	martin	Regen for ticket #406 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1452.2.8	13-Oct-2023	martin	Regen for ticket #403 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1452.2.7	08-Oct-2023	martin	Regen for ticket #393 (Add missing I225/I226 series devices)
1.1452.2.6	21-Aug-2023	martin	regen (for ticket #333)
1.1452.2.5	29-Jul-2023	martin	regen (ticket #252)
1.1452.2.4	21-Jun-2023	martin	Regen for ticket #209
1.1452.2.3	21-Jun-2023	martin	Regen for ticket #209
1.1452.2.2	30-Jan-2023	martin	Regen for ticket #61
1.1452.2.1	18-Jan-2023	martin	Regen for ticket #50
1.1486.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1506.2.1	26-Oct-2025	martin	Regen for ticket #71
1.1509	27-Oct-2025	andvar	regen.
1.1508	21-Oct-2025	pgoyette	regen
1.1507	21-Oct-2025	pgoyette	Regen for vendor MediaTek
1.1506	17-Oct-2025	andvar	regen.
1.1505	03-May-2025	rin	branches: 1.1505.2; pcidevs{,_data}.h: Regen for EXAR XR17V358
1.1504	01-May-2025	andvar	regen.
1.1503	01-May-2025	andvar	regen.
1.1502	28-Apr-2025	andvar	regen.
1.1501	28-Apr-2025	andvar	regen.
1.1500	20-Apr-2025	andvar	regen.
1.1499	05-Apr-2025	andvar	regen.
1.1498	03-Apr-2025	andvar	regen.
1.1497	03-Mar-2025	andvar	regen.
1.1496	23-Jan-2025	andvar	regen.
1.1495	23-Jan-2025	mrg	regen
1.1494	18-Jan-2025	mrg	regen.
1.1493	03-Jan-2025	mrg	regen.
1.1492	12-Dec-2024	msaitoh	regen.
1.1491	11-Nov-2024	msaitoh	Regen.
1.1490	19-Oct-2024	mlelstv	Regen
1.1489	16-Oct-2024	msaitoh	Regen.
1.1488	20-Aug-2024	dogcow	regen
1.1487	17-Jul-2024	mrg	regen after intel arc.
1.1486	11-Jul-2024	msaitoh	Regen.
1.1485	23-Jun-2024	jakllsch	branches: 1.1485.2; Regen. For PR 58358
1.1484	12-May-2024	msaitoh	Regen.
1.1483	06-Apr-2024	andvar	regen.
1.1482	28-Feb-2024	jakllsch	regen
1.1481	18-Jan-2024	msaitoh	regen.
1.1480	28-Dec-2023	msaitoh	pcidevs: Regen.
1.1479	13-Nov-2023	jnemeth	regen for newer Areca boards
1.1478	26-Oct-2023	msaitoh	pcidevs: regen.
1.1477	11-Oct-2023	msaitoh	Regen.
1.1476	10-Oct-2023	abs	Regenerate for ASMEDIA 0x0624
1.1475	08-Oct-2023	msaitoh	Regen.
1.1474	04-Oct-2023	rin	Regen for pcidevs rev 1.1494
1.1473	04-Oct-2023	msaitoh	Regen.
1.1472	10-Aug-2023	andvar	regen.
1.1471	09-Aug-2023	msaitoh	Regen.
1.1470	09-Aug-2023	msaitoh	Regen.
1.1469	05-Aug-2023	andvar	regen.
1.1468	05-Aug-2023	andvar	Regen.
1.1467	04-Aug-2023	msaitoh	Regen.
1.1466	04-Aug-2023	msaitoh	Regen.
1.1465	31-Jul-2023	tnn	pcidevs: regen
1.1464	28-Jul-2023	msaitoh	Regen.
1.1463	23-Jul-2023	msaitoh	Regen.
1.1462	22-Jul-2023	msaitoh	regen.
1.1461	22-Jul-2023	msaitoh	Add Alder Lake-N and Raptor Lake devices.
1.1460	06-Apr-2023	msaitoh	Regen.
1.1459	14-Feb-2023	msaitoh	Regen.
1.1458	03-Feb-2023	msaitoh	Regen.
1.1457	29-Jan-2023	thorpej	Regen for EXAR XR17V354. PR kern/57202
1.1456	18-Jan-2023	msaitoh	Regen.
1.1455	14-Jan-2023	ryo	regen
1.1454	04-Jan-2023	msaitoh	Regen.
1.1453	30-Dec-2022	msaitoh	Regen.
1.1452	28-Dec-2022	msaitoh	Regen.
1.1451	22-Nov-2022	msaitoh	branches: 1.1451.2; Regen.
1.1450	26-Oct-2022	msaitoh	Regen.
1.1449	26-Oct-2022	msaitoh	Regen.
1.1448	07-Oct-2022	snj	regen
1.1447	02-Oct-2022	martin	regen (AMD FCH SATA D added)
1.1446	01-Oct-2022	msaitoh	Regen.
1.1445	28-Sep-2022	msaitoh	Regen.
1.1444	13-Sep-2022	msaitoh	Regen.
1.1443	11-Sep-2022	mrg	regen
1.1442	27-Jun-2022	msaitoh	Regen.
1.1441	17-Jun-2022	msaitoh	Regen.
1.1440	02-Jun-2022	mrg	regen.
1.1439	11-May-2022	markd	Regen
1.1438	04-May-2022	martin	Regen
1.1437	04-May-2022	nia	regen
1.1436	27-Apr-2022	msaitoh	Regen.
1.1435	22-Apr-2022	skrll	Regen
1.1434	24-Mar-2022	yamaguchi	regen
1.1433	16-Mar-2022	yamaguchi	Regen.
1.1432	15-Feb-2022	msaitoh	Regen.
1.1431	02-Feb-2022	msaitoh	Regen.
1.1430	25-Jan-2022	msaitoh	Regen.
1.1429	01-Jan-2022	msaitoh	Regen.
1.1428	30-Nov-2021	msaitoh	Regen.
1.1427	03-Nov-2021	msaitoh	Regen.
1.1426	27-Oct-2021	msaitoh	Regen.
1.1425	27-Oct-2021	msaitoh	Regen.
1.1424	13-Oct-2021	msaitoh	Regen.
1.1423	12-Oct-2021	msaitoh	Regen.
1.1422	11-Oct-2021	msaitoh	Regen.
1.1421	02-Oct-2021	mrg	regen.
1.1420	23-Sep-2021	msaitoh	Regen.
1.1419	06-Sep-2021	mrg	regen.
1.1418	03-Sep-2021	mrg	regen.
1.1417	30-Aug-2021	andvar	regen.
1.1416	30-Aug-2021	mrg	regen.
1.1415	30-Aug-2021	mrg	regen.
1.1414	17-Jul-2021	mrg	regen.
1.1413	12-Jul-2021	msaitoh	Regen.
1.1412	29-Jun-2021	pgoyette	Regen for new sys/dev/devlist2h.awk
1.1411	21-Apr-2021	jakllsch	branches: 1.1411.2; regenerate
1.1410	02-Apr-2021	rin	regen
1.1409	30-Jan-2021	jmcneill	branches: 1.1409.2; regen
1.1408	29-Dec-2020	skrll	regen
1.1407	20-Nov-2020	msaitoh	Regen.
1.1406	19-Nov-2020	reinoud	Regenated from pcidevs
1.1405	14-Jul-2020	jdolecek	branches: 1.1405.2; regen
1.1404	09-Jul-2020	msaitoh	Regen.
1.1403	24-Jun-2020	thorpej	Regen.
1.1402	13-Jun-2020	ryo	regen
1.1401	11-Jun-2020	jdolecek	regen
1.1400	26-May-2020	martin	regen
1.1399	01-May-2020	jmcneill	regen
1.1398	23-Apr-2020	msaitoh	Regen.
1.1397	22-Apr-2020	christos	regen
1.1396	22-Apr-2020	prlw1	regen
1.1395	22-Apr-2020	msaitoh	Regen.
1.1394	18-Apr-2020	simonb	Regen.
1.1393	16-Apr-2020	msaitoh	Regen.
1.1392	10-Apr-2020	christos	regen
1.1391	07-Apr-2020	christos	branches: 1.1391.2; regen
1.1390	01-Apr-2020	msaitoh	Regen.
1.1389	11-Mar-2020	msaitoh	Regen.
1.1388	11-Mar-2020	fcambus	Regen.
1.1387	08-Mar-2020	msaitoh	Regen.
1.1386	05-Mar-2020	msaitoh	Regen.
1.1385	19-Feb-2020	mrg	regen.
1.1384	28-Jan-2020	msaitoh	Regen.
1.1383	21-Jan-2020	msaitoh	Regen.
1.1382	01-Jan-2020	ryo	branches: 1.1382.2; regen
1.1381	29-Dec-2019	sevan	regen
1.1380	18-Dec-2019	msaitoh	Regen.
1.1379	27-Nov-2019	msaitoh	Regen.
1.1378	14-Nov-2019	msaitoh	Regen.
1.1377	08-Nov-2019	msaitoh	Regen.
1.1376	29-Oct-2019	msaitoh	Regen.
1.1375	12-Oct-2019	skrll	regen
1.1374	21-Sep-2019	msaitoh	Regen.
1.1373	21-Sep-2019	msaitoh	Regen.
1.1372	24-Aug-2019	jmcneill	regen
1.1371	21-Aug-2019	mrg	regen for new micron/crucial/amd entries, and PCIe spelling.
1.1370	28-Jul-2019	mlelstv	branches: 1.1370.2; regen
1.1369	24-Jul-2019	msaitoh	Regen.
1.1368	18-Jul-2019	msaitoh	Regen.
1.1367	12-Jul-2019	msaitoh	Regen.
1.1366	12-Jul-2019	msaitoh	Regen.
1.1365	09-Jul-2019	msaitoh	Regen.
1.1364	05-Jul-2019	msaitoh	Regen.
1.1363	02-Jul-2019	msaitoh	Regen.
1.1362	04-Jun-2019	msaitoh	Regen.
1.1361	01-Jun-2019	jdolecek	regen
1.1360	30-May-2019	mlelstv	regen
1.1359	15-Apr-2019	msaitoh	Regen.
1.1358	15-Apr-2019	msaitoh	Regen.
1.1357	12-Apr-2019	msaitoh	Regen.
1.1356	11-Apr-2019	mrg	regen.
1.1355	08-Mar-2019	msaitoh	Regen.
1.1354	05-Mar-2019	msaitoh	Regen.
1.1353	20-Feb-2019	msaitoh	Regen.
1.1352	20-Feb-2019	msaitoh	Regen.
1.1351	14-Feb-2019	msaitoh	Regen.
1.1350	08-Feb-2019	msaitoh	Regen.
1.1349	06-Feb-2019	msaitoh	Regen.
1.1348	28-Jan-2019	msaitoh	Regen.
1.1347	26-Jan-2019	msaitoh	Regen.
1.1346	21-Jan-2019	msaitoh	Regen.
1.1345	26-Dec-2018	msaitoh	Regen.
1.1344	03-Dec-2018	bouyer	Regen for new Symbios devices
1.1343	29-Nov-2018	jmcneill	regen
1.1342	24-Nov-2018	skrll	Regen
1.1341	12-Nov-2018	msaitoh	Regen.
1.1340	08-Nov-2018	msaitoh	Regen.
1.1339	16-Oct-2018	jmcneill	regen
1.1338	15-Oct-2018	uwe	Adaptec RAID 2445 and 2805
1.1337	23-Aug-2018	msaitoh	Regen.
1.1336	30-Jul-2018	msaitoh	Regen.
1.1335	30-Jul-2018	msaitoh	Regen.
1.1334	06-Jul-2018	prlw1	regen
1.1333	20-Jun-2018	msaitoh	branches: 1.1333.2; Regen.
1.1332	14-Jun-2018	msaitoh	Regen.
1.1331	12-Jun-2018	jdolecek	regen
1.1330	09-Jun-2018	jdolecek	regen
1.1329	31-May-2018	mrg	regen.
1.1328	31-May-2018	msaitoh	Regen.
1.1327	28-May-2018	msaitoh	Regen.
1.1326	26-May-2018	jakllsch	regen
1.1325	24-May-2018	msaitoh	Regen.
1.1324	21-May-2018	jnemeth	regen
1.1323	03-May-2018	maya	regen
1.1322	28-Apr-2018	mlelstv	regen
1.1321	24-Apr-2018	maya	regen
1.1320	17-Apr-2018	nonaka	regen
1.1319	16-Apr-2018	msaitoh	Regen.
1.1318	09-Apr-2018	msaitoh	Regen.
1.1317	09-Apr-2018	msaitoh	Regen.
1.1316	04-Apr-2018	msaitoh	Regen.
1.1315	02-Apr-2018	msaitoh	Regen.
1.1314	31-Mar-2018	sevan	regen
1.1313	26-Mar-2018	msaitoh	Regen.
1.1312	20-Mar-2018	nonaka	regen
1.1311	16-Mar-2018	msaitoh	Regen.
1.1310	04-Mar-2018	jdolecek	branches: 1.1310.2; regen
1.1309	04-Mar-2018	jdolecek	regen
1.1308	02-Mar-2018	msaitoh	Regen.
1.1307	24-Feb-2018	mlelstv	Regen.
1.1306	22-Feb-2018	msaitoh	Regen.
1.1305	22-Feb-2018	msaitoh	Regen.
1.1304	22-Feb-2018	msaitoh	Regen.
1.1303	22-Feb-2018	msaitoh	Regen.
1.1302	21-Feb-2018	msaitoh	Regen.
1.1301	17-Feb-2018	sevan	regen
1.1300	16-Feb-2018	sevan	regen
1.1299	25-Jan-2018	sborrill	regen
1.1298	24-Jan-2018	sevan	regen
1.1297	21-Jan-2018	sevan	regen
1.1296	09-Jan-2018	msaitoh	Regen.
1.1295	09-Jan-2018	msaitoh	Regen.
1.1294	18-Dec-2017	msaitoh	Regen.
1.1293	09-Nov-2017	msaitoh	Regen.
1.1292	30-Oct-2017	sevan	regen
1.1291	20-Oct-2017	christos	regen
1.1290	15-Sep-2017	hikaru	Regen.
1.1289	17-Aug-2017	msaitoh	Regen.
1.1288	02-Aug-2017	msaitoh	regen.
1.1287	26-Jul-2017	msaitoh	Regen.
1.1286	22-Jul-2017	msaitoh	Regen.
1.1285	07-Jul-2017	msaitoh	Regen.
1.1284	04-Jul-2017	msaitoh	Regen.
1.1283	27-Jun-2017	wiz	regen (comment change only)
1.1282	17-Jun-2017	jdolecek	regen
1.1281	06-Jun-2017	msaitoh	Regen.
1.1280	23-May-2017	nonaka	branches: 1.1280.2; regen
1.1279	08-May-2017	msaitoh	- Add RICOH R5C576 SD Card Controller. - Add some SiS devices. Taken from OpenBSD. - Change SiS 0x6325 from 6325 AGP VGA to 650 AGP VGA.
1.1278	27-Apr-2017	msaitoh	regen.
1.1277	19-Apr-2017	msaitoh	branches: 1.1277.2; regen.
1.1276	13-Apr-2017	msaitoh	regen.
1.1275	31-Mar-2017	skrll	regen
1.1274	31-Mar-2017	msaitoh	regen.
1.1273	23-Mar-2017	msaitoh	regen.
1.1272	14-Mar-2017	nonaka	regen
1.1271	27-Feb-2017	msaitoh	regen.
1.1270	16-Jan-2017	msaitoh	regen.
1.1269	15-Jan-2017	msaitoh	branches: 1.1269.2; regen.
1.1268	15-Jan-2017	msaitoh	regen.
1.1267	10-Jan-2017	nonaka	regen
1.1266	27-Dec-2016	msaitoh	Regen.
1.1265	26-Dec-2016	msaitoh	Regen.
1.1264	01-Dec-2016	msaitoh	Regen.
1.1263	23-Nov-2016	christos	regen
1.1262	28-Oct-2016	msaitoh	regen.
1.1261	26-Oct-2016	msaitoh	regen.
1.1260	24-Oct-2016	msaitoh	Regen.
1.1259	20-Oct-2016	msaitoh	regen.
1.1258	19-Oct-2016	msaitoh	regen.
1.1257	04-Oct-2016	kiyohara	Regen.
1.1256	16-Sep-2016	msaitoh	regen.
1.1255	01-Sep-2016	msaitoh	regen.
1.1254	23-Aug-2016	msaitoh	regen.
1.1253	23-Aug-2016	msaitoh	regen.
1.1252	23-Aug-2016	msaitoh	regen.
1.1251	19-Aug-2016	msaitoh	regen.
1.1250	17-Aug-2016	msaitoh	regen.
1.1249	17-Aug-2016	msaitoh	regen.
1.1248	08-Aug-2016	msaitoh	regen.
1.1247	29-Jul-2016	nonaka	regen
1.1246	28-Jul-2016	nonaka	regen
1.1245	06-Jul-2016	christos	branches: 1.1245.2; regen
1.1244	27-Jun-2016	pgoyette	Regen
1.1243	01-Jun-2016	pgoyette	Regen
1.1242	30-May-2016	nonaka	regen
1.1241	27-Apr-2016	christos	regen
1.1240	26-Apr-2016	msaitoh	regen.
1.1239	12-Feb-2016	mrg	regen.
1.1238	01-Feb-2016	msaitoh	Regen.
1.1237	22-Jan-2016	msaitoh	Regen.
1.1236	01-Dec-2015	msaitoh	regen.
1.1235	01-Dec-2015	msaitoh	regen.
1.1234	16-Nov-2015	msaitoh	Regen.
1.1233	16-Nov-2015	msaitoh	Regen.
1.1232	06-Nov-2015	nonaka	regen
1.1231	22-Oct-2015	msaitoh	Regen.
1.1230	21-Oct-2015	nonaka	regen
1.1229	19-Oct-2015	tnn	regen
1.1228	28-Aug-2015	nonaka	regen
1.1227	27-Aug-2015	nonaka	regen
1.1226	26-Aug-2015	msaitoh	regen.
1.1225	23-Aug-2015	jakllsch	pcidevs .h regen
1.1224	14-Aug-2015	nonaka	regen
1.1223	14-Aug-2015	knakahara	regen
1.1222	13-Aug-2015	msaitoh	regen.
1.1221	26-Jun-2015	msaitoh	regen.
1.1220	02-Jun-2015	nonaka	regen
1.1219	16-May-2015	msaitoh	regen.
1.1218	13-May-2015	nonaka	regen
1.1217	11-May-2015	msaitoh	regen.
1.1216	09-May-2015	christos	regen
1.1215	04-May-2015	ryo	regen
1.1214	03-May-2015	msaitoh	regen.
1.1213	02-Apr-2015	tnn	regen
1.1212	20-Mar-2015	msaitoh	regen.
1.1211	20-Mar-2015	msaitoh	regen.
1.1210	18-Mar-2015	msaitoh	regen.
1.1209	04-Mar-2015	knakahara	regen.
1.1208	15-Feb-2015	msaitoh	regen.
1.1207	25-Jan-2015	msaitoh	Regen.
1.1206	18-Jan-2015	christos	regen
1.1205	16-Jan-2015	msaitoh	regen.
1.1204	14-Jan-2015	msaitoh	regen.
1.1203	31-Dec-2014	veego	Regen
1.1202	23-Dec-2014	msaitoh	Regen.
1.1201	17-Dec-2014	hikaru	Regen.
1.1200	15-Dec-2014	msaitoh	Regen.
1.1199	15-Dec-2014	msaitoh	regen.
1.1198	30-Oct-2014	nonaka	branches: 1.1198.2; regen
1.1197	29-Oct-2014	nonaka	regen
1.1196	02-Oct-2014	msaitoh	Regen.
1.1195	25-Sep-2014	msaitoh	Regen.
1.1194	21-Sep-2014	christos	regen
1.1193	16-Sep-2014	msaitoh	Regen.
1.1192	01-Sep-2014	msaitoh	Regen.
1.1191	30-Jul-2014	msaitoh	branches: 1.1191.2; regen.
1.1190	25-Jul-2014	msaitoh	Regen.
1.1189	08-Jul-2014	msaitoh	Regen.
1.1188	03-Jul-2014	msaitoh	Regen.
1.1187	02-Jul-2014	msaitoh	Regen.
1.1186	23-Jun-2014	msaitoh	Regen.
1.1185	09-Jun-2014	msaitoh	regen.
1.1184	30-May-2014	msaitoh	Regen.
1.1183	28-May-2014	msaitoh	Regen.
1.1182	28-May-2014	msaitoh	regen.
1.1181	25-May-2014	njoly	Regen for GeForce GT 640M.
1.1180	26-Apr-2014	mrg	regen.
1.1179	09-Apr-2014	msaitoh	regen
1.1178	09-Apr-2014	msaitoh	regen.
1.1177	07-Apr-2014	njoly	Regen for Intel Ivy Bridge PCI Express Root Port.
1.1176	15-Mar-2014	kiyohara	branches: 1.1176.2; Regen.
1.1175	14-Feb-2014	njoly	Regen for GeForce 210 High Definition Audio Controller.
1.1174	26-Jan-2014	msaitoh	Regen.
1.1173	21-Jan-2014	msaitoh	Regen.
1.1172	10-Jan-2014	msaitoh	regen.
1.1171	09-Jan-2014	christos	regen
1.1170	09-Jan-2014	msaitoh	regen.
1.1169	04-Jan-2014	msaitoh	regen.
1.1168	01-Jan-2014	msaitoh	regen.
1.1167	30-Dec-2013	msaitoh	regen.
1.1166	27-Dec-2013	msaitoh	regen.
1.1165	25-Dec-2013	msaitoh	regen.
1.1164	23-Dec-2013	msaitoh	regen.
1.1163	22-Dec-2013	nonaka	regen
1.1162	21-Dec-2013	msaitoh	regen.
1.1161	15-Dec-2013	pgoyette	Regen
1.1160	13-Dec-2013	pedro	regen
1.1159	12-Dec-2013	nonaka	regen
1.1158	08-Dec-2013	msaitoh	regen.
1.1157	13-Nov-2013	msaitoh	regen.
1.1156	07-Nov-2013	msaitoh	Regen.
1.1155	07-Nov-2013	msaitoh	Add Quark X1000 devices.
1.1154	07-Nov-2013	msaitoh	Regen.
1.1153	14-Sep-2013	soren	Regen.
1.1152	04-Sep-2013	christos	regen
1.1151	23-Aug-2013	christos	regen
1.1150	17-Jul-2013	soren	Regen.
1.1149	13-Jun-2013	matt	branches: 1.1149.2; 1.1149.4; Regen.
1.1148	05-Jun-2013	msaitoh	regen.
1.1147	04-Jun-2013	msaitoh	regen
1.1146	03-Jun-2013	msaitoh	Regen.
1.1145	20-Apr-2013	rkujawa	Regen PCI devs database.
1.1144	18-Apr-2013	msaitoh	regen.
1.1143	18-Apr-2013	msaitoh	regen.
1.1142	06-Mar-2013	msaitoh	regen.
1.1141	03-Mar-2013	msaitoh	regen.
1.1140	27-Feb-2013	msaitoh	regen.
1.1139	20-Feb-2013	msaitoh	regen.
1.1138	30-Jan-2013	msaitoh	regen.
1.1137	06-Dec-2012	msaitoh	Regen.
1.1136	05-Dec-2012	christos	Regen
1.1135	29-Nov-2012	msaitoh	regen.
1.1134	21-Oct-2012	matt	Regen.
1.1133	20-Oct-2012	matt	Regen.
1.1132	19-Oct-2012	prlw1	regen
1.1131	29-Sep-2012	gsutre	regen
1.1130	25-Sep-2012	gson	regen
1.1129	23-Sep-2012	chs	regen
1.1128	18-Sep-2012	matt	Regen.
1.1127	17-Sep-2012	matt	Regen.
1.1126	17-Sep-2012	tsutsui	Regen from pcidevs rev 1.1134: > Add Broadcom BCM57762 Gigabit Ethernet, per PR kern/46961.
1.1125	13-Sep-2012	nonaka	regen.
1.1124	10-Aug-2012	bouyer	branches: 1.1124.2; Regen: add some PCI AHCI controllers.
1.1123	07-Aug-2012	msaitoh	regen.
1.1122	06-Aug-2012	msaitoh	regen.
1.1121	05-Aug-2012	msaitoh	regen.
1.1120	05-Aug-2012	bouyer	Regen: add MegaRAID SAS2208
1.1119	27-Jul-2012	matt	Regen.
1.1118	26-Jul-2012	uwe	Regen for VirtualBox ids.
1.1117	25-Jul-2012	msaitoh	regen.
1.1116	23-Jul-2012	msaitoh	regen.
1.1115	22-Jul-2012	matt	Regen.
1.1114	20-Jul-2012	msaitoh	regen.
1.1113	19-Jul-2012	msaitoh	regen.
1.1112	12-Jul-2012	kiyohara	Regen.
1.1111	26-Jun-2012	drochner	regen
1.1110	25-May-2012	msaitoh	regen for Intel I350 (wm)
1.1109	18-Apr-2012	bouyer	Regen: Add various newer LSI RAID/SAS controllers
1.1108	13-Apr-2012	cegger	regen.
1.1107	09-Apr-2012	mlelstv	regen
1.1106	08-Apr-2012	jakllsch	regen.
1.1105	08-Apr-2012	jakllsch	regen.
1.1104	07-Apr-2012	christos	regen
1.1103	04-Apr-2012	uwe	Regen for Marvell Yukon 88e8040 ethernet.
1.1102	28-Mar-2012	pgoyette	Regen for JMB362
1.1101	21-Mar-2012	sborrill	Regen
1.1100	18-Mar-2012	tsutsui	Regen from pcidevs rev 1.1106: > Add VIA VX900 media system processor IDE controller. > PR/46166 from Hiroshi Tokuda.
1.1099	03-Mar-2012	nonaka	regen.
1.1098	02-Mar-2012	nonaka	regen.
1.1097	02-Mar-2012	nonaka	regen.
1.1096	15-Feb-2012	tsutsui	branches: 1.1096.2; Regen from pcidevs rev 1.1102: > Add VIA VX900 Host Bridge, to handle its quirk. PR/46018 > Ok releng@
1.1095	18-Oct-2011	jmcneill	branches: 1.1095.2; 1.1095.6; regen
1.1094	06-Oct-2011	jmcneill	regen
1.1093	04-Oct-2011	njoly	Regen for 3WARE 9750 Series RAID.
1.1092	23-Sep-2011	njoly	Regen for NVIDIA GeForce 210.
1.1091	05-Sep-2011	msaitoh	regen.
1.1090	02-Sep-2011	msaitoh	regen.
1.1089	15-Aug-2011	msaitoh	regen.
1.1088	11-Aug-2011	bouyer	Regen: add SiS 315 Pro VGA
1.1087	09-Aug-2011	mbalmer	re-gen after PWDOG1 addition.
1.1086	06-Aug-2011	jakllsch	regen
1.1085	06-Aug-2011	mbalmer	re-gen.
1.1084	04-Aug-2011	jakllsch	regen
1.1083	23-Jul-2011	matt	Regen.
1.1082	28-Jun-2011	sborrill	Regen
1.1081	27-Jun-2011	nonaka	regen.
1.1080	07-Jun-2011	cegger	regen.
1.1079	06-Jun-2011	msaitoh	regen.
1.1078	28-May-2011	phx	branches: 1.1078.2; regen
1.1077	28-May-2011	phx	regen
1.1076	28-May-2011	ryo	Regen.
1.1075	27-May-2011	matt	Regen.
1.1074	22-May-2011	mrg	regenerate.
1.1073	21-May-2011	msaitoh	regen
1.1072	20-May-2011	msaitoh	regen.
1.1071	15-May-2011	msaitoh	regen.
1.1070	14-May-2011	martin	Regen
1.1069	13-May-2011	msaitoh	regen.
1.1068	02-May-2011	manu	Add support for new SUNIX PCI seriel and/or parallel adapters, with new SUNIX vendor Id 0x1fd4 and product Id 0x1999. Constants were taken from Linux open source driver distributed by SUNIX. The following models should be supported now: 5008 1 parallel port 5016 16 serial ports, only 8 supported right now (PUC_MAX_PORTS limit) 5027 1 serial port 5037 2 serial ports 5056 4 serial ports 5066 8 serial ports, this model is the only one that was actually tested. 5069 1 serial port and 1 parallel port 5079 2 serial ports and 1 parallel port 5099 4 serial ports and 1 parallel port
1.1067	13-Apr-2011	wiz	regen (+ ATI Radeon HD4650)
1.1066	04-Apr-2011	bouyer	Regen: Add RDC PCI-ISA bridge
1.1065	04-Apr-2011	dyoung	Regen.
1.1064	16-Mar-2011	matt	regen.
1.1063	05-Mar-2011	matt	Regen.
1.1062	25-Feb-2011	jmcneill	regen
1.1061	25-Feb-2011	jmcneill	regen
1.1060	24-Feb-2011	matt	Regen.
1.1059	24-Feb-2011	matt	Regen.
1.1058	23-Feb-2011	jmcneill	regen
1.1057	22-Feb-2011	jmcneill	regen
1.1056	21-Feb-2011	njoly	Regen: Add two more Intel 82Q45 devices.
1.1055	20-Feb-2011	msaitoh	regen.
1.1054	18-Feb-2011	jmcneill	regen
1.1053	03-Feb-2011	njoly	Regen for Nvidia devices addition.
1.1052	26-Jan-2011	bouyer	Regen: Add RDC Semiconductor R6040 10/100 Ethernet
1.1051	11-Jan-2011	dyoung	branches: 1.1051.2; 1.1051.4; Regen.
1.1050	16-Dec-2010	wiz	regen.
1.1049	11-Dec-2010	matt	Regen
1.1048	11-Nov-2010	cegger	regen.
1.1047	06-Nov-2010	jakllsch	regen.
1.1046	14-Oct-2010	taca	Regen.
1.1045	26-Aug-2010	dogcow	Regen for BCM43xx update.
1.1044	01-Aug-2010	msaitoh	regen.
1.1043	30-Jul-2010	njoly	Regen for ICH10 devices update.
1.1042	29-Jul-2010	njoly	Regen for Intel LPC updates.
1.1041	29-Jul-2010	njoly	Regen for Matrox MGA G200eW.
1.1040	28-Jul-2010	msaitoh	regen.
1.1039	13-Jul-2010	msaitoh	regen.
1.1038	13-Jul-2010	kiyohara	Regen.
1.1037	08-Jul-2010	msaitoh	regen.
1.1036	08-Jul-2010	msaitoh	regen
1.1035	07-Jul-2010	msaitoh	regen.
1.1034	06-Jul-2010	msaitoh	regen.
1.1033	06-Jul-2010	msaitoh	regen.
1.1032	02-Jul-2010	mrg	regen.
1.1031	16-Jun-2010	riz	Regen for added Intel IDs.
1.1030	08-Jun-2010	pgoyette	Regen
1.1029	08-Jun-2010	mrg	regenerate
1.1028	15-May-2010	jakllsch	Regen.
1.1027	12-May-2010	christos	regen
1.1026	10-May-2010	matt	Regen.
1.1025	08-May-2010	pgoyette	Regen
1.1024	07-May-2010	kiyohara	Regen.
1.1023	17-Apr-2010	christos	regen
1.1022	02-Apr-2010	jakllsch	regen
1.1021	14-Mar-2010	mrg	branches: 1.1021.2; regenerate
1.1020	25-Feb-2010	macallan	regen
1.1019	16-Feb-2010	msaitoh	regen.
1.1018	09-Feb-2010	hubertf	regen
1.1017	09-Feb-2010	msaitoh	regen.
1.1016	08-Feb-2010	msaitoh	regen.
1.1015	03-Feb-2010	macallan	branches: 1.1015.2; regen
1.1014	30-Jan-2010	jakllsch	regen.
1.1013	24-Jan-2010	msaitoh	regen.
1.1012	22-Jan-2010	hubertf	regen - thanks for the reminder @ veego
1.1011	19-Jan-2010	mbalmer	regenerate
1.1010	14-Jan-2010	sborrill	Regen
1.1009	12-Jan-2010	msaitoh	regen
1.1008	12-Jan-2010	nonaka	regen (Added RENESAS SH7780/SH7785 PCI Controller.)
1.1007	11-Jan-2010	msaitoh	regen
1.1006	07-Jan-2010	msaitoh	regen
1.1005	07-Jan-2010	jdc	Regenerate for: Add Sun Cassini/NS Saturn devices.
1.1004	03-Jan-2010	jnemeth	regen
1.1003	29-Dec-2009	msaitoh	regen
1.1002	30-Nov-2009	sborrill	Regen
1.1001	21-Nov-2009	jakllsch	regen.
1.1000	18-Nov-2009	bouyer	Regen: Complete BROADCOM BCM5709 family, and add BCM5716 family.
1.999	18-Nov-2009	jakllsch	regen.
1.998	13-Oct-2009	cegger	regen.
1.997	13-Oct-2009	cegger	regen.
1.996	05-Oct-2009	cegger	regen.
1.995	20-Sep-2009	christos	Regen
1.994	01-Aug-2009	matt	Regen.
1.993	07-Jul-2009	cegger	regen.
1.992	29-Jun-2009	hubertf	regen
1.991	12-Jun-2009	rjs	Regen.
1.990	11-Jun-2009	cegger	regen.
1.989	29-May-2009	skrll	Regen.
1.988	18-May-2009	skrll	Regen
1.987	13-May-2009	cegger	regen.
1.986	09-May-2009	sborrill	regen
1.985	30-Apr-2009	skrll	Regen (again)
1.984	30-Apr-2009	skrll	Regen.
1.983	28-Apr-2009	ginsbach	Regen for Mellanox additions.
1.982	25-Apr-2009	tsutsui	Regen from pcidevs rev 1.983: > Tweak some Realtek re(4) device names.
1.981	07-Apr-2009	msaitoh	regen
1.980	10-Mar-2009	msaitoh	regen
1.979	26-Feb-2009	mjf	Regen.
1.978	19-Feb-2009	sborrill	Regen
1.977	11-Feb-2009	markd	regen
1.976	06-Feb-2009	jakllsch	branches: 1.976.2; regen.
1.975	27-Jan-2009	markd	regen
1.974	19-Jan-2009	jmcneill	regen for Attansic L1E Gigabit Ethernet Adapter
1.973	07-Jan-2009	jnemeth	regen
1.972	27-Dec-2008	tsutsui	Regen from pcidevs rev 1.971: > Fix tyop in previous.
1.971	27-Dec-2008	jnemeth	regen for et (Agere) driver
1.970	21-Dec-2008	nonaka	Regen.
1.969	16-Dec-2008	christos	regen
1.968	29-Nov-2008	christos	regen
1.967	18-Nov-2008	matt	Regen.
1.966	17-Nov-2008	matt	Regen.
1.965	08-Nov-2008	christos	regen.
1.964	04-Nov-2008	cegger	regen.
1.963	04-Nov-2008	cegger	regen.
1.962	15-Oct-2008	bouyer	branches: 1.962.2; 1.962.4; Regen: fix entry description.
1.961	15-Oct-2008	bouyer	Regen: add more ICH9 ethernet devices.
1.960	01-Oct-2008	christos	regen
1.959	01-Oct-2008	christos	ICH10 support from Brad du Plessis
1.958	28-Sep-2008	bouyer	Regen for JMicron Ethernet Controllers.
1.957	18-Sep-2008	cegger	regen.
1.956	18-Sep-2008	cegger	regen.
1.955	13-Sep-2008	njoly	Regen for Nvidia GeForce 6150 devices addition.
1.954	08-Sep-2008	cegger	regen.
1.953	08-Sep-2008	cegger	regen.
1.952	06-Sep-2008	rmind	Regen.
1.951	05-Sep-2008	cegger	regen.
1.950	27-Aug-2008	christos	regen
1.949	23-Aug-2008	tsutsui	Regen from pcidevs rev 1.949: > - add a Realtek 8139D variant (from FreeBSD) > - remove a duplicated entry of Realtek 0x8138
1.948	22-Aug-2008	tnn	regen (Intel 945GME)
1.947	19-Aug-2008	matthias	Regen.
1.946	19-Aug-2008	matthias	Regen.
1.945	06-Aug-2008	dyoung	Regen.
1.944	29-Jul-2008	tron	Regen.
1.943	15-Jul-2008	tonnerre	Regen.
1.942	25-May-2008	chs	branches: 1.942.2; 1.942.4; regen
1.941	20-May-2008	cegger	regen.
1.940	13-May-2008	rjs	Regen.
1.939	08-May-2008	sborrill	Regen
1.938	30-Apr-2008	joerg	branches: 1.938.2; Regen.
1.937	17-Apr-2008	christos	branches: 1.937.2; 1.937.4; regen
1.936	15-Apr-2008	cegger	regen.
1.935	09-Apr-2008	joerg	Regen.
1.934	03-Apr-2008	riz	Regen.
1.933	01-Apr-2008	cegger	regen.
1.932	26-Mar-2008	jmcneill	Regen for PR#38301
1.931	24-Feb-2008	xtraeme	regen
1.930	22-Feb-2008	tsutsui	Regen from pcidevs rev 1.930: > Add I-O DATA RSA-PCI 2 port serial board.
1.929	13-Feb-2008	tron	branches: 1.929.2; 1.929.6; Regen.
1.928	10-Feb-2008	ichiro	regen
1.927	05-Feb-2008	ichiro	regen
1.926	11-Jan-2008	tnn	regen (add ServeRAID 6/7)
1.925	11-Jan-2008	ragge	Regen
1.924	03-Jan-2008	jnemeth	regen -- Symbios Logic 0x0030 is used for both 53c1020/53c1030 from Dave Huang
1.923	30-Dec-2007	jdc	Regenerate for "Add Apple Shasta GMAC (From FreeBSD)".
1.922	28-Dec-2007	dholland	Regen.
1.921	24-Dec-2007	hamajima	regen.
1.920	09-Dec-2007	jmcneill	branches: 1.920.2; Merge jmcneill-pm branch.
1.919	26-Nov-2007	bjs	branches: 1.919.2; 1.919.4; Add PCI IDs for the ATI Radeon X800 GTO (R430). Most sources I have consulted indicated that this ID is an R430 (including xf86-video-ati). PCI_PRODUCT_ATI_RADEON_R430_554F (primary) PCI_PRODUCT_ATI_RADEON_R430_556F (secondary) vga0 at pci1 dev 0 function 0: ATI Technologies product 0x554f (rev. 0x00) ATI Technologies product 0x556f (miscellaneous display) at pci1 dev 0 function 1 not configured
1.918	24-Nov-2007	markd	regen.
1.917	23-Nov-2007	xtraeme	regen
1.916	19-Nov-2007	ad	Regen.
1.915	15-Nov-2007	manu	Regen
1.914	14-Nov-2007	xtraeme	regen
1.913	12-Nov-2007	joerg	Regen.
1.912	30-Oct-2007	xtraeme	branches: 1.912.2; regen
1.911	30-Oct-2007	xtraeme	regen
1.910	30-Oct-2007	xtraeme	regen
1.909	30-Oct-2007	jnemeth	regen
1.908	24-Oct-2007	xtraeme	regen
1.907	22-Oct-2007	manu	Regen
1.906	22-Oct-2007	manu	Add support for Marvell 88CO55. Newer iMacs come with that chip.
1.905	22-Oct-2007	nisimura	regen
1.904	14-Oct-2007	xtraeme	branches: 1.904.2; regen
1.903	09-Oct-2007	garbled	Add a bunch of IBM PCI dev id's for graphics cards and PCI bridges
1.902	20-Sep-2007	xtraeme	branches: 1.902.2; regen
1.901	17-Sep-2007	msaitoh	regen
1.900	11-Sep-2007	jnemeth	regen
1.899	09-Sep-2007	xtraeme	regen
1.898	08-Sep-2007	ober	regen. added Intel Pro/Wifi 4965AGN
1.897	07-Sep-2007	cube	Regen.
1.896	02-Sep-2007	ober	regen. added Intel LPC 82801HEM and 82801IO
1.895	01-Sep-2007	ober	* empty log message *
1.894	31-Aug-2007	xtraeme	regen
1.893	30-Aug-2007	xtraeme	regen
1.892	28-Aug-2007	msaitoh	regen
1.891	26-Aug-2007	xtraeme	branches: 1.891.2; regen
1.890	19-Aug-2007	riz	Regen.
1.889	14-Aug-2007	jnemeth	regen
1.888	11-Aug-2007	jnemeth	regem
1.887	08-Aug-2007	tron	Regen.
1.886	06-Aug-2007	martin	branches: 1.886.2; regen
1.885	04-Aug-2007	kiyohara	Regen.
1.884	07-Jul-2007	jklos	branches: 1.884.2; 1.884.6; Added Chelsio vendor and various Chelsio hardware.
1.883	06-Jul-2007	hpeyerl	regen
1.882	27-Jun-2007	xtraeme	regen
1.881	26-Jun-2007	xtraeme	regen
1.880	28-May-2007	mrg	regenerate.
1.879	19-May-2007	tsutsui	Regen from pcidevs rev 1.880: > Add IBM GXT250P video.
1.878	16-May-2007	mrg	regenerate
1.877	02-May-2007	bouyer	Regen: added JMicron Technology products
1.876	29-Apr-2007	bouyer	Regen: remove extra "Intel" from previous commit.
1.875	29-Apr-2007	bouyer	Regen: add Intel i82801H IFE G/GT LAN Controller
1.874	27-Mar-2007	pooka	regen
1.873	25-Mar-2007	isaki	Regen; ATI ES1000.
1.872	17-Mar-2007	dyoung	Regen.
1.871	10-Mar-2007	christos	branches: 1.871.2; 1.871.4; regen
1.870	09-Mar-2007	xtraeme	branches: 1.870.2; regen
1.869	03-Mar-2007	jmcneill	Regen.
1.868	01-Mar-2007	jmcneill	Regen.
1.867	16-Feb-2007	markd	branches: 1.867.2; regen for i965Q
1.866	10-Feb-2007	mlelstv	Get the correct RCS-Id pulled in
1.865	10-Feb-2007	mlelstv	Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.864	30-Jan-2007	msaitoh	regen
1.863	28-Jan-2007	chs	regen
1.862	27-Jan-2007	isaki	Regen.
1.861	21-Jan-2007	isaki	Regen; AMD Geode LX and CS5536.
1.860	14-Jan-2007	rpaulo	regen
1.859	06-Jan-2007	jmcneill	Regen.
1.858	04-Jan-2007	jmcneill	Regen.
1.857	04-Jan-2007	jmcneill	Regen.
1.856	31-Dec-2006	ichiro	regen
1.855	10-Dec-2006	bouyer	Regen: another PCI id for Dell PERC 5
1.854	09-Dec-2006	manu	Add support for 4 and 8 ports modem boards from Perle systems
1.853	26-Nov-2006	tsutsui	branches: 1.853.2; Regen from pcidevs rev 1.855: > - add more Broadcom product IDs, and fix two alternative IDs > (from FreeBSD and OpenBSD) > - sort by product ID
1.852	26-Nov-2006	enami	Regen.
1.851	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.853: > Add an ID of Broadcom BCM5752M. From Matt Kolb on tech-kern.
1.850	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.852: > - add Realtek RTL8100 (from FreeBSD) > - fix entry of RTL8169SC > - remove vague comment in Realtek entry
1.849	18-Nov-2006	tsutsui	Regen from pcidevs rev 1.851: > Add and note more Realtek Ethernet chips.
1.848	13-Nov-2006	xtraeme	regen
1.847	13-Nov-2006	bouyer	Regen: add NetMos 9855 Quad UART and 1284 Printer port
1.846	08-Nov-2006	rpaulo	regen
1.845	08-Nov-2006	martti	Regenerated.
1.844	07-Nov-2006	martti	Regenerated.
1.843	03-Nov-2006	abs	Regenerate after adding Broadcom BCM5754 (0x167a)
1.842	31-Oct-2006	manu	regen
1.841	25-Oct-2006	bouyer	Regen: Add VIA VT8237A Integrated SATA Controller
1.840	23-Oct-2006	rjs	Regen.
1.839	21-Oct-2006	bouyer	Regen: added some Intel gigabit ethernet products
1.838	09-Oct-2006	cube	Regen.
1.837	02-Oct-2006	cube	Regen (fix pasto).
1.836	02-Oct-2006	cube	Regen (MCP55 ISA brige).
1.835	29-Sep-2006	cube	Regen.
1.834	27-Sep-2006	cube	Regen (Intel/0x1094).
1.833	27-Sep-2006	sketch	Regenerate for TI ACX changes.
1.832	26-Sep-2006	heas	Regen for pcidevs 1.833
1.831	24-Sep-2006	briggs	Regen with new IDs for devices in the last-generation Apple PowerBook G4.
1.830	07-Sep-2006	riz	branches: 1.830.2; Regen.
1.829	07-Sep-2006	itohy	branches: 1.829.2; Regen from pcidevs rev 1.830.
1.828	03-Sep-2006	xtraeme	regen
1.827	03-Sep-2006	xtraeme	regen
1.826	03-Sep-2006	xtraeme	regen
1.825	03-Sep-2006	xtraeme	regen
1.824	03-Sep-2006	xtraeme	regen
1.823	03-Sep-2006	xtraeme	regen
1.822	27-Aug-2006	christos	Regen
1.821	24-Aug-2006	riz	Regen (Marvell product IDs)
1.820	23-Aug-2006	cube	Regen (Broadcom NetXtreme II).
1.819	22-Aug-2006	riz	Regen for GALILEO -> MARVELL
1.818	13-Aug-2006	dsainty	Regen: Add the Promise PDC20771, present on the Promise FastTrak TX2300
1.817	10-Aug-2006	simonb	Regen.
1.816	09-Aug-2006	jnemeth	regen for Token Ring additions
1.815	08-Aug-2006	cube	Regen (Intel 63xxESB devices).
1.814	06-Aug-2006	jmcneill	branches: 1.814.2; Regen.
1.813	03-Aug-2006	bsh	regen
1.812	01-Aug-2006	xtraeme	regen
1.811	01-Aug-2006	bsh	regen
1.810	30-Jul-2006	jmcneill	regen
1.809	30-Jul-2006	oster	Regen
1.808	30-Jul-2006	simonb	Regen.
1.807	21-Jul-2006	gdamore	Regen.
1.806	21-Jul-2006	gdamore	Update ADMtek pci device ids to match data sheets. While here, add the 9513 device (although I've never seen one, I have a datasheet for it. :-)
1.805	18-Jul-2006	itohy	Regen from pcidevs rev 1.807.
1.804	13-Jul-2006	kent	regen for VT82xx HDA
1.803	07-Jul-2006	simonb	Regen.
1.802	30-Jun-2006	nisimura	Regen after Micrel KSZ8842/8841 product addtion.
1.801	30-Jun-2006	xtraeme	regen
1.800	30-Jun-2006	xtraeme	regen
1.799	27-Jun-2006	kent	regen for NVIDA MCP55_HDA
1.798	26-Jun-2006	xtraeme	regen
1.797	26-Jun-2006	xtraeme	regen
1.796	25-Jun-2006	soren	Regen.
1.795	18-Jun-2006	christos	Regen
1.794	18-Jun-2006	xtraeme	regen
1.793	15-Jun-2006	heinz	Regen (LSI SAS, parallel SCSI and FC devices).
1.792	14-Jun-2006	grant	branches: 1.792.2; regen.
1.791	06-Jun-2006	rpaulo	regen
1.790	05-Jun-2006	martin	Regen (DWL-610 added)
1.789	05-Jun-2006	msaitoh	regen
1.788	31-May-2006	riz	Regen.
1.787	19-May-2006	markd	branches: 1.787.2; regen
1.786	30-Apr-2006	xtraeme	regen
1.785	30-Apr-2006	xtraeme	regen
1.784	29-Apr-2006	gdamore	Regen.
1.783	25-Apr-2006	wrstuden	Regen to catch change for 3ware cards.
1.782	20-Apr-2006	gdamore	Regen.
1.781	15-Apr-2006	jmcneill	Regen.
1.780	06-Apr-2006	simonb	Regen.
1.779	05-Apr-2006	nonaka	Regen.
1.778	26-Mar-2006	jonathan	Regen pcidevs{,_data}.h from pcidevs rev 1.778 (fixes cut-and-paste typo in description of bcm5780S).
1.777	25-Mar-2006	rpaulo	regen
1.776	21-Mar-2006	he	Regenerated files after adding IDs for HP/Compaq Smart Array controllers.
1.775	12-Mar-2006	chs	branches: 1.775.2; regen
1.774	09-Mar-2006	he	Commit regenerated files after adding Nortel Baystack 21 info.
1.773	09-Mar-2006	hamajima	Regen: Add EXAR devices, and fix duplicated NVIDIA product id 0x0057.
1.772	07-Mar-2006	bouyer	Regen: added more serverwork SATA controllers.
1.771	07-Mar-2006	bouyer	branches: 1.771.2; Regen for serverworks K2 SATA addition.
1.770	06-Mar-2006	martin	Regen (MOXA devices added)
1.769	01-Mar-2006	dogcow	branches: 1.769.2; add a bunch of Nvidia nForce ethernet IDs (via OpenBSD.)
1.768	28-Feb-2006	gdamore	Regen.
1.767	21-Feb-2006	chuck	regen.
1.766	16-Feb-2006	kochi	regen.
1.765	12-Feb-2006	tron	Regen.
1.764	12-Feb-2006	tron	Regen.
1.763	12-Feb-2006	tron	Regen.
1.762	07-Feb-2006	cube	Regen, correctly this time. Sorry for the mess.
1.761	07-Feb-2006	xtraeme	regen
1.760	06-Feb-2006	abs	regen to add SIS 190 ethernet, as found in ASUS T2-AE1
1.759	02-Feb-2006	christos	branches: 1.759.2; regen
1.758	26-Jan-2006	bouyer	branches: 1.758.2; Regen: Add Promise PDC2057x
1.757	16-Jan-2006	christos	regen
1.756	10-Jan-2006	christos	Regen
1.755	02-Jan-2006	manu	regen
1.754	02-Jan-2006	tron	Regen.
1.753	02-Jan-2006	tron	Regen.
1.752	01-Jan-2006	xtraeme	regen
1.751	29-Dec-2005	silke	branches: 1.751.2; Regen from revision 1.751 of pcidevs.
1.750	29-Dec-2005	jonathan	Regen from src/sys/dev/pci/pcidevs rev 1.751, which adds an entry for the integral PCIe-to-PCIX bridge found in broadcom 5714/5715 PCI-Express dual-port gigabit ethernet chips.
1.749	19-Dec-2005	jonathan	Regem from src/sys/dev/pci/pcidevs rev 1.750, adding an entry for SK PCi-Express gigabit ethernet boards based on Marvell's Yukon-2 chipset.
1.748	19-Dec-2005	jonathan	regen after sys/dev/pci/pcidevs 1.749 (addition of SiI 3124)
1.747	18-Dec-2005	sekiya	Regen.
1.746	11-Dec-2005	jonathan	Regen after pcidevs rev 1.746, 1.747 (Add device IDs for LSI Megaraid, and Intel IDs for software-compatible Intel RAID products).
1.745	10-Dec-2005	jonathan	regen from pcidevs rev.1.745 with IDs for broadcom bcm5714, bcm5715, bcm5780.
1.744	09-Dec-2005	xtraeme	regen
1.743	08-Dec-2005	xtraeme	regen
1.742	07-Dec-2005	jonathan	regen from sys/dev/pci/pcidevs rev 1.742 (add bcm5714 et al.).
1.741	06-Dec-2005	fvdl	Regen for BCM5789.
1.740	04-Dec-2005	christos	Regen.
1.739	04-Dec-2005	christos	Regen.
1.738	27-Nov-2005	jdolecek	regen - 3ware ATA RAID description/ID fixes
1.737	22-Nov-2005	rpaulo	regen
1.736	18-Nov-2005	rpaulo	regen
1.735	18-Nov-2005	rpaulo	regen
1.734	01-Nov-2005	abs	branches: 1.734.2; Regen after adding Radeon 9000 PCI (0x4966)
1.733	05-Oct-2005	jonathan	branches: 1.733.2; Regen with PCI ids for Intel pro/1000 PT pci-e gigabit ethernet.
1.732	04-Sep-2005	tron	Regen.
1.731	21-Aug-2005	chs	regen
1.730	21-Aug-2005	gavan	regen
1.729	30-Jul-2005	grant	regen
1.728	30-Jul-2005	grant	regen
1.727	29-Jul-2005	cube	Regen.
1.726	25-Jul-2005	christos	regen.
1.725	01-Jul-2005	drochner	branches: 1.725.2; regen
1.724	01-Jul-2005	augustss	Regen.
1.723	28-Jun-2005	thorpej	Regen -- static'ify pci_vendors[] and pci_products[]
1.722	26-Jun-2005	seb	Regen.
1.721	24-Jun-2005	seb	Regen.
1.720	20-Jun-2005	briggs	Regen.
1.719	19-Jun-2005	chs	regen
1.718	15-Jun-2005	bouyer	Regen: add Intel 82801FBM Serial ATA Controller.
1.717	14-Jun-2005	bouyer	Regen: add SiS 965.
1.716	13-Jun-2005	tron	Regen.
1.715	13-Jun-2005	nonaka	Regen.
1.714	13-Jun-2005	nonaka	Regen.
1.713	07-Jun-2005	skrll	Regen.
1.712	28-May-2005	kent	regen
1.711	27-May-2005	augustss	Regen.
1.710	18-May-2005	riz	Regen.
1.709	12-May-2005	cube	Regen
1.708	09-May-2005	martin	Regen
1.707	04-May-2005	cube	Regen.
1.706	02-May-2005	rh	regen
1.705	20-Apr-2005	bouyer	Regen for "Compaq Netelligent 10 T/2 PCI UTP/Coax Controller".
1.704	06-Apr-2005	martti	Regen for Dell CERC SATA RAID 1.5/6ch
1.703	24-Mar-2005	tron	Regen.
1.702	19-Mar-2005	tron	Regen.
1.701	19-Mar-2005	tron	Regen.
1.700	16-Mar-2005	xtraeme	branches: 1.700.2; regen
1.699	16-Mar-2005	uwe	Regen for ATI Radeon 9100 IGP chipset.
1.698	15-Mar-2005	xtraeme	regen
1.697	15-Mar-2005	xtraeme	regen
1.696	15-Mar-2005	uwe	Regen for ENE CB710, CB720 CardBus controllers.
1.695	07-Mar-2005	christos	Regen
1.694	27-Feb-2005	perry	regen
1.693	20-Feb-2005	jdolecek	regen: change the VIATECH VT61XX to VT612X
1.692	02-Feb-2005	christos	Regen
1.691	02-Feb-2005	abs	Regen to pick up corrected code for VIA VT6421
1.690	01-Feb-2005	abs	regen to pick up the 'VT642 SATA RAID Controller'.
1.689	27-Jan-2005	jdolecek	regen: add id for nForce2 Ultra 400 audio
1.688	27-Jan-2005	kent	regenerate for nForce 4 devices
1.687	25-Jan-2005	matt	branches: 1.687.2; Regen.
1.686	23-Jan-2005	fredb	Regen.
1.685	18-Jan-2005	tsubai	Regen.
1.684	15-Jan-2005	skrll	branches: 1.684.2; Regen.
1.683	15-Jan-2005	jdolecek	regen: Fix Geforce 6800GT entry
1.682	15-Jan-2005	jdolecek	regen: add bunch of NVIDIA GeForce 6800 devices
1.681	15-Jan-2005	jdolecek	regen: use full 'nForce2 Ultra 400' tag for the NFORCE2_400* enties
1.680	13-Jan-2005	jdolecek	regen: add bunch of NVIDIA nForce2 Ultra 400 chipset devices
1.679	13-Jan-2005	jdolecek	regen: VIA Velocity
1.678	08-Jan-2005	fredb	Regen.
1.677	08-Jan-2005	ginsbach	regen
1.676	06-Jan-2005	he	Regen after adding entry for LSI Logic FC919X.
1.675	27-Dec-2004	reinoud	Update generated files from pcidevs (ENE update)
1.674	18-Dec-2004	xtraeme	regen
1.673	16-Dec-2004	heas	Add entry for the Radeon Mobility T2.
1.672	11-Dec-2004	reinoud	Add ATI IXP audio controllers
1.671	07-Dec-2004	jdarrow	Regen from pcidevs 1.672: Add a new vendor and product ID for a cardbus NIC I've got Belkin 0x1799 F5D6020v3 0x6020
1.670	06-Dec-2004	cube	Regen (ATI IXP IDE).
1.669	05-Dec-2004	bouyer	Regen: Cleanup in VIA chipset names. From Lloyd Parkes in kern/15466.
1.668	30-Nov-2004	grant	regen
1.667	22-Nov-2004	rumble	regen
1.666	19-Nov-2004	tacha	regen.
1.665	17-Nov-2004	jdarrow	Regen from pcidevs 1.666 (add two wireless cards)
1.664	17-Nov-2004	junyoung	Regen.
1.663	16-Nov-2004	skrll	Regen
1.662	14-Nov-2004	soren	Regen.
1.661	06-Nov-2004	xtraeme	Regen for: Add Nvidia nForce3 250 chipset products, from Kouichirou Hiratsuka in PR 27843.
1.660	28-Oct-2004	martti	Regen for Dell PERC 4/SC
1.659	26-Oct-2004	augustss	Regen.
1.658	22-Oct-2004	cube	Regen.
1.657	20-Oct-2004	martti	Regen for DELL PERC 4e/Si
1.656	13-Oct-2004	augustss	Regen.
1.655	13-Oct-2004	drochner	regen
1.654	11-Oct-2004	jmmv	Regen after addition of ATI Technologies Radeon 9200SE Secondary identifier.
1.653	05-Oct-2004	augustss	Regen.
1.652	30-Sep-2004	cube	Regen.
1.651	18-Sep-2004	martin	Regen (FC929X added)
1.650	17-Sep-2004	xtraeme	Regen for GeForce FX 5700 LE.
1.649	17-Sep-2004	xtraeme	Add GeForceFX 5700 LE and regen.
1.648	16-Sep-2004	ragge	Regen.
1.647	10-Sep-2004	darrenr	Regen'd to add Sierra Wireless AirCard 580
1.646	26-Aug-2004	itohy	Regen from pcidevs rev 1.647.
1.645	05-Aug-2004	mycroft	Regen.
1.644	05-Aug-2004	cube	Regen.
1.643	04-Aug-2004	mycroft	Regen.
1.642	04-Aug-2004	mycroft	Regen.
1.641	04-Aug-2004	mycroft	Regen.
1.640	02-Aug-2004	mycroft	Regen.
1.639	02-Aug-2004	mycroft	Regen.
1.638	02-Aug-2004	mycroft	Regen.
1.637	02-Aug-2004	mycroft	Regen.
1.636	29-Jul-2004	cube	Regen.
1.635	27-Jul-2004	christos	regen
1.634	27-Jul-2004	rumble	regen.
1.633	25-Jul-2004	christos	regen
1.632	21-Jul-2004	mycroft	Regen.
1.631	17-Jul-2004	mycroft	Regen.
1.630	15-Jul-2004	junyoung	Regen for 3c920B-EMB-WNM Integrated Fast Ethernet.
1.629	05-Jul-2004	mycroft	Regen.
1.628	04-Jul-2004	wiz	regen from pcidevs, so that this file has the correct CVS Id.
1.627	04-Jul-2004	mycroft	Update Serverworks chip IDs, per: http://www.plasma-online.de/english/identify/picture/serverworks.html This seems remarkably more sane.
1.626	15-Jun-2004	bouyer	Regen for pcidevs 1.628: Add the Netelligent 10/100 TX variant found on the Compaq Proliant Professional Workstation 5100.
1.625	27-May-2004	christos	Regen
1.624	27-May-2004	abs	regen from pcidevs 1.626: Add ALI 'M5257 PCI Software Modem'
1.623	12-May-2004	tron	Regen for PR kern/25166.
1.622	11-May-2004	thorpej	Regen: Add i6300ESB IDE / SATA controller PCI IDs.
1.621	10-May-2004	lukem	regen for: add Nvidia GeForce4 Ti 4200 Go AGP 8x (0x286)
1.620	10-May-2004	gendalia	regen (Add support for Adaptec AAR 2810SA raid controller.)
1.619	07-May-2004	sekiya	regen.
1.618	22-Apr-2004	skd	regen.
1.617	17-Apr-2004	matt	Regen.
1.616	03-Apr-2004	martti	Added another SB Live! 5.1 soundcard and joystick version
1.615	03-Apr-2004	mycroft	Add RealTek RT8180.
1.614	01-Apr-2004	tron	Regen from "pcidevs".
1.613	26-Mar-2004	gendalia	branches: 1.613.2; regen for Adaptect 2200S (Vulcan 2M) / Dell PERC 320/DC
1.612	25-Mar-2004	drochner	regen
1.611	12-Mar-2004	drochner	regen
1.610	03-Mar-2004	gavan	Add Creative Labs SB Audigy LS
1.609	26-Feb-2004	drochner	regen
1.608	28-Jan-2004	chs	regen
1.607	26-Jan-2004	blymn	Added note for updating these files.
1.606	26-Jan-2004	blymn	Regen with correct rcs id.
1.605	26-Jan-2004	blymn	Add amd8151 agp bridge id.
1.604	19-Jan-2004	ragge	Err, get correct RCS IDs. As Love said.
1.603	19-Jan-2004	ragge	Add "Intel PRO/10GbE LR Server Adapter". Regen.
1.602	06-Jan-2004	matt	Regen.
1.601	05-Jan-2004	matt	Regen.
1.600	20-Dec-2003	thorpej	Regen: Add product ID for the SiI 3124 SATALink.
1.599	19-Dec-2003	thorpej	Regen: Add product ID for the Silicon Image 3114 SATALink.
1.598	17-Dec-2003	tls	Add support for some more Marvell-based adapters to the sk driver: D-Link DGE-530T Linksys EG1032v2 Linksys EG1064v2 The Linux driver treats all of these the same. Tested with a DGE-530T; it works, though performance on at least one of my systems is atrocious. Fix setting of ethernet capabilities into interface capabilites word in attach routine. Manual page: note that jumbo support doesn't work and checksum support doesn't exist.
1.597	14-Dec-2003	tsutsui	Regen: added ATI Radeon 9200SE.
1.596	09-Dec-2003	ad	Regen.
1.595	08-Dec-2003	pooka	regen for broadcom 5703a3
1.594	08-Dec-2003	scw	Regen
1.593	06-Dec-2003	jdolecek	regen: add NetMos NM9805 and NM9815
1.592	05-Dec-2003	grant	regen
1.591	28-Nov-2003	chs	regen
1.590	28-Nov-2003	cube	Regen.
1.589	26-Nov-2003	matt	Regen.
1.588	22-Nov-2003	fvdl	Regenerate after rearranging the Promise ids.
1.587	03-Nov-2003	cube	Regen.
1.586	25-Oct-2003	chs	regen
1.585	23-Oct-2003	fvdl	Regen.
1.584	23-Oct-2003	briggs	Regenerate for ENE Cardbus product IDs.
1.583	22-Oct-2003	fvdl	Regen.
1.582	18-Oct-2003	matt	Regen.
1.581	18-Oct-2003	lukem	regen for more ServerWorks devices
1.580	15-Oct-2003	itojun	regen
1.579	14-Oct-2003	martti	Regen: Dell PERC 4/Di
1.578	10-Oct-2003	thorpej	Regen: Add several more Intel Gig-E product IDs.
1.577	08-Oct-2003	jonathan	Regen from pcidevs rev 1.581 with Hifn 7954/5 and 7956.
1.576	26-Sep-2003	matt	Regen.
1.575	25-Sep-2003	joda	regen
1.574	21-Sep-2003	nisimura	Regen.
1.573	21-Sep-2003	nisimura	Regen to fix the company name.
1.572	15-Sep-2003	bouyer	Regen: rename VT8237_RAID to VT8237_SATA
1.571	13-Sep-2003	jdolecek	regen - fix description of one Intel PRO/100 VM variant, as reported by Lars Heidieker on current-users@
1.570	28-Aug-2003	thorpej	Regen; spell Hifn correctly, add more Hifn products.
1.569	27-Aug-2003	kleink	Regen.
1.568	26-Aug-2003	jdolecek	regen: add SysKonnect 9821
1.567	26-Aug-2003	tron	Regen.
1.566	23-Aug-2003	dogcow	regen (add VIA VT6105)
1.565	20-Aug-2003	jonathan	regen from pcidevs 1.567 with fix for typo in bcm5823 product-id.
1.564	17-Aug-2003	chs	regen
1.563	17-Aug-2003	bouyer	Regen: Add 3com 3c940 Gigabit Ethernet, and VIA KT600 PCI devices.
1.562	14-Aug-2003	itojun	regen
1.561	07-Aug-2003	wiz	regen (fix ID of Radeon 9100 Secondary)
1.560	01-Aug-2003	wiz	regen (avoid duplicate definition, PR 22331)
1.559	01-Aug-2003	jonathan	regen from pcidevs rev 560 (adding Sun's bcm582x-compatible crypto accelerators)
1.558	29-Jul-2003	thorpej	Regen: Add the product ID used for the i82546EB chips present on the Intel PRO/1000 MT Quad Port Server Adapter.
1.557	29-Jul-2003	martin	Regen.
1.556	27-Jul-2003	jonathan	regen from pcidevs rev 1.556 with vendor Cavium and product Nitrox.
1.555	27-Jul-2003	jonathan	Regen from pcidevs 1.555 (Bluesteel,Broadcom security coprocessors)
1.554	23-Jul-2003	kleink	Regen from 1.554.
1.553	17-Jul-2003	hannken	Regen.
1.552	17-Jul-2003	hannken	Add support for the BCM5705: - Only BCM5705M asic rev A1 was tested. Thanks to Bill Paul (wpaul@freebsd.org) for help and support. Approved by: Frank van der Linden <fvdl@netbsd.org>
1.551	14-Jul-2003	jmmv	Regen (+ ATI Radeon 9200)
1.550	14-Jul-2003	wiz	regen (+MGA G550).
1.549	12-Jul-2003	hubertf	Regen for Toshiba SD and SM cards
1.548	06-Jul-2003	dyoung	Add driver atw for PCI/Mini-PCI/Cardbus 802.11b NICs based on the ADMtek ADM8211. Read the man page for bugs and other outstanding issues.
1.547	06-Jul-2003	simonb	Regen: Trailing white-space and TABs nits.
1.546	06-Jul-2003	simonb	Regen; E7500 and E7505 updates.
1.545	06-Jul-2003	tron	Regen.
1.544	06-Jul-2003	simonb	Regen: Add an entry for the Intel E7505 Chipset Memory Controller Hub.
1.543	06-Jul-2003	tron	Regen.
1.542	06-Jul-2003	tron	Regen.
1.541	28-Jun-2003	bouyer	branches: 1.541.2; Regen: some more Intel devices.
1.540	27-Jun-2003	christos	Regen
1.539	27-Jun-2003	drochner	regen
1.538	26-Jun-2003	ichiro	regen.
1.537	14-Jun-2003	ichiro	regen.
1.536	13-Jun-2003	kent	Regen
1.535	08-Jun-2003	ichiro	regen.
1.534	07-Jun-2003	wiz	regen (Add some Brooktree device IDs).
1.533	07-Jun-2003	wiz	regen (+ATI 9100, 9600TX, 9500/9700)
1.532	07-Jun-2003	ichiro	regen.
1.531	07-Jun-2003	ichiro	regen.
1.530	07-Jun-2003	ichiro	regen.
1.529	28-Apr-2003	haya	Regenerate for pcidevs 1.529 and 1.530 changes.
1.528	24-Apr-2003	nathanw	Regen.
1.527	22-Apr-2003	fvdl	Regenerated after AMD 8131 additions.
1.526	15-Apr-2003	augustss	Regen.
1.525	15-Apr-2003	rjs	Regen.
1.524	13-Apr-2003	explorer	typo, s/TX4/TX2/
1.523	12-Apr-2003	explorer	add an older pci id for SATA150TX2PLUS, too
1.522	12-Apr-2003	christos	regen
1.521	11-Apr-2003	explorer	add Promise ATA/150 TX4 id
1.520	05-Apr-2003	lha	Regenerate. (add Radeon Mobility M7 LW (AGP))
1.519	03-Apr-2003	kent	regen
1.518	29-Mar-2003	jdc	Regenerate. (Add support for USRobotics Wireless Access PCI adapter (model 132415).)
1.517	25-Mar-2003	tsubai	Regen.
1.516	20-Mar-2003	thorpej	Regen; add more LSI MPT IDs.
1.515	18-Mar-2003	thorpej	Regen; added Silicon Image SATALink 3112.
1.514	13-Mar-2003	joda	regen: more nvidia devices
1.513	06-Mar-2003	bouyer	Regen: Remove duplicate in previous.
1.512	06-Mar-2003	bouyer	Regen: Add several SiS host bridges, from FreeBSD.
1.511	03-Mar-2003	soren	Regen.
1.510	02-Mar-2003	salo	Regen.
1.509	28-Feb-2003	enami	Regen.
1.508	26-Feb-2003	salo	Regen.
1.507	18-Feb-2003	grant	regen
1.506	03-Feb-2003	pooka	regen
1.505	03-Feb-2003	joda	regen
1.504	02-Feb-2003	bouyer	Regen: add SiS 651 host bridge.
1.503	01-Feb-2003	matt	Regen.
1.502	24-Jan-2003	thorpej	Regen; Tweak a Promise description.
1.501	24-Jan-2003	thorpej	Regen; add some more Promise products.
1.500	24-Jan-2003	thorpej	Regen; add Nvidia nForce2 ATA133 IDE.
1.499	17-Jan-2003	jonathan	regen from pcidevs rev 1.497: Add pcidevs entries for more Broadcom 570x family variants.
1.498	16-Jan-2003	gendalia	regen after pcidevs is committed so the correct version number is embedded. Thanks to wiz.
1.497	16-Jan-2003	gendalia	Add 3com 3c905CX-TX and Nvidia nForce2 MCP-T ac-97, on the ASUS A7N8X Deluxe motherboard, Chipset: NVIDIA nForce2 MCP-T
1.496	10-Jan-2003	augustss	Regen.
1.495	30-Dec-2002	explorer	update
1.494	26-Dec-2002	matt	Regen.
1.493	26-Dec-2002	matt	Regen.
1.492	11-Dec-2002	jdolecek	regen: add Madge Networks Smart 16/4 PCI Ringnode Mk2 entry
1.491	03-Dec-2002	explorer	add Ricoh 5C552 Cardbus + Firewire product ID
1.490	26-Nov-2002	thorpej	Regen: Add IBM 133 PCI-X Bridge.
1.489	26-Nov-2002	thorpej	Regen: Add Intel i80321 I/O Processor.
1.488	19-Nov-2002	bsh	Regen. for additional vendor IDs for Samsung Electronics Co. Ltd.
1.487	15-Nov-2002	tsutsui	Regen from pcidevs rev 1.486: > Remove an entry for AMD PCNETS_PCI. > > AMD Am79C974 PCnet-SCSI Ethernet and SCSI conrtoller is > a multi-function PCI device which has two device IDs > of Am53C974(PCscsi-PCI) and Am79C970(PCnet-PCI), > and there is no its own ID.
1.486	15-Nov-2002	ad	Regen.
1.485	09-Nov-2002	grant	regen for ATI Rage Mobility M1.
1.484	08-Nov-2002	msaitoh	regen
1.483	07-Nov-2002	thorpej	Regen; add Intel i31244 Serial ATA controller.
1.482	07-Nov-2002	martin	Regen (Myson MTD804 added)
1.481	04-Nov-2002	wiz	regen (typo fix from PR 18921)
1.480	25-Oct-2002	kent	regen
1.479	24-Oct-2002	grant	regen
1.478	21-Oct-2002	soren	Regen.
1.477	21-Oct-2002	grant	regen
1.476	18-Oct-2002	yamt	regen; fix simple typos
1.475	16-Oct-2002	leo	Regen.
1.474	05-Oct-2002	kent	regen
1.473	24-Sep-2002	pooka	regen
1.472	20-Sep-2002	pooka	regen
1.471	19-Sep-2002	nathanw	Regen.
1.470	11-Sep-2002	bouyer	Regen: add 2 acard pciide controllers.
1.469	06-Sep-2002	uwe	Regen for Integraphics Systems CyberPro 2000 and 2010.
1.468	23-Aug-2002	bouyer	Regen: Add Promise Ultra133TX2 and Ultra133TX2v2 PCI ID
1.467	16-Aug-2002	heinz	regen
1.466	10-Aug-2002	toshii	Regen; added SiS745 host bridge.
1.465	08-Aug-2002	drochner	regen
1.464	04-Aug-2002	thorpej	Regen; added ADI BECC product ID.
1.463	03-Aug-2002	kent	regen
1.462	02-Aug-2002	soren	Regen.
1.461	02-Aug-2002	ad	Regen.
1.460	31-Jul-2002	bouyer	Regen: remove duplicate entry for 82801DB_LPC.
1.459	30-Jul-2002	bouyer	Regen: Add Intel 845G PCI IDs
1.458	29-Jul-2002	augustss	Regen.
1.457	26-Jul-2002	onoe	regen: Add Silicon Image (CMDTECH) 0680
1.456	22-Jul-2002	msaitoh	regen
1.455	12-Jul-2002	enami	Regen.
1.454	09-Jul-2002	matt	Regen.
1.453	09-Jul-2002	thorpej	Regen: Clean up Intel i8254x product names, and add i82540, i82545, and i82546 product IDs.
1.452	07-Jul-2002	augustss	Regen.
1.451	05-Jul-2002	enami	Regen.
1.450	04-Jul-2002	rjs	Regenerate.
1.449	03-Jul-2002	msaitoh	regen
1.448	27-Jun-2002	pooka	regen
1.447	26-Jun-2002	cjs	Regenerate.
1.446	26-Jun-2002	cjs	Regenerate.
1.445	22-Jun-2002	fvdl	Regenerate.
1.444	22-Jun-2002	cjs	Regenerate.
1.443	21-Jun-2002	thorpej	Regenerate: Add product ID for Intel 82452KX/GX PCI-Host bridge.
1.442	17-Jun-2002	jdolecek	regen: add VIA Technologies VT8653 and VT8633 entries move VT82C686A_MC97 entry to where it numerically belongs
1.441	10-Jun-2002	cjs	Regenerate.
1.440	09-Jun-2002	taca	Regen.
1.439	04-Jun-2002	fvdl	Regen.
1.438	03-Jun-2002	briggs	Regen. for MPC8245 host bridge.
1.437	02-Jun-2002	seb	regen for Titan PCI-800L.
1.436	28-May-2002	christos	regen
1.435	28-May-2002	tron	Regen from "pcidevs".
1.434	27-May-2002	drochner	regen
1.433	25-May-2002	ichiro	regen
1.432	24-May-2002	lukem	regen for NetMos Quad UART and 1284 Printer port
1.431	19-May-2002	bouyer	branches: 1.431.2; Regen: add HPT372 and 74 PCI IDE controllers.
1.430	19-May-2002	sommerfeld	Regen
1.429	19-May-2002	sommerfeld	Regen
1.428	18-May-2002	matt	Regen.
1.427	15-May-2002	augustss	branches: 1.427.2; Regen.
1.426	14-May-2002	augustss	Regen.
1.425	03-May-2002	augustss	Regen.
1.424	29-Apr-2002	nonaka	Regen. (Added two IBM devices.)
1.423	26-Apr-2002	ad	Regen.
1.422	18-Apr-2002	thorpej	Regen: Added Symbios 53c875A product ID.
1.421	18-Apr-2002	tshiozak	add PCI/USB vendor ID for Kurusugawa Electronics, Inc. PCI-ID: 0x6809, USB-ID: 0x0b7e.
1.420	16-Apr-2002	drochner	regen
1.419	15-Apr-2002	mycroft	Regen.
1.418	15-Apr-2002	ad	Regen.
1.417	05-Apr-2002	mycroft	Regen.
1.416	27-Mar-2002	abs	Regen for 'CONEXANT 56KFAXMODEM 0x1085 HW 56K Fax Modem'
1.415	19-Mar-2002	bouyer	Regen: add AMD768 devices
1.414	11-Mar-2002	joda	regen
1.413	01-Mar-2002	kleink	Regen: add Motorola Raven, Falcon and Hawk product IDs.
1.412	28-Feb-2002	junyoung	Regen.
1.411	28-Feb-2002	junyoung	Regen; fix for duplicate ATI Rage Pro.
1.410	27-Feb-2002	martin	regen
1.409	27-Feb-2002	junyoung	Regen.
1.408	24-Feb-2002	kleink	Regen: add IBM 82660 PowerPC to PCI Bridge and Memory Controller.
1.407	21-Feb-2002	simonb	regen: Use BCM prefix instead of SB prefix in previous.
1.406	20-Feb-2002	simonb	regen: Add the SiByte/Broadcom SB1250 PCI and LDT bridge IDs.
1.405	20-Feb-2002	simonb	regen: Rename the SIBYTE entry to "Broadcom Corp. (SiByte)".
1.404	16-Feb-2002	thorpej	Regen: Add the Elan SC520 System Controller.
1.403	04-Feb-2002	junyoung	Regen.
1.402	02-Feb-2002	tsutsui	Regen; add Initio INIC-850/1060 SCSI and variants.
1.401	02-Feb-2002	tsutsui	Regen; add HiNT Corp. HB1 PCI-PCI bridge.
1.400	30-Jan-2002	tron	Regen.
1.399	29-Jan-2002	junyoung	Regen.
1.398	28-Jan-2002	thorpej	Regen: Add product ID for the Intel i82544GC 64-bit variant.
1.397	26-Jan-2002	tron	Regen.
1.396	26-Jan-2002	ichiro	regen
1.395	22-Jan-2002	junyoung	Regen.
1.394	22-Jan-2002	junyoung	Regen. Add Mach64 GS.
1.393	21-Jan-2002	ichiro	regen
1.392	21-Jan-2002	eeh	Regen.
1.391	21-Jan-2002	thorpej	Regen; Broadcom BCM570x-related IDs.
1.390	15-Jan-2002	thorpej	Regen; add Intel i82544GC Gig-E.
1.389	14-Jan-2002	augustss	Regen.
1.388	14-Jan-2002	augustss	Regen.
1.387	13-Jan-2002	augustss	Regen.
1.386	09-Jan-2002	augustss	Regen.
1.385	05-Jan-2002	junyoung	Regen.
1.384	05-Jan-2002	augustss	Regen.
1.383	03-Jan-2002	rafal	Regen (added Rainbow Technologies & Rainbow's CryptoSwift 200).
1.382	01-Jan-2002	augustss	Regen.
1.381	30-Dec-2001	ichiro	regen
1.380	29-Dec-2001	ichiro	regen
1.379	20-Dec-2001	thorpej	Regen; added SiS 630 PCI-Host bridge.
1.378	16-Dec-2001	thorpej	Regen: Add Middle Digital, Inc. vendor ID. Add product IDs for the Middle Digital PCI-Weasel Virtual VGA, Serial Port, and Control registers.
1.377	15-Dec-2001	soren	Regen.
1.376	13-Dec-2001	bouyer	Regen: Add some promise controllers, from FreeBSD
1.375	11-Dec-2001	ad	Regen.
1.374	11-Dec-2001	kleink	Regen: add product ID for the Motorola MPC107.
1.373	07-Dec-2001	onoe	Regen: Add HPB and LPC functions of 82801BAM.
1.372	07-Nov-2001	thorpej	Regen; correct i80312 description.
1.371	07-Nov-2001	thorpej	Regen; add Cyclone PCI-700 (IQ80310).
1.370	25-Oct-2001	augustss	Regen.
1.369	24-Oct-2001	thorpej	Regen; Add product IDs for the Powerhouse Systems PowerPro and PowerTop PowerPC system controllers.
1.368	24-Oct-2001	thorpej	Regen; more Intel Gig-E fixup, and add Intel i82544 IDs.
1.367	16-Oct-2001	tron	branches: 1.367.2; Regen.
1.366	13-Oct-2001	ichiro	regen.
1.365	11-Oct-2001	tron	Regen.
1.364	03-Oct-2001	simonb	Regen: Minor correction to the name of the 405GP PCI bridge.
1.363	01-Oct-2001	thorpej	Regen; added product ID for RealTek 8139[BC].
1.362	28-Sep-2001	thorpej	Regen; more corrections to Intel Gig-e product IDs.
1.361	27-Sep-2001	itojun	regen
1.360	15-Sep-2001	chris	Regen
1.359	10-Sep-2001	eeh	Regen.
1.358	05-Sep-2001	drochner	branches: 1.358.2; regen
1.357	04-Sep-2001	thorpej	Regen; added Intel IXP1200 Network Processor.
1.356	04-Sep-2001	pooka	regen.. again.. sigh
1.355	04-Sep-2001	pooka	regen
1.354	30-Aug-2001	soren	Regen.
1.353	28-Aug-2001	tsubai	Regen.
1.352	27-Aug-2001	enami	Regen.
1.351	27-Aug-2001	ad	Regen.
1.350	23-Aug-2001	thorpej	Regen; take part numbers out of PCnet-PCI product descriptions.
1.349	23-Aug-2001	thorpej	Regen; Add Promise ATA/100 TX2.
1.348	20-Aug-2001	fvdl	Regen.
1.347	20-Aug-2001	gehenna	Regen
1.346	18-Aug-2001	thorpej	Regen; added Avance Logic ALS4000.
1.345	17-Aug-2001	matt	regen.
1.344	31-Jul-2001	matt	regen
1.343	28-Jul-2001	pooka	regen
1.342	25-Jul-2001	thorpej	Regen; Added Antares Gigabit Ethernet.
1.341	23-Jul-2001	bouyer	Regen for TITAN PCI100H.
1.340	30-Jun-2001	thorpej	branches: 1.340.2; Regen.
1.339	28-Jun-2001	thorpej	Add Tamarack Microelectronics, Inc. vendor ID, and product ID for Tamarack TC9021 Gigabit Ethernet.
1.338	27-Jun-2001	mcr	added Solidum PAX.ware 1100 and Lucent ORCA PCI targets
1.337	20-Jun-2001	thorpej	Regen; add/update 3Com 3CR990 products.
1.336	20-Jun-2001	thorpej	Regen; add SysKonnect SKNET-GE.
1.335	20-Jun-2001	imp	Regen.
1.334	18-Jun-2001	onoe	regen: Add Cisco/Aironet 350
1.333	17-Jun-2001	tsubai	Regen.
1.332	13-Jun-2001	simonb	Regen; add IBM 405GP PCI Bridge & ROB-in i960RP Microprocessor.
1.331	11-Jun-2001	thorpej	Regen; Add Level One vendor ID, and LXT-1001 10/100/1000 Ethernet product ID.
1.330	11-Jun-2001	tron	Regen from "pcidevs".
1.329	31-May-2001	nisimura	Regen; product IDs for IDT RC32334 and RC32332 systeml controllers.
1.328	21-May-2001	thorpej	Regen; add ICH3 network controller IDs.
1.327	18-May-2001	enami	Regen.
1.326	17-May-2001	drochner	regen
1.325	14-May-2001	thorpej	Regen; added DP83820 Gigabit Ethernet chip.
1.324	06-May-2001	fvdl	Regen.
1.323	25-Apr-2001	simonb	Regen.
1.322	21-Apr-2001	tsutsui	Regen from pcidevs rev. 1.318: > Correct an entry for Acard UDMA IDE controller. > The chip is actually marked as ATP860-A, not ATP860R.
1.321	13-Apr-2001	hubertf	Regen for moved Trident CyberBlade entry
1.320	12-Apr-2001	itojun	regen.
1.319	12-Apr-2001	matt	Regen.
1.318	09-Apr-2001	ad	Regen.
1.317	31-Mar-2001	minoura	Regen.
1.316	21-Mar-2001	soren	Regen.
1.315	21-Mar-2001	hubertf	regen for trident cyberclade
1.314	20-Mar-2001	mrg	regen
1.313	17-Mar-2001	thorpej	Regen.
1.312	13-Mar-2001	bouyer	Regen.
1.311	07-Mar-2001	matt	Regen.
1.310	05-Mar-2001	matt	branches: 1.310.2; Regen
1.309	04-Mar-2001	msaitoh	regen
1.308	02-Mar-2001	thorpej	Regen -- Add Mylex AcceleRAID and eXtremeRAID product and subsystem IDs.
1.307	02-Mar-2001	thorpej	Regen; added Xircom WinGlobal modem product ID.
1.306	18-Feb-2001	bouyer	Regen.
1.305	18-Feb-2001	matt	regen
1.304	12-Feb-2001	tshiozak	add CMI8738B.
1.303	11-Feb-2001	ad	Regen.
1.302	04-Feb-2001	briggs	Regen.
1.301	02-Feb-2001	hubertf	Regen: Add M5243 PCI-AGP Bridge Fixes PR 11466 by Jared D. McNeill <jmcneill@invisible.yi.org>
1.300	31-Jan-2001	thorpej	Regen; add SEGA Dreamcast Broadband Adapter.
1.299	26-Jan-2001	rh	regen
1.298	26-Jan-2001	rh	regen
1.297	22-Jan-2001	jdolecek	regen: US Robotics (3Com) 56K Voice Internal PCI Modem (WinModem) entry added
1.296	22-Jan-2001	mjl	Regen for stallion.
1.295	18-Jan-2001	jdolecek	regen: make pci_knowndevs[] const
1.294	17-Jan-2001	thorpej	Regen; Intel Gigabit Ethernet update.
1.293	15-Jan-2001	thorpej	Regen.
1.292	12-Jan-2001	itojun	regen
1.291	09-Jan-2001	rh	regen
1.290	09-Jan-2001	thorpej	Regen; added Symbios 53c1510D.
1.289	08-Jan-2001	thorpej	Regen; added Accton EN2242.
1.288	08-Jan-2001	thorpej	Regen; added DEC SWXCR RAID.
1.287	08-Jan-2001	rh	regen
1.286	03-Jan-2001	mycroft	Regen.
1.285	31-Dec-2000	ad	Regen.
1.284	30-Dec-2000	eeh	Regen.
1.283	27-Dec-2000	tsutsui	Regen. (Fix Acard UDMA IDE)
1.282	27-Dec-2000	augustss	Regen.
1.281	23-Dec-2000	tsutsui	Regen. (Adding Acard UDMA IDE controllers)
1.280	14-Dec-2000	onoe	regen
1.279	05-Dec-2000	drochner	regen
1.278	20-Nov-2000	onoe	regen. CXD32222 -> CXD3222 (Sony OHCI Controller)
1.277	18-Nov-2000	thorpej	Regen.
1.276	14-Nov-2000	bouyer	Regen.
1.275	04-Nov-2000	thorpej	Regen.
1.274	03-Nov-2000	ad	regen.
1.273	03-Nov-2000	ad	Regen.
1.272	27-Oct-2000	thorpej	Regen.
1.271	27-Oct-2000	thorpej	Regen.
1.270	19-Oct-2000	thorpej	Regen.
1.269	14-Oct-2000	simonb	Regen
1.268	12-Oct-2000	itojun	regen
1.267	03-Oct-2000	thorpej	Regen: ADMtek AN985
1.266	27-Sep-2000	augustss	Regen.
1.265	20-Sep-2000	thorpej	Regen.
1.264	19-Sep-2000	fvdl	Regen.
1.263	13-Sep-2000	ad	Regen.
1.262	13-Sep-2000	tron	Regen.
1.261	13-Sep-2000	tron	Regen.
1.260	13-Sep-2000	tron	Regen.
1.259	07-Sep-2000	soren	Regen.
1.258	04-Sep-2000	soren	Regen.
1.257	04-Sep-2000	itojun	regen
1.256	02-Sep-2000	cgd	regen for changed pcidevs
1.255	29-Aug-2000	wiz	regen
1.254	21-Aug-2000	enami	Regen.
1.253	21-Aug-2000	haya	Regen.
1.252	07-Aug-2000	billc	Manually edit the headers to show which version of pcidevs this was generated from. XXX - this really should be automated
1.251	07-Aug-2000	billc	Added PCI IDs for 3Com's Home Connect (3c450) and Office Connect (3cSOHO100). They both are register compat with later versions of the 3c905 (tornado/ vortex) series.
1.250	04-Aug-2000	ad	Regen.
1.249	23-Jul-2000	jonathan	* Add entries for older Yamaha YMF chips (724, 740) plus the newer steppings (740C, 724F) which use the same microcode as the DS-1S/DS-1E. Regen.
1.248	20-Jul-2000	ad	Regen.
1.247	10-Jul-2000	ad	Regen.
1.246	09-Jul-2000	mycroft	Regen.
1.245	08-Jul-2000	mrg	regen
1.244	04-Jul-2000	enami	Regen.
1.243	03-Jul-2000	gmcgarry	Regen.
1.242	24-Jun-2000	joda	regen
1.241	21-Jun-2000	tron	Regen.
1.240	16-Jun-2000	soren	branches: 1.240.2; Regen.
1.239	13-Jun-2000	tron	Regen.
1.238	12-Jun-2000	nathanw	Regen.
1.237	12-Jun-2000	bouyer	Regen.
1.236	12-Jun-2000	bouyer	Regen.
1.235	09-Jun-2000	augustss	Regen.
1.234	07-Jun-2000	ad	Regen.
1.233	07-Jun-2000	mrg	regen.
1.232	07-Jun-2000	soren	Regen.
1.231	06-Jun-2000	kleink	Regen.
1.230	05-Jun-2000	kleink	Regen.
1.229	05-Jun-2000	kleink	Regen.
1.228	30-May-2000	matt	Regen
1.227	28-May-2000	drochner	regen
1.226	27-May-2000	scw	Regen.
1.225	27-May-2000	soren	branches: 1.225.2; Regen.
1.224	26-May-2000	matt	Regen.
1.223	18-May-2000	drochner	regen
1.222	18-May-2000	garbled	regen
1.221	14-May-2000	soren	Regen.
1.220	12-May-2000	thorpej	Regen.
1.219	07-May-2000	jonathan	PCI product codes for Yamaha YMF744B (DS-1S) YMF754 (DS-1E) audio chips.
1.218	30-Apr-2000	dante	Add ASB-3940U3W host adapters (AdvanSys)
1.217	28-Apr-2000	augustss	Regen.
1.216	26-Apr-2000	thorpej	Regen.
1.215	22-Apr-2000	augustss	Regen.
1.214	19-Apr-2000	bouyer	Regen
1.213	18-Apr-2000	cgd	regen for changed pcidevs
1.212	17-Apr-2000	thorpej	Regen.
1.211	03-Apr-2000	joda	regen
1.210	31-Mar-2000	tsarna	Add auvia, a driver for the integrated AC'97 audio on the VIA VT82C686A southbridge. Tested on a ASUS K7M (w/ AD1881 codec). Please report any problems.
1.209	30-Mar-2000	joda	regen
1.208	27-Mar-2000	ad	Regen.
1.207	23-Mar-2000	joda	regen
1.206	23-Mar-2000	ad	Regen.
1.205	21-Mar-2000	joda	regen
1.204	16-Mar-2000	thorpej	Regen.
1.203	14-Mar-2000	tsarna	Regen
1.202	13-Mar-2000	ad	Regen.
1.201	11-Mar-2000	jdolecek	regen
1.200	09-Mar-2000	jdolecek	regen
1.199	06-Mar-2000	mark	Regenerate from pcidevs.
1.198	04-Mar-2000	tron	Regen from "pcidevs".
1.197	16-Feb-2000	soren	Regen.
1.196	08-Feb-2000	bouyer	regen
1.195	03-Feb-2000	dante	Add AdvanSys U2W host adapters
1.194	03-Feb-2000	joda	regen
1.193	31-Jan-2000	thorpej	Regen.
1.192	25-Jan-2000	haya	Regen.
1.191	24-Jan-2000	joda	regen
1.190	23-Jan-2000	tsubai	Regen.
1.189	12-Jan-2000	haya	Regen.
1.188	11-Jan-2000	matt	Regen.
1.187	09-Jan-2000	soren	Regen.
1.186	05-Jan-2000	drochner	regen
1.185	31-Dec-1999	hubertf	Add an entry for the Trident Microsystems CYBER 9525 (as found in an Toshiba Satellite 2650 XDVD).
1.184	17-Dec-1999	drochner	regen
1.183	17-Dec-1999	augustss	Regen.
1.182	17-Dec-1999	soren	Regen.
1.181	17-Dec-1999	soren	Regen.
1.180	15-Dec-1999	kleink	Regen.
1.179	04-Dec-1999	castor	Regen
1.178	04-Dec-1999	thorpej	Regen.
1.177	01-Dec-1999	drochner	regen
1.176	30-Nov-1999	drochner	regen
1.175	17-Nov-1999	augustss	Regen.
1.174	12-Nov-1999	is	Identification of DOMEX PCI-SCSI board, by Thilo Manske, PR 8772.
1.173	07-Nov-1999	cgd	regen
1.172	06-Nov-1999	rh	regen.
1.171	06-Nov-1999	augustss	Regen.
1.170	04-Nov-1999	tsarna	regen
1.169	04-Nov-1999	augustss	Regen.
1.168	04-Nov-1999	joda	regen
1.167	01-Nov-1999	augustss	Regen.
1.166	29-Oct-1999	matt	Regen.
1.165	29-Oct-1999	haya	Maintainance the name of Ricoh PCI-CardBus bridges.
1.164	11-Oct-1999	hannken	branches: 1.164.2; 1.164.4; Regen.
1.163	29-Sep-1999	thorpej	branches: 1.163.2; Regen.
1.162	29-Sep-1999	thorpej	Regen.
1.161	27-Sep-1999	ad	Regen.
1.160	11-Sep-1999	soren	Regen.
1.159	10-Sep-1999	augustss	Regen.
1.158	08-Sep-1999	thorpej	Regen.
1.157	08-Sep-1999	thorpej	Regen.
1.156	08-Sep-1999	thorpej	Regen.
1.155	01-Sep-1999	drochner	regen
1.154	28-Aug-1999	augustss	Regen.
1.153	27-Aug-1999	thorpej	Regen.
1.152	24-Aug-1999	soren	Regenerate.
1.151	17-Aug-1999	drochner	regen
1.150	16-Aug-1999	mcr	Added Solidum Systems PCI vendor ID and Device IDs
1.149	12-Aug-1999	bouyer	Regen.
1.148	01-Aug-1999	veego	Regen.
1.147	29-Jul-1999	thorpej	Regen.
1.146	28-Jul-1999	soren	Regen.
1.145	27-Jul-1999	castor	regen
1.144	27-Jul-1999	thorpej	Regen.
1.143	20-Jul-1999	drochner	regen
1.142	15-Jul-1999	ross	Regen.
1.141	15-Jul-1999	augustss	Regen.
1.140	14-Jul-1999	fvdl	Regen.
1.139	12-Jul-1999	kleink	Regen.
1.138	01-Jul-1999	itojun	IPv6 kernel code, based on KAME/NetBSD 1.4, SNAP kit 19990628. (Sorry for a big commit, I can't separate this into several pieces...) Pls check sys/netinet6/TODO and sys/netinet6/IMPLEMENTATION for details. - sys/kern: do not assume single mbuf, accept chained mbuf on passing data from userland to kernel (or other way round). - "midway" ATM card: ATM PVC pseudo device support, like those done in ALTQ package (ftp://ftp.csl.sony.co.jp/pub/kjc/). - sys/netinet/tcp: IPv4/v6 dual stack tcp support. - sys/netinet/{ip6,icmp6}.h, sys/net/pfkeyv2.h: IETF document assumes those file to be there so we patch it up. - sys/netinet: IPsec additions are here and there. - sys/netinet6/: most of IPv6 code sits here. - sys/netkey: IPsec key management code - dev/pci/pcidevs: regen In my understanding no code here is subject to export control so it should be safe.
1.137	27-Jun-1999	drochner	regen
1.136	19-Jun-1999	augustss	Regen.
1.135	18-Jun-1999	tsubai	Regen.
1.134	05-Jun-1999	tron	Regen.
1.133	05-Jun-1999	tron	Regen.
1.132	02-Jun-1999	drochner	regen
1.131	29-May-1999	mrg	regen.
1.130	29-May-1999	mrg	regen.
1.129	29-May-1999	fvdl	Regen.
1.128	27-May-1999	mrg	regen.
1.127	23-May-1999	veego	Regen.
1.126	20-May-1999	augustss	Regen.
1.125	17-May-1999	drochner	regen
1.124	07-May-1999	hwr	Regen.
1.123	07-May-1999	hwr	Regen.
1.122	07-May-1999	drochner	regen
1.121	07-May-1999	matt	Regen
1.120	06-May-1999	mrg	regen
1.119	12-Apr-1999	jonathan	Add entries for TI, Ricoh, Toshiba PCI-to-cardbus bridge chips, so we can identify them as cardbus chips supported by the cardbus patches. Add entry for OPTI chipsets whose interrupts arent properly set up by some BIOSes. From cardbus patches<ftp://nandra.iri.co.jp/pub/NetBSD/CardBus by HAYAKAWA Koichi <haya@tcad.ulsi.sony.co.jp>.
1.118	25-Mar-1999	explorer	branches: 1.118.2; Add LAN Media Corporation info
1.117	19-Mar-1999	mycroft	Regen.
1.116	17-Mar-1999	thorpej	Regen.
1.115	16-Mar-1999	tron	Regen from "pcidevs".
1.114	11-Mar-1999	sakamoto	Regen.
1.113	09-Mar-1999	drochner	regen
1.112	19-Feb-1999	thorpej	Regen.
1.111	19-Feb-1999	thorpej	Regen.
1.110	19-Feb-1999	thorpej	Regen.
1.109	14-Feb-1999	tron	Regen from pcidevs.
1.108	05-Feb-1999	thorpej	Regen.
1.107	28-Jan-1999	augustss	Regen.
1.106	16-Jan-1999	drochner	regen
1.105	30-Dec-1998	hwr	Regen
1.104	21-Dec-1998	drochner	regen
1.103	20-Dec-1998	drochner	regen
1.102	18-Dec-1998	thorpej	Regen.
1.101	09-Dec-1998	thorpej	Regen.
1.100	05-Dec-1998	matt	Add Sigma Designs REALmagic Hollywood Plus MPEG-2 decoder
1.99	30-Nov-1998	jtk	regen
1.98	29-Nov-1998	thorpej	Regen.
1.97	29-Nov-1998	thorpej	Regen.
1.96	27-Nov-1998	drochner	regen
1.95	26-Nov-1998	mcr	regenerated pcidevs.h and pcidevs_data.h due to kern/6280 (not kern/5847 as originally noted)
1.94	20-Oct-1998	thorpej	Regen.
1.93	19-Oct-1998	bouyer	Regen.
1.92	05-Oct-1998	tron	Regen.
1.91	04-Oct-1998	tron	Regen.
1.90	01-Oct-1998	fvdl	Regen.
1.89	26-Sep-1998	dante	Change product ADVSYS 2300 to ADVSYS WIDE
1.88	22-Sep-1998	mark	Regenerate from pcidevs.
1.87	18-Sep-1998	drochner	regen
1.86	18-Sep-1998	mark	Regen.
1.85	06-Sep-1998	veego	Regen.
1.84	26-Aug-1998	drochner	regen
1.83	26-Aug-1998	dante	Changed ADVSYSPROD to ADVSYS vendor ID Added 1200A, 1200B, ULTRA, 2300 AdvanSys products ID
1.82	07-Aug-1998	thorpej	Regen.
1.81	28-Jul-1998	thorpej	Regen.
1.80	27-Jul-1998	tron	Regen from pcidevs.
1.79	27-Jul-1998	tron	Regen from pcidevs.
1.78	27-Jul-1998	augustss	Regen.
1.77	25-Jul-1998	explorer	commit this so it gets the right generated from tag
1.76	25-Jul-1998	explorer	define RISCom and RISCom N2 product
1.75	22-Jul-1998	drochner	regen
1.74	19-Jul-1998	tron	Regen from pcidevs.
1.73	13-Jul-1998	mark	Regen.
1.72	10-Jul-1998	tron	Regen from pcidevs.
1.71	30-Jun-1998	augustss	Regen.
1.70	15-Jun-1998	bouyer	Regen.
1.69	15-Jun-1998	bouyer	Regen.
1.68	15-Jun-1998	bouyer	Regen.
1.67	25-May-1998	mark	Regen from pcidevs.
1.66	14-May-1998	augustss	Regen.
1.65	13-May-1998	augustss	Regen.
1.64	05-May-1998	thorpej	Regen.
1.63	04-May-1998	thorpej	Regen.
1.62	30-Apr-1998	mycroft	Regen.
1.61	16-Apr-1998	drochner	regen
1.60	28-Mar-1998	thorpej	Regen.
1.59	28-Mar-1998	cgd	regen
1.58	07-Mar-1998	augustss	Regen.
1.57	04-Mar-1998	cgd	regen
1.56	04-Mar-1998	cgd	regen
1.55	27-Feb-1998	mycroft	Regen.
1.54	07-Feb-1998	augustss	Newly generated version.
1.53	07-Feb-1998	augustss	Add the AGP device in the 82443LX.
1.52	27-Jan-1998	thorpej	Regen: add Packet Engines G-NIC gigabit Ethernet.
1.51	22-Jan-1998	mycroft	Regen.
1.50	09-Jan-1998	thorpej	Regen.
1.49	08-Jan-1998	thorpej	Regen: back out RCS ID related changes.
1.48	05-Jan-1998	perry	regened with RCSIDs in place
1.47	16-Nov-1997	christos	Regen.
1.46	27-Oct-1997	thorpej	Regen.
1.45	14-Oct-1997	thorpej	branches: 1.45.2; Regen.
1.44	08-Oct-1997	mycroft	Regen.
1.43	06-Oct-1997	lonhyn	pcidevs change regen.
1.42	01-Oct-1997	drochner	regen
1.41	23-Sep-1997	mycroft	Regen.
1.40	15-Sep-1997	bouyer	Regen.
1.39	12-Sep-1997	drochner	regen
1.38	28-Aug-1997	thorpej	Regen: added SMC83C170 entry.
1.37	23-Aug-1997	drochner	regen
1.36	14-Aug-1997	drochner	Regen with new RCS ID. (thanks, Bernd)
1.35	14-Aug-1997	drochner	Include some more stuff.
1.34	03-Aug-1997	fvdl	Regen.
1.33	19-Jul-1997	cgd	branches: 1.33.2; regen for new pcidevs
1.32	24-Jun-1997	thorpej	Correct Intel Triton MX PCI IDs, from Tom Yu <tlyu@mit.edu>, PR #3478.
1.31	24-Jun-1997	kml	Added definitions for Essential Communications RoadRunner HIPPI and Gigabit Ethernet interfaces.
1.30	13-May-1997	jtk	regen
1.29	04-Apr-1997	augustss	Regen.
1.28	20-Mar-1997	chuck	regen
1.27	16-Feb-1997	veego	regen.
1.26	01-Jan-1997	veego	branches: 1.26.4; Add the Matrox Millenium and regen the header files.
1.25	30-Nov-1996	jtk	forgot to regenerate these after committing pcidevs for its NetBSD header tag
1.24	28-Nov-1996	jtk	add Cirrus CL-7541, 7543 and ACC 2188; regen .h files
1.23	28-Nov-1996	thorpej	Regen.
1.22	19-Oct-1996	jonathan	regen from pcidevs 1.20
1.21	15-Oct-1996	christos	regen
1.20	13-Oct-1996	christos	regen
1.19	08-Oct-1996	gibbs	dev/microcode/aic7xxx/aic7xxx.seq, dev/microcode/aic7xxx_seq.h, dev/ic/aic7xxxreg.h: Remove intrinsic knowledge about SDTR and WDTR messages and replace it with a generic message system that allows the kernel driver to handle SDTR, WDTR and any other type of extended message it chooses too. This makes the sequencer code much simpler, makes extended message handling debuggable since the bulk of the work is in the kernel driver, and saves lots of instruction space. Regen microcode header file. dev/ic/aic7xxx.c, dev/ic/aic7xxxvar.h: Add code to handle WDTR and SDTR negotiation in light of the changes in the message interface to the sequencer. Don't reject targets that negotiate async by sending an SDTR with a 0 offset. Use an sdtr message with 0,0 to negotiate async when a target suggests a period that is too long for us to handle. Some tape and cdrom drives don't like us doing the message reject that we did in the past. Fix a problem with handing the QUEUE FULL condition. Fix a race condition (most likely the cause of the SCB paging problems) that might allow the sequencer to get unpaused before the condition that caused it to be paused (a SEQINT) was handled. Race condition pointed out by Doug Ledford <dledford@dialnet.net> and by "Dan Willis" <dan@plutotech.com>. dev/pci/ahc_pci.c: Add support for the 2940AU, an aic7860 based controller. dev/pci/pcidevs.h, dev/pci/pcidevs_data.h: Add product IDs for the 2940AU, aic7860 and aic7855. Regen data file. scsi/scsi_message.h: Add MSG_EXT_SDTR_LEN and MSG_EXT_WDTR_LEN - the length of bytes in these extended messages. Thanks to Chuck Cranor <chuck@maria.wustl.edu> for testing these changes out for me.
1.18	24-Sep-1996	christos	New cyclades driver from Timo Rossi.
1.17	29-Aug-1996	cgd	regen for new pcidevs
1.16	26-Aug-1996	cgd	regen for new pcidevs
1.15	23-Aug-1996	cgd	regen for new pcidevs
1.14	15-Jul-1996	cgd	regen for updated pcidevs
1.13	22-Jun-1996	chuck	regenerate to fix "generated from" comment (next commit pcidevs first, then regenerate)
1.12	22-Jun-1996	chuck	add product code for efficient networks eni155p atm card and regenerate
1.11	16-May-1996	mycroft	branches: 1.11.4; Regen.
1.10	13-May-1996	thorpej	Regenerate.
1.9	07-May-1996	thorpej	Regenerate; pcidevs changed.
1.8	04-Apr-1996	mycroft	Regen.
1.7	19-Feb-1996	christos	pcidevs was changed.
1.6	31-Jan-1996	cgd	update for new version of 'pcidevs'
1.5	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.4	10-Nov-1995	christos	pcidevs was changed
1.3	10-Nov-1995	christos	- pcidevs: add 3c590 and adaptec ultra - aic7xxx.c: fix the probe routines to use pcidevs.h
1.2	20-Jun-1995	cgd	vendor IDs for Acer and SIS
1.1	18-Jun-1995	cgd	built from pcidevs list
1.11.4.1	04-Mar-1997	mycroft	Regen.
1.26.4.1	12-Mar-1997	is	Merge in changes from The Trunk
1.33.2.9	14-Oct-1997	thorpej	Sync w/ trunk.
1.33.2.8	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.7	14-Oct-1997	thorpej	Regen.
1.33.2.6	29-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.5	22-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.4	16-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.3	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.2	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.33.2.1	23-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.45.2.2	17-Nov-1997	thorpej	Regen.
1.45.2.1	27-Oct-1997	thorpej	Regen.
1.118.2.13	11-Sep-2000	he	Pull up regenerated files (requested by billc): Recognize 3Com's Home Connect (3c450) and Office Connect (3cSOHO100) Ethernet cards.
1.118.2.12	23-Jul-2000	he	Pull up regenerated files (requested by jonathan): Add product codes for various Yamaha audio chips.
1.118.2.11	07-Jul-2000	he	Pull up regenerated files (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0)
1.118.2.10	03-Jun-2000	he	Pull up regenerated files (requested by tsubai): Add support for PCI variants of Cyclom-4Y and -8Y cards. Also fixes PR#7175 (new serial controller).
1.118.2.9	22-May-2000	he	Pull up regenerated files (requested by bouyer): Add product IDs for some VIA Technologies devices.
1.118.2.8	10-Feb-2000	he	Pull up regenerated file (requeste by bouyer): Add support for US Robotics 3CP5609 PCI 16550 Modem.
1.118.2.7	15-Jan-2000	he	Regenerate after fixing DPT SCSI boards description.
1.118.2.6	20-Oct-1999	he	Regenerate these after adding Realtek to pcidevs.
1.118.2.5	05-Oct-1999	he	Regenerate these from recent pcidevs pullup.
1.118.2.4	24-Sep-1999	cgd	regenerate _after_ pcidevs is checked in
1.118.2.3	29-Aug-1999	he	Regenerated. Recognize 3c905C-TX card. (bouyer)
1.118.2.2	22-Jun-1999	perry	regen (fvdl): recognize another 390xB card
1.118.2.1	13-Apr-1999	jonathan	branches: 1.118.2.1.2; 1.118.2.1.4; Pull up sys/dev/pci/pcidevs 1.113->1.114 (CardBus bridge chips) to 1.4. pcidevs.h, pcidevs_data.h: regen.
1.118.2.1.4.2	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.118.2.1.4.1	28-Jun-1999	itojun	KAME/NetBSD 1.4 SNAP kit, dated 19990628. NOTE: this branch (kame) is used just for refernce. this may not compile due to multiple reasons.
1.118.2.1.2.3	02-Aug-1999	thorpej	Regen.
1.118.2.1.2.2	01-Jul-1999	thorpej	Sync w/ -current.
1.118.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.163.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.164.4.1	15-Nov-1999	fvdl	Sync with -current
1.164.2.10	23-Apr-2001	bouyer	Sync with HEAD.
1.164.2.9	21-Apr-2001	bouyer	Sync with HEAD
1.164.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.164.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.164.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.164.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.164.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.164.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.164.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.164.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.225.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.240.2.27	07-Apr-2004	jmc	Regen (requested by itojun in ticket #71) Add National Semiconductors SC1100 (GEODE) PCI devices
1.240.2.26	06-Jun-2002	he	Pull up regenerated files (requested by mycroft): Add Intel ICH3 network controller IDs. Part of fix to fxp driver to work around a bug which can cause some machines to randomly hang, and to drag along performance improvements and improved hardware support.
1.240.2.25	25-Mar-2002	he	Pull up regenerated files (requested by bouyer): Add support for the pciide part of the AMD-768 chipset.
1.240.2.24	06-Feb-2002	he	Pull up regenerated files (requested by tsutsui): Add support for Acard ATP-850/860 pciide controllers.
1.240.2.23	29-Jan-2002	he	Pull up regenerated files (requested by bouyer): Add support for the 53c1010-33, and bring documentation up to date.
1.240.2.22	24-Jan-2002	he	Pull up regenerated files (requested by he): Add driver for the ESS Allegro-1 / Maestro-3 audio hardware.
1.240.2.21	25-Oct-2001	he	Pull up regenerated files (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.240.2.20	18-Jun-2001	jhawk	Pull up regenerated files (requested by tsubai): Support the new dual USB iBook.
1.240.2.19	11-Jun-2001	he	Pull up regenerated files (requested by tron): Support the D-Link DFE-530TX+. Fixes PR#13161.
1.240.2.18	15-May-2001	he	Pull up regenerated files (requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.240.2.17	06-May-2001	he	Pull up regenerated files (requested by he): Add driver for Cirrus Logic CrystalClear PCI Audio CS4281.
1.240.2.16	03-May-2001	he	Pull up regenerated files (requested by skrll): Add a driver for the ESS Technology Maestro-1/2/2E AC97 audio chips, ES1968 and ES1978.
1.240.2.15	23-Apr-2001	he	Pull up regenerated files (requested by hubertf): Support Accton EN2242 and other AmdTek AN985 compatible cards with the tlp(4) driver.
1.240.2.14	22-Mar-2001	he	Pull up regenerated files (requested by hubertf): Make the Intel Pro/100 as e.g. found on Toshiba Satellite Pro 4600s work.
1.240.2.13	20-Mar-2001	he	Pull up regenerated files (requested by chuck): Add support for the 3Com 3c555, 3c556 and 3c556B MiniPCI Ethernet cards.
1.240.2.12	26-Feb-2001	he	Pull up regenerated files (requested by bouyer): Add support for the Intel 82801BAM IDE controller.
1.240.2.11	31-Dec-2000	jhawk	Regen.
1.240.2.10	30-Oct-2000	tv	Regen.
1.240.2.9	19-Oct-2000	he	Pull up regenerated files (requested by billc): Add PCI IDs for 3Com Home Connect (3c450) and Office Connect (3cSOHO100), register compatible with later versions of 3c905.
1.240.2.8	21-Sep-2000	thorpej	Regen: Add National Semiconductor DP83815 10/100 Ethernet.
1.240.2.7	13-Sep-2000	tron	Regen.
1.240.2.6	22-Aug-2000	enami	Regen.
1.240.2.5	23-Jul-2000	jonathan	Pull up pcidevs.h rev 1.251 and pcidevs_data.h rev 249, by pulling up pcidevs rev 1.245 (YMF 7xx audio product codes), then regen.
1.240.2.4	16-Jul-2000	jhawk	Pullup pcidevs rev 1.242 and regen pcidevs_data.h, pcidevs.h, approved by jhawk: Add ID for i82559ER.
1.240.2.3	06-Jul-2000	enami	Regen.
1.240.2.2	27-Jun-2000	thorpej	Regen.
1.240.2.1	26-Jun-2000	bouyer	Regen.
1.310.2.29	17-Jan-2003	thorpej	Sync with HEAD.
1.310.2.28	15-Jan-2003	thorpej	Sync with HEAD.
1.310.2.27	03-Jan-2003	thorpej	Sync with HEAD.
1.310.2.26	29-Dec-2002	thorpej	Sync with HEAD.
1.310.2.25	19-Dec-2002	thorpej	Sync with HEAD.
1.310.2.24	11-Dec-2002	thorpej	Sync with HEAD.
1.310.2.23	26-Nov-2002	thorpej	Sync with HEAD.
1.310.2.22	11-Nov-2002	nathanw	Catch up to -current
1.310.2.21	18-Oct-2002	nathanw	Catch up to -current.
1.310.2.20	17-Sep-2002	nathanw	Catch up to -current.
1.310.2.19	27-Aug-2002	nathanw	Catch up to -current.
1.310.2.18	13-Aug-2002	nathanw	Catch up to -current.
1.310.2.17	01-Aug-2002	nathanw	Catch up to -current.
1.310.2.16	20-Jun-2002	nathanw	Catch up to -current.
1.310.2.15	17-Apr-2002	nathanw	Catch up to -current.
1.310.2.14	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.310.2.13	28-Feb-2002	nathanw	Regen.
1.310.2.12	28-Feb-2002	nathanw	Regen.
1.310.2.11	28-Feb-2002	nathanw	Catch up to -current.
1.310.2.10	11-Jan-2002	nathanw	More catchup.
1.310.2.9	08-Jan-2002	nathanw	Catch up to -current.
1.310.2.8	14-Nov-2001	nathanw	Catch up to -current.
1.310.2.7	22-Oct-2001	nathanw	Catch up to -current.
1.310.2.6	08-Oct-2001	nathanw	Catch up to -current.
1.310.2.5	21-Sep-2001	nathanw	Catch up to -current.
1.310.2.4	24-Aug-2001	nathanw	Catch up with -current.
1.310.2.3	21-Jun-2001	nathanw	Catch up to -current.
1.310.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.310.2.1	05-Mar-2001	nathanw	file pcidevs_data.h was added on branch nathanw_sa on 2001-04-09 01:57:07 +0000
1.340.2.9	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.340.2.8	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.340.2.7	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.340.2.6	16-Mar-2002	jdolecek	Catch up with -current.
1.340.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.340.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.340.2.3	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.340.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.340.2.1	03-Aug-2001	lukem	update to -current
1.358.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.358.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.367.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.427.2.4	29-Aug-2002	gehenna	catch up with -current.
1.427.2.3	15-Jul-2002	gehenna	catch up with -current.
1.427.2.2	20-Jun-2002	gehenna	catch up with -current.
1.427.2.1	30-May-2002	gehenna	Catch up with -current.
1.431.2.36	26-Oct-2005	jmc	regen
1.431.2.35	01-Jun-2005	riz	Regen for ticket 5721.
1.431.2.34	01-May-2005	tron	Regen for ticket 5756.
1.431.2.33	19-Mar-2005	tron	Regen for ticket 1755.
1.431.2.32	11-Nov-2004	he	Pull up regenerated files (requested by itohy in ticket #1741): Add support for the Workbit NinjaSCSI-32 PCI/Cardbus SCSI driver, njs(4).
1.431.2.31	12-Jul-2004	he	Pull up regenerated files (requested by grant in ticket #1720): Add support for SATALink 3112.
1.431.2.30	22-Jun-2004	tron	Regen for ticket #1718.
1.431.2.29	06-Apr-2004	jmc	Regen (requested by briggs in ticket #1648) Pull in changes to support the Intel 82801EB/ER (ICH5) Controller. PR#25002
1.431.2.28	28-Mar-2004	jmc	Regen for ticket #1631
1.431.2.27	15-Mar-2004	jmc	Regen for ticket #1612
1.431.2.26	20-Oct-2003	cyber	Pull up revision 1.569 (requested by tron in ticket #1440): Add entry for Intel i865G Integrated Graphics Device.
1.431.2.25	10-Sep-2003	grant	Regenerate (requested by tron in ticket #1417): Correct names and descriptions for i875 entries to be consistent with other Intel chipsets.
1.431.2.24	15-Aug-2003	tron	Regen.
1.431.2.23	10-Jul-2003	tron	Regen for ticket #1365.
1.431.2.22	30-Jun-2003	grant	Regen (requested by drochner in ticket #1343): add Intel PRO/100M.
1.431.2.21	19-Jun-2003	grant	Regen (requested by hamajima in ticket #1323): Support the PowerBook G4 12-inch (and maybe 17-inch). Kauai ATA, new obio chip, and Japanese keymap for WSDISPLAY_COMPAT_RAWKBD
1.431.2.20	16-Jun-2003	grant	Regenerate (requested by bouyer in ticket #1234): Add several SiS host bridges, from FreeBSD. Remove duplicate in previous.
1.431.2.19	02-Jun-2003	tron	Regen for ticket 1143.
1.431.2.18	28-Jan-2003	jmc	Regen (requested by fvdl in ticket #1126) Pullup support for bge device. (Broadcom Gigabit Ethernet)
1.431.2.17	27-Jan-2003	jmc	Regen from pcidevs commits to Add Intel 82801E Communications I/O Controller Hub (C-ICH)
1.431.2.16	12-Dec-2002	he	Pull up regenerated file (requested by ad in ticket #1030): Add support for Adaptec 2005S. Handle DPT ioctls() from native binaries. Fix SCSI probe problem.
1.431.2.15	28-Nov-2002	tron	Regen for ticket #1019.
1.431.2.14	26-Nov-2002	tron	Regen for tickets #1017 and #1018.
1.431.2.13	21-Nov-2002	tron	Regen for ticket #925.
1.431.2.12	16-Nov-2002	he	Pull up regenerated files (requested by ad in ticket #646): Add additional Adaptec RAID controller types. Fixes PR#17719.
1.431.2.11	07-Nov-2002	tron	Regen for ticket #452.
1.431.2.10	01-Nov-2002	tron	Regen for ticket #724.
1.431.2.9	02-Aug-2002	lukem	regen (for ticket #603)
1.431.2.8	01-Aug-2002	lukem	regen (for ticket #589)
1.431.2.7	12-Jul-2002	thorpej	Regen. pullup-1-6 ticket #488 (matt).
1.431.2.6	25-Jun-2002	lukem	regen for ticket #362 from cjs
1.431.2.5	21-Jun-2002	lukem	regen for tickes #344 / #345
1.431.2.4	11-Jun-2002	lukem	regen for ticket #249
1.431.2.3	11-Jun-2002	lukem	regen for ticket #166
1.431.2.2	11-Jun-2002	lukem	regen for ticket #139
1.431.2.1	10-Jun-2002	tv	Regen.
1.541.2.18	11-Dec-2005	christos	Sync with head.
1.541.2.17	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.541.2.16	01-Apr-2005	skrll	Sync with HEAD.
1.541.2.15	08-Mar-2005	skrll	Sync with HEAD.
1.541.2.14	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.541.2.13	04-Feb-2005	skrll	Sync with HEAD.
1.541.2.12	24-Jan-2005	skrll	Sync with HEAD.
1.541.2.11	17-Jan-2005	skrll	Sync with HEAD.
1.541.2.10	18-Dec-2004	skrll	Sync with HEAD.
1.541.2.9	29-Nov-2004	skrll	Sync with HEAD.
1.541.2.8	14-Nov-2004	skrll	Sync with HEAD.
1.541.2.7	02-Nov-2004	skrll	Sync with HEAD.
1.541.2.6	19-Oct-2004	skrll	Sync with HEAD
1.541.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.541.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.541.2.3	03-Sep-2004	skrll	Sync with HEAD
1.541.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.541.2.1	03-Aug-2004	skrll	Sync with HEAD
1.613.2.17	12-Nov-2004	jmc	branches: 1.613.2.17.2; Regen (requested by martti in ticket #953) Regen
1.613.2.16	12-Nov-2004	jmc	Regen (requested by martti in ticket #947) Regen
1.613.2.15	08-Oct-2004	jmc	Regen for ticket #896
1.613.2.14	18-Sep-2004	he	Pull up regenerated files (requested by martin in ticket #856): Add support for PCI-X version of LSI FC929, from PR#26836.
1.613.2.13	30-Aug-2004	tron	Regen for ticket 800.
1.613.2.12	30-Aug-2004	tron	Regen for ticket 800.
1.613.2.11	22-Aug-2004	tron	Regen for ticket 781.
1.613.2.10	28-Jul-2004	tron	Regen for ticket #721.
1.613.2.9	28-Jul-2004	tron	Regen for ticket #718.
1.613.2.8	23-Jul-2004	he	Pull up regenerated files (requested by mycroft in ticket #691): Add support for SiS 741. Fixes PR#26394.
1.613.2.7	05-Jul-2004	he	Pull up regenerated files (requested by mycroft in ticket #606): Update ServerWorks chip IDs, per http://www.plasma-online.de/english/identify/picture/serverworks.html Seems remarkably more sane.
1.613.2.6	22-Jun-2004	tron	Regen for ticket #537.
1.613.2.5	13-May-2004	jdc	Regnerate (requested by tron in ticket #316)
1.613.2.4	11-May-2004	tron	Regen for ticket 284, this time with correct revision id for "pcidevs".
1.613.2.3	11-May-2004	tron	Regen for ticket 284.
1.613.2.2	05-Apr-2004	tron	Regen for ticket 54.
1.613.2.1	01-Apr-2004	jmc	Regen (requested by tron in ticket #38) Correct entry for PCI vendor id 0x11af. It is assigned to Avid Technology Inc. and not to Pro-Log Corporation. PR#25004
1.613.2.17.2.19	21-Nov-2006	ghen	Regen for ticket #10749.
1.613.2.17.2.18	11-Sep-2006	tron	Regen for ticket #10693.
1.613.2.17.2.17	04-Jun-2006	tron	Regen for ticket #10635.
1.613.2.17.2.16	11-Apr-2006	riz	Regen for ticket 10409.
1.613.2.17.2.15	17-Mar-2006	riz	Regen for ticket 10364.
1.613.2.17.2.14	01-Dec-2005	tron	Regen for ticket #10158.
1.613.2.17.2.13	01-Dec-2005	riz	Regen for ticket 5990.
1.613.2.17.2.12	21-Jul-2005	snj	Regen for ticket 5507.
1.613.2.17.2.11	05-Jul-2005	riz	Regen for ticket 2029.
1.613.2.17.2.10	21-Jun-2005	riz	Regen for ticket 2007.
1.613.2.17.2.9	19-May-2005	tron	Regen for ticket 1532.
1.613.2.17.2.8	19-May-2005	tron	Regen for ticket 1531.
1.613.2.17.2.7	19-May-2005	tron	Regen for ticket 1531.
1.613.2.17.2.6	11-May-2005	riz	Regen for ticket 1402.
1.613.2.17.2.5	10-May-2005	riz	Regen for ticket 1341.
1.613.2.17.2.4	05-May-2005	riz	Regen for ticket 1153.
1.613.2.17.2.3	01-May-2005	tron	Regen for ticket 1438.
1.613.2.17.2.2	17-Apr-2005	tron	Regen for ticket 1120.
1.613.2.17.2.1	07-Apr-2005	tron	Regen for ticket 1103.
1.684.2.1	29-Apr-2005	kent	sync with -current
1.687.2.4	26-Mar-2005	yamt	sync with head.
1.687.2.3	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.687.2.2	12-Feb-2005	yamt	sync with head.
1.687.2.1	25-Jan-2005	yamt	file pcidevs_data.h was added on branch yamt-km on 2005-02-12 18:17:48 +0000
1.700.2.48	10-Nov-2008	snj	Regen for ticket 1976.
1.700.2.47	08-Apr-2008	jdc	Regenerate for pcidevs 1.701.2.49. (requested by bouyer in ticket #1909).
1.700.2.46	19-Nov-2007	bouyer	Regen for ticket #1863
1.700.2.45	19-Nov-2007	bouyer	Regen for ticket 1862
1.700.2.44	14-Oct-2007	riz	Regen for ticket 1770.
1.700.2.43	14-Oct-2007	riz	Regen for ticket 1769.
1.700.2.42	13-Sep-2007	bouyer	Regen for ticket #1822
1.700.2.41	31-Mar-2007	bouyer	regen for ticket 1681
1.700.2.40	03-Mar-2007	bouyer	regen for ticket #1673
1.700.2.39	23-Dec-2006	ghen	Regen for ticket #1609.
1.700.2.38	23-Dec-2006	ghen	Regen for ticket #1608.
1.700.2.37	20-Nov-2006	ghen	Regen for ticket #1584.
1.700.2.36	27-Aug-2006	ghen	branches: 1.700.2.36.2; Regen for ticket #1479.
1.700.2.35	15-Aug-2006	ghen	Regen for ticket #1466.
1.700.2.34	03-Aug-2006	tron	Regenerate for ticket #1446.
1.700.2.33	31-Jul-2006	tron	Regenerate for ticket #1441.
1.700.2.32	30-Jul-2006	tron	Regen for ticket #1432.
1.700.2.31	04-Jun-2006	tron	Regen for ticket #1351.
1.700.2.30	04-Jun-2006	tron	Regen for ticket #1349.
1.700.2.29	04-May-2006	tron	Regen for ticket #1298.
1.700.2.28	30-Apr-2006	riz	Regen for ticket 1295.
1.700.2.27	20-Apr-2006	snj	Regen for ticket 1266.
1.700.2.26	08-Apr-2006	tron	Regen for ticket #1247.
1.700.2.25	17-Mar-2006	tron	Regen for ticket #1203.
1.700.2.24	15-Feb-2006	riz	Regen for ticket 1169.
1.700.2.23	14-Feb-2006	riz	Sigh. Regen _again_ for ticket 1168.
1.700.2.22	14-Feb-2006	riz	Regen for ticket 1168.
1.700.2.21	05-Feb-2006	riz	Regen for ticket 1156.
1.700.2.20	21-Jan-2006	snj	Regen for ticket 1126.
1.700.2.19	10-Jan-2006	snj	Regen for ticket 1112.
1.700.2.18	05-Jan-2006	riz	Regen for ticket 1086.
1.700.2.17	15-Dec-2005	tron	branches: 1.700.2.17.2; Regen for ticket #1047.
1.700.2.16	24-Nov-2005	tron	Regen for ticket #994.
1.700.2.15	21-Nov-2005	tron	Regen for ticket #957.
1.700.2.14	08-Sep-2005	tron	Regen for ticket #752.
1.700.2.13	06-Sep-2005	riz	Regen for ticket 748.
1.700.2.12	24-Jul-2005	riz	Regen for ticket 617.
1.700.2.11	28-Jun-2005	tron	Regen for ticket 485.
1.700.2.10	22-Jun-2005	tron	Regen for ticket 478.
1.700.2.9	21-Jun-2005	tron	Regen for ticket 476.
1.700.2.8	18-Jun-2005	riz	Regen for ticket 450.
1.700.2.7	28-May-2005	snj	Regen for ticket 323.
1.700.2.6	10-May-2005	tron	Regen for ticket 272.
1.700.2.5	21-Apr-2005	tron	Regen for ticket 180.
1.700.2.4	07-Apr-2005	jmc	Regen for ticket #59
1.700.2.3	07-Apr-2005	jmc	Regen for ticket #16
1.700.2.2	07-Apr-2005	jmc	Regen for ticket #15
1.700.2.1	07-Apr-2005	tron	Regen for ticket 116.
1.700.2.36.2.1	04-Oct-2007	bouyer	regen for ticket #1822
1.700.2.17.2.1	04-Oct-2007	bouyer	regen for ticket #1822
1.725.2.10	27-Feb-2008	yamt	sync with head.
1.725.2.9	11-Feb-2008	yamt	sync with head.
1.725.2.8	21-Jan-2008	yamt	sync with head
1.725.2.7	07-Dec-2007	yamt	sync with head
1.725.2.6	15-Nov-2007	yamt	sync with head.
1.725.2.5	27-Oct-2007	yamt	sync with head.
1.725.2.4	03-Sep-2007	yamt	sync with head.
1.725.2.3	26-Feb-2007	yamt	sync with head.
1.725.2.2	30-Dec-2006	yamt	sync with head.
1.725.2.1	21-Jun-2006	yamt	sync with head.
1.733.2.1	02-Nov-2005	yamt	sync with head.
1.734.2.2	29-Nov-2005	yamt	sync with head.
1.734.2.1	22-Nov-2005	yamt	sync with head.
1.751.2.4	01-Mar-2006	yamt	sync with head.
1.751.2.3	18-Feb-2006	yamt	sync with head.
1.751.2.2	01-Feb-2006	yamt	sync with head.
1.751.2.1	15-Jan-2006	yamt	sync with head.
1.758.2.1	09-Sep-2006	rpaulo	sync with head
1.759.2.3	07-Jun-2006	kardel	Sync with head.
1.759.2.2	01-Jun-2006	kardel	Sync with head.
1.759.2.1	22-Apr-2006	simonb	Sync with head.
1.769.2.8	14-Sep-2006	yamt	sync with head.
1.769.2.7	03-Sep-2006	yamt	sync with head.
1.769.2.6	11-Aug-2006	yamt	sync with head
1.769.2.5	26-Jun-2006	yamt	sync with head.
1.769.2.4	24-May-2006	yamt	sync with head.
1.769.2.3	11-Apr-2006	yamt	sync with head
1.769.2.2	01-Apr-2006	yamt	sync with head.
1.769.2.1	13-Mar-2006	yamt	sync with head.
1.771.2.3	11-May-2006	elad	sync with head
1.771.2.2	19-Apr-2006	elad	sync with head.
1.771.2.1	07-Mar-2006	elad	file pcidevs_data.h was added on branch elad-kernelauth on 2006-04-19 03:25:36 +0000
1.775.2.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.775.2.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.787.2.3	22-Jun-2006	chap	Complete a sync sys/ with head.
1.787.2.2	19-Jun-2006	chap	Sync with head.
1.787.2.1	19-May-2006	chap	file pcidevs_data.h was added on branch chap-midi on 2006-06-19 04:01:36 +0000
1.792.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.814.2.7	15-Sep-2006	tron	Regen for ticket #153.
1.814.2.6	09-Sep-2006	rpaulo	Regen for ticket #139.
1.814.2.5	07-Sep-2006	riz	Regen for ticket #115.
1.814.2.4	07-Sep-2006	riz	Regen for ticket #113
1.814.2.3	07-Sep-2006	riz	Regen for ticket #112.
1.814.2.2	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #28): sys/dev/pci/pcidevs: revision 1.819 sys/dev/pci/pcidevs.h: revision 1.819 sys/dev/pci/pcidevs_data.h: revision 1.818 (via patch) sys/dev/pci/pdcsata.c: revision 1.10 Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.814.2.1	11-Aug-2006	riz	Regen for ticket #8.
1.829.2.3	01-Feb-2007	ad	Sync with head.
1.829.2.2	12-Jan-2007	ad	Sync with head.
1.829.2.1	18-Nov-2006	ad	Sync with head.
1.830.2.3	18-Dec-2006	yamt	sync with head.
1.830.2.2	10-Dec-2006	yamt	sync with head.
1.830.2.1	22-Oct-2006	yamt	sync with head
1.853.2.25	31-Oct-2008	snj	Regen for ticket 1215.
1.853.2.24	29-Sep-2008	jdc	Regenerate for pullup ticket #1192. (addd SYMBIOS SAS1068_2, SAS1064E_2 and SAS1068E_2)
1.853.2.23	17-May-2008	bouyer	Regen for ticket 1145
1.853.2.22	17-May-2008	bouyer	Regen for ticket 1143
1.853.2.21	19-Apr-2008	bouyer	Regen for ticket 1128
1.853.2.20	09-Mar-2008	bouyer	Regen for ticket #1097
1.853.2.19	23-Feb-2008	bouyer	Regen for ticket #1061
1.853.2.18	22-Feb-2008	bouyer	Regen for ticket #1074
1.853.2.17	11-Jan-2008	xtraeme	Regen for ticket #1037.
1.853.2.16	15-Nov-2007	pavel	Backout accidental commit to the release branch.
1.853.2.15	15-Nov-2007	manu	regen
1.853.2.14	04-Nov-2007	xtraeme	Regen for ticket #954.
1.853.2.13	24-Oct-2007	pavel	regen for ticket #935.
1.853.2.12	22-Oct-2007	pavel	regen for ticket #931.
1.853.2.11	07-Oct-2007	xtraeme	Regen for ticket #923.
1.853.2.10	31-Aug-2007	pavel	regen for ticket #678
1.853.2.9	29-Aug-2007	liamjfoy	Regen for ticket #846
1.853.2.8	29-Aug-2007	liamjfoy	Pull up following revision(s) (requested by masanobu in ticket #850): sys/dev/pci/if_wm.c 1.137 sys/dev/pci/if_wm.c 1.138 sys/dev/pci/if_wm.c 1.141 sys/dev/pci/if_wm.c 1.144 sys/dev/pci/if_wmreg.h 1.23 sys/dev/pci/pcidevs 1.893 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphy.c 1.12 sys/dev/mii/miidevs 1.72 sys/dev/mii/miidevs.h regen sys/dev/mii/miidevs_data.h regen share/man/man4/wm.4 1.18 Fix for 82541 and 82547's reset bug. Modify wm_reset() to make some device stable. Add ICH9 support to wm. Add I82566 support to igphy. Pulled done via patch.
1.853.2.7	26-Aug-2007	liamjfoy	Regen pcidevs.h pcidevs_data.h
1.853.2.6	10-Jun-2007	bouyer	Regen for ticket 713
1.853.2.5	02-May-2007	snj	branches: 1.853.2.5.2; Regen for ticket 597.
1.853.2.4	21-Feb-2007	tron	Regenerate for ticket #446.
1.853.2.3	17-Feb-2007	tron	Regen for ticket #449.
1.853.2.2	17-Feb-2007	tron	Regen for ticket #423.
1.853.2.1	12-Dec-2006	tron	Regen for ticket for #267.
1.853.2.5.2.4	03-Jun-2008	skrll	Sync with netbsd-4.
1.853.2.5.2.3	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.853.2.5.2.2	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.853.2.5.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.867.2.5	17-May-2007	yamt	sync with head.
1.867.2.4	07-May-2007	yamt	sync with head.
1.867.2.3	15-Apr-2007	yamt	sync with head.
1.867.2.2	24-Mar-2007	yamt	sync with head.
1.867.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.870.2.9	23-Oct-2007	ad	Sync with head.
1.870.2.8	12-Oct-2007	ad	Sync with head.
1.870.2.7	09-Oct-2007	ad	Sync with head.
1.870.2.6	20-Aug-2007	ad	Sync with HEAD.
1.870.2.5	15-Jul-2007	ad	Sync with head.
1.870.2.4	09-Jun-2007	ad	Sync with head.
1.870.2.3	27-May-2007	ad	Sync with head.
1.870.2.2	10-Apr-2007	ad	Sync with head.
1.870.2.1	13-Mar-2007	ad	Sync with head.
1.871.4.2	29-Mar-2007	reinoud	Pullup to -current
1.871.4.1	18-Mar-2007	reinoud	First attempt to bring branch in sync with HEAD
1.871.2.1	11-Jul-2007	mjf	Sync with head.
1.884.6.13	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.884.6.12	21-Nov-2007	joerg	Sync with HEAD.
1.884.6.11	14-Nov-2007	joerg	Sync with HEAD.
1.884.6.10	31-Oct-2007	joerg	Sync with HEAD.
1.884.6.9	28-Oct-2007	joerg	Sync with HEAD.
1.884.6.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.884.6.7	02-Oct-2007	joerg	Sync with HEAD.
1.884.6.6	06-Sep-2007	jmcneill	Regen.
1.884.6.5	03-Sep-2007	jmcneill	Sync with HEAD.
1.884.6.4	27-Aug-2007	jmcneill	Regen.
1.884.6.3	16-Aug-2007	jmcneill	Sync with HEAD.
1.884.6.2	09-Aug-2007	jmcneill	Sync with HEAD.
1.884.6.1	04-Aug-2007	jmcneill	Sync with HEAD.
1.884.2.3	10-Sep-2007	skrll	Sync with HEAD.
1.884.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.884.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.886.2.2	06-Aug-2007	martin	regen
1.886.2.1	06-Aug-2007	martin	file pcidevs_data.h was added on branch matt-mips64 on 2007-08-06 22:40:09 +0000
1.891.2.3	23-Mar-2008	matt	sync with HEAD
1.891.2.2	09-Jan-2008	matt	sync with HEAD
1.891.2.1	06-Nov-2007	matt	sync with HEAD
1.902.2.1	14-Oct-2007	yamt	sync with head.
1.904.2.4	21-Nov-2007	bouyer	Sync with HEAD
1.904.2.3	18-Nov-2007	bouyer	Sync with HEAD
1.904.2.2	13-Nov-2007	bouyer	Sync with HEAD
1.904.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.912.2.4	18-Feb-2008	mjf	Sync with HEAD.
1.912.2.3	27-Dec-2007	mjf	Sync with HEAD.
1.912.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.912.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.919.4.1	11-Dec-2007	yamt	sync with head.
1.919.2.1	26-Dec-2007	ad	Sync with head.
1.920.2.3	11-Jan-2008	bouyer	Sync with HEAD
1.920.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.920.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.929.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.929.6.4	05-Oct-2008	mjf	Sync with HEAD.
1.929.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.929.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.929.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.929.2.1	24-Mar-2008	keiichi	sync with head.
1.937.4.9	09-Oct-2010	yamt	sync with head
1.937.4.8	11-Aug-2010	yamt	sync with head.
1.937.4.7	11-Mar-2010	yamt	sync with head
1.937.4.6	19-Aug-2009	yamt	sync with head.
1.937.4.5	18-Jul-2009	yamt	sync with head.
1.937.4.4	20-Jun-2009	yamt	sync with head
1.937.4.3	16-May-2009	yamt	sync with head
1.937.4.2	04-May-2009	yamt	sync with head.
1.937.4.1	16-May-2008	yamt	sync with head.
1.937.2.2	04-Jun-2008	yamt	sync with head
1.937.2.1	18-May-2008	yamt	sync with head.
1.938.2.4	10-Oct-2008	skrll	Sync with HEAD.
1.938.2.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.938.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.938.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.942.4.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.942.4.1	19-Oct-2008	haad	Sync with HEAD.
1.942.2.2	31-Jul-2008	simonb	Sync with head.
1.942.2.1	18-Jul-2008	simonb	Sync with head.
1.962.4.21	09-Nov-2015	sborrill	Regen
1.962.4.20	30-Apr-2015	snj	regen for ticket 1962
1.962.4.19	07-Sep-2013	bouyer	Regen for ticket 1867
1.962.4.18	19-Jun-2013	bouyer	regen for ticket #1850
1.962.4.17	20-Jan-2013	bouyer	regen for ticket #1842
1.962.4.16	17-Mar-2012	bouyer	Regen for ticket 1739
1.962.4.15	22-Nov-2010	riz	branches: 1.962.4.15.2; Regen for ticket 1470.
1.962.4.14	20-Nov-2010	riz	Regen for ticket 1362.
1.962.4.13	19-Nov-2010	riz	Regen for ticket 1359.
1.962.4.12	19-Nov-2010	riz	Regen for ticket #1358.
1.962.4.11	28-Mar-2010	snj	Regen for ticket 1326.
1.962.4.10	27-Jan-2010	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1277): sys/dev/pci/if_wm.c 1.184-1.192, 1.194 sys/dev/pci/if_wmreg.h 1.29-1.35 sys/dev/pci/if_wmvar.h 1.5-1.8 sys/dev/pci/pcidevs 1.1006,1.1009-1.1010, 1.1012-1.1013 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/mii/igphyreg.h 1.5 sys/dev/mii/inbmphyreg.h 1.1 - Add support for i82583V. - Add some ICH9 and ICH10 devices. - Add support for PCH. - Fix the bug that ICH9 can't found a PHY. Fixes PR#42237 - Fix an incorrect test for WM_F_EEPROM_INVALID since rev. 1.183. Some old chips don't set EECD_EE_PRES. - Fix a bug that both WM_F_EEPROM_SPI and WM_F_EEPROM_FLASH are set. - Add a missing decrement for a timeout reported by Wolfgang Stukenbrock in PR#42422. - PBA setting for i82574 is not 12K but 20K. - Enable checking the management mode on 82574. - Fix the length of the delay() in wm_gmii_reset(). It fixed the problem that sometimes the driver misunderstood PHYs in mii_attach(). It was reported by MATSUI Yoshihiro. We observed it on ICH9. - Fix the checking of jumbo frame function - Remove the extra macro definition for the offset 0x1a in EEPROM. - Add missing break in wm_reset()... - Fix the offset of WMREG_PBS... - Make wm_reset() and wm_gmii_reset() close to e1000 driver. At least, this change make wm_attach() stable on ICH9. - Reset GMII interface after wm_reset() in wm_init(). - Rework for assigning mii_{read,write}reg(). Use PCI product ID to identify the PHY. - Add code about LPLU(Low Power Link Up) function. It seems that we have to do the same work for ICH9. - Fixes the rx stall problem on 82578 by MANY workaround code. We need more work for 82577.
1.962.4.9	27-Jan-2010	sborrill	Regen
1.962.4.8	23-Jan-2010	bouyer	Regen for ticket 1254
1.962.4.7	18-Dec-2009	snj	Regen for ticket 1175.
1.962.4.6	28-Oct-2009	bouyer	Regen for ticket 1114
1.962.4.5	26-Sep-2009	snj	Regen for ticket 1021.
1.962.4.4	21-May-2009	snj	Regen for ticket 771.
1.962.4.3	05-May-2009	bouyer	Regen for ticket 737
1.962.4.2	03-May-2009	snj	Regen for ticket 722.
1.962.4.1	20-Mar-2009	msaitoh	branches: 1.962.4.1.4; Regen for ticket 587.
1.962.4.15.2.2	14-Oct-2011	matt	Regen.
1.962.4.15.2.1	07-Jan-2011	matt	Regen.
1.962.4.1.4.9	05-Nov-2013	matt	Regen.
1.962.4.1.4.8	05-Nov-2013	matt	Pull down xhci support from HEAD
1.962.4.1.4.7	03-Jan-2012	matt	Regen.
1.962.4.1.4.6	31-Dec-2011	matt	Regen.
1.962.4.1.4.5	24-Dec-2011	matt	Regen.
1.962.4.1.4.4	10-May-2010	matt	Regen.
1.962.4.1.4.3	21-Apr-2010	matt	sync to netbsd-5
1.962.4.1.4.2	07-Apr-2010	cliff	add RMI XLR PCI-X bridge
1.962.4.1.4.1	09-Nov-2009	cliff	in pcidevs: - add vendor entry for RMI - add product entry for RMI XLS integrated PCIe Interface regenerate (make) pcidevs.h and pcidevs_data.h
1.962.2.3	28-Apr-2009	skrll	Sync with HEAD.
1.962.2.2	03-Mar-2009	skrll	Sync with HEAD.
1.962.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.976.2.2	23-Jul-2009	jym	Sync with HEAD.
1.976.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1015.2.3	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.1015.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.1015.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1021.2.7	12-Jun-2011	rmind	sync with head
1.1021.2.6	31-May-2011	rmind	sync with head
1.1021.2.5	21-Apr-2011	rmind	sync with head
1.1021.2.4	06-Mar-2011	rmind	sync with head (and fix few botches with this)
1.1021.2.3	05-Mar-2011	rmind	sync with head
1.1021.2.2	03-Jul-2010	rmind	sync with head
1.1021.2.1	30-May-2010	rmind	sync with head
1.1051.4.2	05-Mar-2011	bouyer	Sync with HEAD
1.1051.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.1051.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1078.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.1095.6.7	02-Jun-2012	mrg	sync to latest -current.
1.1095.6.6	29-Apr-2012	mrg	sync to latest -current.
1.1095.6.5	05-Apr-2012	mrg	sync to latest -current.
1.1095.6.4	06-Mar-2012	mrg	sync to -current
1.1095.6.3	06-Mar-2012	mrg	sync to -current
1.1095.6.2	04-Mar-2012	mrg	sync to latest -current.
1.1095.6.1	18-Feb-2012	mrg	merge to -current.
1.1095.2.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1095.2.4	16-Jan-2013	yamt	sync with (a bit old) head
1.1095.2.3	30-Oct-2012	yamt	sync with head
1.1095.2.2	23-May-2012	yamt	sync with head.
1.1095.2.1	17-Apr-2012	yamt	sync with head
1.1096.2.18	30-Apr-2015	snj	regen for ticket 1296
1.1096.2.17	19-Apr-2015	riz	Regen for ticket 1293.
1.1096.2.16	07-Dec-2014	martin	regen
1.1096.2.15	04-Dec-2014	snj	regen for ticket 1202
1.1096.2.14	09-Nov-2014	martin	Regen for ticket #1188
1.1096.2.13	07-Sep-2013	bouyer	regen for ticket 939
1.1096.2.12	14-Jul-2013	riz	Regen for ticket #906.
1.1096.2.11	22-Nov-2012	riz	Regen for ticket 691.
1.1096.2.10	18-Nov-2012	msaitoh	Regen for ticket #658
1.1096.2.9	31-Oct-2012	riz	Regen for ticket 643.
1.1096.2.8	24-Oct-2012	riz	Regen for ticket #631
1.1096.2.7	24-Oct-2012	riz	Regen for ticket 628.
1.1096.2.6	28-Jun-2012	riz	branches: 1.1096.2.6.2; Regen for ticket #362.
1.1096.2.5	23-Apr-2012	riz	Regen for ticket #191.
1.1096.2.4	16-Apr-2012	riz	Regen for ticket 180.
1.1096.2.3	03-Apr-2012	riz	Regen for ticket #151.
1.1096.2.2	03-Apr-2012	riz	Regen for ticket #149.
1.1096.2.1	22-Mar-2012	riz	Regen for ticket #137.
1.1096.2.6.2.1	18-Nov-2012	matt	Merge tickets 658 and 659.
1.1124.2.5	03-Dec-2017	jdolecek	update from HEAD
1.1124.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1124.2.3	23-Jun-2013	tls	resync from head
1.1124.2.2	25-Feb-2013	tls	resync with head
1.1124.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.1149.4.1	23-Jul-2013	riastradh	sync with HEAD
1.1149.2.2	18-May-2014	rmind	sync with head
1.1149.2.1	28-Aug-2013	rmind	sync with head
1.1176.2.1	10-Aug-2014	tls	Rebase.
1.1191.2.12	14-Aug-2018	martin	Regen for ticket #1628
1.1191.2.11	03-Jan-2018	snj	regen for ticket 1547
1.1191.2.10	25-Jul-2017	snj	regen for ticket 1413
1.1191.2.9	09-Dec-2016	snj	regen for ticket 1300
1.1191.2.8	09-Dec-2016	snj	regen for ticket 1298
1.1191.2.7	08-Dec-2016	snj	regen for ticket 1291
1.1191.2.6	19-Apr-2015	riz	branches: 1.1191.2.6.4; Regen for ticket #699
1.1191.2.5	17-Feb-2015	martin	Regen (ticket #528)
1.1191.2.4	04-Feb-2015	snj	regen for ticket 473
1.1191.2.3	28-Jan-2015	snj	regen for ticket 467
1.1191.2.2	17-Jan-2015	martin	regen (ticket #432)
1.1191.2.1	07-Nov-2014	snj	regen for ticket 190
1.1191.2.6.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.1198.2.11	28-Aug-2017	skrll	Sync with HEAD
1.1198.2.10	05-Feb-2017	skrll	Sync with HEAD
1.1198.2.9	05-Dec-2016	skrll	Sync with HEAD
1.1198.2.8	05-Oct-2016	skrll	Sync with HEAD
1.1198.2.7	09-Jul-2016	skrll	Sync with HEAD
1.1198.2.6	29-May-2016	skrll	Sync with HEAD
1.1198.2.5	19-Mar-2016	skrll	Sync with HEAD
1.1198.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1198.2.3	21-Sep-2015	skrll	Regen
1.1198.2.2	06-Jun-2015	skrll	Sync with HEAD
1.1198.2.1	06-Apr-2015	skrll	Sync with HEAD
1.1245.2.6	26-Apr-2017	pgoyette	Regen
1.1245.2.5	26-Apr-2017	pgoyette	Sync with HEAD
1.1245.2.4	20-Mar-2017	pgoyette	Sync with HEAD
1.1245.2.3	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1245.2.2	04-Nov-2016	pgoyette	Sync with HEAD
1.1245.2.1	06-Aug-2016	pgoyette	Sync with HEAD
1.1269.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.1277.2.2	11-May-2017	pgoyette	Sync with HEAD
1.1277.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.1280.2.31	03-Feb-2024	martin	regen (ticket #1931)
1.1280.2.30	18-Oct-2023	martin	regen (for ticket #1915)
1.1280.2.29	13-Oct-2023	martin	Regen for ticket #1913 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1280.2.28	13-Oct-2023	martin	Regen for ticket #1911 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1280.2.27	21-Aug-2023	martin	regen (for ticket #1887)
1.1280.2.26	29-Jul-2023	martin	regen (ticket #1855)
1.1280.2.25	21-Jun-2023	martin	Regen for ticket #1846
1.1280.2.24	18-Jan-2023	martin	Regen for ticket #1783
1.1280.2.23	11-Oct-2022	martin	regen (Ticket #1771)
1.1280.2.22	02-Aug-2022	snj	regen for ticket 1754
1.1280.2.21	03-Dec-2021	martin	regen (ticket #1710)
1.1280.2.20	23-Oct-2021	martin	Regen for ticket #1700: - Add modern QUMRANET/Red Hat VIRTIO range PCI devices. - Add some more product IDs for mcx(4). - Add current generation NVIDIA graphics cards (3050-3090 etc.). - Add RDC R6022 PCI-Host bridge. - Add a whole bunch of radeon devices. - Add Realtek RTL8821CE. - Add Intel 660p SSD, and expand the 760p description. - Add Intel I219's version number. - Add Intel I219V 15-19 and I219LM 16-19. - Add Intel I225V, I225LM and WiFi 6 AX201. - Add newer Intel PCH internal devices. - Add Intel Jasper Lake devices. - PDC20265 is Ultra/100, not 66. - Fix Realtek RTL8125 description. - Sort Cavium devices.
1.1280.2.19	20-Jul-2020	martin	Regen for ticket #1579
1.1280.2.18	20-Jul-2020	martin	regen for ticket #1578
1.1280.2.17	19-Nov-2019	martin	regen for ticket #1448
1.1280.2.16	06-Aug-2019	martin	regen (ticket #1326)
1.1280.2.15	17-Jul-2019	martin	regen
1.1280.2.14	07-Mar-2019	martin	Regen for ticket #1205
1.1280.2.13	11-Feb-2019	martin	Regen for ticket #1188
1.1280.2.12	04-Dec-2018	martin	Regen for ticket #1123
1.1280.2.11	04-Dec-2018	martin	Regen for #1119
1.1280.2.10	17-Oct-2018	martin	Regen for ticket #1064
1.1280.2.9	26-Jul-2018	snj	regen for ticket 934
1.1280.2.8	15-Jun-2018	msaitoh	Regen for ticket #878.
1.1280.2.7	07-Jun-2018	martin	Regen (ticket #846)
1.1280.2.6	18-Apr-2018	martin	regen for ticket #780
1.1280.2.5	14-Apr-2018	martin	regen (for ticket #751): Add Nvidia GeForce 9400M as found in the late 2009 13" MacBook (MacBook6,1) Add new I219 devices (using with 300 series chipset).
1.1280.2.4	09-Apr-2018	martin	Regen (ticket #723)
1.1280.2.3	08-Mar-2018	martin	Regen for ticket #616
1.1280.2.2	21-Nov-2017	martin	regen
1.1280.2.1	25-Jun-2017	snj	regen for ticket 55
1.1310.2.15	26-Jan-2019	pgoyette	Sync with HEAD
1.1310.2.14	18-Jan-2019	pgoyette	Synch with HEAD
1.1310.2.13	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1310.2.12	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.1310.2.11	20-Oct-2018	pgoyette	Sync with head
1.1310.2.10	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1310.2.9	28-Jul-2018	pgoyette	Sync with HEAD
1.1310.2.8	25-Jun-2018	pgoyette	Sync with HEAD
1.1310.2.7	21-May-2018	pgoyette	Sync with HEAD
1.1310.2.6	02-May-2018	pgoyette	Synch with HEAD
1.1310.2.5	22-Apr-2018	pgoyette	Sync with HEAD
1.1310.2.4	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.1310.2.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1310.2.2	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.1310.2.1	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.1333.2.3	21-Apr-2020	martin	Sync with HEAD
1.1333.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1333.2.1	10-Jun-2019	christos	Sync with HEAD
1.1370.2.26	26-Oct-2025	martin	Regen for ticket #1978
1.1370.2.25	08-May-2025	martin	regen for ticket #1944
1.1370.2.24	08-May-2025	martin	Regen for ticket #1944: Various VIA IDE controller related additions/changes.
1.1370.2.23	03-Jul-2024	martin	Regen for ticket #1852 (Brainboxes devices)
1.1370.2.22	03-Feb-2024	martin	regen (ticket #1790)
1.1370.2.21	18-Oct-2023	martin	Regen for ticket #1915
1.1370.2.20	13-Oct-2023	martin	Regen for ticket #1750 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1370.2.19	13-Oct-2023	martin	Regen for ticket #1748 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1370.2.18	21-Aug-2023	martin	regen (for ticket #1719)
1.1370.2.17	29-Jul-2023	martin	regen (ticket #1671)
1.1370.2.16	21-Jun-2023	martin	Regen for ticket #1654
1.1370.2.15	30-Jan-2023	martin	Regen for ticket #1580
1.1370.2.14	18-Jan-2023	martin	Regen for ticket #1566
1.1370.2.13	11-Oct-2022	martin	regen (Ticket #1537)
1.1370.2.12	27-Jul-2022	martin	regen for ticket #1482
1.1370.2.11	03-Dec-2021	martin	regen (ticket #1380)
1.1370.2.10	21-Oct-2021	martin	Regen (for ticket #1362)
1.1370.2.9	10-Jul-2020	martin	Regen for ticket #991
1.1370.2.8	07-Jul-2020	martin	regen
1.1370.2.7	28-Apr-2020	martin	Regen for ticket #864
1.1370.2.6	19-Mar-2020	martin	Regen (for ticket #784)
1.1370.2.5	31-Jan-2020	martin	regen (for ticket #670)
1.1370.2.4	19-Nov-2019	martin	Regen for ticket #449
1.1370.2.3	24-Oct-2019	martin	Regen for ticket #373
1.1370.2.2	24-Sep-2019	martin	regen for ticket #237 (Mellanox device IDs added)
1.1370.2.1	26-Aug-2019	martin	Regen for ticket #112
1.1382.2.2	29-Feb-2020	ad	Sync with head.
1.1382.2.1	25-Jan-2020	ad	Sync with head.
1.1391.2.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.1391.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.1405.2.3	03-Apr-2021	thorpej	Sync with HEAD.
1.1405.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.1405.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.1409.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.1411.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.1451.2.18	26-Oct-2025	martin	Regen for ticket #1184
1.1451.2.17	08-May-2025	martin	regen for ticket #1101
1.1451.2.16	08-May-2025	martin	Regen for ticket #1101: Various VIA IDE controller related additions/changes.
1.1451.2.15	06-Dec-2024	snj	regen for ticket 1024
1.1451.2.14	24-Aug-2024	martin	Regen for ticket #805
1.1451.2.13	03-Jul-2024	martin	Regen for ticket #730 (Brainboxes devices)
1.1451.2.12	03-Feb-2024	martin	regen (ticket #560)
1.1451.2.11	03-Nov-2023	martin	Regen for ticket #447 (Add Mellanox ConnectX-6 Lx)
1.1451.2.10	18-Oct-2023	martin	Regen for ticket #421
1.1451.2.9	13-Oct-2023	martin	Regen for ticket #406 (Intel 82599 LS and Pericom(Diodes) PCIe switches added)
1.1451.2.8	13-Oct-2023	martin	Regen for ticket #403 (add ASMEDIA 0x0624 - ASM1062 + JMB575 Port Multiplier)
1.1451.2.7	08-Oct-2023	martin	Regen for ticket #393 (Add missing I225/I226 series devices)
1.1451.2.6	21-Aug-2023	martin	regen (for ticket #333)
1.1451.2.5	29-Jul-2023	martin	regen (ticket #252)
1.1451.2.4	21-Jun-2023	martin	Regen for ticket #209
1.1451.2.3	21-Jun-2023	martin	Regen for ticket #209
1.1451.2.2	30-Jan-2023	martin	Regen for ticket #61
1.1451.2.1	18-Jan-2023	martin	Regen for ticket #50
1.1485.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1505.2.1	26-Oct-2025	martin	Regen for ticket #71
1.219	06-Nov-2010	jakllsch	branches: 1.219.18; Rework pciide(4) detachment to take the legacy interrupt mapping into consideration and avoid future code duplication. Ports wanting to enable detachment of controllers with compatibility-mapped channels will need to supply a pciide_machdep_compat_intr_disestablish() function.
1.218	05-Nov-2010	jakllsch	pciide(4): Sprinkle static, add detachment support.
1.217	19-Oct-2009	bouyer	branches: 1.217.2; 1.217.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.216	18-Mar-2008	cube	branches: 1.216.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.215	16-Nov-2006	christos	branches: 1.215.24; 1.215.44; 1.215.48; __unused removal on arguments; approved by core.
1.214	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.213	11-Dec-2005	christos	branches: 1.213.20; 1.213.22; merge ktrace-lwp.
1.212	27-Feb-2005	perry	branches: 1.212.4; nuke trailing whitespace
1.211	04-Feb-2005	perry	de-__P
1.210	17-Sep-2004	enami	branches: 1.210.4; 1.210.6; Whitespace nit.
1.209	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.208	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.207	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.206	25-Sep-2003	mycroft	Hide the use of config_interrupts() in one place.
1.205	23-Sep-2003	mycroft	Fix more probe delay and/or failure problems: 1) Don't wait for DRQ on an IDENTIFY command -- if it's not set when we see BSY clear, abort the command and ignore the drive. (Do this by testing for DRQ in the read/write cases in __wdccommand_intr().) 2) Don't wait for DRQ to deassert when we finish an IDENTIFY (or any other non-block command that reads data) -- we don't do this for block I/O, and empirically it doesn't clear on my CF cards at all, causing a pointless 1s delay. 3) Add comments to some of the delay()s, and add missing ones in wdcreset() and the WDCC_RECAL in the so-called "pre-ATA" probe. 4) Slightly simplify the reset sequence -- we were doing an extra I/O. 5) Modify the register writability test to make sure that registers are not overlapped -- this can happen in some weird cases with a missing device 1. 6) Check the error register value after the reset -- if it's not 01h or 81h, as appropriate (see ATA spec), punt. Tested with a number of ATA-only, ATAPI-only, mixed ATA-ATAPI, CF, and IDE disk configurations. Also remove the SINGLE_DRIVE nonsense again.
1.204	21-Sep-2003	bouyer	Ops, remove a line that escaped out of my local tree.
1.203	21-Sep-2003	bouyer	The return value from pciide_mapregs_* are ignored; make them return void. Re-add support for HPT366 in compat mode.
1.202	21-Sep-2003	bouyer	Since we can't detect ghost drives in the wdc back-end, resurect WDC_CAPABILITY_SINGLE_DRIVE.
1.201	20-Sep-2003	enami	Make sure the message from pciide_chipen() starts at the beginning of line.
1.200	19-Sep-2003	mycroft	1) Use config_interrupts() to attach IDE and ATAPI drives. This eliminates most polling. 2) Clean up some goofiness in pciide -- get rid of the whole "candisable" path (it's gratuitous) and simplify the code by calling pciide_map_compat_intr(), *_set_modes() and wdc_print_modes() from central locations. 3) Add a register writability and register ghost test to eliminate phantom drives more quickly.
1.199	17-Sep-2003	bouyer	Add back support for the legacy VT8237 IDE controller. This wasn't an error, on this chipset we have the SATA controller on function 0 of the IDE controller, not the pcib bridge. Fix provided by Stephen Degler.
1.198	15-Sep-2003	bouyer	Add support for Intel 82801EB Serial ATA. Not tested with a drive yet, Matthias Scheler tested that the controller attaches properly.
1.197	15-Sep-2003	bouyer	Add support for VIA 8237 Serial ATA. From Stephen Degler in kern/22727, with some cleanup by me.
1.196	17-Aug-2003	bouyer	Add support for VIA VT8237 (KT600) IDE controller. From Stephen Degler in PR port-i386/22453.
1.195	10-Aug-2003	bouyer	Corect offset for the promise "magic registers" (it depends on channel number). This makes cable detect work properly for newer promise (pdc268 and newer).
1.194	28-Jun-2003	bouyer	branches: 1.194.2; Correct setup of General Purpose Register for PDC20265 (Ultra/100) and newer. From Andreas Johansson in private mail.
1.193	28-Jun-2003	bouyer	Add support for ICH5, from Quentin Garnier in private mail.
1.192	17-May-2003	thorpej	* Use aprint(). Add Ultra/133 to wdc_print_modes().
1.191	28-Apr-2003	nakayama	Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.190	19-Apr-2003	christos	PR/21236: Simon Hitzemann: Fix missing space in error format. While I am there make a single copy of the error string instead of 6 and share it.
1.189	05-Apr-2003	kent	Add '(ICH3)' to the names of 82801CA for consistency with other 82801xx names
1.188	04-Apr-2003	kent	Support for ICH4M IDE Controller
1.187	20-Mar-2003	thorpej	Add support for the Silicon Image SATALink 3112 Serial ATA controller.
1.186	19-Mar-2003	thorpej	Fix pasto.
1.185	18-Mar-2003	thorpej	Add register definitions for the Intel i31244 Serial ATA controller.
1.184	18-Mar-2003	thorpej	Add very basic support for the Intel i31244 Serial ATA controller. We only support legacy (i.e. PCI IDE compatible) mode, for now. Also note that DMA is disabled for rev 0 chips unless explicitly enabled with PCIIDE_I31244_ENABLEDMA.
1.183	15-Mar-2003	bouyer	Ops, I commited the previous change (improved SiS support) from thr wrong tree. Sync with the latest diff I made available, and also include changes from kern/20710 which is much better than the hack I've done.
1.182	14-Mar-2003	bouyer	Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver.
1.181	28-Feb-2003	enami	Match ServerWorks CSB6 IDE controller.
1.180	27-Jan-2003	thorpej	Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.179	25-Jan-2003	thorpej	Add a couple more Promise ATA133 controllers.
1.178	24-Jan-2003	thorpej	* Enable Ultra133 on the NVIDIA nForce 2. * Make sure to test for PCI_VENDOR_AMD before testing for any AMD products.
1.177	24-Jan-2003	thorpej	Add support for the NVIDIA nForce ATA100 and nForce2 ATA133 IDE controllers, which are more-or-less compatible with the AMD controllers. XXX Need to determine the correct timing value for the nForce2 XXX at Ultra133, so we cap it at Ultra100, for now.
1.176	26-Dec-2002	matt	Recognize Promise Serial ATA/150 TX2plus
1.175	20-Nov-2002	bouyer	For sis_setup_channel, acer_setup_channel, hpt_setup_channel, pdc202xx_setup_channel, pdc20268_setup_channel: Properly compute the address of the DMA control register for channel 1. I think the controllers ignore these bits, I suspect it's only there so that the BIOS can tell the OS is has configured DMA, but better be correct. Thanks to Alexander Yurchenko for pointing this out.
1.174	05-Oct-2002	kent	Add support for VIA VT8235
1.173	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.172	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.171	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.170	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.169	15-Sep-2002	bouyer	Fix typo in comment. From Alexander Yurchenko <grange@rt.mipt.ru>
1.168	25-Aug-2002	bouyer	Support Ultra/133 on promise controllers that can do it.
1.167	25-Aug-2002	bouyer	Correct setup for Ultra133 capable VIA chipsets, From Matthias Drochner on current-users, with cross-check and some improvement from linux-2.4.19 and FreeBSD-current. Also don't set the APO_UDMA_CLK66 bit for Ultra/100 capable chipset, and support Ultra/133 for the VT8233A.
1.166	23-Aug-2002	bouyer	Add support for HPT372. From sdegler@kashmir.degler.net in kern/17908, with some improvement from me, and from FreeBSD.
1.165	23-Aug-2002	bouyer	Add support for Promise Ultra133TX2 and Ultra133TX2v2, from FreeBSD and stephen@degler.net in kern/17509.
1.164	10-Aug-2002	toshii	Use UDMA/100 on SiS745 chipset.
1.163	30-Jul-2002	bouyer	Add support for the 82801DB ICH4 IDE controller, from kern/17757 by Andreas Wrede <andreas@planix.com>
1.162	26-Jul-2002	wiz	Spell '[Rr]ight' correctly. From Jim Bernard.
1.161	26-Jul-2002	onoe	Add support of Silicon Image 0680 Ultra ATA/133 ATA Controller. It's ugly that all register values are written in numeric, but I can't find any definition of the registers to be written in literal.
1.160	22-Jul-2002	bouyer	Ops, this had already been commited by cjs. Still, call the controller ATA133, even though we use it at ATA100 for now ...
1.159	22-Jul-2002	bouyer	Add VIA VT8233A support (at Ultra/100 until someone can test 133 for me). From Ryo HAYASAKA, PR kern/17206.
1.158	10-Jun-2002	cjs	Recognise the VT8233A ATA100 controller and use Ultra-DMA mode 5 with it.
1.157	09-Jun-2002	taca	Add support for VIA Technologies's VT8231 IDE Controller to ATA/100.
1.156	08-Jun-2002	bouyer	Ops, use the proper setup_channel() function for promise adapters. Using pdc202xx_setup_channel() for PDC20268 and newer is wrong, and will cause trap trying to read from a non-existent register on some arches (e.g. macppc). pointed out by Makoto Fujiwara on port-macppc.
1.155	04-Jun-2002	fvdl	AMD 8111 DMA support.
1.154	01-Jun-2002	bouyer	Remove the "set south-bridge's enable bit" code. It doens't seems to make any difference for rev 0xc3 controllers, and hang on 0xc4 controllers. Fixes kern/16954
1.153	19-May-2002	bouyer	branches: 1.153.2; Add support for HPT374. Submited by Matthew Rezny in kern/16897, with minor adjustements by me.
1.152	24-Apr-2002	aymeric	branches: 1.152.2; OPTi changes: o keep WDC_CAPABILITY_DATA32 regardless of the version of the chip. o correct typo to disable DMA correctly
1.151	23-Apr-2002	bouyer	More copyright fixes, pointed out by Thomas. Thanks !
1.150	14-Apr-2002	bouyer	bus_space_unmap() ctl_baseioh instead of ctl_ioh for the native case. Pointed out by enami, and should fix kern/16307.
1.149	12-Apr-2002	mycroft	Add support for the ServerWorks OSB4 and CSB5 chipsets. Note: The code is written a little more cruftily than it should be. It's also only tested on the OSB4. I'm not sure it even makes sense to have support for `native' mode, but I put it in just in case.
1.148	05-Apr-2002	thorpej	Reenable DMA on the Symphony/Winbond chips. The lossage I saw was due to a broken pmap_extract() on ARM.
1.147	04-Apr-2002	thorpej	Hm, Symphony/Winbond support doens't seem to work quite right yet; disable it for now.
1.146	03-Apr-2002	thorpej	Add bus master DMA support for the Symphony Labs 82C105 PCI IDE controller. This part is also found in the Winbond 83C553 Southbrige.
1.145	24-Mar-2002	bouyer	Add support for the AMD768. Tested by Murray Armfield <murray@river-styx.org>.
1.144	29-Jan-2002	bouyer	Support Ultra/100 on Intel 830M. From ks@ub.uni-mainz.de on -current-users.
1.143	28-Jan-2002	bouyer	In hpt_pci_intr(): There is interrupt pending once IDEDMA_CTL_ACT is turned off. It seems that IDEDMA_CTL_INTR is asserted before DMA transfer is complete, leading to race condition in case of interrupt sharing. Discovered reading the FreeBSD code.
1.142	14-Jan-2002	augustss	Add 82201CA to the chip table.
1.141	08-Jan-2002	bouyer	Fix typo in comment. Reported by Alexander Yurchenko in private mail.
1.140	18-Dec-2001	bouyer	Ops, a Ultra100>=TX2 is always in native mode !
1.139	18-Dec-2001	bouyer	The CMD0649 has 2 truly independant channel. tested by and patch from Love <lha@stacken.kth.se>. I suspect this is true for the 648 too; if someone with a 648 and one device on each channel could test that all works with one_channel = 1, it would be great !
1.138	16-Dec-2001	bouyer	dd support for the Promise Ultra100TX2, Promise Ultra100TX2v2 and Ultra133 IDE controllers. In part from OpenBSD via ragge, in part from FreeBSD.
1.137	13-Dec-2001	bouyer	pdc202xx_setup_channel: remove a duplicate line.
1.136	21-Nov-2001	wiz	Explicitly write explicitly without a second e.
1.135	15-Nov-2001	bouyer	Support Ultra/100 on CMD 0649.
1.134	13-Nov-2001	lukem	add RCSID
1.133	26-Oct-2001	augustss	Add Via KT266 as an ATA100 controller.
1.132	21-Oct-2001	thorpej	branches: 1.132.2; Add a config flag (0x02) that can force DMA to be disabled on chipsets for which it would otherwise be enabled by default.
1.131	16-Oct-2001	tron	Add UDMA/100 support for SiS chipsets 645, 650 and 730.
1.130	14-Oct-2001	tron	Use UDMA/100 on SiS 735 chipset.
1.129	24-Sep-2001	bouyer	Ali M5229 IDE controllers: tweak the rigth 0x79 register (the one from the pci/isa bridge, not the one from the IDE controller).
1.128	23-Aug-2001	tsutsui	branches: 1.128.2; Check 80 pins cable in acard_setup_channel() for Acard ATP860.
1.127	03-Aug-2001	tsutsui	Fix some typos in newer ALi M5229 support. Approved by bouyer.
1.126	01-Aug-2001	tsutsui	Remove unused pciide_print() declaration. From OpenBSD.
1.125	28-Jul-2001	tsutsui	Enable the Acard UIDE controllers by default. My WDC MDMA-only (non-UDMA) drives did not work on the Acard controllers, but it turns out that the problem was not Acard specific. These WDC drives do not work on the ESS ISAPnP wdc port nor on macppc obio wdc port neither, and another Quantum MDMA-only drive works fine on the Acard. These WDC drives work fine on my i386 pciide (which is initialized by the BIOS), so maybe we have to do something in MI wdc to initialize such drives properly...
1.124	26-Jul-2001	bouyer	Better support for newer ALI M5229 chipsets: support Ultra/66 for rev >= 0xC2, Ultra/100 for revs >= 0xC4. The the generic PCIIDE interupt routine for chipsets rev >= 0xC2 in native mode, it seems that newer chipsets don't have the ACER_CHIDS register :( From Linux and FreeBSD.
1.123	23-Jul-2001	bouyer	Handle HPT-370A controllers, from skanto@sjk-software.fi (kern/13540) but with some differences to the original patch: don't assume all controllers with rev >= HPT370_REV are HPT370, and explicitely print if we have a chip with a rev the driver does't know.
1.122	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.121	04-Jul-2001	bouyer	branches: 1.121.2; Better fix for rev 1.94 (disable UDMA for SiS controllers with rev 0xd0): Different chipsets have the same vendor/device/rev ID for the IDE controller, but only one of them is buggy. So check dev/rev ID of the function 0 (pchb on the buggy one) of the same device to detect the buggy controller.
1.120	13-Jun-2001	scw	Workaround for buggy OPTi pciide chipset revisions, as reported in PR/11644.
1.119	08-Jun-2001	simonb	For ports that wire up pciide in compatibility mode, have them define __HAVE_PCIIDE_MACHDEP_COMPAT_INTR_ESTABLISH in pci_machdep.h and pciide_map_compat_intr() only calls pciide_machdep_compat_intr_establish() if that preprocessor define exists. Ports that don't need to do this no longer need to supply a dummy function.
1.118	30-May-2001	bouyer	Uninitialised variable (from Krister Walfridsson): in acard_chip_map() drop compatchan and use the real channel number instead.
1.117	14-May-2001	matt	Add latent vendor entries for ServerWorks and Winbond IDE controllers. No code to use them as yet.
1.116	06-May-2001	fvdl	Add amd766 support. Rename some variables to 7x6 now that it supports both 756 and 766.
1.115	06-May-2001	fvdl	Add missing break statement.
1.114	04-May-2001	bouyer	Now that we can do it, print "HPT366" or "HPT370" depending on PCI revision.
1.113	04-May-2001	bouyer	Better support for VIA chipsets: look at the product/rev ID of the ISA bridge to guess the pciide capabilities, rather than trying to guess it by ourselve. Add preliminary support for the 686b (Ultra/100) guessed from FreeBSD/linux driver (datasheet not publically available, I contacted via). Let chip-specific map routine do the autoconf printf if ide_name is NULL (they may have more details about the controller than we have in pciide_attach)
1.112	21-Apr-2001	tsutsui	Preliminary support for Acard ATP850/860 UDMA PCI IDE controllers. XXX Currently disabled by default because it has some problems on macppc. XXX Maybe some more initialization is needed, but there is few information XXX about the chips.
1.111	21-Apr-2001	tsutsui	space -> tab
1.110	20-Mar-2001	bouyer	Don't set up PDC2xx_SCR the same way for promise Ultra/66 & Ultra/100, this special setup seems to break some adapters. Reported by Kazushi Marukawa.
1.109	20-Mar-2001	bouyer	Add BUS_DMA_STREAMING for the data bus_dmamap_load().
1.108	12-Mar-2001	bouyer	Support the Promise Ultra/100 at full speed, and fixes the "bogus intr" problem some peoples reported for Promise Ultra/100. Thanks to Chuck Silvers for giving me a Promise Ultra/100 so I could do some tests.
1.107	18-Feb-2001	bouyer	branches: 1.107.2; ALi controller: can only do UltraDMA for revs >= 0x20. From OpenBSD.
1.106	18-Feb-2001	bouyer	Add support for INTEL 82801BAM IDE controller, from OpenBSD.
1.105	12-Jan-2001	bouyer	Fix bug which affects pciide controllers in native mode (found by Paul Kranenburg, many thanks !): the control register I/O is 4 byte long although only one is used, but the control register is at offset 2, and not 0 as expected by IC code. Use bus_space_subregion() to get a handle which points to the control register, and is one byte long.
1.104	05-Jan-2001	bouyer	Run at Ultra/66 on VIA controllers that can do it. PCI vendor/device/revision is the same for the Ultra/33 and Ultra/66 version, so test writability of the U66 enable bit (idea from Chris Cappuccio). Thanks to David Carrel for testings.
1.103	05-Jan-2001	bouyer	Ops, make it compile again.
1.102	05-Jan-2001	bouyer	Add support for Ultra/100 on intel ICH2; from Tomokazu HARADA in kern/11747.
1.101	05-Jan-2001	bouyer	Add support for Ultra/100 on the HPT370 XXX the datasheet for the HPT370 is wrong While I'm there clear the DRIVE_DMA flag when we're going to use Ultra/DMA where it was missed, so that wdc_print_modes() only prints what's used.
1.100	29-Dec-2000	tsutsui	s/AP0_UDMA_MASK/APO_UDMA_MASK/
1.99	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.98	21-Dec-2000	mycroft	Disable the 16-bit I/O space hack for ALL PROMISE CONTROLLERS. This code is TOTALLY BROKEN and causes a NULL POINTER DEREFERENCE whenever it is fired.
1.97	17-Dec-2000	pk	Cast a `bus_addr_t' argument to `u_long' for `%lx' format (XXX..)
1.96	04-Dec-2000	fvdl	Avoid format warnings.
1.95	30-Nov-2000	thorpej	Don't restrict I/O space to 16-bits on the Promise Ultra/33 boards. Empirical evidence suggests that the board works just fine with an I/O address >= 0x10000.
1.94	26-Nov-2000	christos	Don't try to use ultra-dma on sis chips with revision 0xd0. It loses immediately with a lost interrupt on udma mode 2, downgrades to mode 1 at which point it silently corrupts data on high disk activity. This happens on two out of 3 machines I own that exhibit high disk activity.
1.93	19-Nov-2000	bouyer	Add support for Intel ICH2
1.92	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.91	08-Nov-2000	matt	Make the test for dmareg >= 0x10000 quirked on IDE_16BIT_IOSPACE
1.90	08-Nov-2000	wrstuden	Move guts of pciide_print_modes() to wdc_print_modes() so that non-pciide wdc drivers (like macppc's obio IDE interface) can use it. Also add support to both wd attach line and to wdc_print_modes() to print Ultra/{33,66,100} for respective UDMA modes (From Manuel Bouyer).
1.89	05-Nov-2000	matt	Don't allow use of mapreg is they are at or above 0x10000 in I/O space. (in other words, if the address uses bits in the top 16 bits, don't use it).
1.88	04-Oct-2000	bouyer	only VT82C586A rev >= 6 supports Ultra-DMA. From OpenBSD.
1.87	21-Aug-2000	enami	Recognize Promise ATA-100 controller as PDC-20262 instead of 20246.
1.86	21-Aug-2000	enami	Match the Promise ATA-100 controller found on the mother board MS-6321 (MSI 694D Pro). Reported in PR#10756 by Kazuki Sakamoto.
1.85	09-Aug-2000	drochner	recognize the i440MX mobile chipset's IDE interface
1.84	02-Aug-2000	bouyer	PCIIDE_CMD0646U_UDMA->PCIIDE_CMD0646U_ENABLEUDMA for consistency with PCIIDE_AMD756_ENABLEDMA defopt PCIIDE_CMD0646U_ENABLEUDMA, PCIIDE_AMD756_ENABLEDMA, PCIIDE_CMD064x_DISABLE Fix a typo pointed out by John Hawkinson
1.83	02-Aug-2000	bouyer	Add support for the CMD PCI646U. Linux claims that this driver is brocken with UDMA, so enable Ultra-DMA only if "options PCIIDE_CMD0646U_UDMA" is set.
1.82	01-Aug-2000	bouyer	Add support for the CMD PCI0646U2, an Ultra/33 version of the 0646. Note: there's also a PCI0646U, for which I don't have docs for now.
1.81	27-Jul-2000	bouyer	HPT370: clear disable interrupt bit; make it works in Ultra/66 mode.
1.80	20-Jul-2000	bouyer	Make it compile when PCIIDE_AMD756_ENABLEDMA is defined. From kern/10555 by MURATA Shuuichirou.
1.79	07-Jul-2000	bouyer	Use the CMD PCI0648/9 IRQ ACK code for the 0646 too, makes the 0646 works in native mode.
1.78	06-Jul-2000	bouyer	Work around a bug in AMD756 rev D2, from patches provided by David Sainty: disable multiword DMA for these chips. multiword DMA can be forced with options PCIIDE_AMD756_ENABLEDMA on rev D2 chips, but use at your own risk ! While I'm there remove a duplicate allocation of sc_wdcdev.nchannels in HPT code.
1.77	05-Jul-2000	bouyer	Back out previous, it has to be done in a different way.
1.76	05-Jul-2000	bouyer	Apply patch from David Sainty <David.Sainty@optimation.co.nz>: Some AMD controllers have a bug which can look up the machine when using DMA, so disable DMA for some revisions (info provided by AMD). "options PCIIDE_AMD756_ENABLEDMA" can be used to force DMA on these chips.
1.75	05-Jul-2000	bouyer	HPT: use pciide_channels[i] not pciide_channels[0]. My HPT370 now probes both channels (but still doesn't work properly; I suspect I got a broken one).
1.74	04-Jul-2000	enami	Match with promise ultra100/ata contoller. I haven't actually test this contoller with ultra100 drive, but it works at least with ultra66 or more older drive
1.73	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.72	27-Jun-2000	tron	Add special IRQ handler for CMD PCI0648 and PCI0649 taken from the experimental version of Manuel Bouyer's driver.
1.71	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.70	26-Jun-2000	bouyer	Add support for the CMD PCI0648 and PCI0649 IDE controllers. Thanks to Matthias Scheler for testing.
1.69	26-Jun-2000	bouyer	Correct 80 pin handling for promise Ultra/66: when the bit is set we don't have a Ultra/66 cable.
1.68	12-Jun-2000	bouyer	branches: 1.68.2; Shorter description for the HPT366
1.67	12-Jun-2000	bouyer	- add a pciide_irqack() callback, which clears the IDE DMA status bit once the IRQ has been cleared on the drive. - use pa->pa_class instead of re-reading PCI_CLASS_REG when possible - Add support for Highpoint HPT366 and HPT370 (370 untested), based on patches from Roger Brooks <R.S.Brooks@liverpool.ac.uk> posted on current-users Mach, 15. Given how Highpoint docs have been wrong for the 366, the 370 is likely to not work. Thanks to Chris Cappuccio <chris@dqc.org> for sending me the Highpoint docs, and to Total Archive (http://www.totalarchive.com/) for sending me hardware.
1.66	07-Jun-2000	scw	The OPTi controller supports a 32-bit dataport after all. Also detect when the chip is sitting on a 25MHz PCIbus and set the timing registers accordingly.
1.65	07-Jun-2000	thorpej	Add missing break;
1.64	06-Jun-2000	thorpej	Improve the Cypress name a little.
1.63	06-Jun-2000	thorpej	In pciide_mapreg_dma(), check to see what type the BAR is before mapping the registers, as suggested by a comment in that function.
1.62	06-Jun-2000	soren	Shorten names of VIA controllers to fit in 80 columns with versions.
1.61	06-Jun-2000	thorpej	Actually program the DMA mode of the drives into the Cypress controller. Fixes a long-standing problem where IDE DMA wasn't working on the AlphaPC 164SX.
1.60	04-Jun-2000	gmcgarry	pciiide -> pciide
1.59	27-May-2000	scw	Add support for the OPTi 82c621 PCIIDE controller and its derivatives. I only have a Compaq laptop on which to test this, so reports of success/failure in other systems would be welcomed.
1.58	15-May-2000	bouyer	branches: 1.58.2; Sync my copyrigth notice.
1.57	12-May-2000	thorpej	Print the revision info from the PCI configuration header. From Dave Sainty, kern/10025.
1.56	01-Apr-2000	bouyer	- DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op if an IRQ was not detected, unless the force flag was given. Use this to detect if the IRQ was for us (closer to shared IRQ for controllers which don't have their own IRQ handler in pciide.c) and to poll for DMA xfer. Also makes the timeout recovery code simpler. - ATAPI cleanup: don't call controller-specific functions from atapiconf.c (wdc_), so that it's possible to attach an atapibus to something else than a wdc/pciide (Hi Lennart :). Overload struct scsi_adapter with struct atapi_adapter, defined as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points to an scsi_adapter, atapi code casts it to atapi_adapter if needed. Move atapi_softc to atapiconf.h so that it can be used by the underlying controller code (e.g. atapi_wdc.c). Add an atapi-specific callback atapi_probedev(), which probe a drive in a controller-specific way, allocate the sc_link and fills in the ataparams if needed. It then calls atapi_probedev() (from atapiconf.c) to do the generic initialisations and attach the device. - While I'm there merge and centralise the state definitions in atavar.h. It should now be possible to use a common ata/atapi routine to set the drive's modes (will do later).
1.55	10-Mar-2000	bouyer	Don't reset cp->hw_ok ro 0 when cp isn't initialised in cy693_chip_map() (used only in failure case). Pointed out by Wolfgang Solfrank. While I'm here correct indentation.
1.54	09-Mar-2000	soren	Move PCIIDE_CHANNEL_NAME macro to pciidereg.h.
1.53	06-Mar-2000	bouyer	Add support for the AMD 756 DMA/UDMA IDE controller, provided in PR kern/9536 by Dave Sainty.
1.52	18-Jan-2000	bouyer	Clean up revision stuff for the sis. Suggested by Chris Cappuccio.
1.51	16-Jan-2000	bouyer	From chris@openbsd.org: "Don't enable UDMA modes for revisions of SiS 5513 < 0xd0 The only revisions I know which don't actually support UDMA are 0x09 and below.. But the only revision I know which does support UDMA is 0xd0 (and presumably above that)"
1.50	26-Dec-1999	soren	Lower-case Bus-Master for consistency.
1.49	12-Dec-1999	thorpej	Use htole32() and le32toh().
1.48	28-Nov-1999	bouyer	Improve Ultra/66 support now that I've got some docs from Promise.
1.47	13-Nov-1999	soren	Export softc.
1.46	03-Nov-1999	mycroft	Fix silly error that caused the secondary channel to be ignored if the primary channel was disabled.
1.45	25-Oct-1999	bouyer	Add a missing '\n' in the cmd0640 attach printfs.
1.44	02-Sep-1999	ross	branches: 1.44.2; 1.44.4; 1.44.6; Make it compile.
1.43	01-Sep-1999	bouyer	Don't try to unmap unmapped space in case of failure in pciide_mapregs_compat(). From OpenBSD.
1.42	30-Aug-1999	bouyer	Add support for Intel 810 chipset (ICH/ICH0). While I'm there merge back piix_channel_map into piix_chip_map.
1.41	29-Aug-1999	bouyer	Add support for the Promise Ultra/33 and /66 pci IDE controller. In addition to chip-dependant code this required the following changes: - Instead of attaching the device in a generic way with some chip-dependant routines, use a chip-dependant attach routine with some common code factored out. The code is marginally bigger, but this allows the CMD64x flag hack to go away. - For chips that report per-channel 'irq triggered', test this before calling wdcintr() for the native-pci irq case (compat intr can't be shared), as wdcintr() has no good way to know if a irq was for it or not, and ends up with irq loss. XXX for chips that don't have this feature irq sharing will not work properly ! - add my copyrigth notice (could have been done some time ago I think :) There are still some issues to be solved with the Promise controller and ATAPI devices. Many thanks to Paul Newhouse for shipping me 2 Ultra/33 boards for doing this work.
1.40	12-Jul-1999	bouyer	Fix typo in a printf, from Soren S. Jorvan.
1.39	08-Jun-1999	mrg	fix a few KNF nits ..
1.38	27-May-1999	bouyer	For the PIIX, make sure the PIO_mode and DMA_mode get reset to the values used by the controller for all drives.
1.37	05-May-1999	bouyer	Fix the way we compute the mode to use: for multiword DMA, the used mode was 2 less than the one we could really use, so for multiword DMA mode 0 or 1, the driver tried to use DMA mode 255 or 254 (0 - 2 with a u_int8_t).
1.36	03-May-1999	ross	Protect WDCDEBUG from multiple definitions.
1.35	28-Apr-1999	thorpej	Make PCI IDE DMA work on big-endian systems.
1.34	06-Apr-1999	bouyer	Kill an extra 'pciide0: ' in a printf
1.33	22-Feb-1999	bouyer	branches: 1.33.2; In cy693_setup_channel(), setup timings for IOR too (they were left to 0, which is a way too higth timing for some devices). Thanks to Ken Wellsch for trying the multiple debug kernels until the problem was located.
1.32	16-Feb-1999	bouyer	Correctly compute PIO/DMA mode for sis and acer chips when the drive support a DMA mode with higther capabilities than PIO mode.
1.31	02-Feb-1999	bouyer	channel_map is called before setup_chip, so whe need to enable the channel status bits in acer_channel_map().
1.30	02-Feb-1999	bouyer	Support for Acerlab M5229 IDE controller. Thanks to Thilo Manske for testing the code, and to Takahiro Kambe who run several tests and finally found the bug by himself :)
1.29	16-Dec-1998	bouyer	Defer mapping of pci interrupt to pciide_mapregs_native(). This way, the native interrupt shouldn't be mapped if a channel is in native mode, but disabled.
1.28	16-Dec-1998	bouyer	Rearange the modes setup to allow these to be dyanmically changed. Fill in the new "set_mode" callback.
1.27	03-Dec-1998	bouyer	Rename pio_mode, etc ... to PIO_cap, etc ... for consistency with the ata_drive_datas struct. Suggested by Soren S. Jorvan.
1.26	03-Dec-1998	bouyer	Now that vtophy() is no longuer used, re-enable WDCDEBUG, with wdcdebug_pciide_mask = 0 (so that one can easily patch this variable and give me more informations :)
1.25	03-Dec-1998	bouyer	add a udma_mode field to wdc_softc, and use it the same way dma_mode is used (higthest ultra-dma mode supported). There may be a higther ultra-dma mode defined ...
1.24	03-Dec-1998	bouyer	Use correct register when disabling the second channel.
1.23	03-Dec-1998	bouyer	Restore changes from revision 1.17: "If a channel has no drives, do not unmap its I/O regions. It's not really safe to use them for anything else, and in legacy mode it will just cause us to probe the channel again as an ISA device."
1.22	03-Dec-1998	bouyer	Correct a few bogons in the SiS chip initialisation.
1.21	03-Dec-1998	bouyer	Ouh ! Correct the 8-bit PCI registers reading/writing functions: need to multiply the register offset by 8.
1.20	02-Dec-1998	bouyer	- change drive_flags from u_int8_t to u_int16_t - keep the modes supported by the drive in struct ata_drive_datas (will be later used for downgrading the DMA/PIO mode on error) - use config flags to force/disable PIO/DMA/UDMA modes - For the CMD PCI0643/6 setup DMA mode to DMA Read multiple.
1.19	24-Nov-1998	drochner	a small optimization for the compat interrupt handling, possible after channel_softc is within pciide_channel
1.18	21-Nov-1998	drochner	some restructuring, more or less to get support for weird IDE controllers, eg the Cypress ISA bridge: -put channel mapping into the chip specific part, unify with channel_probe() into channel_map() -use pointer to channel data as function call argument wherever possible instead of the channel number -allow the "compat" channel number to differ from the per-controller channel number - for mapping and interrupt functions -add support for SiS5598 and Cypress 82C693 chips Mostly done by Manuel, I only contributed to the first 2 items.
1.17	17-Nov-1998	mycroft	If a channel has no drives, do not unmap its I/O regions. It's not really safe to use them for anything else, and in legacy mode it will just cause us to probe the channel again as an ISA device.
1.16	12-Nov-1998	bouyer	Force PCI_COMMAND_MASTER_ENABLE if DMA has been setup properly. The BIOS is supposed to do it but who knows ...
1.15	11-Nov-1998	bouyer	- clearify the boot messages (features supported vs features used). Thanks to Havard Eidnes for his complains about this :) - fix some typo in comments - hoppefully better detection of drives reporting bogus PIO modes.
1.14	09-Nov-1998	bouyer	Support for the CMD PCI064{3,6}. Tested on a 0646 with a "wd0: PIO mode 4, DMA mode 2, UDMA mode 2" device.
1.13	22-Oct-1998	bouyer	Add config flags for pciide: 0x001 forces the use of DMA when the driver don't know how to set the controller's modes.
1.12	20-Oct-1998	bouyer	Fix for Apollo DMA mode (not UDMA): DMA mode = PIO mode - 2, not PIO mode + 2 !
1.11	19-Oct-1998	bouyer	Add support for the second flavour of the VIA IDE chip (which has UDMA). Don't claim DMA support if we don't have explicit support for this chip. They're to may way to loose when trying to use DMA without configuring the controller and disks.
1.10	13-Oct-1998	bouyer	pciide.c: don't define WDCDEBUG, so it compiles on alpha. Correct a bogon in the printing of DMA mode (piix3/4 only) others: set the debug_mask to 0, so that debug messages are turned off by default but can be easily turned on. Reset drive_flags to 0 for unconfigured devices, so that they are ignored later. For configured devices, reset state to 0 after probe/attach.
1.9	12-Oct-1998	bouyer	Merge bouyer-ide
1.8	14-Aug-1998	drochner	2 changes to pciide_probe_wdc() (used in compatibility mode to check if i drive is responding): -if the reset succeeds, check some registers to make sure there is really a drive (and not a chipset which echoes back the last written value) -explicitely select the master before trying to read the master's status
1.7	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.6	12-Mar-1998	cgd	branches: 1.6.2; when considering attaching compatibility-mode channels, try a quick reset and see if anything responds. if nothing (that's attributable to the PCI IDE controller) responds, then that channel either has no devices on it or has been disabled (via a non-standard mechanism) by the BIOS. If nothing responds, don't map the compat.-mode interrupt or attach the wdc to that channel, because the BIOS is likely to assign that IRQ to a different PCI device. If that happens, the kernel will panic because that device will try to map the IRQ level-triggered, but the compat interrupt will have been mapped edge-triggered. (One possible way around this is to map the compat interrupt edge-triggered, but it's not clear reading the spec that this is correct or desirable.)
1.5	06-Mar-1998	cgd	reorganize mapping/attachment of wdc channels so that it'll be easier to insert a check to see whether a channel appears to be enabled. Shouldn't be necessary, according to the spec, but some PC chipsets allow individual compatibility channels to be disabled. "I hate PCs."
1.4	06-Mar-1998	cgd	despite the spec, some people map the bus master IDE registers into memory space. Note that in a comment, but don't try to fix it (for now).
1.3	04-Mar-1998	cgd	clean up a printf
1.2	04-Mar-1998	cgd	slight cleanup (consistency, make a few comments better). add support for recognizing bus-master DMA interface and mapping the regs (but still no support for DMA).
1.1	04-Mar-1998	cgd	PCI IDE glue. Right now, just glues 'wdc's to PCI IDE controller channels. Eventually should do things like support PCI IDE DMA (it _DOES NOT_ do that now).
1.6.2.17	05-Oct-1998	bouyer	Rename WDC_CAPABILITY_PIO to WDC_CAPABILITY_MODE, as it's used to tell wether the controller's driver can tell which timing mode it uses. Check this before setting DMA modes too. This allow the generic DMA code to work again.
1.6.2.16	04-Oct-1998	bouyer	atavar.h: drv_softc is a struct device * instead of void*, as it's mostly used for dv_xname wd.c: convert for drv_softc type change, printf cleanup wdc.c: always call ata_get_params() (params was used initialised with non-32bit controllers, leaving to bogus PIO/DMA mode report). Cleaup of the PIO/DMA mode message. pciide_piix_reg.h: Fix definition of PIIX_IDETIM_CLEAR (unused before) pciide.c: add a method do disable a channel on know device. If a channel doesn't have any drive, we disable it and free its resources if disable was successfull. This should help with laptops where the second channel of the PIIX4 is unused but not disabled by BIOS. On such laptops, irq15 can be used for PCMCIA but it was claimed by pciide. Misc printf cleanup. wdc_isa.c: printf cleanup.
1.6.2.15	02-Oct-1998	bouyer	Cleanup a few error message, remove some dead code. Re-add a reset at end of attach, unless we have WDC_NO_EXTRA_RESETS. Add support for data32iot/data32ioh from -current: each controller pass WDC_CAPABILITY_DATA32 if they can do 32bit, and WDC_CAPABILITY_DATA16 if they can do 16bit. For controller that support both, the usual autodetect mechanism is still used.
1.6.2.14	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.6.2.13	21-Aug-1998	bouyer	Sync with -current
1.6.2.12	13-Aug-1998	bouyer	- sync with -current - full support of CMD PCI0640 chips. This needs a special probe routine, as this chip doesn't honnor the "PCI_COMMAND_IO_ENABLE" bit. sigh. - do not map the native pci interrupts if we don't have a native channel. some chips (e.g. PCI0640) have a valid interrupt value which gets established, which may prevent the compat channel or some other ISA device from working properly later. - b* -> mem*
1.6.2.11	25-Jun-1998	bouyer	Add support for VIA's apollo VP chipset.
1.6.2.10	17-Jun-1998	bouyer	in piix3_4_setup(), drvp is not a pointer to an array, but a pointer to the rigth element of the array. This was breaking more-or-less badly on configs with a slave drive on a channel.
1.6.2.9	13-Jun-1998	bouyer	Coorect a comment.
1.6.2.8	12-Jun-1998	bouyer	Correct a bug where the PIO mode would not be set up for the controller, if UDMA mode is used. Ultra-DMA/33 works !
1.6.2.7	11-Jun-1998	bouyer	pciide: when setting up DMA mode, also enable fast timings for PIO modes if possible. For ATAPI devices, commands are always send via PIO. atapi: matain a drive state (like wd drives) and use it to set up the timings if needed.
1.6.2.6	10-Jun-1998	bouyer	- Fix timing settings for DMA: the controller was always set up to use DMA mode 0 (compatible). - The 2 Ultra-dma registers are in fact one 32-bit register. Change the macros and setup in a way it may now work (but still untested, thus not enabled by default). - For DMA mode 1, use a more efficient timing than the one suggested by intel. Some work is still needed here to get ATAPI DMA working (should be done soon).
1.6.2.5	09-Jun-1998	bouyer	- sync with trunck - for the piix, don't set timings for nonexistent drives.
1.6.2.4	06-Jun-1998	bouyer	Renamed pciide_pIIx_reg.h to pciide_piix_reg.h via repository copy, as suggested by Soren S. Jorvan.
1.6.2.3	05-Jun-1998	bouyer	Disable Ultra DMA for now, as it does not work.
1.6.2.2	05-Jun-1998	bouyer	A few bux fixes (untested): - If the PIO mode of the controller is unknown, put the drive in the hightest mode it supports. A drive in a faster mode than the controller should work. - Create DMA maps even for unknown controllers in pciide.
1.6.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.33.2.11	14-Aug-2000	he	Pull up revision 1.84 (via patch, requested by bouyer): Change PCIIDE_CMD0646U_UDMA to PCIIDE_CMD0646U_ENABLEUDMA for consistency with PCIIDE_AMD756_ENABLEDMA.
1.33.2.10	14-Aug-2000	he	Pull up revisions 1.82-1.83 (requested by bouyer): Add UltraDMA support for the CMD PCI0646U and PCI0646U2 controllers; normally disabled on PCI0646U due to a chip bug, but can be enabled with a kernel option.
1.33.2.9	27-Jul-2000	he	Pull up revision 1.81 (requested by bouyer): Clear the interrupt disable bit of the HPT370, and document HPT370 support.
1.33.2.8	27-Jul-2000	he	Pull up revision 1.80 (requested by bouyer): Make pciide compile with ``options PCIIDE_AMD756_ENABLEDMA.'' Fixes PR#10555.
1.33.2.7	23-Jul-2000	he	Pull up revision 1.79 (requested by bouyer): Handle CMD PCI0646 interrupts as for the PCI0648; makes the PCI0646 work in native mode with shared interrupts.
1.33.2.6	23-Jul-2000	he	Pull up revision 1.74 (requested by bouyer): Add support for the Promise Ultra/100 PCI IDE controller (in Ultra/66 mode only).
1.33.2.5	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.33.2.4	24-Jun-1999	perry	pullup 1.37->1.38 (bouyer)
1.33.2.3	05-May-1999	perry	branches: 1.33.2.3.2; 1.33.2.3.4; pullup 1.36->1.37 (bouyer)
1.33.2.2	28-Apr-1999	perry	sync to 1.35 (thorpej): fixes PCIIDE DMA on big-endian systems
1.33.2.1	06-Apr-1999	bouyer	Kill an extra 'pciide0:' in a printf.
1.33.2.3.4.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.33.2.3.2.2	02-Aug-1999	thorpej	Update from trunk.
1.33.2.3.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.44.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.44.4.1	15-Nov-1999	fvdl	Sync with -current
1.44.2.9	23-Apr-2001	bouyer	Sync with HEAD.
1.44.2.8	27-Mar-2001	bouyer	Sync with HEAD.
1.44.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.44.2.6	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.44.2.5	15-Jan-2001	bouyer	req_sense_length is dead.
1.44.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.44.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.44.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.44.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.58.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.68.2.33	22-Apr-2002	he	Pull up revision 1.150 (requested by bouyer): Fix use of bus_space_unmap() for the native case. Fixes PR#16307.
1.68.2.32	25-Mar-2002	he	Pull up revision 1.145 (requested by bouyer): Add support for the pciide part of the AMD-768 chipset.
1.68.2.31	09-Feb-2002	he	Pull up revision 1.143 (requested by bouyer): In hpt_pci_intr(), accept interrupt only once IDEDMA_CTL_ACT is turned off. Fixes DMA errors, lost interrupts, and other nasty errors with some HPT730s in native mode.
1.68.2.30	06-Feb-2002	he	Pull up revisions 1.112,1.118,1.125,1.128 (via patch, requested by tsutsui): Add support for Acard ATP-850/860 pciide controllers.
1.68.2.29	05-Jan-2002	he	Pull up revision 1.135 (requested by bouyer): Support Ultra-DMA 100 on the CMD 0649.
1.68.2.28	13-Nov-2001	he	Pull up revision 1.123 (requested by bouyer): Support HPT-370A controllers. Fixes PR#13540.
1.68.2.27	13-Nov-2001	he	Pull up revision 1.121 (requested by bouyer): Disable UDMA only for SiS controllers which are really buggy, not all the ones with revision 0xd0.
1.68.2.26	29-Jul-2001	he	Pull up revision 1.120 (requested by scw): Workaround for buggy OPTi pciide chipset revisions, as reported in PR#11644.
1.68.2.25	02-Jul-2001	jhawk	Pull up revision 1.94 (requested by bouyer): Don't try to use ultra-dma on sis chips with revision 0xd0. It loses immediately with a lost interrupt on udma mode 2, downgrades to mode 1 at which point it silently corrupts data on high disk activity.
1.68.2.24	15-May-2001	he	Pull up revisions 1.113-1.116 (via patch, requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.68.2.23	24-Apr-2001	he	Pull up revision 1.110 (requested by bouyer): Revert setting of PDC2xx_SCR for Promise Ultra/100, it seems to break with some adapters.
1.68.2.22	13-Mar-2001	he	Pull up revision 1.108 (requested by bouyer): Support Ultra/100 speed on Promise Ultra/100, and fix ``bogus intr'' messages generated under some conditions.
1.68.2.21	26-Feb-2001	he	Pull up revision 1.107 (requested by bouyer): Disable Ultra-DMA on Ali controllers which don't support it.
1.68.2.20	26-Feb-2001	he	Pull up revision 1.106 (requested by bouyer): Add support for the Intel 82801BAM IDE controller.
1.68.2.19	04-Feb-2001	he	Pull up revision 1.105 (requested by bouyer): Fix bug which affects pciide controllers in native mode. Main effect was a non-functionnal IDE controller on some sparc64 and macppc when booted from IDE disk.
1.68.2.18	04-Feb-2001	he	Pull up revisions 1.102-1.103 (requested by bouyer): Add support for Ultra/100 on Intel ICH2.
1.68.2.17	04-Feb-2001	he	Pull up revision 1.104 (requested by bouyer): Run at Ultra/66 on VIA controllers that can do it.
1.68.2.16	04-Feb-2001	he	Pull up revision 1.101 (requested by bouyer): Add support for Ultra/100 on the HPT370 controller.
1.68.2.15	15-Dec-2000	he	Pull up revision 1.93 (requested by bouyer): Add support for Intel ICH2 IDE controller.
1.68.2.14	04-Oct-2000	bouyer	Pull up 1.87->1.88, approved by thorpej: only VT82C586A rev >= 6 supports Ultra-DMA. From OpenBSD.
1.68.2.13	22-Aug-2000	enami	Pullup rev. 1.87 (approved by jhawk): Recognize Promise ATA-100 controller as PDC-20262 instead of 20246.
1.68.2.12	22-Aug-2000	enami	Pullup rev. 1.86 (approved by jhawk and thorpej): Match the Promise ATA-100 controller found on the mother board MS-6321 (MSI 694D Pro). Reported in PR#10756 by Kazuki Sakamoto.
1.68.2.11	03-Aug-2000	bouyer	Pull up (approved by thorpej): sys/dev/pci/pciide.c 1.82 -> 1.84 sys/dev/pci/pciide_cmd_reg.h 1.8 -> 1.9 sys/dev/pci/files.pci 1.100 -> 1.101 share/man/man4/pciide.4 1.23 -> 1.24 Add support for the CMD PCI646U. Linux claims that this driver is brocken with UDMA, so enable Ultra-DMA only if "options PCIIDE_CMD0646U_ENABLEUDMA" is set. defopt PCIIDE_CMD0646U_ENABLEUDMA, PCIIDE_AMD756_ENABLEDMA, PCIIDE_CMD064x_DISABLE Fix a typo pointed out by John Hawkinson Update documentation.
1.68.2.10	02-Aug-2000	bouyer	Pull up (approved by thorpej): pciide_cmd_reg.h 1.7 -> 1.8 pciide.c 1.81 -> 1.82 Add supports for the PCI0646U2, Ultra/33 version of the PCI0646.
1.68.2.9	27-Jul-2000	bouyer	Pull up (approved by thorpej): sys/dev/pci/pciide.c 1.80 -> 1.81 sys/dev/pci/pciide_hpt_reg.h 1.1 -> 1.2 distrib/notes/i386/hardware 1.74 -> 1.75 share/man/man4/pciide.4 1.22 -> 1.23 Make sure the HPT370's interrupts disable bit is cleared. Now that HPT370 supports works, document it.
1.68.2.8	20-Jul-2000	bouyer	Pull up 1.80 (approved by thorpej): Make is compile when PCIIDE_AMD756_ENABLEDMA is defined; close PR kern/10555.
1.68.2.7	19-Jul-2000	bouyer	Pull up 1.79, approved by thorpej: Use the CMD PCI0648/9 IRQ ACK code for the 0646 too, makes the 0646 works in native mode.
1.68.2.6	07-Jul-2000	bouyer	Pull up pciide.c 1.77 -> 1.78 pciide_amd_reg.h 1.1 -> 1.2 (approved by thorpej): workaround for a bug in some revs of the AMD IDE controller, which can lock up the machine when multiword DMA is used.
1.68.2.5	06-Jul-2000	enami	Pull up rev. 1.74 (approved by releng-1-5): Match with promise ultra100/ata contoller. I haven't actually test this contoller with ultra100 drive, but it works at least with ultra66 or more older drive
1.68.2.4	05-Jul-2000	bouyer	Pull up rev 1.75, approved by thorpej: Fix a error where we would always try to attach channel 0 of an HPT controller (only aftects HPT370, the 366 has only one channel per pciide function).
1.68.2.3	27-Jun-2000	tron	Really pull up revision 1.72.
1.68.2.2	27-Jun-2000	bouyer	Pull up pciide.c 1.69->1.70, 1.71->1.72 pciide_cmd_reg.h 1.6->1.7 Approved by thorpej: Add support for the CMD PCI0648/0649 IDE controller.
1.68.2.1	27-Jun-2000	bouyer	Pull up 1.69, approved by thorpej: Fix promise Ultra/66 cable detection: the Ultra/66 cable detection bit was interpreted backward.
1.107.2.20	29-Dec-2002	thorpej	Sync with HEAD.
1.107.2.19	11-Dec-2002	thorpej	Sync with HEAD.
1.107.2.18	18-Oct-2002	nathanw	Catch up to -current.
1.107.2.17	17-Sep-2002	nathanw	Catch up to -current.
1.107.2.16	27-Aug-2002	nathanw	Catch up to -current.
1.107.2.15	13-Aug-2002	nathanw	Catch up to -current.
1.107.2.14	01-Aug-2002	nathanw	Catch up to -current.
1.107.2.13	20-Jun-2002	nathanw	Catch up to -current.
1.107.2.12	17-Apr-2002	nathanw	Catch up to -current.
1.107.2.11	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.107.2.10	28-Feb-2002	nathanw	Catch up to -current.
1.107.2.9	11-Jan-2002	nathanw	More catchup.
1.107.2.8	08-Jan-2002	nathanw	Catch up to -current.
1.107.2.7	14-Nov-2001	nathanw	Catch up to -current.
1.107.2.6	22-Oct-2001	nathanw	Catch up to -current.
1.107.2.5	26-Sep-2001	nathanw	Catch up to -current. Again.
1.107.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.107.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.107.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.107.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.121.2.7	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.121.2.6	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.121.2.5	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.121.2.4	11-Feb-2002	jdolecek	Sync w/ -current.
1.121.2.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.121.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.121.2.1	03-Aug-2001	lukem	update to -current
1.128.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.132.2.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.152.2.3	29-Aug-2002	gehenna	catch up with -current.
1.152.2.2	20-Jun-2002	gehenna	catch up with -current.
1.152.2.1	30-May-2002	gehenna	Catch up with -current.
1.153.2.17	12-Jul-2004	he	Pull up revision 1.187 (via patch, requested by grant in ticket #1720): Add support for SATALink 3112.
1.153.2.16	28-Mar-2004	jmc	Pullup rev 1.196 via patch (requested by bouyer in ticket #1632) Add support for VIA VT8237 (KT600) IDE controller. PR#22453
1.153.2.15	28-Mar-2004	jmc	Pullup rev 1.195 (requested by bouyer in ticket #1630) Correct offset for the promise "magic registers" (it depends on channel number). This makes cable detect work properly for newer promise (pdc268 and newer).
1.153.2.14	16-Aug-2003	tron	Pull up revision 1.194 (requested by bouyer in ticket #1369): Correct setup of General Purpose Register for PDC20265 (Ultra/100) and newer. From Andreas Johansson in private mail.
1.153.2.13	16-Aug-2003	tron	Pull up revision 1.193 via patch (requested by bouyer in ticket #1367): Add support for ICH5, from Quentin Garnier in private mail.
1.153.2.12	16-Jun-2003	grant	Apply patch (requested by bouyer in ticket #1234): Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver. Ops, I commited the previous change (improved SiS support) from the wrong tree. Sync with the latest diff I made available, and also include changes from kern/18015 which is much better than the hack I've done.
1.153.2.11	28-Apr-2003	tron	Pull up revision 1.191 (requested by nakayama in ticket #1275): Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.153.2.10	28-Nov-2002	tron	Pull up revision 1.174 (requested by kent in ticket #1019): Add support for VIA VT8235
1.153.2.9	01-Nov-2002	tron	Pull up revision 1.168 (requested by bouyer in ticket #728): Support Ultra/133 on promise controllers that can do it.
1.153.2.8	01-Nov-2002	tron	Pull up revision 1.167 (requested by bouyer in ticket #727): Correct setup for Ultra133 capable VIA chipsets, From Matthias Drochner on current-users, with cross-check and some improvement from linux-2.4.19 and FreeBSD-current. Also don't set the APO_UDMA_CLK66 bit for Ultra/100 capable chipset, and support Ultra/133 for the VT8233A.
1.153.2.7	01-Nov-2002	tron	Pull up revision 1.166 (requested by bouyer in ticket #725): Add support for HPT372. From sdegler@kashmir.degler.net in kern/17908, with some improvement from me, and from FreeBSD.
1.153.2.6	01-Nov-2002	tron	Pull up revision 1.165 (requested by bouyer in ticket #724): Add support for Promise Ultra133TX2 and Ultra133TX2v2, from FreeBSD and stephen@degler.net in kern/17509.
1.153.2.5	02-Aug-2002	lukem	pull up revisions 1.158-1.160 (requested by bouyer in ticket #603): 1.160: Ops, this had already been commited by cjs. Still, call the controller ATA133, even though we use it at ATA100 for now ... 1.159: Add VIA VT8233A support (at Ultra/100 until someone can test 133 for me). From Ryo HAYASAKA, PR kern/17206. 1.158: Recognise the VT8233A ATA100 controller and use Ultra-DMA mode 5 with it.
1.153.2.4	01-Aug-2002	lukem	Pull up revision 1.163 (requested by bouyer in ticket #590): Add support for the 82801DB ICH4 IDE controller, from kern/17757 by Andreas Wrede <andreas@planix.com>
1.153.2.3	10-Jun-2002	tv	Pull up revision 1.157 (requested by taca in ticket #228): Add support for VIA Technologies's VT8231 IDE Controller to ATA/100.
1.153.2.2	10-Jun-2002	tv	Pull up revision 1.156 (requested by bouyer in ticket #214): Ops, use the proper setup_channel() function for promise adapters. Using pdc202xx_setup_channel() for PDC20268 and newer is wrong, and will cause trap trying to read from a non-existent register on some arches (e.g. macppc). pointed out by Makoto Fujiwara on port-macppc.
1.153.2.1	04-Jun-2002	lukem	pull up rev 1.154 (requested by bouyer in #129): Remove the "set south-bridge's enable bit" code. It doens't seems to make any difference for rev 0xc3 controllers, and hang on 0xc4 controllers. Fixes kern/16954
1.194.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.194.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.194.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.194.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.194.2.1	03-Aug-2004	skrll	Sync with HEAD
1.210.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.210.6.1	12-Feb-2005	yamt	sync with head.
1.210.4.1	29-Apr-2005	kent	sync with -current
1.212.4.2	24-Mar-2008	yamt	sync with head.
1.212.4.1	30-Dec-2006	yamt	sync with head.
1.213.22.2	10-Dec-2006	yamt	sync with head.
1.213.22.1	22-Oct-2006	yamt	sync with head
1.213.20.1	18-Nov-2006	ad	Sync with head.
1.215.48.1	03-Apr-2008	mjf	Sync with HEAD.
1.215.44.1	24-Mar-2008	keiichi	sync with head.
1.215.24.1	23-Mar-2008	matt	sync with HEAD
1.216.4.1	11-Mar-2010	yamt	sync with head
1.217.4.1	05-Mar-2011	rmind	sync with head
1.217.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.219.18.2	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.219.18.1	12-Sep-2012	tls	Initial snapshot of work to eliminate 64K MAXPHYS. Basically works for physio (I/O to raw devices); needs more doing to get it going with the filesystems, but it shouldn't damage data. All work's been done on amd64 so far. Not hard to add support to other ports. If others want to pitch in, one very helpful thing would be to sort out when and how IDE disks can do 128K or larger transfers, and adjust the various PCI IDE (or at least ahcisata) drivers and wd.c accordingly -- it would make testing much easier. Another very helpful thing would be to implement a smart minphys() for RAIDframe along the lines detailed in the MAXPHYS-NOTES file.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file pciide.c.orig was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8	14-May-2008	tsutsui	Normalize my licenses.
1.7	25-Dec-2007	perry	branches: 1.7.6; 1.7.8; 1.7.10; 1.7.12; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.6	11-Dec-2005	christos	branches: 1.6.46; 1.6.52; 1.6.56; 1.6.60; merge ktrace-lwp.
1.5	08-May-2005	tsutsui	branches: 1.5.2; ACARD ATP865-A (used on AEC-6280/6880) supports UltraDMA/133 too.
1.4	27-Feb-2005	perry	nuke trailing whitespace
1.3	24-Oct-2003	tsutsui	branches: 1.3.8; 1.3.10; Add support for Acard ATP865/A Ultra100 IDE Controllers. From Tsubai Masanari.
1.2	21-Oct-2001	thorpej	branches: 1.2.4; 1.2.20; Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.1	21-Apr-2001	tsutsui	branches: 1.1.2; 1.1.4; 1.1.6; Preliminary support for Acard ATP850/860 UDMA PCI IDE controllers. XXX Currently disabled by default because it has some problems on macppc. XXX Maybe some more initialization is needed, but there is few information XXX about the chips.
1.1.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.3	22-Oct-2001	nathanw	Catch up to -current.
1.1.4.2	21-Jun-2001	nathanw	Catch up to -current.
1.1.4.1	21-Apr-2001	nathanw	file pciide_acard_reg.h was added on branch nathanw_sa on 2001-06-21 20:05:10 +0000
1.1.2.2	23-Apr-2001	bouyer	Sync with HEAD.
1.1.2.1	21-Apr-2001	bouyer	file pciide_acard_reg.h was added on branch thorpej_scsipi on 2001-04-23 09:42:30 +0000
1.2.20.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.20.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.20.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.20.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.20.1	03-Aug-2004	skrll	Sync with HEAD
1.2.4.2	06-Feb-2002	he	Pull up revision 1.1 (new, requested by tsutsui): Add support for Acard ATP-850/860 pciide controllers.
1.2.4.1	21-Oct-2001	he	file pciide_acard_reg.h was added on branch netbsd-1-5 on 2002-02-06 14:18:08 +0000
1.3.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.3.8.1	29-Apr-2005	kent	sync with -current
1.5.2.1	21-Jan-2008	yamt	sync with head
1.6.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.6.56.1	26-Dec-2007	ad	Sync with head.
1.6.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.6.46.1	09-Jan-2008	matt	sync with HEAD
1.7.12.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.7.10.1	16-May-2008	yamt	sync with head.
1.7.8.1	18-May-2008	yamt	sync with head.
1.7.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.13	21-Jul-2017	nakayama	Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.12	19-Oct-2009	bouyer	branches: 1.12.18; 1.12.22; 1.12.38; 1.12.40; 1.12.56; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.11	25-Dec-2007	perry	branches: 1.11.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.10	11-Dec-2005	christos	branches: 1.10.46; 1.10.52; 1.10.56; 1.10.60; merge ktrace-lwp.
1.9	06-Aug-2005	bouyer	Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.8	27-Feb-2005	perry	branches: 1.8.2; 1.8.4; nuke trailing whitespace
1.7	05-Oct-2003	bouyer	branches: 1.7.6; 1.7.8; 1.7.10; Remove references to University of California from my copyright notices.
1.6	23-Apr-2002	bouyer	branches: 1.6.10; More copyright fixes, pointed out by Thomas. Thanks !
1.5	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.4	26-Jul-2001	bouyer	Better support for newer ALI M5229 chipsets: support Ultra/66 for rev >= 0xC2, Ultra/100 for revs >= 0xC4. The the generic PCIIDE interupt routine for chipsets rev >= 0xC2 in native mode, it seems that newer chipsets don't have the ACER_CHIDS register :( From Linux and FreeBSD.
1.3	15-May-2000	bouyer	branches: 1.3.6; 1.3.8; Sync my copyrigth notice.
1.2	29-Aug-1999	bouyer	branches: 1.2.2; Add support for the Promise Ultra/33 and /66 pci IDE controller. In addition to chip-dependant code this required the following changes: - Instead of attaching the device in a generic way with some chip-dependant routines, use a chip-dependant attach routine with some common code factored out. The code is marginally bigger, but this allows the CMD64x flag hack to go away. - For chips that report per-channel 'irq triggered', test this before calling wdcintr() for the native-pci irq case (compat intr can't be shared), as wdcintr() has no good way to know if a irq was for it or not, and ends up with irq loss. XXX for chips that don't have this feature irq sharing will not work properly ! - add my copyrigth notice (could have been done some time ago I think :) There are still some issues to be solved with the Promise controller and ATAPI devices. Many thanks to Paul Newhouse for shipping me 2 Ultra/33 boards for doing this work.
1.1	02-Feb-1999	bouyer	branches: 1.1.2; Support for Acerlab M5229 IDE controller. Thanks to Thilo Manske for testing the code, and to Takahiro Kambe who run several tests and finally found the bug by himself :)
1.1.2.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.2.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.8.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.8.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.8.1	03-Aug-2001	lukem	update to -current
1.3.6.3	20-Jun-2002	nathanw	Catch up to -current.
1.3.6.2	22-Oct-2001	nathanw	Catch up to -current.
1.3.6.1	24-Aug-2001	nathanw	Catch up with -current.
1.6.10.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.10.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.10.1	03-Aug-2004	skrll	Sync with HEAD
1.7.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.7.6.1	07-Aug-2005	riz	Pull up revision 1.9 via patch (requested by bouyer in ticket #5555): Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.8.4.2	21-Jan-2008	yamt	sync with head
1.8.4.1	21-Jun-2006	yamt	sync with head.
1.8.2.1	18-Aug-2005	tron	Pull up revision 1.9 (requested by bouyer in ticket #653): Add an optionnal controller callback for channel reset. If the callback is set to NULL, use the generic reset code. Use this to work around a bug in some Acer IDE controllers (like the one found in some sparc systems) where a controller disable/enable is required after a reset to avoid data corruption when Ultra-DMA is used. Workaround from opensolaris, thanks to Hiroki Sato for testing.
1.10.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.10.56.1	26-Dec-2007	ad	Sync with head.
1.10.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.10.46.1	09-Jan-2008	matt	sync with HEAD
1.11.10.1	11-Mar-2010	yamt	sync with head
1.12.56.1	25-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #153): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.12.40.1	28-Aug-2017	skrll	Sync with HEAD
1.12.38.1	26-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #1458): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.12.22.1	03-Dec-2017	jdolecek	update from HEAD
1.12.18.1	23-Jul-2017	snj	Pull up following revision(s) (requested by nakayama in ticket #1463): sys/dev/pci/aceride.c: revision 1.37 sys/dev/pci/pciide_acer_reg.h: revision 1.13 Apply workaround from FreeBSD to fix read data corruption observed on Fire V100 and mSATA-SSD with mSATA to IDE adapter. The patch is from port-sparc64@.
1.7	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.6	24-Jan-2003	thorpej	branches: 1.6.2; * Enable Ultra133 on the NVIDIA nForce 2. * Make sure to test for PCI_VENDOR_AMD before testing for any AMD products.
1.5	24-Jan-2003	thorpej	Add support for the NVIDIA nForce ATA100 and nForce2 ATA133 IDE controllers, which are more-or-less compatible with the AMD controllers. XXX Need to determine the correct timing value for the nForce2 XXX at Ultra133, so we cap it at Ultra100, for now.
1.4	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.3	06-May-2001	fvdl	branches: 1.3.2; Add amd766 support. Rename some variables to 7x6 now that it supports both 756 and 766.
1.2	06-Jul-2000	bouyer	branches: 1.2.2; 1.2.4; 1.2.6; Work around a bug in AMD756 rev D2, from patches provided by David Sainty: disable multiword DMA for these chips. multiword DMA can be forced with options PCIIDE_AMD756_ENABLEDMA on rev D2 chips, but use at your own risk ! While I'm there remove a duplicate allocation of sc_wdcdev.nchannels in HPT code.
1.1	06-Mar-2000	bouyer	branches: 1.1.4; Add support for the AMD 756 DMA/UDMA IDE controller, provided in PR kern/9536 by Dave Sainty.
1.1.4.2	15-May-2001	he	Pull up revision 1.3 (requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.1.4.1	07-Jul-2000	bouyer	Pull up pciide.c 1.77 -> 1.78 pciide_amd_reg.h 1.1 -> 1.2 (approved by thorpej): workaround for a bug in some revs of the AMD IDE controller, which can lock up the machine when multiword DMA is used.
1.2.6.2	22-Oct-2001	nathanw	Catch up to -current.
1.2.6.1	21-Jun-2001	nathanw	Catch up to -current.
1.2.4.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.4.1	06-Jul-2000	bouyer	file pciide_amd_reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:34 +0000
1.2.2.2	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.2.2.1	06-Jul-2000	he	file pciide_amd_reg.h was added on branch netbsd-1-4 on 2000-07-07 17:33:50 +0000
1.3.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.21	03-Apr-2025	andvar	viaide(4): Add support for VIA VT6415/VT6330 single-channel IDE controllers. Beyond adding the usual identification code, several adjustments were required: 1) Introduced 'single_channel` to set 'nchannels` to 1 for this controller. 2) Added 'APO_IDECONF_ALWAYS_EN` definition and the 'no_ideconf` flag to mimic enabled channel bits, as this controller's enable chip register is unset. Applied the same to VT6410, following OpenBSD and Linux, since some controllers are known to not set this register as well. 3) Combined all VIA controllers identified by PCI ID (not ISA bus) using switch fallthrough, as they currently share the same UDMA setting. Consequently, the VX900 name printing was moved to the device description and adjusted to better reflect the device. 4) Moved setting interface bits for RAID controllers under the RAID capability check, enabling the above fallthrough and paving the way for support of other RAID-capable controllers in upcoming changes. 5) The VT6330 is a combo FireWire/IDE controller. Added its FireWire controller PCI ID to pcidevs. Tested on ASRock P5B-DE. Reviewed by jak and bad. Addresses PR kern/45917
1.20	10-Jul-2011	jakllsch	branches: 1.20.60; 1.20.86; 1.20.92; Add complete support for all channels on VT6421A SATA/PATA controller to viaide(4).
1.19	30-Jun-2011	wiz	dependant -> dependent
1.18	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.17	25-Dec-2007	perry	branches: 1.17.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.16	11-Dec-2005	christos	branches: 1.16.46; 1.16.52; 1.16.56; 1.16.60; merge ktrace-lwp.
1.15	27-Feb-2005	perry	branches: 1.15.4; nuke trailing whitespace
1.14	08-Oct-2003	bouyer	branches: 1.14.8; 1.14.10; Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.13	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.12	25-Aug-2002	bouyer	branches: 1.12.6; Correct setup for Ultra133 capable VIA chipsets, From Matthias Drochner on current-users, with cross-check and some improvement from linux-2.4.19 and FreeBSD-current. Also don't set the APO_UDMA_CLK66 bit for Ultra/100 capable chipset, and support Ultra/133 for the VT8233A.
1.11	23-Apr-2002	bouyer	branches: 1.11.2; 1.11.4; More copyright fixes, pointed out by Thomas. Thanks !
1.10	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.9	04-May-2001	bouyer	branches: 1.9.2; Better support for VIA chipsets: look at the product/rev ID of the ISA bridge to guess the pciide capabilities, rather than trying to guess it by ourselve. Add preliminary support for the 686b (Ultra/100) guessed from FreeBSD/linux driver (datasheet not publically available, I contacted via). Let chip-specific map routine do the autoconf printf if ide_name is NULL (they may have more details about the controller than we have in pciide_attach)
1.8	05-Jan-2001	bouyer	branches: 1.8.2; Run at Ultra/66 on VIA controllers that can do it. PCI vendor/device/revision is the same for the Ultra/33 and Ultra/66 version, so test writability of the U66 enable bit (idea from Chris Cappuccio). Thanks to David Carrel for testings.
1.7	29-Dec-2000	tsutsui	s/AP0_UDMA_MASK/APO_UDMA_MASK/
1.6	15-May-2000	bouyer	branches: 1.6.4; Sync my copyrigth notice.
1.5	18-Jan-2000	bouyer	Correct URLs.
1.4	16-Dec-1998	bouyer	branches: 1.4.2; 1.4.8; Rearange the modes setup to allow these to be dyanmically changed. Fill in the new "set_mode" callback.
1.3	19-Oct-1998	bouyer	Fix a comment.
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	25-Jun-1998	bouyer	branches: 1.1.2; file pciide_apollo_reg.h was initially added on branch bouyer-ide.
1.1.2.1	25-Jun-1998	bouyer	Add support for VIA's apollo VP chipset.
1.4.8.3	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.4.8.2	05-Jan-2001	bouyer	Sync with HEAD
1.4.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.6.4.2	15-May-2001	he	Pull up revision 1.9 (via patch, requested by bouyer): Add support for newer VIA pciide controllers. Add support for the AMD 766 pciide controller. Properly distinguish between HPT366 and HPT370 controllers.
1.6.4.1	04-Feb-2001	he	Pull up revision 1.8 (requested by bouyer): Run at Ultra/66 on VIA controllers that can do it.
1.8.2.4	27-Aug-2002	nathanw	Catch up to -current.
1.8.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.8.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.8.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.9.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.9.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.11.4.1	01-Nov-2002	tron	Pull up revision 1.12 (requested by bouyer in ticket #727): Correct setup for Ultra133 capable VIA chipsets, From Matthias Drochner on current-users, with cross-check and some improvement from linux-2.4.19 and FreeBSD-current. Also don't set the APO_UDMA_CLK66 bit for Ultra/100 capable chipset, and support Ultra/133 for the VT8233A.
1.11.2.1	29-Aug-2002	gehenna	catch up with -current.
1.12.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.12.6.1	03-Aug-2004	skrll	Sync with HEAD
1.14.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.14.8.1	29-Apr-2005	kent	sync with -current
1.15.4.1	21-Jan-2008	yamt	sync with head
1.16.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.16.56.1	26-Dec-2007	ad	Sync with head.
1.16.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.16.46.1	09-Jan-2008	matt	sync with HEAD
1.17.10.1	11-Mar-2010	yamt	sync with head
1.20.92.1	02-Aug-2025	perseant	Sync with HEAD
1.20.86.1	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1101): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.20.60.1	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1944): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file pciide_apollo_reg.h.orig was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.18	05-Dec-2021	msaitoh	s/followings bits/following bits/ in comment.
1.17	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.16	25-Dec-2007	perry	branches: 1.16.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.15	11-Dec-2005	christos	branches: 1.15.46; 1.15.52; 1.15.56; 1.15.60; merge ktrace-lwp.
1.14	27-Feb-2005	perry	branches: 1.14.4; nuke trailing whitespace
1.13	05-Oct-2003	bouyer	branches: 1.13.8; 1.13.10; Remove references to University of California from my copyright notices.
1.12	23-Apr-2002	bouyer	branches: 1.12.10; More copyright fixes, pointed out by Thomas. Thanks !
1.11	15-Nov-2001	bouyer	Support Ultra/100 on CMD 0649.
1.10	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.9	02-Aug-2000	bouyer	branches: 1.9.2; 1.9.4; Add support for the CMD PCI646U. Linux claims that this driver is brocken with UDMA, so enable Ultra-DMA only if "options PCIIDE_CMD0646U_UDMA" is set.
1.8	01-Aug-2000	bouyer	Add support for the CMD PCI0646U2, an Ultra/33 version of the 0646. Note: there's also a PCI0646U, for which I don't have docs for now.
1.7	26-Jun-2000	bouyer	Add support for the CMD PCI0648 and PCI0649 IDE controllers. Thanks to Matthias Scheler for testing.
1.6	15-May-2000	bouyer	branches: 1.6.4; Sync my copyrigth notice.
1.5	29-Aug-1999	bouyer	branches: 1.5.2; Add support for the Promise Ultra/33 and /66 pci IDE controller. In addition to chip-dependant code this required the following changes: - Instead of attaching the device in a generic way with some chip-dependant routines, use a chip-dependant attach routine with some common code factored out. The code is marginally bigger, but this allows the CMD64x flag hack to go away. - For chips that report per-channel 'irq triggered', test this before calling wdcintr() for the native-pci irq case (compat intr can't be shared), as wdcintr() has no good way to know if a irq was for it or not, and ends up with irq loss. XXX for chips that don't have this feature irq sharing will not work properly ! - add my copyrigth notice (could have been done some time ago I think :) There are still some issues to be solved with the Promise controller and ATAPI devices. Many thanks to Paul Newhouse for shipping me 2 Ultra/33 boards for doing this work.
1.4	02-Dec-1998	bouyer	branches: 1.4.4; - change drive_flags from u_int8_t to u_int16_t - keep the modes supported by the drive in struct ata_drive_datas (will be later used for downgrading the DMA/PIO mode on error) - use config flags to force/disable PIO/DMA/UDMA modes - For the CMD PCI0643/6 setup DMA mode to DMA Read multiple.
1.3	09-Nov-1998	bouyer	Support for the CMD PCI064{3,6}. Tested on a 0646 with a "wd0: PIO mode 4, DMA mode 2, UDMA mode 2" device.
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	13-Aug-1998	bouyer	branches: 1.1.2; file pciide_cmd_reg.h was initially added on branch bouyer-ide.
1.1.2.1	13-Aug-1998	bouyer	- sync with -current - full support of CMD PCI0640 chips. This needs a special probe routine, as this chip doesn't honnor the "PCI_COMMAND_IO_ENABLE" bit. sigh. - do not map the native pci interrupts if we don't have a native channel. some chips (e.g. PCI0640) have a valid interrupt value which gets established, which may prevent the compat channel or some other ISA device from working properly later. - b* -> mem*
1.4.4.2	14-Aug-2000	he	Pull up revisions 1.8-1.9 (requested by bouyer): Add UltraDMA support for the CMD PCI0646U and PCI0646U2 controllers; normally disabled on PCI0646U due to a chip bug, but can be enabled with a kernel option.
1.4.4.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.5.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.6.4.4	05-Jan-2002	he	Pull up revision 1.11 (via patch, requested by bouyer): Support Ultra-DMA 100 on the CMD 0649.
1.6.4.3	03-Aug-2000	bouyer	Pull up (approved by thorpej): sys/dev/pci/pciide.c 1.82 -> 1.84 sys/dev/pci/pciide_cmd_reg.h 1.8 -> 1.9 sys/dev/pci/files.pci 1.100 -> 1.101 share/man/man4/pciide.4 1.23 -> 1.24 Add support for the CMD PCI646U. Linux claims that this driver is brocken with UDMA, so enable Ultra-DMA only if "options PCIIDE_CMD0646U_ENABLEUDMA" is set. defopt PCIIDE_CMD0646U_ENABLEUDMA, PCIIDE_AMD756_ENABLEDMA, PCIIDE_CMD064x_DISABLE Fix a typo pointed out by John Hawkinson Update documentation.
1.6.4.2	02-Aug-2000	bouyer	Pull up (approved by thorpej): pciide_cmd_reg.h 1.7 -> 1.8 pciide.c 1.81 -> 1.82 Add supports for the PCI0646U2, Ultra/33 version of the PCI0646.
1.6.4.1	27-Jun-2000	bouyer	Pull up pciide.c 1.69->1.70, 1.71->1.72 pciide_cmd_reg.h 1.6->1.7 Approved by thorpej: Add support for the CMD PCI0648/0649 IDE controller.
1.9.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.9.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.9.2.1	22-Oct-2001	nathanw	Catch up to -current.
1.12.10.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.12.10.1	03-Aug-2004	skrll	Sync with HEAD
1.13.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.8.1	29-Apr-2005	kent	sync with -current
1.14.4.1	21-Jan-2008	yamt	sync with head
1.15.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.15.56.1	26-Dec-2007	ad	Sync with head.
1.15.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.15.46.1	09-Jan-2008	matt	sync with HEAD
1.16.10.1	11-Mar-2010	yamt	sync with head
1.73	03-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "pciide-disable-dma".
1.72	16-Apr-2025	andvar	Introduce the ATADEBUG_PCIIDE_MASK option to allow setting the atadebug_pciide_mask variable, similar to the existing ATADEBUG_MASK and ATADEBUG_WD_MASK options used in ata/ata.c and ata/wdc.c No functional changes intended by default. approved by bouyer.
1.71	31-Mar-2024	thorpej	branches: 1.71.2; As the final step in configuraing bus-master DMA, consult the boolean device property "pciide-disable-dma" and disable bus-master DMA if that property is present and true.
1.70	20-Nov-2023	thorpej	pciide_dma_dmamap_setup(): If we end up with a DMA segment with an odd length or odd starting address, unload the map and return EINVAL. Some controllers get really upset if a DMA segment has an odd address or length. This can happen if a physio user performs a virtually-contiguous I/O that starts at an odd address and spans a page boundary where the resulting physical pages are discontiguous. The EINVAL return will cause the upper layers in the ATA code to re-try the I/O using PIO, which should (will in all of my tests) succeed. PR port-alpha/56434
1.69	20-Nov-2023	thorpej	Revert previous. I'm going to make a slight change to the patch, and want it collapsed into a single commit to make it easier to pull into netbsd-10.
1.68	20-Nov-2023	thorpej	pciide_dma_dmamap_setup(): If we end up with a DMA segment with an odd length, unload the map and return EIO. Some controllers get really upset if a DMA segment has an odd length. This can happen if a physio user performs a virtually-contiguous I/O that starts at an odd address and spans a page boundary where the resulting physical pages are discontiguous. Ultimately, it's up to the physio user to paint inside the lines, but this will prevent the disk controller from wandering off into the weeds, at least. PR port-alpha/56434
1.67	24-Aug-2020	msaitoh	branches: 1.67.20; s/ressource/resource/. No functional change.
1.66	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.65	20-Oct-2017	jdolecek	branches: 1.65.2; 1.65.4; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.64	17-Oct-2017	jdolecek	reintroduce ATACH_IRQ_WAIT flag for attachments using wdcintr(), only process the interrupt when the flag is set - this fixes spurious interrupt during post-reset drive setup in wdc_ata_bio_start(), and wdc_atapi_start() while those functions set WDCTL_IDS, this seems to be ignored by certain (maybe all) PCI-IDE controllers; usually the implicit KERNEL_LOCK() would prevent the interrupt anyway, but not when the start routine is started from the atabus thread, which doesn't take it fixes 'panic: wdc_ata_bio_intr: bad state' reported on current-users by Chavdar Ivanov
1.63	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.62	13-Oct-2016	jdolecek	branches: 1.62.4; provide intr xname
1.61	07-Jul-2016	msaitoh	branches: 1.61.2; KNF. Remove extra spaces. No functional change.
1.60	29-Mar-2014	christos	branches: 1.60.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.59	22-Jun-2013	matt	branches: 1.59.2; When allocating the channel queue, make sure the return memory is zeroed.
1.58	14-Nov-2012	jakllsch	don't attempt to re-unmap the interrupt on detach
1.57	31-Jul-2012	bouyer	branches: 1.57.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.56	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.55	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.54	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.53	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.52	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.51	17-May-2011	dyoung	branches: 1.51.4; 1.51.8; PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.50	10-May-2011	dyoung	Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.49	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.48	17-Nov-2010	dholland	branches: 1.48.2; Fix build when NATA_DMA is 0.
1.47	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.46	06-Nov-2010	jakllsch	Rework pciide(4) detachment to take the legacy interrupt mapping into consideration and avoid future code duplication. Ports wanting to enable detachment of controllers with compatibility-mapped channels will need to supply a pciide_machdep_compat_intr_disestablish() function.
1.45	05-Nov-2010	jakllsch	Correct copy/paste error in as-of-yet-unused pciide_common_detach().
1.44	05-Nov-2010	jakllsch	Pave the way for detachment of pciide(4)-family ATA controllers.
1.43	19-Oct-2009	bouyer	branches: 1.43.2; 1.43.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.42	23-Aug-2009	jmcneill	Print device description on the same line as locators.
1.41	15-Mar-2009	cegger	ansify function definitions
1.40	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.39	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.38	18-Mar-2008	cube	branches: 1.38.4; 1.38.12; 1.38.18; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.37	04-Mar-2007	christos	branches: 1.37.16; 1.37.32; 1.37.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.36	09-Feb-2007	ad	branches: 1.36.2; Merge newlock2 to head.
1.35	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.34	17-Oct-2006	itohy	Silence unused variable warning.
1.33	17-Oct-2006	itohy	Make pciide(4)-only configurations (without other DMA-capable driver) compile.
1.32	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.31	29-Mar-2006	thorpej	branches: 1.31.8; 1.31.10; Use device_cfdata().
1.30	17-Jan-2006	bouyer	branches: 1.30.2; 1.30.4; 1.30.6; 1.30.8; 1.30.10; Add missing ch_ndrive initialisations. Pointed out by Juan RP.
1.29	04-Jan-2006	bouyer	Fix forcing use of DMA mode for the generic pciide driver: default_chip_map() is called from pciide_attach() and at this point we don't know which drives are here. Just assume all drives are there and allocate DMA ressources for all of them.
1.28	11-Dec-2005	christos	branches: 1.28.2; merge ktrace-lwp.
1.27	29-May-2005	christos	branches: 1.27.2; Add a const.
1.26	27-Feb-2005	perry	branches: 1.26.2; nuke trailing whitespace
1.25	15-Feb-2005	briggs	Allow MAXPHYS-sized instead of IDEDMA_BYTE_COUNT_MAX-sized transfers to be described by a transfer table.
1.24	15-Feb-2005	briggs	Move the definition of NIDEDMA_TABLES from pciidereg.h to be closer to its only user in pciide_common.c. Also redefine NIDEDMA_TABLES to match the max DMA transfer size specified in the call to bus_dmamap_create() (IDEDMA_BYTE_COUNT_MAX instead of MAXPHYS). The macro is also redefined to handle devices that have a PAGE_SIZE greater than sc_dma_maxsegsz (buggy revision of satalink 3112 on ibm4xx).
1.23	04-Feb-2005	perry	de-__P
1.22	24-Nov-2004	bouyer	branches: 1.22.4; 1.22.6; Move bus_dma setup out of pciide_dma_init() in a new function, pciide_dma_dmamap_setup(), for the benefit of drivers that needs special registers setup in dmainit().
1.21	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.20	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.19	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.18	16-Aug-2004	enami	Dereference ch_wdc after filled.
1.17	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.16	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.15	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.14	02-Aug-2004	bouyer	branches: 1.14.2; Make it possible for (*dma_finish)() to abort quietly a DMA op. Use this in wdc_reset_channel().
1.13	02-Aug-2004	bouyer	Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. Fix kern/26502 from Christian Biere.
1.12	04-Jun-2004	bouyer	Don't try to bus_space_subregion() registers that didn't get mapped. Problem reported and fix tested by Kurt Schreiner on port-alpha.
1.11	25-May-2004	thorpej	Add the notion of "shadow registers" to the wdc driver. These shadow registers are registers that overlap with others on many controllers, but which may actually be distinct on some controllers. Right now, the two shadows are: - wd_status (usually overlaps wd_command) - wd_features (usually overlaps wd_error) Add a new helper function, wdc_init_shadow_regs(), used to initialize the shadow register handles on controllers where they do actually overlap. Partially from Jordan Rhody @ Wasabi Systems, Inc.
1.10	05-May-2004	bouyer	default_chip_map(): We can't use pciide_mapchan() here, because pciide_mapchan() will attach an atabus, and we don't know yet if this channel is enabled or not. Fix kern/25455 by Dave Huang.
1.9	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.8	03-Jan-2004	thorpej	branches: 1.8.2; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.7	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.6	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.5	19-Dec-2003	thorpej	Some controllers must have certain bits always be set in the IDEDMA_CMD register for proper operation. Add a prototype field in the pciide_channel for this register and use it as necessary when writing the IDEDMA_CMD register.
1.4	19-Dec-2003	thorpej	Move the PCIIDE_OPTIONS_* constants into pciidevar.h
1.3	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.2	23-Oct-2003	bouyer	If we found a compat channel disabled, unmap the registers. There may be an ISA IDE adapter at the same address.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.8.2.3	11-Aug-2004	jmc	branches: 1.8.2.3.2; Pullup rev 1.13 (requested by bouyer in ticket #734) Don't map the interrupt in pciide_mapregs_compat(). In default_chip_map() we'll do further checks to see if the channel should really be enabled, and in case it's not we would keep the interrupt mapped. PR#26502
1.8.2.2	05-Jun-2004	jmc	Pullup rev 1.12 (requested by bouyer in ticket #443) Don't try to bus_space_subregion() registers that didn't get mapped.
1.8.2.1	09-May-2004	jdc	Pull up revision 1.10 (requested by bouyer in ticket #267) default_chip_map(): We can't use pciide_mapchan() here, because pciide_mapchan() will attach an atabus, and we don't know yet if this channel is enabled or not. Fix kern/25455 by Dave Huang.
1.8.2.3.2.2	05-Jan-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10212): sys/dev/pci/pciide_common.c: revision 1.29 via patch Fix forcing use of DMA mode for the generic pciide driver: default_chip_map() is called from pciide_attach() and at this point we don't know which drives are here. Just assume all drives are there and allocate DMA ressources for all of them.
1.8.2.3.2.1	16-Mar-2005	tron	Pull up revision 1.22 (requested by bouyer in ticket #1016): Move bus_dma setup out of pciide_dma_init() in a new function, pciide_dma_dmamap_setup(), for the benefit of drivers that needs special registers setup in dmainit().
1.14.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.14.2.8	15-Feb-2005	skrll	Sync with HEAD.
1.14.2.7	04-Feb-2005	skrll	Sync with HEAD.
1.14.2.6	29-Nov-2004	skrll	Sync with HEAD.
1.14.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.14.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.14.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.14.2.2	03-Aug-2004	skrll	Sync with HEAD
1.14.2.1	02-Aug-2004	skrll	file pciide_common.c was added on branch ktrace-lwp on 2004-08-03 10:49:11 +0000
1.22.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.22.6.1	12-Feb-2005	yamt	sync with head.
1.22.4.1	29-Apr-2005	kent	sync with -current
1.26.2.1	05-Jan-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1087): sys/dev/pci/pciide_common.c: revision 1.29 Fix forcing use of DMA mode for the generic pciide driver: default_chip_map() is called from pciide_attach() and at this point we don't know which drives are here. Just assume all drives are there and allocate DMA ressources for all of them.
1.27.2.5	24-Mar-2008	yamt	sync with head.
1.27.2.4	03-Sep-2007	yamt	sync with head.
1.27.2.3	26-Feb-2007	yamt	sync with head.
1.27.2.2	30-Dec-2006	yamt	sync with head.
1.27.2.1	21-Jun-2006	yamt	sync with head.
1.28.2.2	01-Feb-2006	yamt	sync with head.
1.28.2.1	15-Jan-2006	yamt	sync with head.
1.30.10.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.30.8.1	19-Apr-2006	elad	sync with head.
1.30.6.1	01-Apr-2006	yamt	sync with head.
1.30.4.1	22-Apr-2006	simonb	Sync with head.
1.30.2.1	09-Sep-2006	rpaulo	sync with head
1.31.10.2	10-Dec-2006	yamt	sync with head.
1.31.10.1	22-Oct-2006	yamt	sync with head
1.31.8.3	06-Feb-2007	ad	Quiten noisy boot messages.
1.31.8.2	06-Feb-2007	ad	Quieten noisy boot messages.
1.31.8.1	18-Nov-2006	ad	Sync with head.
1.36.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.37.36.1	03-Apr-2008	mjf	Sync with HEAD.
1.37.32.1	24-Mar-2008	keiichi	sync with head.
1.37.16.1	23-Mar-2008	matt	sync with HEAD
1.38.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.38.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.38.4.3	11-Mar-2010	yamt	sync with head
1.38.4.2	16-Sep-2009	yamt	sync with head
1.38.4.1	04-May-2009	yamt	sync with head.
1.43.4.3	31-May-2011	rmind	sync with head
1.43.4.2	21-Apr-2011	rmind	sync with head
1.43.4.1	05-Mar-2011	rmind	sync with head
1.43.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.48.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.51.8.1	18-Feb-2012	mrg	merge to -current.
1.51.4.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.51.4.3	16-Jan-2013	yamt	sync with (a bit old) head
1.51.4.2	30-Oct-2012	yamt	sync with head
1.51.4.1	17-Apr-2012	yamt	sync with head
1.57.2.4	03-Dec-2017	jdolecek	update from HEAD
1.57.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.57.2.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.57.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.59.2.1	18-May-2014	rmind	sync with head
1.60.6.2	05-Dec-2016	skrll	Sync with HEAD
1.60.6.1	09-Jul-2016	skrll	Sync with HEAD
1.61.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.62.4.3	27-Sep-2017	jdolecek	change wdc_init_shadow_regs() to accept only struct wdc_regs, it doesn't touch anything else factor out the probe-only struct ata_channel initialization to wdcprobe(), to reduce duplication of logic in individual drivers, and to actually work now that more init is needed beyond the memset()
1.62.4.2	12-Aug-2017	jdolecek	remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status, so that there is less hidden state shared by commands; primary intent is to make the NCQ and non-NCQ paths more similar, and remove possibility of incorrect handling for the NCQ commands tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4), ahcisata(4), mvsata(4) on real hw
1.62.4.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.65.4.1	10-Jun-2019	christos	Sync with HEAD
1.65.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.67.20.1	26-Nov-2023	bouyer	Pull up following revision(s) (requested by thorpej in ticket #470): sys/dev/pci/pciide_common.c: revision 1.70 pciide_dma_dmamap_setup(): If we end up with a DMA segment with an odd length or odd starting address, unload the map and return EINVAL. Some controllers get really upset if a DMA segment has an odd address or length. This can happen if a physio user performs a virtually-contiguous I/O that starts at an odd address and spans a page boundary where the resulting physical pages are discontiguous. The EINVAL return will cause the upper layers in the ATA code to re-try the I/O using PIO, which should (will in all of my tests) succeed. PR port-alpha/56434
1.71.2.1	02-Aug-2025	perseant	Sync with HEAD
1.12	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.11	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.10	25-Dec-2007	perry	branches: 1.10.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.9	11-Dec-2005	christos	branches: 1.9.46; 1.9.52; 1.9.56; 1.9.60; merge ktrace-lwp.
1.8	27-Feb-2005	perry	branches: 1.8.4; nuke trailing whitespace
1.7	05-Oct-2003	bouyer	branches: 1.7.8; 1.7.10; Remove references to University of California from my copyright notices.
1.6	23-Apr-2002	bouyer	branches: 1.6.10; More copyright fixes, pointed out by Thomas. Thanks !
1.5	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.4	15-May-2000	bouyer	branches: 1.4.6; 1.4.8; Sync my copyrigth notice.
1.3	29-Aug-1999	bouyer	branches: 1.3.2; Add definitions of the DMA mode control registers. They should be used by pciide, but I have yet to find how to map them.
1.2	03-Dec-1998	bouyer	branches: 1.2.4; Add URLs of documentation source (from Matthias Drochner).
1.1	21-Nov-1998	drochner	headers for new hardware support in pciide.c. from Manuel.
1.2.4.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.3.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.4.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.4.6.1	22-Oct-2001	nathanw	Catch up to -current.
1.6.10.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.10.1	03-Aug-2004	skrll	Sync with HEAD
1.7.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.8.4.1	21-Jan-2008	yamt	sync with head
1.9.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.9.56.1	26-Dec-2007	ad	Sync with head.
1.9.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.46.1	09-Jan-2008	matt	sync with HEAD
1.10.10.1	11-Mar-2010	yamt	sync with head
1.6	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.5	25-Dec-2007	perry	branches: 1.5.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.4	11-Dec-2005	christos	branches: 1.4.46; 1.4.52; 1.4.56; 1.4.60; merge ktrace-lwp.
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	27-Jul-2004	rumble	branches: 1.2.2; 1.2.6; 1.2.8; Add support for the controller found on the AMD (previously NatSemi) Geode SC1100 controller (as found on the Soekris NET4801). The chip has two bugs: the first requires dword alignment, and the second cannot handle exact 64K transfers. Also, fix a few typos while we're here. Timings from FreeBSD. Reviewed by Manuel Bouyer.
1.1	09-Jul-2004	bouyer	branches: 1.1.2; Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.1.2.3	28-Jul-2004	tron	Pull up revision 1.2 (requested by rumble in ticket #718): Add support for the controller found on the AMD (previously NatSemi) Geode SC1100 controller (as found on the Soekris NET4801). The chip has two bugs: the first requires dword alignment, and the second cannot handle exact 64K transfers. Also, fix a few typos while we're here. Timings from FreeBSD. Reviewed by Manuel Bouyer.
1.1.2.2	14-Jul-2004	tron	Pull up revision 1.1 (requested by bouyer in ticket #644): Add geodeide(4), a driver for the AMD Geode CS5530A IDE controller. Tested by Ian Zagorskih (ianzag at megasignal.com).
1.1.2.1	09-Jul-2004	tron	file pciide_geode_reg.h was added on branch netbsd-2-0 on 2004-07-14 09:08:28 +0000
1.2.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.6.1	29-Apr-2005	kent	sync with -current
1.2.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.2.2.2	03-Aug-2004	skrll	Sync with HEAD
1.2.2.1	27-Jul-2004	skrll	file pciide_geode_reg.h was added on branch ktrace-lwp on 2004-08-03 10:49:11 +0000
1.3.4.1	21-Jan-2008	yamt	sync with head
1.4.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.4.56.1	26-Dec-2007	ad	Sync with head.
1.4.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.46.1	09-Jan-2008	matt	sync with HEAD
1.5.10.1	11-Mar-2010	yamt	sync with head
1.16	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.15	25-Dec-2007	perry	branches: 1.15.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.14	05-Aug-2006	xtraeme	branches: 1.14.30; 1.14.36; 1.14.40; 1.14.44; HPT368 support. Patch from Jonathan A. Kollasch via PR kern/34097. Thanks.
1.13	11-Dec-2005	christos	branches: 1.13.4; 1.13.8; merge ktrace-lwp.
1.12	27-Feb-2005	perry	branches: 1.12.2; 1.12.4; nuke trailing whitespace
1.11	10-Mar-2004	bouyer	branches: 1.11.8; 1.11.10; Add the HPT370 DPLL output frequency register.
1.10	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.9	23-Aug-2002	bouyer	branches: 1.9.6; Add support for HPT372. From sdegler@kashmir.degler.net in kern/17908, with some improvement from me, and from FreeBSD.
1.8	01-Jun-2002	simonb	Add "__attribute__((__unused__))" to hpt374_pio, hpt374_dma and hpt374_udma.
1.7	19-May-2002	bouyer	branches: 1.7.2; Add support for HPT374. Submited by Matthew Rezny in kern/16897, with minor adjustements by me.
1.6	23-Apr-2002	bouyer	branches: 1.6.2; More copyright fixes, pointed out by Thomas. Thanks !
1.5	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.4	23-Jul-2001	bouyer	Handle HPT-370A controllers, from skanto@sjk-software.fi (kern/13540) but with some differences to the original patch: don't assume all controllers with rev >= HPT370_REV are HPT370, and explicitely print if we have a chip with a rev the driver does't know.
1.3	05-Jan-2001	bouyer	branches: 1.3.2; 1.3.4; Add support for Ultra/100 on the HPT370 XXX the datasheet for the HPT370 is wrong While I'm there clear the DRIVE_DMA flag when we're going to use Ultra/DMA where it was missed, so that wdc_print_modes() only prints what's used.
1.2	27-Jul-2000	bouyer	branches: 1.2.2; Ops, forgot this one: Add interrupt disable bit #define for HPT370
1.1	12-Jun-2000	bouyer	branches: 1.1.2; 1.1.4; 1.1.6; - add a pciide_irqack() callback, which clears the IDE DMA status bit once the IRQ has been cleared on the drive. - use pa->pa_class instead of re-reading PCI_CLASS_REG when possible - Add support for Highpoint HPT366 and HPT370 (370 untested), based on patches from Roger Brooks <R.S.Brooks@liverpool.ac.uk> posted on current-users Mach, 15. Given how Highpoint docs have been wrong for the 366, the 370 is likely to not work. Thanks to Chris Cappuccio <chris@dqc.org> for sending me the Highpoint docs, and to Total Archive (http://www.totalarchive.com/) for sending me hardware.
1.1.6.3	27-Jul-2000	he	Pull up revision 1.2 (requested by bouyer): Clear the interrupt disable bit of the HPT370, and document HPT370 support.
1.1.6.2	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.1.6.1	12-Jun-2000	he	file pciide_hpt_reg.h was added on branch netbsd-1-4 on 2000-07-07 17:33:50 +0000
1.1.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.1.4.1	12-Jun-2000	minoura	file pciide_hpt_reg.h was added on branch minoura-xpg4dl on 2000-06-22 17:07:35 +0000
1.1.2.3	13-Nov-2001	he	Pull up revision 1.4 (requested by bouyer): Support HPT-370A controllers. Fixes PR#13540.
1.1.2.2	04-Feb-2001	he	Pull up revision 1.3 (requested by bouyer): Add support for Ultra/100 on the HPT370 controller.
1.1.2.1	27-Jul-2000	bouyer	Pull up (approved by thorpej): sys/dev/pci/pciide.c 1.80 -> 1.81 sys/dev/pci/pciide_hpt_reg.h 1.1 -> 1.2 distrib/notes/i386/hardware 1.74 -> 1.75 share/man/man4/pciide.4 1.22 -> 1.23 Make sure the HPT370's interrupts disable bit is cleared. Now that HPT370 supports works, document it.
1.2.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.1	27-Jul-2000	bouyer	file pciide_hpt_reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:34 +0000
1.3.4.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.3.4.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.4.1	03-Aug-2001	lukem	update to -current
1.3.2.4	27-Aug-2002	nathanw	Catch up to -current.
1.3.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.3.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.3.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.6.2.3	29-Aug-2002	gehenna	catch up with -current.
1.6.2.2	20-Jun-2002	gehenna	catch up with -current.
1.6.2.1	30-May-2002	gehenna	Catch up with -current.
1.7.2.1	01-Nov-2002	tron	Pull up revision 1.8-1.9 (requested by bouyer in ticket #725): Add support for HPT372. From sdegler@kashmir.degler.net in kern/17908, with some improvement from me, and from FreeBSD.
1.9.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.9.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.9.6.1	03-Aug-2004	skrll	Sync with HEAD
1.11.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.8.1	29-Apr-2005	kent	sync with -current
1.12.4.2	21-Jan-2008	yamt	sync with head
1.12.4.1	30-Dec-2006	yamt	sync with head.
1.12.2.1	06-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #1455): sys/dev/pci/hptide.c: revision 1.21 sys/dev/pci/pciide_hpt_reg.h: revision 1.14 HPT368 support. Patch from Jonathan A. Kollasch via PR kern/34097. Thanks.
1.13.8.1	11-Aug-2006	yamt	sync with head
1.13.4.1	09-Sep-2006	rpaulo	sync with head
1.14.44.1	02-Jan-2008	bouyer	Sync with HEAD
1.14.40.1	26-Dec-2007	ad	Sync with head.
1.14.36.1	18-Feb-2008	mjf	Sync with HEAD.
1.14.30.1	09-Jan-2008	matt	sync with HEAD
1.15.10.1	11-Mar-2010	yamt	sync with head
1.3	11-Dec-2005	christos	merge ktrace-lwp.
1.2	11-Feb-2005	rearnsha	Add support for the Artisea device operating in DPA mode. Approved by briggs.
1.1	18-Mar-2003	thorpej	branches: 1.1.2; 1.1.10; 1.1.12; Add register definitions for the Intel i31244 Serial ATA controller.
1.1.12.1	12-Feb-2005	yamt	sync with head.
1.1.10.1	29-Apr-2005	kent	sync with -current
1.1.2.1	15-Feb-2005	skrll	Sync with HEAD.
1.2	11-Dec-2005	christos	merge ktrace-lwp.
1.1	01-Dec-2004	grant	branches: 1.1.4; add iteide(4) driver for ITE8212-based IDE controllers. from OpenBSD, ported to NetBSD by me. ok'd by bouyer@, thorpej@.
1.1.4.2	18-Dec-2004	skrll	Sync with HEAD.
1.1.4.1	01-Dec-2004	skrll	file pciide_ite_reg.h was added on branch ktrace-lwp on 2004-12-18 09:32:10 +0000
1.5	29-Apr-2008	martin	Convert to new 2 clause license
1.4	10-Sep-2007	cube	branches: 1.4.20; 1.4.22; 1.4.24; Remove 3rd clause and my name from all the licences which were only in my name.
1.3	11-Dec-2005	christos	branches: 1.3.30; 1.3.44; 1.3.46; merge ktrace-lwp.
1.2	27-Feb-2005	perry	branches: 1.2.4; nuke trailing whitespace
1.1	06-Dec-2004	cube	branches: 1.1.4; 1.1.6; 1.1.8; Add ixpide(4), a driver for the IDE controller found in the ATI IXP series of integrated chipsets. Documented from the Linux driver. Tested by reinoud@, ok'd by bouyer@.
1.1.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.6.1	29-Apr-2005	kent	sync with -current
1.1.4.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.4.2	18-Dec-2004	skrll	Sync with HEAD.
1.1.4.1	06-Dec-2004	skrll	file pciide_ixp_reg.h was added on branch ktrace-lwp on 2004-12-18 09:32:10 +0000
1.2.4.1	27-Oct-2007	yamt	sync with head.
1.3.46.1	06-Nov-2007	matt	sync with HEAD
1.3.44.1	02-Oct-2007	joerg	Sync with HEAD.
1.3.30.1	09-Oct-2007	ad	Sync with head.
1.4.24.1	16-May-2008	yamt	sync with head.
1.4.22.1	18-May-2008	yamt	sync with head.
1.4.20.1	02-Jun-2008	mjf	Sync with HEAD.
1.1	10-Nov-2010	skrll	branches: 1.1.6; Add nside(4) - a driver for the National Semiconductor PC87415 IDE controller as found in many HP PA-RISC machines. From OpenBSD. Reviewed by Manuel Bouyer.
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	10-Nov-2010	rmind	file pciide_natsemi_reg.h was added on branch rmind-uvmplock on 2011-03-05 20:53:56 +0000
1.12	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.11	25-Dec-2007	perry	branches: 1.11.6; 1.11.8; 1.11.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.10	16-Feb-2006	perry	branches: 1.10.40; 1.10.46; 1.10.50; 1.10.54; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.9	24-Dec-2005	perry	branches: 1.9.2; 1.9.4; 1.9.6; __inline__ -> inline
1.8	11-Dec-2005	christos	merge ktrace-lwp.
1.7	19-Aug-2004	thorpej	branches: 1.7.12; - Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.6	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.5	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.4	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.3	21-Oct-2001	thorpej	branches: 1.3.18; Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.2	07-Jun-2000	scw	branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12; The OPTi controller supports a 32-bit dataport after all. Also detect when the chip is sitting on a 25MHz PCIbus and set the timing registers accordingly.
1.1	27-May-2000	scw	Add support for the OPTi 82c621 PCIIDE controller and its derivatives. I only have a Compaq laptop on which to test this, so reports of success/failure in other systems would be welcomed.
1.2.12.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.10.1	22-Oct-2001	nathanw	Catch up to -current.
1.2.8.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.8.1	07-Jun-2000	bouyer	file pciide_opti_reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:35 +0000
1.2.6.2	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.2.6.1	07-Jun-2000	he	file pciide_opti_reg.h was added on branch netbsd-1-4 on 2000-07-07 17:33:50 +0000
1.2.4.2	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.2.4.1	07-Jun-2000	minoura	file pciide_opti_reg.h was added on branch minoura-xpg4dl on 2000-06-22 17:07:36 +0000
1.3.18.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.18.3	18-Sep-2004	skrll	Sync with HEAD.
1.3.18.2	25-Aug-2004	skrll	Sync with HEAD.
1.3.18.1	03-Aug-2004	skrll	Sync with HEAD
1.7.12.2	21-Jan-2008	yamt	sync with head
1.7.12.1	21-Jun-2006	yamt	sync with head.
1.9.6.1	22-Apr-2006	simonb	Sync with head.
1.9.4.1	09-Sep-2006	rpaulo	sync with head
1.9.2.1	18-Feb-2006	yamt	sync with head.
1.10.54.1	02-Jan-2008	bouyer	Sync with HEAD
1.10.50.1	26-Dec-2007	ad	Sync with head.
1.10.46.1	18-Feb-2008	mjf	Sync with HEAD.
1.10.40.1	09-Jan-2008	matt	sync with HEAD
1.11.10.1	16-May-2008	yamt	sync with head.
1.11.8.1	18-May-2008	yamt	sync with head.
1.11.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file pciide_pIIx_reg.h was initially added on branch bouyer-ide.
1.1.2.2	06-Jun-1998	bouyer	Renamed pciide_pIIx_reg.h to pciide_piix_reg.h via repository copy, as suggested by Soren S. Jorvan.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.17	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.16	15-Oct-2022	rin	Fix style and typo in comments. No binary changes.
1.15	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.14	25-Dec-2007	perry	branches: 1.14.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.13	11-Dec-2005	christos	branches: 1.13.46; 1.13.52; 1.13.56; 1.13.60; merge ktrace-lwp.
1.12	27-Feb-2005	perry	branches: 1.12.4; nuke trailing whitespace
1.11	05-Oct-2003	bouyer	branches: 1.11.8; 1.11.10; Remove references to University of California from my copyright notices.
1.10	28-Jun-2003	bouyer	branches: 1.10.2; Correct setup of General Purpose Register for PDC20265 (Ultra/100) and newer. From Andreas Johansson in private mail.
1.9	28-Apr-2003	nakayama	Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.8	26-Jul-2002	wiz	Spell '[Rr]ight' correctly. From Jim Bernard.
1.7	23-Apr-2002	bouyer	branches: 1.7.2; 1.7.4; More copyright fixes, pointed out by Thomas. Thanks !
1.6	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.5	05-Jul-2001	toshii	branches: 1.5.2; Fix typo. s/extention/extension/
1.4	12-Mar-2001	bouyer	Correct definition of PDC262_U66_EN (from FreeBSD). Correct values of MW DMA 2, and add values for UDMA5 (from linux).
1.3	15-May-2000	bouyer	branches: 1.3.4; 1.3.6; 1.3.8; Sync my copyrigth notice.
1.2	28-Nov-1999	bouyer	Improve Ultra/66 support now that I've got some docs from Promise.
1.1	29-Aug-1999	bouyer	branches: 1.1.2; 1.1.8; Add support for the Promise Ultra/33 and /66 pci IDE controller. In addition to chip-dependant code this required the following changes: - Instead of attaching the device in a generic way with some chip-dependant routines, use a chip-dependant attach routine with some common code factored out. The code is marginally bigger, but this allows the CMD64x flag hack to go away. - For chips that report per-channel 'irq triggered', test this before calling wdcintr() for the native-pci irq case (compat intr can't be shared), as wdcintr() has no good way to know if a irq was for it or not, and ends up with irq loss. XXX for chips that don't have this feature irq sharing will not work properly ! - add my copyrigth notice (could have been done some time ago I think :) There are still some issues to be solved with the Promise controller and ATAPI devices. Many thanks to Paul Newhouse for shipping me 2 Ultra/33 boards for doing this work.
1.1.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.1.2.2	12-Mar-2001	bouyer	Sync with HEAD.
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.8.5	01-Aug-2002	nathanw	Catch up to -current.
1.3.8.4	20-Jun-2002	nathanw	Catch up to -current.
1.3.8.3	22-Oct-2001	nathanw	Catch up to -current.
1.3.8.2	24-Aug-2001	nathanw	Catch up with -current.
1.3.8.1	09-Apr-2001	nathanw	Catch up with -current.
1.3.6.2	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.3.6.1	15-May-2000	he	file pciide_pdc202xx_reg.h was added on branch netbsd-1-4 on 2000-07-07 17:33:50 +0000
1.3.4.1	13-Mar-2001	he	Pull up revision 1.4 (requested by bouyer): Support Ultra/100 speed on Promise Ultra/100, and fix ``bogus intr'' messages generated under some conditions.
1.5.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.5.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.5.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.4.2	16-Aug-2003	tron	Pull up revision 1.10 (requested by bouyer in ticket #1369): Correct setup of General Purpose Register for PDC20265 (Ultra/100) and newer. From Andreas Johansson in private mail.
1.7.4.1	28-Apr-2003	tron	Pull up revision 1.9 (requested by nakayama in ticket #1275): Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.7.2.1	29-Aug-2002	gehenna	catch up with -current.
1.10.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.11.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.8.1	29-Apr-2005	kent	sync with -current
1.12.4.1	21-Jan-2008	yamt	sync with head
1.13.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.13.56.1	26-Dec-2007	ad	Sync with head.
1.13.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.13.46.1	09-Jan-2008	matt	sync with HEAD
1.14.10.1	11-Mar-2010	yamt	sync with head
1.1	20-Sep-2009	christos	branches: 1.1.4; PR/42100: Dave J. Barnes: Support for old toshiba PICCOLO IDE controllers.
1.1.4.2	11-Mar-2010	yamt	sync with head
1.1.4.1	20-Sep-2009	yamt	file pciide_piccolo_reg.h was added on branch yamt-nfs-mp on 2010-03-11 15:03:58 +0000
1.16	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.15	04-Jul-2018	kamil	Avoid undefined behavior in pciiide macros Cast the 'bytes' argument in PIIX_IDETIM_SET() and PIIX_IDETIM_CLEAR() to unsigned int. This prevents UB because of shifting the bits and changing the bit of signedness. sys/dev/pci/piixide.c:714:11, left shift of 65535 by 16 places cannot be represented in type 'int' sys/dev/pci/piixide.c:720:11, left shift of 32768 by 16 places cannot be represented in type 'int' Detected with Kernel Undefined Behavior Sanitizer. Reported by <Harry Pantazis>
1.14	19-Oct-2009	bouyer	branches: 1.14.62; 1.14.64; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.13	25-Dec-2007	perry	branches: 1.13.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.12	03-Sep-2006	xtraeme	branches: 1.12.28; 1.12.34; 1.12.38; 1.12.42; * Add support for ICH8 and ICH8M SATA/RAID controllers. * If the controller is in AHCI, ask for SATA IDE mode of operation. jsg@openbsd says: "X60/T60 Thinkpads are shipped in AHCI configuration by default, this makes them work without changing a BIOS option." Tested by eye of the beholder. From OpenBSD. Ok'ed tls.
1.11	17-Jun-2006	jmcneill	branches: 1.11.2; Forgot to commit the register definition for PIIX_UDMATIM.
1.10	11-Dec-2005	christos	branches: 1.10.4; 1.10.8; 1.10.14; 1.10.16; merge ktrace-lwp.
1.9	27-Feb-2005	perry	branches: 1.9.4; nuke trailing whitespace
1.8	05-Oct-2003	bouyer	branches: 1.8.8; 1.8.10; Remove references to University of California from my copyright notices.
1.7	23-Apr-2002	bouyer	branches: 1.7.10; More copyright fixes, pointed out by Thomas. Thanks !
1.6	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.5	05-Jan-2001	bouyer	branches: 1.5.2; 1.5.4; Add support for Ultra/100 on intel ICH2; from Tomokazu HARADA in kern/11747.
1.4	15-May-2000	bouyer	branches: 1.4.4; Sync my copyrigth notice.
1.3	30-Aug-1999	bouyer	branches: 1.3.2; Add support for Intel 810 chipset (ICH/ICH0). While I'm there merge back piix_channel_map into piix_chip_map.
1.2	12-Oct-1998	bouyer	branches: 1.2.6; Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file pciide_pIIx_reg.h was initially added on branch bouyer-ide.
1.1.2.4	04-Oct-1998	bouyer	atavar.h: drv_softc is a struct device * instead of void*, as it's mostly used for dv_xname wd.c: convert for drv_softc type change, printf cleanup wdc.c: always call ata_get_params() (params was used initialised with non-32bit controllers, leaving to bogus PIO/DMA mode report). Cleaup of the PIO/DMA mode message. pciide_piix_reg.h: Fix definition of PIIX_IDETIM_CLEAR (unused before) pciide.c: add a method do disable a channel on know device. If a channel doesn't have any drive, we disable it and free its resources if disable was successfull. This should help with laptops where the second channel of the PIIX4 is unused but not disabled by BIOS. On such laptops, irq15 can be used for PCMCIA but it was claimed by pciide. Misc printf cleanup. wdc_isa.c: printf cleanup.
1.1.2.3	10-Jun-1998	bouyer	Fix confusion between binary and hex in DMA modes: 0x10 -> 0x02.
1.1.2.2	10-Jun-1998	bouyer	- Fix timing settings for DMA: the controller was always set up to use DMA mode 0 (compatible). - The 2 Ultra-dma registers are in fact one 32-bit register. Change the macros and setup in a way it may now work (but still untested, thus not enabled by default). - For DMA mode 1, use a more efficient timing than the one suggested by intel. Some work is still needed here to get ATAPI DMA working (should be done soon).
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.2.6.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.3.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.3.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.4.1	04-Feb-2001	he	Pull up revision 1.5 (requested by bouyer): Add support for Ultra/100 on Intel ICH2.
1.5.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.5.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.5.2.2	20-Jun-2002	nathanw	Catch up to -current.
1.5.2.1	22-Oct-2001	nathanw	Catch up to -current.
1.7.10.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.10.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.10.2	18-Sep-2004	skrll	Sync with HEAD.
1.7.10.1	03-Aug-2004	skrll	Sync with HEAD
1.8.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.8.8.1	29-Apr-2005	kent	sync with -current
1.9.4.3	21-Jan-2008	yamt	sync with head
1.9.4.2	30-Dec-2006	yamt	sync with head.
1.9.4.1	21-Jun-2006	yamt	sync with head.
1.10.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.10.14.1	19-Jun-2006	chap	Sync with head.
1.10.8.2	14-Sep-2006	yamt	sync with head.
1.10.8.1	26-Jun-2006	yamt	sync with head.
1.10.4.1	09-Sep-2006	rpaulo	sync with head
1.11.2.1	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #117): sys/dev/pci/piixide.c: revision 1.29 sys/dev/pci/pciide_piix_reg.h: revision 1.12 * Add support for ICH8 and ICH8M SATA/RAID controllers. * If the controller is in AHCI, ask for SATA IDE mode of operation. jsg@openbsd says: "X60/T60 Thinkpads are shipped in AHCI configuration by default, this makes them work without changing a BIOS option." Tested by eye of the beholder. From OpenBSD. Ok'ed tls.
1.12.42.1	02-Jan-2008	bouyer	Sync with HEAD
1.12.38.1	26-Dec-2007	ad	Sync with head.
1.12.34.1	18-Feb-2008	mjf	Sync with HEAD.
1.12.28.1	09-Jan-2008	matt	sync with HEAD
1.13.10.1	11-Mar-2010	yamt	sync with head
1.14.64.1	10-Jun-2019	christos	Sync with HEAD
1.14.62.1	28-Jul-2018	pgoyette	Sync with HEAD
1.1	06-Nov-2010	jakllsch	branches: 1.1.6; Add Intel SCH IDE controller driver. From OpenBSD via Tomokazu HARADA in PR#42310.
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	06-Nov-2010	rmind	file pciide_sch_reg.h was added on branch rmind-uvmplock on 2011-03-05 20:53:56 +0000
1.3	11-Dec-2005	christos	merge ktrace-lwp.
1.2	20-Dec-2003	thorpej	branches: 1.2.4; Add support for the Silicon Image 3114 SATALink 4-port SATA controller.
1.1	20-Mar-2003	thorpej	branches: 1.1.2; Add support for the Silicon Image SATALink 3112 Serial ATA controller.
1.1.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.1	03-Aug-2004	skrll	Sync with HEAD
1.2.4.2	12-Jul-2004	he	Pull up revision 1.1 (new, requested by grant in ticket #1720): Add support for SATALink 3112.
1.2.4.1	20-Dec-2003	he	file pciide_sii3112_reg.h was added on branch netbsd-1-6 on 2004-07-12 21:24:08 +0000
1.16	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.15	25-Dec-2007	perry	branches: 1.15.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.14	11-Dec-2005	christos	branches: 1.14.46; 1.14.52; 1.14.56; 1.14.60; merge ktrace-lwp.
1.13	16-Jun-2005	bouyer	branches: 1.13.2; Add quirk for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.12	27-Feb-2005	perry	branches: 1.12.2; nuke trailing whitespace
1.11	05-Oct-2003	bouyer	branches: 1.11.6; 1.11.8; 1.11.10; Remove references to University of California from my copyright notices.
1.10	14-Mar-2003	bouyer	branches: 1.10.2; Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver.
1.9	23-Apr-2002	bouyer	branches: 1.9.4; More copyright fixes, pointed out by Thomas. Thanks !
1.8	21-Oct-2001	thorpej	Make the various timing, etc. tables const, and add the __unused__ attribute to them, just in case something other than the pciide driver proper needs to pull in the header.
1.7	16-Oct-2001	tron	Add missing timings for Ultra DMA modes 3 to 5 on SiS hostadapters. Problem pointed out by IWAMOTO Toshihiro on "tech-kern@netbsd.org".
1.6	15-May-2000	bouyer	branches: 1.6.6; 1.6.8; Sync my copyrigth notice.
1.5	04-Dec-1998	drochner	branches: 1.5.4; 1.5.10; The chipset manual doesn't tell the truth about the channel enable register bits. Replace by experimental result.
1.4	03-Dec-1998	bouyer	Correct definition for SIS_TIM_UDMA_EN macro. Ultra-DMA should now work ...
1.3	03-Dec-1998	bouyer	Add URLs of documentation source (from Matthias Drochner).
1.2	03-Dec-1998	bouyer	Correct a few bogons in the SiS chip initialisation.
1.1	21-Nov-1998	drochner	headers for new hardware support in pciide.c. from Manuel.
1.5.10.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.4.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.6.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.6.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.6.6.1	22-Oct-2001	nathanw	Catch up to -current.
1.9.4.1	16-Jun-2003	grant	Apply patch (requested by bouyer in ticket #1234): Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver.
1.10.2.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.10.2.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.11.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.8.1	29-Apr-2005	kent	sync with -current
1.11.6.1	05-Jul-2005	riz	Pull up revision 1.13 via patch (requested by bouyer in ticket #2032): Add quick for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.12.2.1	26-Jun-2005	tron	Pull up revision 1.13 (requested by bouyer in ticket #498): Add quirk for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.13.2.1	21-Jan-2008	yamt	sync with head
1.14.60.1	02-Jan-2008	bouyer	Sync with HEAD
1.14.56.1	26-Dec-2007	ad	Sync with head.
1.14.52.1	18-Feb-2008	mjf	Sync with HEAD.
1.14.46.1	09-Jan-2008	matt	sync with HEAD
1.15.10.1	11-Mar-2010	yamt	sync with head
1.6	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.5	25-Dec-2007	perry	branches: 1.5.6; 1.5.8; 1.5.10; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.4	11-Jun-2007	nisimura	branches: 1.4.8; 1.4.14; 1.4.16; 1.4.20; - fix errors in register number and bit field definitions. - put a bit better comment about how LEGIRQ bit works.
1.3	11-Dec-2005	christos	branches: 1.3.30; 1.3.32; merge ktrace-lwp.
1.2	27-Feb-2005	perry	branches: 1.2.4; nuke trailing whitespace
1.1	03-Apr-2002	thorpej	branches: 1.1.2; 1.1.8; 1.1.14; 1.1.22; 1.1.24; Add bus master DMA support for the Symphony Labs 82C105 PCI IDE controller. This part is also found in the Winbond 83C553 Southbrige.
1.1.24.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.22.1	29-Apr-2005	kent	sync with -current
1.1.14.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.1.8.1	03-Apr-2002	jdolecek	file pciide_sl82c105_reg.h was added on branch kqueue on 2002-06-23 17:48:02 +0000
1.1.2.2	17-Apr-2002	nathanw	Catch up to -current.
1.1.2.1	03-Apr-2002	nathanw	file pciide_sl82c105_reg.h was added on branch nathanw_sa on 2002-04-17 00:06:06 +0000
1.2.4.2	21-Jan-2008	yamt	sync with head
1.2.4.1	03-Sep-2007	yamt	sync with head.
1.3.32.1	11-Jul-2007	mjf	Sync with head.
1.3.30.1	15-Jul-2007	ad	Sync with head.
1.4.20.1	02-Jan-2008	bouyer	Sync with HEAD
1.4.16.1	26-Dec-2007	ad	Sync with head.
1.4.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.8.1	09-Jan-2008	matt	sync with HEAD
1.5.10.1	16-May-2008	yamt	sync with head.
1.5.8.1	18-May-2008	yamt	sync with head.
1.5.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.2	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.1	06-Mar-2006	bouyer	branches: 1.1.4; 1.1.8; 1.1.10; 1.1.16; 1.1.22; Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.1.22.2	09-Sep-2006	rpaulo	sync with head
1.1.22.1	06-Mar-2006	rpaulo	file pciide_svwsata_reg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.16.2	21-Jun-2006	yamt	sync with head.
1.1.16.1	06-Mar-2006	yamt	file pciide_svwsata_reg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:06 +0000
1.1.10.2	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): share/man/man4/pciide.4: revision 1.60 share/man/man4/Makefile: revision 1.376 via patch sys/dev/pci/svwsata.c: revision 1.1 sys/dev/pci/files.pci: revision 1.243 sys/arch/i386/conf/INSTALL: revision 1.281 sys/dev/pci/pcidevs: revision 1.771 via patch sys/dev/pci/pciidevar.h: revision 1.33 sys/arch/i386/conf/XEN0: revision 1.34 sys/arch/i386/conf/GENERIC: revision 1.732 sys/dev/pci/pciide_svwsata_reg.h: revision 1.1 Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.1.10.1	06-Mar-2006	tron	file pciide_svwsata_reg.h was added on branch netbsd-3 on 2006-05-04 12:12:08 +0000
1.1.8.2	22-Apr-2006	simonb	Sync with head.
1.1.8.1	06-Mar-2006	simonb	file pciide_svwsata_reg.h was added on branch simonb-timecounters on 2006-04-22 11:39:15 +0000
1.1.4.2	13-Mar-2006	yamt	sync with head.
1.1.4.1	06-Mar-2006	yamt	file pciide_svwsata_reg.h was added on branch yamt-pdpolicy on 2006-03-13 09:07:27 +0000
1.11	24-Nov-2006	wiz	s/independant/independent/, from Zafer.
1.10	11-Dec-2005	christos	branches: 1.10.20; 1.10.22; merge ktrace-lwp.
1.9	27-Feb-2005	perry	branches: 1.9.4; nuke trailing whitespace
1.8	15-Feb-2005	briggs	Move the definition of NIDEDMA_TABLES from pciidereg.h to be closer to its only user in pciide_common.c. Also redefine NIDEDMA_TABLES to match the max DMA transfer size specified in the call to bus_dmamap_create() (IDEDMA_BYTE_COUNT_MAX instead of MAXPHYS). The macro is also redefined to handle devices that have a PAGE_SIZE greater than sc_dma_maxsegsz (buggy revision of satalink 3112 on ibm4xx).
1.7	27-Nov-2003	fvdl	branches: 1.7.8; 1.7.10; There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.6	14-Nov-2000	thorpej	branches: 1.6.24; NBPG -> PAGE_SIZE
1.5	09-Mar-2000	soren	Move PCIIDE_CHANNEL_NAME macro to pciidereg.h.
1.4	02-Feb-1999	bouyer	branches: 1.4.2; 1.4.8; Support for Acerlab M5229 IDE controller. Thanks to Thilo Manske for testing the code, and to Takahiro Kambe who run several tests and finally found the bug by himself :)
1.3	12-Oct-1998	bouyer	Merge bouyer-ide
1.2	04-Mar-1998	cgd	branches: 1.2.2; add def'ns for PCI IDE bus-master DMA interface recognition and register mapping.
1.1	04-Mar-1998	cgd	PCI IDE glue. Right now, just glues 'wdc's to PCI IDE controller channels. Eventually should do things like support PCI IDE DMA (it _DOES NOT_ do that now).
1.2.2.2	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.2.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.4.8.2	22-Nov-2000	bouyer	Sync with HEAD.
1.4.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.6.24.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.24.4	15-Feb-2005	skrll	Sync with HEAD.
1.6.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.24.1	03-Aug-2004	skrll	Sync with HEAD
1.7.10.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.7.8.1	29-Apr-2005	kent	sync with -current
1.9.4.1	30-Dec-2006	yamt	sync with head.
1.10.22.1	10-Dec-2006	yamt	sync with head.
1.10.20.1	12-Jan-2007	ad	Sync with head.
1.51	26-Dec-2021	andvar	fix various typos, mainly in comments.
1.50	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.49	22-Oct-2017	jdolecek	branches: 1.49.2; do not share queue between the non-indepedant channels; instead make sure only one of the channels is ever active on the same controller fixes PR kern/52606 by Martin Husemann, thanks for report and testing
1.48	19-Oct-2017	jdolecek	replace the chek for the shared channel of cmdide(4) a flag of the product array, rather than switch inside attach routine XXX judging from product name, Silicon Image 0680 might be newer than 0649 XXX and hence have actually independant channels, but I don't have the hw XXX so keeping as-is no functional change, just to improve visibility in course of fixing PR kern/52606
1.47	24-Aug-2015	pooka	add some _KERNEL_OPT as the finishing touch
1.46	31-Jul-2012	bouyer	branches: 1.46.2; 1.46.16; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.45	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.44	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.43	04-Apr-2011	dyoung	branches: 1.43.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.42	06-Nov-2010	jakllsch	branches: 1.42.2; Rework pciide(4) detachment to take the legacy interrupt mapping into consideration and avoid future code duplication. Ports wanting to enable detachment of controllers with compatibility-mapped channels will need to supply a pciide_machdep_compat_intr_disestablish() function.
1.41	05-Nov-2010	jakllsch	Pave the way for detachment of pciide(4)-family ATA controllers.
1.40	14-Nov-2009	cegger	branches: 1.40.2; 1.40.4; include <sys/device_if.h> for device_t. There is no use of struct device.
1.39	05-Oct-2009	dyoung	It's a big job to write detachment hooks for pciide(4) and all of its derivatives, and I haven't the time myself. In struct pciide_product_desc, provide a tiny bit of commented-out code to guide an enterprising developer who takes on the immense task.
1.38	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.37	15-Mar-2009	cegger	ansify function definitions
1.36	04-Jan-2008	joerg	branches: 1.36.10; 1.36.18; 1.36.24; Allow saving of up to 4 PCI registers in the IDE code.
1.35	17-Oct-2006	itohy	branches: 1.35.24; 1.35.30; 1.35.38; Make pciide(4)-only configurations (without other DMA-capable driver) compile.
1.34	17-Jun-2006	jmcneill	branches: 1.34.4; 1.34.6; Restore IDE and UDMA timings on resume for the PIIX4.
1.33	06-Mar-2006	bouyer	branches: 1.33.6; 1.33.8; Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.32	24-Dec-2005	perry	branches: 1.32.4; 1.32.6; 1.32.8; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.31	11-Dec-2005	christos	merge ktrace-lwp.
1.30	27-Feb-2005	perry	branches: 1.30.2; 1.30.4; nuke trailing whitespace
1.29	11-Feb-2005	rearnsha	Add support for the Artisea device operating in DPA mode. Approved by briggs.
1.28	04-Feb-2005	perry	de-__P
1.27	26-Jan-2005	jmcneill	Add powerhooks for PIIX IDE driver.
1.26	24-Nov-2004	bouyer	branches: 1.26.4; 1.26.6; Move bus_dma setup out of pciide_dma_init() in a new function, pciide_dma_dmamap_setup(), for the benefit of drivers that needs special registers setup in dmainit().
1.25	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.24	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.23	19-Aug-2004	thorpej	Protect against multiple inclusion.
1.22	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.21	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.20	03-Jan-2004	thorpej	branches: 1.20.4; Rename "struct channel_softc" to "struct wdc_channel".
1.19	19-Dec-2003	thorpej	While legacy PCI IDE controllers have only 2 channels, it is quite common for modern PCI SATA controllers to have more. Define a new PCIIDE_MAX_CHANNELS which to specify the maximum number of channels we support, and use it to allocate channel resources.
1.18	19-Dec-2003	thorpej	Some controllers must have certain bits always be set in the IDEDMA_CMD register for proper operation. Add a prototype field in the pciide_channel for this register and use it as necessary when writing the IDEDMA_CMD register.
1.17	19-Dec-2003	thorpej	Const poison the channel name.
1.16	19-Dec-2003	thorpej	Move the PCIIDE_OPTIONS_* constants into pciidevar.h
1.15	17-Dec-2003	thorpej	Add BA5 space tag/handles for Silicon Image SATALink.
1.14	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.13	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.12	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.11	23-Sep-2003	mycroft	GC a structure element.
1.10	20-Mar-2003	thorpej	branches: 1.10.2; Add support for the Silicon Image SATALink 3112 Serial ATA controller.
1.9	14-Mar-2003	bouyer	Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver.
1.8	24-Jan-2003	thorpej	Add support for the NVIDIA nForce ATA100 and nForce2 ATA133 IDE controllers, which are more-or-less compatible with the AMD controllers. XXX Need to determine the correct timing value for the nForce2 XXX at Ultra133, so we cap it at Ultra100, for now.
1.7	08-Jun-2001	simonb	branches: 1.7.16; For ports that wire up pciide in compatibility mode, have them define __HAVE_PCIIDE_MACHDEP_COMPAT_INTR_ESTABLISH in pci_machdep.h and pciide_map_compat_intr() only calls pciide_machdep_compat_intr_establish() if that preprocessor define exists. Ports that don't need to do this no longer need to supply a dummy function.
1.6	12-Jan-2001	bouyer	branches: 1.6.2; Fix bug which affects pciide controllers in native mode (found by Paul Kranenburg, many thanks !): the control register I/O is 4 byte long although only one is used, but the control register is at offset 2, and not 0 as expected by IC code. Use bus_space_subregion() to get a handle which points to the control register, and is one byte long.
1.5	06-Jun-2000	thorpej	branches: 1.5.2; Actually program the DMA mode of the drives into the Cypress controller. Fixes a long-standing problem where IDE DMA wasn't working on the AlphaPC 164SX.
1.4	01-Apr-2000	bouyer	branches: 1.4.2; - DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op if an IRQ was not detected, unless the force flag was given. Use this to detect if the IRQ was for us (closer to shared IRQ for controllers which don't have their own IRQ handler in pciide.c) and to poll for DMA xfer. Also makes the timeout recovery code simpler. - ATAPI cleanup: don't call controller-specific functions from atapiconf.c (wdc_), so that it's possible to attach an atapibus to something else than a wdc/pciide (Hi Lennart :). Overload struct scsi_adapter with struct atapi_adapter, defined as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points to an scsi_adapter, atapi code casts it to atapi_adapter if needed. Move atapi_softc to atapiconf.h so that it can be used by the underlying controller code (e.g. atapi_wdc.c). Add an atapi-specific callback atapi_probedev(), which probe a drive in a controller-specific way, allocate the sc_link and fills in the ataparams if needed. It then calls atapi_probedev() (from atapiconf.c) to do the generic initialisations and attach the device. - While I'm there merge and centralise the state definitions in atavar.h. It should now be possible to use a common ata/atapi routine to set the drive's modes (will do later).
1.3	13-Nov-1999	soren	Export softc.
1.2	12-Oct-1998	bouyer	branches: 1.2.6; 1.2.12; 1.2.14; 1.2.18; Merge bouyer-ide
1.1	04-Mar-1998	cgd	branches: 1.1.2; PCI IDE glue. Right now, just glues 'wdc's to PCI IDE controller channels. Eventually should do things like support PCI IDE DMA (it _DOES NOT_ do that now).
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.2.18.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.2.14.1	15-Nov-1999	fvdl	Sync with -current
1.2.12.2	15-Jan-2001	bouyer	req_sense_length is dead.
1.2.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.4.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.5.2.1	04-Feb-2001	he	Pull up revision 1.6 (requested by bouyer): Fix bug which affects pciide controllers in native mode. Main effect was a non-functionnal IDE controller on some sparc64 and macppc when booted from IDE disk.
1.6.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.7.16.2	12-Jul-2004	he	Pull up revision 1.10 (via patch, requested by grant in ticket #1720): Add support for SATALink 3112.
1.7.16.1	16-Jun-2003	grant	Apply patch (requested by bouyer in ticket #1234): Rework SiS support: more controller supported (up to Ultra/133) and better support for the older ones. Information for this work extracted from Soeren Schmidt's FreeBSD driver.
1.10.2.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.10.2.7	15-Feb-2005	skrll	Sync with HEAD.
1.10.2.6	04-Feb-2005	skrll	Sync with HEAD.
1.10.2.5	29-Nov-2004	skrll	Sync with HEAD.
1.10.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.10.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.10.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.10.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.4.1	16-Mar-2005	tron	Pull up revision 1.26 (requested by bouyer in ticket #1016): Move bus_dma setup out of pciide_dma_init() in a new function, pciide_dma_dmamap_setup(), for the benefit of drivers that needs special registers setup in dmainit().
1.26.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.26.6.1	12-Feb-2005	yamt	sync with head.
1.26.4.1	29-Apr-2005	kent	sync with -current
1.30.4.3	21-Jan-2008	yamt	sync with head
1.30.4.2	30-Dec-2006	yamt	sync with head.
1.30.4.1	21-Jun-2006	yamt	sync with head.
1.30.2.1	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): share/man/man4/pciide.4: revision 1.60 share/man/man4/Makefile: revision 1.376 via patch sys/dev/pci/svwsata.c: revision 1.1 sys/dev/pci/files.pci: revision 1.243 sys/arch/i386/conf/INSTALL: revision 1.281 sys/dev/pci/pcidevs: revision 1.771 via patch sys/dev/pci/pciidevar.h: revision 1.33 sys/arch/i386/conf/XEN0: revision 1.34 sys/arch/i386/conf/GENERIC: revision 1.732 sys/dev/pci/pciide_svwsata_reg.h: revision 1.1 Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.32.8.2	26-Jun-2006	yamt	sync with head.
1.32.8.1	13-Mar-2006	yamt	sync with head.
1.32.6.1	22-Apr-2006	simonb	Sync with head.
1.32.4.1	09-Sep-2006	rpaulo	sync with head
1.33.8.1	13-Jul-2006	gdamore	Merge from HEAD.
1.33.6.1	19-Jun-2006	chap	Sync with head.
1.34.6.1	22-Oct-2006	yamt	sync with head
1.34.4.1	18-Nov-2006	ad	Sync with head.
1.35.38.1	08-Jan-2008	bouyer	Sync with HEAD
1.35.30.1	18-Feb-2008	mjf	Sync with HEAD.
1.35.24.1	09-Jan-2008	matt	sync with HEAD
1.36.24.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.36.18.1	28-Apr-2009	skrll	Sync with HEAD.
1.36.10.3	11-Mar-2010	yamt	sync with head
1.36.10.2	16-May-2009	yamt	sync with head
1.36.10.1	04-May-2009	yamt	sync with head.
1.40.4.2	21-Apr-2011	rmind	sync with head
1.40.4.1	05-Mar-2011	rmind	sync with head
1.40.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.42.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.43.4.1	30-Oct-2012	yamt	sync with head
1.46.16.1	22-Sep-2015	skrll	Sync with HEAD
1.46.2.1	03-Dec-2017	jdolecek	update from HEAD
1.49.2.1	22-Apr-2018	pgoyette	Sync with HEAD
1.5	24-Sep-2016	mrg	introduce a better pci_drvname() and PCI_IOC_DRVNAME as pciio_drvnameonbus() and PCI_IOC_DRVNAMEONBUS. the new ones also take a (autoconf) PCI bus number, which allows lookups for any device on any pci bus node. use this in pcictl which current reports the wrong values sometimes. up next: use these in libpciaccess.
1.4	25-Jul-2014	mrg	branches: 1.4.4; 1.4.8; introduce a new pci_drvname(3) to libpci. implement it in the backend via PCI_IOC_DRVNAME. update manual and set lists (and remove a couple of doubled entries.)o this will be used in libpciaccess() to implement the has_kernel_driver() method.
1.3	06-Jun-2009	cegger	branches: 1.3.22; 1.3.36; typo in comment: autconfiguration -> autoconfiguration
1.2	14-Sep-2001	simonb	branches: 1.2.2; 1.2.4; 1.2.8; 1.2.126; 1.2.142; Fix tyop.
1.1	13-Sep-2001	thorpej	Add an ioctl interface to the PCI bus. Add ioctls to read/write PCI configuration space registers, and to fetch bus info.
1.2.142.1	23-Jul-2009	jym	Sync with HEAD.
1.2.126.1	20-Jun-2009	yamt	sync with head
1.2.8.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.8.1	14-Sep-2001	thorpej	file pciio.h was added on branch kqueue on 2002-01-10 19:57:03 +0000
1.2.4.2	01-Oct-2001	fvdl	Catch up with -current.
1.2.4.1	14-Sep-2001	fvdl	file pciio.h was added on branch thorpej-devvp on 2001-10-01 12:46:03 +0000
1.2.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.2.2.1	14-Sep-2001	nathanw	file pciio.h was added on branch nathanw_sa on 2001-09-21 22:36:02 +0000
1.3.36.1	10-Aug-2014	tls	Rebase.
1.3.22.2	03-Dec-2017	jdolecek	update from HEAD
1.3.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.4.8.1	04-Nov-2016	pgoyette	Sync with HEAD
1.4.4.1	05-Oct-2016	skrll	Sync with HEAD
1.172	31-Dec-2024	skrll	Use C99 types
1.171	14-Jun-2024	msaitoh	branches: 1.171.2; PCI_CLASS_MASK: Use unsigned to avoid undefined behavior. Found by kUBSan.
1.170	19-Apr-2024	andvar	s/Resorce/Resource/ in comment and log message.
1.169	02-Feb-2024	andvar	s/Staus/Status/ in comments.
1.168	17-Oct-2022	mrg	branches: 1.168.2; add pcie capability and read request size linux compat, some pci root support implement support for: - pcie_capability_read_dword() - pcie_capability_read_word() - pcie_capability_write_dword() - pcie_capability_write_word() - pcie_get_readrq() - pcie_set_readrq() implement the "struct pci_dev" bus->self member by creating a minimal fake "struct pci_dev" for the pci bus itself. this is kind of gross. it checks that the current device's parent is a netbsd "pci" device, and that it has a (grand) parent "ppb" device, and then fills in the fake device based upon the pci and ppb devices. add some PCIE_LCSR2_TGT_LSPEED encodings, and map them to linux names. map several other PCIE_LCSR and PCIE_LCAP names. uncomment several pcie code segments in radeon and amdgpu. (not sure that we can test the amdgpu_si.c change, as we use the radeon version and the amdgpu version hangs on the one machine i have.) tested on amdgpu (RX550) and radeon (7750 & 3650). ok @riastradh
1.167	01-Oct-2022	rin	No need to mangle argument of macro.
1.166	20-Sep-2022	mrg	fill out more of the linux pci API compat - implement pcie_get_speed_cap(), pcie_bandwidth_available(), and pci_is_root_bus(). - expand "enum pci_bus_speed" to add PCIe 5.x and 6.x speeds. - add "enum pcie_link_width". - add defines for PCIE_LCSR_LINKSPEED (PCIe generation) and PCIE_LCSR_NLW (negotiated lane width) to pcireg.h - enable amdgpu_device_get_pcie_info() code now it works. ok riastradh@
1.165	31-Jan-2022	msaitoh	Decode SCSI programming interface. Whitespace fix.
1.164	31-Jan-2022	msaitoh	Add three new extended capabilities: - Physical Layer 32.0 GT/s - Alternate Protocol - System Firmware Intermediary
1.163	01-Jan-2022	msaitoh	Print Physical Layer 16.0 GT/s and Lane Margining at the Receiver extended cap. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Rename pcie_link_compliance_preset_deemphasis to pcie_link_preset_preshoot_deemphasis because the table is referenced from multiple places. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Rename PCI_EXTCAP_PYSLAY_16GT to PCI_EXTCAP_PL16G.
1.162	28-Dec-2021	msaitoh	Decode link control2's Compliance Preset/De-emphasis more. Fix typo.
1.161	10-Oct-2021	msaitoh	Use PCI-SIG official acronyms: - RP stands for Root Port. - RC stands for Root Complex. - RCIEP stands for Root Complex Integrated End Point.
1.160	10-Oct-2021	msaitoh	Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet)
1.159	10-Oct-2021	msaitoh	Whitespace fix. No functional change.
1.158	09-Sep-2021	mrg	fix part of the previous: Link Capability Max Speed is a value not a bitmask. pointed out by msaitoh@.
1.157	09-Sep-2021	mrg	add some bits in the pci Link Capabilities Register, and also the similar set in the Link Capabilities 2 Register.
1.156	17-Aug-2021	andvar	fix multiplei repetitive typos in comments, messages and documentation. mainly because copy paste code big amount of files are affected.
1.155	24-Jun-2021	thorpej	From the "Should have been done years ago" department, add some macros describing the format of PCI Type 0 and Type 1 Configuration Cycle addresses.
1.154	15-May-2021	jakllsch	Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE() Should fix PR 56176.
1.153	28-Dec-2020	skrll	branches: 1.153.4; 1.153.6; Add second space indentation for bit field values for Enhanced Allocation capabilities. From thorpej.
1.152	28-Dec-2020	skrll	Trailing whitespace
1.151	18-Feb-2020	msaitoh	branches: 1.151.6; Add comment.
1.150	25-Jan-2020	msaitoh	Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet.
1.149	22-Jan-2020	msaitoh	Remove unused shift and mask definitions.
1.148	11-Dec-2019	msaitoh	branches: 1.148.2; s/enalbe/enable/
1.147	01-Mar-2019	msaitoh	branches: 1.147.4; - Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.146	30-Nov-2018	msaitoh	Add new PCIE_HAS_LINKREGS(pcie_devtype) and use it. No functional change.
1.145	30-Nov-2018	msaitoh	Add new macro PCIE_HAS_ROOTREGS(pcie_devtype) and use it. No functional change.
1.144	28-Nov-2018	msaitoh	The register offset of the mask and pending register is depend on the 64bit address capable bit, so fix the definition of PCI MSI vector mask and pending register. This problem was not a real bug because PCI_MSI{MASK,PENDING} were not used from anywhere.
1.143	05-Nov-2018	msaitoh	Decode PCI Enhanced Allocation.
1.142	03-Oct-2018	msaitoh	- Don't print TPH requester's ST Table Size if the ST table location field is not PCI_TPH_REQ_STTBLLOC_TPHREQ because the size field is only applicaple for PCI_TPH_REQ_STTBLLOC_TPHREQ case. - Add comment.
1.141	27-Sep-2018	msaitoh	Root Complex Event Collector Bus Number Association ECN. - If capability version is 2 (or greater), decode RCEC Associated Bus Numbers register.
1.140	12-Sep-2018	msaitoh	Add ATS Relaxed Ordering supported bit described in Address Translation Relaxed Ordering ECN.
1.139	02-Jul-2018	msaitoh	VGA 16 bit decode bit is not bit 3 but bit 4.
1.138	09-May-2018	msaitoh	branches: 1.138.2; Fix typo. s/TPL/TLP/
1.137	01-Feb-2018	msaitoh	branches: 1.137.2; - Add PCie Link Activation ECN. - Use macro. - KNF.
1.136	18-Dec-2017	msaitoh	Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified.
1.135	19-Oct-2017	msaitoh	Fix a bug that the TPH ST table is decoded even if it's not in the TPH Requester extended capability structure.
1.134	10-Oct-2017	msaitoh	Decode IOMMU capability of PCI secure device capability. From "AMD I/O Virtualization Technology(IOMMU) Specification (#48882) Revision 3.00".
1.133	05-Oct-2017	msaitoh	- Add PCI_MAPREG_ROM_ADDR_MASK macro and PCI_MAPREG_ROM_ADDR() macro. - print PCI_MAPREG_ROM_ENABLE bit. - Decode Expansion ROM Validation ECN. - Add Native PCIe Enclosure Management ECN's extended capability type. Not decoded yet.
1.132	13-Jul-2017	msaitoh	- Official shortname of LN Requester is LNR, so change PCI_EXTCAP_LN_REQ to PCI_EXTCAP_LNR - Use macro.
1.131	15-Jun-2017	msaitoh	Fix a bug that LTR's latency in L1 PM Substates capability and Latency Tolerance Reporting capability isn't printed correctly.
1.130	29-May-2017	msaitoh	branches: 1.130.2; Print MSI Message data in 32bits when the Extended Message Data Capable bit is set.
1.129	24-May-2017	msaitoh	Decode TPH Requester Control register.
1.128	21-Apr-2017	msaitoh	Add Flattening Portal Bridge capability ID and Hierarchy ID extended capability ID.
1.127	20-Apr-2017	msaitoh	Add Downstream Port Containment (DPC) ECN and Enhanced DPC(eDPC) ECN.
1.126	17-Apr-2017	msaitoh	Use macro. No functional change.
1.125	28-Mar-2017	msaitoh	Sort & indent. No functional change.
1.124	28-Mar-2017	msaitoh	Indent. No functional change.
1.123	28-Mar-2017	msaitoh	Lowercase for consistency. No functional change.
1.122	14-Mar-2017	msaitoh	Add two new PCI classes: - processing accelerators - non-essential instrumentation
1.121	27-Feb-2017	msaitoh	Decode AGP capability.
1.120	15-Feb-2017	msaitoh	- Print Data Select and Data Scale in pci_conf_print_pcipm_cap(). - The Message Data register of MSI cap is not 32bit but 16bit. - When the PCIE_LCAP_MAX_SPEED bitfiled is 0, it means it supports 2.5GT/s only. - Print link de-emphasis value by "-X dB". - Print Completion Timeout Ranges Supported filed with alphabets. - Print TPH Completer Supported fileld's meaning. - Print PCIE_DCAP2_MAX_EETLP correctly. 0 means 4 End-End TLP Prefixes. - If the Supported Link Speed Vector is 0, the Link Capability 2 register is not implemented. Don't decode LCAP2 when the vector is 0. - The ACS's Egress Control Vector is 32bit, so print with 0x%08x. - Print SR-IOV's device ID. - Use __SHIFTOUT() to avoid using magic number. - Prefix "0x" for hexadecimal value.
1.119	28-Dec-2016	msaitoh	branches: 1.119.2; The Power Controller Control bit (PCIE_SLCSR_PCC) in the Slot Control & Status Register is 0 on power on. Print "Power <on\|off>" instead of "<on\|off>".
1.118	27-Dec-2016	msaitoh	Fix PASID Control Register.
1.117	31-Oct-2016	msaitoh	Decode Resizable BAR.
1.116	20-Oct-2016	msaitoh	- pci_conf_print_pwrbdgt_base_power(): From 0xf3 to 0xff of the Base power are reserved above 300W (PCI 3.0 Errata). - Emergency Power Reduction mechanism with PWRBRK signal ECN. - Extended Message Data for MSI ECN. - Fix typo in comment.
1.115	19-Oct-2016	msaitoh	Add VF Resizable BARs ECN.
1.114	17-Aug-2016	msaitoh	Add Dynamic Power Allocation (DPA) ECN support.
1.113	11-May-2016	msaitoh	branches: 1.113.2; Add Precision Time Management (PTM) ECN.
1.112	18-Nov-2015	msaitoh	- Add the Auto Slot Power Limit Disable bit in Slot Control register and the Completion Timeout Prefix/Header Log Capable bit in the AER capability and control register (ECN: Downstream Port Containment (DPC)). - Add the Poisoned TLP Egress Block bit (ECN: Enhanced DPC). - Update Link Capabilities 2 register and Link Control 3 register (ECN: Separate Refclk Independent SSC Architecture (SRIS)) - ECN: Readiness Notifications (RN) - Add the Retimer Presence Detect Supported bit in the Link Capabilities 2 register and the Retimer Presence Detected bit in the Link Status 2 register (ECN: Extension Devices)
1.111	17-Nov-2015	msaitoh	- ARI's function group is not bit 32-24 but 22-20. - Add the Structure Length field in AF capability register. - Add Enhanced Allocation extended capability ID (ECN: Enhanced Allocation (EA) for Memory and I/O Resources). - Add LN System CLS (ECN: Lightweight Notification (LN) Protocol). - Add ST Upper and Lower bit definitions (ECN: TLP Processiong Hints). - Add the Global Invalidate bit in the ATS capability register and the PRG Response PASID Required bit in the Page Request status register (ECN: PASID Translation) - Decode ASPM support bit more (ECN: ASPM Optionally) - Use __BITS()
1.110	17-Nov-2015	msaitoh	No functional change: - Add comments. - Remove obsolete comment. - Move definitions to better location. - Rename bit definition. - KNF. - Indent.
1.109	16-Nov-2015	msaitoh	Define PCIE_XCAP_{VER,TYPE}(x) and use them.
1.108	12-Nov-2015	msaitoh	- Restore pci_subr.c rev. 1.135's change in pci_conf_print_caplist(). As wrote in the comment, HyperTransport capability appears multiple times. pci_conf_cap() reruns only the first entry, so it can't be used here. - Try to decode HyperTransport capability. Currently, the capability type of each HyperTransport capability is printed and only the MSI Mapping capability is decoded. - Style change.
1.107	12-Nov-2015	msaitoh	PCI_HT_CAP() is right shifted value, so PCI_HT_CAP_* should not use __SHIFTIN(). This change fixes a bug that HyperTransport system misunderstand whether MSI/MSI-X can be used or not.
1.106	30-Oct-2015	msaitoh	- Move PCI_INTRSTR_LEN from pcireg.h to pcivar.h. - In PCI-X cap, print 2nd bus's PCI-X mode, error protection type, Max clock frequency and Max clock period. - In SATA cap, print register location correctly. - In Virtual Channel cap, print reference clock with "ns". - In Root Complex Link Declaration, print Link Entry number.
1.105	21-Oct-2015	msaitoh	Decode SATA Capability and Multicast Extendeded Capability.
1.104	02-Oct-2015	msaitoh	PCI Extended Configuration stuff written by nonaka@: - Add PCI Extended Configuration Space support into x86. - Check register offset of pci_conf_read() in MD part. It returns (pcireg_t)-1 if it isn't accessible. - Decode Extended Capability in PCI Extended Configuration Space. Currently the following extended capabilities are decoded: - Advanced Error Reporting - Virtual Channel - Device Serial Number - Power Budgeting - Root Complex Link Declaration - Root Complex Event Collector Association - Access Control Services - Alternative Routing-ID Interpretation - Address Translation Services - Single Root IO Virtualization - Page Request - TPH Requester - Latency Tolerance Reporting - Secondary PCI Express - Process Address Space ID - LN Requester - L1 PM Substates The following extended capabilities are not decoded yet: - Root Complex Internal Link Control - Multi-Function Virtual Channel - RCRB Header - Vendor Unique - Configuration Access Correction - Multiple Root IO Virtualization - Multicast - Resizable BAR - Dynamic Power Allocation - Protocol Multiplexing - Downstream Port Containment - Precision Time Management - M-PCIe - Function Reading Status Queueing - Readiness Time Reporting - Designated Vendor-Specific
1.103	27-Jul-2015	msaitoh	Add NVMe.
1.102	27-Apr-2015	knakahara	add x86 MD MSI/MSI-X support code.
1.101	23-Feb-2015	knakahara	- add macros to calculate MSI MME and MMC. - add MSI-X table offset macros. - fix MSI-X table entry name.
1.100	24-Nov-2014	msaitoh	branches: 1.100.2; Add PCIe CRS Software Visibility bit.
1.99	23-Oct-2014	msaitoh	Add some HyperTransport related defines. It's required for the MSI.
1.98	23-Oct-2014	msaitoh	Add comment.
1.97	06-Oct-2014	msaitoh	- Add some PCI subclass and interfaces. - The interface of PCI_SUBCLASS_BRIDGE_RACEWAY is not decoded yet. - Fix typo in a message. - Add comment. - Modify comment.
1.96	24-Sep-2014	msaitoh	Rename PCIE_XCAP_VER_* macros to avoid confusion.
1.95	09-Jun-2014	msaitoh	branches: 1.95.2; Add IOMMU and the Root Complex Event Collector.
1.94	30-May-2014	msaitoh	- Add PCI-X capability stuff. - remove extra ':' in pci_conf_print_pcie_cap() - Add comments.
1.93	27-May-2014	msaitoh	- Add some register definition for MSI and MSI-X - print MSI-X capability
1.92	27-May-2014	msaitoh	- Fix incorrect calculation in PCI_MSIX_CTL_TBLSIZE(). - The PCI_MSIX_CTL_TBLSIZE bit field is in N-1, so add +1.
1.91	24-May-2014	msaitoh	Print some PCI Capabilities: - Vendor specific (ID:0x09) - Debugport (ID:0x0a) - Subsystem (ID:0x0d) - PCI Advanced Features (ID:0x13)
1.90	24-May-2014	msaitoh	- Decode the programming interface field in the Class Code register and print it. - Print the cache line size in bytes. - Print the Link Status 2 register itself. - Some bits were not printed if the bit is 0. Always print them using with onoff() macro. - Print more bits. - KNF. - Use macro. - Add comments.
1.89	23-May-2014	msaitoh	- Add some register definitions (subclass, power management, etc.) - Print some information (subclass, power management) - Use macro.
1.88	23-May-2014	msaitoh	PME# clock is not bit 2 but bit 3. Use the macro!
1.87	23-May-2014	msaitoh	No functional change: - sort in PCI capability ID order. - add comments.
1.86	09-May-2014	msaitoh	Print the CRS Software Visibility Enable bit and the Crosslink Supported bit.
1.85	29-Mar-2014	christos	branches: 1.85.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.84	21-Apr-2013	msaitoh	branches: 1.84.4; - Add some PCIe 2.0 or higher capability register definitions. - Rename some registers. - Add comments. - Indent.
1.83	21-Apr-2013	msaitoh	Delete "PCI_" from PCIX and PICE capability registers.
1.82	17-Apr-2013	msaitoh	- Add slot related registers - Add root port related registers - Fix the definition of PCI_PCIE_SLCAP_PSN - Cleanup
1.81	17-Apr-2013	msaitoh	Add PCI_CAP_SUBVENDOR (= 0x0d).
1.80	16-Apr-2013	msaitoh	Decode some PCIe capability register bits.
1.79	16-Apr-2013	msaitoh	Fix a bug that IRQ(MSI) bits in PCIe capability register is incorrectly decoded. The bit field is not 0x4e000000 but 0x3e000000.
1.78	15-Apr-2013	msaitoh	Add some PCIE capability register bit definitions from FreeBSD.
1.77	12-Apr-2013	msaitoh	Add comments for some PCI Express Capability registers.
1.76	03-Mar-2013	msaitoh	Add PCI Express's relax ordering bit.
1.75	20-Oct-2012	matt	Add PCI_SUBCLASS_MASS_STORAGE_NVM
1.74	02-Sep-2012	matt	branches: 1.74.2; Finish devices the bits in PCIE capability LCSR field.
1.73	17-Aug-2011	dyoung	branches: 1.73.2; 1.73.8; 1.73.10; Redefine PCI_MSI_* and PCI_PCIE_* constants in terms of bits(3). Use named constants and more conventional variable names in pci_msi_establish() and pci_msi_disestablish(). Fix a couple of bugs: pci_msi_establish() returned a pointer to the struct intrhand instead of to the struct msi_hdl as it was intended to, and pci_msi_disestablish() did not free(9) the msi_hdl.
1.72	06-Jun-2011	msaitoh	Add two new capabilities(PCI_CAP_SATA and PCI_CAP_PCIAF).
1.71	05-Apr-2011	dyoung	branches: 1.71.2; Use PCI_MAPREG_START instead of the anonymous constant 0x10.
1.70	26-Jan-2011	dyoung	Define masks for subsystem vendor and subsystem ID fields, plus a couple of macros for extracting them.
1.69	10-Jan-2011	jmcneill	branches: 1.69.2; 1.69.4; ppb_fix_pcix changes: - rename to ppb_fix_pcie - support version PCI-E 2.0 - print version and device/port type information - use constants from pcireg.h instead of magic numbers changes: ppb2 at pci0 dev 21 function 0: vendor 0x15ad product 0x07a0 (rev. 0x01) ppb2: unsupported PCI Express version to: ppb2 at pci0 dev 21 function 0: vendor 0x15ad product 0x07a0 (rev. 0x01) ppb2: PCI Express 2.0 <Root Port of PCI-E Root Complex>
1.68	11-Dec-2010	matt	Add some PCI Express definitions, MSI, MSIX, etc.
1.67	20-Mar-2010	dyoung	Add PCI_BAR(n) for the nth Base Address Register.
1.66	26-Feb-2010	dyoung	branches: 1.66.2; Move the definitions for PCI_BAR0, PCI_BAR1, PCI_BAR2, PCI_BAR3, PCI_BAR4, and PCI_BAR5 to pcireg.h for re-use.
1.65	23-Feb-2010	dyoung	Remove unused functions pci_disable_retry() and cardbus_disable_retry().
1.64	12-Feb-2010	msaitoh	- Add some PCIe config registers. - The register at 0x1c is not Control Status register. It contains Control bits only. Rename it.
1.63	01-Feb-2010	msaitoh	branches: 1.63.2; Fix the bug that unaligned access occurs on amd64. It also fixes the bug that error bits aren't cleard because these bits are W2C (in other word, W1C). Reported by Michael van Elst.
1.62	28-Jan-2010	msaitoh	Add some register definitions.
1.61	11-Sep-2009	christos	Add support for the 5100 and newer firmware. Tested also with the 4900. Thanks to everyone for helping and testing.
1.60	17-Aug-2009	jakllsch	A few new PCI register #defines: - size of config and extended config space - EFI ROM code type number - extended capability list register bits
1.59	16-Jan-2009	cegger	rename PCI_PMCSR_PME to PCI_PMCSR_PME_STS. Per request from jmcneill
1.58	16-Jan-2009	cegger	add Power Management flag. From FreeBSD.
1.57	25-Dec-2007	perry	branches: 1.57.6; 1.57.10; 1.57.18; 1.57.20; 1.57.26; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.56	28-Nov-2007	briggs	branches: 1.56.2; 1.56.6; Correct a typo. To create the command/status word, shift the components left, not right.
1.55	12-Nov-2007	joerg	Merge pci_disable_retry function from jmcneill-pm as it is found in various drivers.
1.54	13-May-2007	kent	branches: 1.54.6; 1.54.8; 1.54.12; 1.54.14; Add PCI_SUBCLASS_MULTIMEDIA_HDAUDIO to pcireg.h and azalia(4) uses it
1.53	26-Jan-2007	dyoung	branches: 1.53.2; 1.53.6; 1.53.8; #Define PCI_PMCSR_PME_EN.
1.52	08-Nov-2006	drochner	branches: 1.52.2; 1.52.4; avoid magic number
1.51	17-Jun-2006	christos	branches: 1.51.4; 1.51.6; re-factor the pci powestate api. reviewed by gimpy
1.50	31-May-2006	drochner	branches: 1.50.2; -add 2 subclasses new in rev. 3.0 of the spec, and fix a pasto in another -get power management rev printing right
1.49	01-Mar-2006	gdamore	branches: 1.49.2; 1.49.8; Add pci_find_rom() API as discussed on tech-kern.
1.48	27-Feb-2006	gdamore	Add PCI_MAPREG_TYPE_ROM and allow it to be used with pci_mapreg_map(). Fix to configure (but do not enable) BARs for expansion ROMs. Reviewed by briggs@
1.47	11-Dec-2005	christos	branches: 1.47.2; 1.47.4; 1.47.6; merge ktrace-lwp.
1.46	02-Aug-2004	joda	branches: 1.46.12; bridge memory and lower prefetch memory ranges are only 12 bits wide
1.45	04-Feb-2004	soren	Use the right bits for the AGP version.
1.44	02-Dec-2003	briggs	Configure PCI-Cardbus bridges, too. Patch from KIYOHARA Takashi on current-users.
1.43	21-Oct-2003	thorpej	- Since we access PCI config space as 32-bit words, redefine the PCI-X command register bits offset and shifted appropriatly (PCI-X command makes up the upper 16 bits of the register that holds the PCI-X cap ID and next-cap-pointer). - Define shift counts for max memory read byte count in the command and status registers.
1.42	05-May-2003	fvdl	branches: 1.42.2; Define PCI32_DMA_BOUNCE_THRESHOLD
1.41	20-Apr-2003	fvdl	Add register definitions for the PCI-X capability.
1.40	25-Mar-2003	thorpej	Add PCI VPD access routines. From psi.cz!freza, PR kern/20889.
1.39	21-Sep-2002	drochner	-correct the "MSI" capability -add some new subclasses and capability IDs
1.38	18-Jun-2002	tshiozak	add support for the per-device power management capability. int pci_set_powerstate(pci_chipset_tag_t pc, pcitag_t tag, int newstate) set power state of the device to newstate. int pci_get_powerstate(pci_chipset_tag_t pc, pcitag_t tag) get current power state of the device. In the future, these functions will be used for ACPI support.
1.37	22-Mar-2002	drochner	branches: 1.37.2; iadd subclass codes from the 2.3 spec
1.36	30-Aug-2001	briggs	Rename PCI_MAPREG_TYPE_ROM to PCI_MAPREG_ROM_ENABLE to keep up with pciconf.?
1.35	06-Jul-2001	thorpej	branches: 1.35.2; Add a bunch of Vital Product Data definitions.
1.34	12-Feb-2001	briggs	branches: 1.34.2; Deal with a lack of the proper type of memory. For example, if a bridge only supports 16-bit I/O addresses, do not configure it with addresses having the upper 16-bits non-zero.
1.33	09-Feb-2001	briggs	Add an option (defopt) PCI_NETBSD_CONFIGURE that provides PCI bus configuration (assignment of bus numbers, BARs, timer values, interrupt lines, etc.). The interface must be called from m.d. code prior to probing the bus. It is meant to be called once for each primary (bus == 0) PCI bus in the system. It will configure any busses behind PCI-PCI bridges. Section 9 man page for pci_configure_bus() will come soon. In the meantime, sample usage is in arch/sandpoint/sandpoint/mainbus.c. [ Reviewed by thorpej ]
1.32	07-Oct-2000	cgd	Update for current PCI device class/subclass and capability codes. (also, tweak the I2O subclass string to be "standard" -- the removal of version info didn't extend that far.)
1.31	02-Oct-2000	ad	I2O subclasses currently have nothing to do with the protocol version.
1.30	21-Aug-2000	castor	Add a bunch of macros of the form PCI_<regname>_CODE similar to existing PCI_ID_CODE to create the PCI appropriate register from parameters. Avoid use of 'class' in macro -- it's a C++ reserved keyword.
1.29	18-Jul-2000	soda	add "#define PCI_INTERRUPT_PIN_MAX 0x04"
1.28	09-Jun-2000	soda	branches: 1.28.2; long long constant needs "LL" suffix.
1.27	12-May-2000	jhawk	branches: 1.27.2; Define some PCI power management CSR constants.
1.26	10-May-2000	thorpej	Add support for mapping 64-bit PCI memory space. If the region is mapped in a way that is inaccessible by a 32-bit bus_addr_t, then print a message to that effect and return failure. Original patches by Bill Studenmund, with a few small changes by me.
1.25	08-May-2000	kleink	Add a register offset for the Capability List Pointer in header type 2. XXX Ideally the PCI-Cardbus Bridge header should be restructured to just XXX present standard register definitions, making it ~safe to be included.
1.24	28-Apr-2000	uch	add PCI_MAPREG_PPB_END (PCI-PCI bridge) PCI_MAPREG_PCB_END (PCI-Cardbus bridge)
1.23	25-Jan-2000	drochner	use BUS_SPACE_MAP_PREFETCHABLE instead BUS_SPACE_MAP_CACHEABLE where the PCI BAR bit is referred to
1.22	16-Nov-1999	enami	Make this file compile again; terminate the continuation line with backslash.
1.21	15-Nov-1999	thorpej	Add a macro to generate a class code given class, subclass, and interface. From UCHIYAMA Yasushi's PCI BIOS patches.
1.20	27-Sep-1999	cgd	branches: 1.20.2; 1.20.8; add classes/subclasses new in PCI 2.2. Needs a bit of cleanup, but then, so does everything involving configuration space headers and if i don't get this out of my source tree i'll go insane.
1.19	21-Dec-1998	drochner	add two of the newer register definitions
1.18	07-Nov-1998	drochner	add support for "extended capabilities" (new in PCI spec 2.2)
1.17	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.16	01-Jun-1998	cgd	according to the PCI 2.1 spec, the low _two_ bits of I/O BARs have defined meaning/value other than specifying the address of the region. (lowest bit is 1, meaning I/O space. second-lowest bit is reserved.)
1.15	18-May-1998	cgd	add a PCI_HDRTYPE_TYPE() macro, to get the 'type' portion of the headertype register (i.e., not including the 'multifunction' bit).
1.14	14-Apr-1998	thorpej	Add UDF and 66MHz capable bit definitions for the PCI status register. From Zubin D. Dittia <zubin@clouseau.arl.wustl.edu>, PR #4249.
1.13	11-Apr-1997	cgd	clean up some constants ([A-F]->[a-f] in hex constants
1.12	19-Mar-1997	cgd	fix multi-function device support, add new known classes/subclasses, and clean up class/subclass printing. From brb@brig.com via PR 3359.
1.11	10-Aug-1996	mycroft	Change PCI_MAPREG_{MEM,IO}_SIZE() to use the standard `x & -x' trick to select the lowest bit set. This isn't any more or less valid according to the PCI spec, but it deals with lame devices that don't implement all of the top bits.
1.10	06-Aug-1996	cgd	revert PCI_MAPREG_IO_ADDR_MASK and PCI_MAPREG_IO_SIZE_MASK to their old values, i.e. 0xfffffffe and 0xffffffff respectively. The changed definitions were incorrect, according to the PCI Local Bus Specification (Revision 2.0). Further rationale and a workaround for the broken devices that instigated the change provided in a message to current-users@netbsd.org, dated Mon, 05 Aug 1996 22:06:58 -0400, message ID 16773.839297218@ux2.sp.cs.cmu.edu>.
1.9	26-Jul-1996	mycroft	Add PCI_MAPREG_{IO,MEM}_SIZE(), and use them.
1.8	26-Jul-1996	mycroft	Changes PCI_MAPREG_IO_ADDR_MASK to 0xfffe.
1.7	27-Mar-1996	cgd	branches: 1.7.4; modify these to provide a new, better-specified PCI interface (soon to be documented on mailing lists; eventually in section 9 manual pages), most importantly: (1) support interrupt pin swizzling on non-i386 systems with PCI-PCI bridges (per PPB spec; done, but meaningless, on i386). (2) provide pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register. (3) provide pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (4) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.6	27-Mar-1996	cgd	add definitions for the BIST/Header Type/Latency Timer/Cache Line Size configuration space register, and use it to determine whether or not a given PCI device uses multiple functions.
1.5	04-Mar-1996	cgd	reorganize mapping register definitions
1.4	27-Jul-1995	mycroft	Add stuff for I/O mapping.
1.3	18-Jun-1995	cgd	macros to split out various parts of PCI registers, adjust constants to match. (now, comparisons are comparisons, code doing them doesn't have to mask.) define types for the various parts of the registers' contents, where practical.
1.2	27-Oct-1994	cgd	new RCS ID format.
1.1	09-Aug-1994	mycroft	Add PCI autoconfiguration support.
1.7.4.1	10-Dec-1996	mycroft	From trunk: Deal with devices that do not allow allow of the I/O address bits to be set.
1.20.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.20.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.20.2.2	11-Feb-2001	bouyer	Sync with HEAD.
1.20.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.27.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.28.2.2	25-Oct-2001	he	Pull up revisions 1.31-1.32 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.28.2.1	10-Aug-2000	soda	Pull up to netbsd-1-5 branch Approved by: thorpej - define PCI_INTERRUPT_PIN_MAX and use it instead of magic number. - define I386_PCI_INTERRUPT_LINE_NO_CONNECTION and use it instead of magic number. Revisions pulled up: > cvs rdiff -r1.28 -r1.29 syssrc/sys/dev/pci/pcireg.h > cvs rdiff -r1.14 -r1.15 syssrc/sys/arch/i386/include/pci_machdep.h > cvs rdiff -r1.38 -r1.39 syssrc/sys/arch/i386/pci/pci_machdep.c > cvs rdiff -r1.1 -r1.2 syssrc/sys/arch/i386/pci/opti82c558.c \ > syssrc/sys/arch/i386/pci/sis85c503.c \ > syssrc/sys/arch/i386/pci/via82c586.c
1.34.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.34.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.34.2.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.34.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.34.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.35.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.35.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.35.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.35.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.37.2.1	15-Jul-2002	gehenna	catch up with -current.
1.42.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.42.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.42.2.1	03-Aug-2004	skrll	Sync with HEAD
1.46.12.7	21-Jan-2008	yamt	sync with head
1.46.12.6	07-Dec-2007	yamt	sync with head
1.46.12.5	15-Nov-2007	yamt	sync with head.
1.46.12.4	03-Sep-2007	yamt	sync with head.
1.46.12.3	26-Feb-2007	yamt	sync with head.
1.46.12.2	30-Dec-2006	yamt	sync with head.
1.46.12.1	21-Jun-2006	yamt	sync with head.
1.47.6.2	01-Jun-2006	kardel	Sync with head.
1.47.6.1	22-Apr-2006	simonb	Sync with head.
1.47.4.1	09-Sep-2006	rpaulo	sync with head
1.47.2.1	01-Mar-2006	yamt	sync with head.
1.49.8.1	19-Jun-2006	chap	Sync with head.
1.49.2.1	26-Jun-2006	yamt	sync with head.
1.50.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.51.6.1	10-Dec-2006	yamt	sync with head.
1.51.4.2	01-Feb-2007	ad	Sync with head.
1.51.4.1	18-Nov-2006	ad	Sync with head.
1.52.4.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.52.2.1	30-Jul-2007	liamjfoy	Pull up following revision(s) (requested by dyoung in ticket #739): sys/dev/pci/pccbb.c: revision 1.140 sys/dev/pci/pccbb.c: revision 1.141 sys/dev/pci/pccbb.c: revision 1.142 sys/dev/pci/pccbb.c: revision 1.143 sys/dev/pci/pccbb.c: revision 1.144 sys/dev/pci/pccbbvar.h: revision 1.28 sys/dev/pci/pcireg.h: revision 1.53 sys/dev/pci/pccbb.c: revision 1.139 Name magic numbers PCI_PMCSR_PME_EN and PCI_PMCSR. Fix grammar in comment. From Patrick Welche. Use the right subroutine name for the debug message. Convert the rather long and backslash-ridden DELAY_MS macro to a much shorter static subroutine, delay_ms(). Cosmetic: KNF indentation, curly braces, and argument declarations. sc_pwrcycle is shared between the interrupt handler and Cardbus event thread, so make it volatile. Fix a bug in Cardbus power activation. Most Cardbus bridges supported by pccbb(4) fire a power-cycle interrupt when the power state of a cardslot changes from 'off' to 'on'. TI bridges fire a power-cycle interrupt on both on->off and off->on changes. When pccbb_power() powered-down a cardslot, it did not wait around for the power-cycle interrupt. When pccbb_power() powered-up a cardslot, it did wait for the interrupt. If a pccbb_power(UP) followed a pccbb_power(DOWN) very closely, pccbb_power() used to interpret the power-cycle interrupt for the up->down transition as "power-up complete," read the power-state bit and, finding that power had NOT been activated, complain, "cbb0: power on failed?" Then pccbb_power() exited before power-activation was complete, falsely indicating that the power-activation was complete. After that, a driver attach/enable routine would blithely configure a card that was not fully powered-up. An operator who ran a command such as 'ifconfig rtw0 down up' or 'ifconfig ath0 down up' would read 'cbb0: power on failed?' in the system log, and their NIC would misbehave. This excerpt from a comment in the source should suffice to explain how I fixed the bug, /* * Wait as long as 200ms for a power-cycle interrupt. If * interrupts are enabled, but the socket has already * changed to the desired status, keep waiting for the * interrupt. "Consuming" the interrupt in this way keeps * the interrupt from prematurely waking some subsequent * pccbb_power call. And this explains why this patch will work for Ricoh bridges that do not fire an interrupt on the on->off transition: * XXX Not every bridge interrupts on the ->OFF transition. * XXX That's ok, we will time-out after 200ms. * * XXX The power cycle event will never happen when attaching * XXX a 16-bit card. That's ok, we will time-out after * XXX 200ms. */ M. Warner Losh and Charles M. Hannum provided valuable input on this patch.
1.53.8.1	11-Jul-2007	mjf	Sync with head.
1.53.6.1	27-May-2007	ad	Sync with head.
1.53.2.1	17-May-2007	yamt	sync with head.
1.54.14.3	18-Feb-2008	mjf	Sync with HEAD.
1.54.14.2	08-Dec-2007	mjf	Sync with HEAD.
1.54.14.1	19-Nov-2007	mjf	Sync with HEAD.
1.54.12.1	13-Nov-2007	bouyer	Sync with HEAD
1.54.8.1	09-Jan-2008	matt	sync with HEAD
1.54.6.3	03-Dec-2007	joerg	Sync with HEAD.
1.54.6.2	14-Nov-2007	joerg	Sync with HEAD.
1.54.6.1	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.56.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.56.2.1	26-Dec-2007	ad	Sync with head.
1.57.26.2	15-Feb-2014	matt	sync pcireg.h with HEAD. update if_wm.c and ppb.c accordingly.
1.57.26.1	24-Dec-2011	matt	Pull down latest from -HEAD.
1.57.20.3	09-Nov-2015	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1983): sys/dev/pci/pcidevs: revisions 1.1079, 1.1134, 1.1148-1.1149, 1.1151 sys/dev/pci/pcidevs.h: regen sys/dev/pci/pcidevs_data.h: regen sys/dev/pci/if_bge.c: revisions 1.183-1.185, 1.187, 1.189-1.193, 1.195-1.199, 1.202-1.226, 1.228-1.237, 1.240-1.264, 1.267-1.276, 1.278-1.280, 1.283-1.287 via patch sys/dev/pci/if_bgereg.h: revisions 1.57-1.74, 1.76-1.90 via patch sys/dev/pci/if_bgevar.h: revisions 1.6, 1.10-1.13, 1.15-1.17 via patch sys/dev/pci/if_bnx.c: revisions 1.32, 1.34-1.43, 1.48-1.49, 1.52 sys/dev/pci/if_bnxreg.h: revisions 1.8, 1.11-1.14 sys/dev/pci/if_bnxvar.h: revisions 1.1-1.3 sys/dev/mii/brgphy.c: revisions 1.53-1.63, 1.65-69, 1.72-1.74 via patch sys/dev/mii/brgphyreg.h: revisions 1.5-1.8 sys/dev/mii/miivar.h: revisions 1.61 sys/dev/pci/pcireg.h: patch Sync bge(4) up to if_bge.c rev. 1.287. Sync brgphy(4) up to 1.74. Fix some bugs on bnx(4). Common: - Add device IDs for Broadcom BCM57710, BCM57711(E), BCM57712(E) and BCM57766 (pcidevs only). - Fix BCM5709 PHY detection. - Fix detection of BGEPHYF_FIBER_{MII\|TBI} - Add BCM5708S support in brgphy(4). - Don't use the WIRESPEED function for fiber devices. bge(4): - Add some Fujitsu's device support from Michael Moll. - Add BCM57762 support (PR#46961 from Ryo Onodera). - Add Altima AC1003, APPLE BCM5701, Broadcom BCM5785F. BCM5785G, BCM5787F, BCM5719, BCM5720, BCM57766, BCM57782 and BCM57786. - Fix DMA setting for read/write on conventional PCI bus devices. This bug was added in rev. 1.166. - Fix printing "discarding oversize frame (len=-4)" message and crash by NULL pointer dereferencing. - The BCM5785 is a PCIe chip but does not report PCIe capabilities. Check for this chip explicitely and enable PCIe. Fixes 'firmware handshake timeout'. - Allow disabling interrupt mitigation. - Workaround for BCM5906 silicon bug. When auto-negotiation results in half-duplex operation, excess collision on the ethernet link may cause internal chip delays that may result in subsequent valid frames being dropped due to insufficient receive buffer resources. (FreeBSD: r214219, r214251, r214292). - Allow write DMA to request larger DMA burst size to get better performance on BCM5785. (FreeBSD r213333: OpenBSD 1.294) - Enable TX MAC state machine lockup fix for both BCM5755 or higher and BCM5906. Publicly available data sheet just says it may happen due to corrupted TxMbuf. (FreeBSD r214216) - Follow Broadcom datasheet: Delay 100 microseconds after enabling transmit MAC. Delay 10 microseconds after enabling receive MAC. (FreeBSD r241220) - Insert the completion barrier between register write and the consecutive delay(). It will fix some device timeout problems we have seen before. - Add DELAY(40) after turning on write DMA state machine. - Add some workarounds for 5717 A0 and 5776[56] to be stable. - Check BGE_RXBDFLAG_IPV6 flag for 5717_PLUS case. Note that {tcp,udp}6csum flag is currently not added in the capability. - Add delay after clearing BGE_MACMODE_TBI_SEND_CFGS for the link checking. - Do not touch the jumbo replenish threshold register on chips that do not have jumbo support. - Wait for the bootcode to complete initialization for 5717 and newer devices. - 5718 and 57785 document say we should wait 100us in init. - Fix a bug that chips which have BCM5906 ASIC touch GPIO wrongly. - Fix the setting of Tx Random Backoff Register. - Check the hardware config words and print them. - Set BGE_MISC_CTL's byte/word swap options before using bge_readmem_ind(). Fixes PR#47716. - For BGE_IS_575X_PLUS() devices, don't set BGE_RXLPSTATCONTROL_DACK_FIX bits because these bits are reserved. - Document says 5717 and newer chips have no BGE_PCISTATE_INTR_NOT_ACTIVE bit, so don't use the bit on those chips. Same as OpenBSD. - Fix a bug that the PHY address bits in MI_MODE register is wrongly cleard. Set the PHY address correctly. - Use BGE_SETBIT() instead of CSR_WRITE_4() for the BGE_MISC_LOCAL_CTL register to not to modify some GPIO bits. - Set DMA watermark depend on the PCI max payload size. - Set BGE_JUMBO_CAPABLE correctly. - Fix a link detect bug on non-autopoll systems. - Change the TX ring size for 5717 series and 57764 series. - Set maximum read byte count to 2048 for PCI-X BCM5703/5704 devices. - For PCI-X BCM5704, set maximum outstanding split transactions to 0. - Add 40bit DMA bug workaround(BGEF_40BIT_BUG) from FreeBSD. This workaround is for 5714/5715 controllers and is not actually a MAC controller bug but an issue with the embedded PCIe to PCI-X bridge in the device. This change uses bus_dmatag_subregion(), so this workaround won't work on some archs which doesn't support bus_dmatag_subregion(). - Add 2500SX support (not tested). - Don't use the PHY Auto Poll Mode on many chips. This fixes a bug that MII Fiber NIC drop packet about 50%. Tested on HP Moonshot. - Add workaround for PR#48451. Some BCM5717-5720 based systems getNMI on boot. This problem doesn't occur when we don't use prefetchable memory in the APE area. Tested with HP MicroServer Gen8. - In the BCM5703, the DMA read watermark should be set to less than or equal to the maximum memory read byte count of the PCI-X command register. - Fix a bug that BGE_PHY_TEST_CTRL_REG isn't set correctly on some PCIe devices. - Use another firmware command in bge_asf_driver_up(). Same as Linux. This change fixes a bug that watchdog timeout occurs every 25-30 minutes on HP ML110 G6 reported enami@ in PR#49657. - Fix mbuf leak on failure. - Remove PCI_PRODUCT_BROADCOM_BCM5724 and PCI_PRODUCT_BROADCOM_BCM5750M. These devices have not released to public. - Add some workaround code for BGE_ASICREV_BCM5784 from Linux. - Change some printf() to aprint_*(). - Fix typo in comments. - Cleanup. brgphy(4): - Fix bit definition of BRGPHY_MRBE_MSG_PG5_NP_T2 from FreeBSD. - Add BCM5481, BCM5709S, BCM5756, BCM5717C, BCM5720C, BCM5785, BCM57765(PR#46961), BCM57780 - In brgphyattach(), set sc_isbge, sc_isbnx and sc_phyflags before PHY_RESET() because brgphy_reset() refers those flags. - Call brgpy specific autonego function in MII_TICK. Before this commit, only MII_MEDIACHG calls brgphy_mii_phy_auto() and MII_TICK calls MI mii_phy_auto(). That was not intended. - Remove extra delay in brgphy_mii_phy_auto. Same as {Free,Open}BSD. bnx(4): - Add missing ifmedia_delete_instance() in bnx_detach(). - Fix a bug that BNX_NO_WOL_FLAG isn't correctly set on some chips. Reported by From Henning Petersen in PR#44151. - Fix SERDES initialization. - Get out of the interrupt handler early if !IFF_RUNNING.
1.57.20.2	11-May-2013	riz	Applied patch (requested by msaitoh in ticket #1844): sys/dev/pci/pcireg.h 1.69 sys/dev/pci/ppb.c 1.44-1.45 Support PCI Express 2.0. Print version and device/port type information [msaitoh, ticket #1844]
1.57.20.1	19-Nov-2010	riz	branches: 1.57.20.1.2; Pull up revisions (requested by msaitoh in ticket #1358): sys/dev/pci/if_wm.c 1.196-1.199,1.202,1.205 sys/dev/pci/if_wmvar.h 1.9 sys/dev/pci/if_wmreg.h 1.36-1.39 sys/dev/pci/pcireg.h 1.61-1.64 sys/dev/pci/pcidevs 1.1023 sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen mii/igphy.c 1.21 mii/igphyvar.h 1.1 mii/inbmphyreg.h 1.2 - Count Receive error, CRC error, Alignment error, Symbol error, Sequence error, Carrier extension error and Receive length error into ierror. Fixes PR#30349 reported by UMEZAWA Takeshi. - Add support for 82575, 82576 and 82580(ER). - Apply the patch for 82575 from Wolfgang Stukenbrock (PR#42422). We use only one RX ring and with the legacy mode. - Add support for 82576. - Partial support for 82580. - Partial support for the serdes systems. - Add two workarounds for ICH8 with igp3. - Workaround for 82566 Kumeran PCS lock loss. - WOL from S5 stops working. - (pcireg.h) Add PCIe config register definitions. - Note that the changes to count Missed packet (rx fifo overflow) and Receive no buffers (rx ring full) into iqdrops in rev. 1.196 of if_wm.c is not pulled up.
1.57.20.1.2.1	07-Jan-2011	matt	Add/define some MSI support
1.57.18.1	19-Jan-2009	skrll	Sync with HEAD.
1.57.10.5	11-Aug-2010	yamt	sync with head.
1.57.10.4	11-Mar-2010	yamt	sync with head
1.57.10.3	16-Sep-2009	yamt	sync with head
1.57.10.2	19-Aug-2009	yamt	sync with head.
1.57.10.1	04-May-2009	yamt	sync with head.
1.57.6.1	17-Jan-2009	mjf	Sync with HEAD.
1.63.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.66.2.4	12-Jun-2011	rmind	sync with head
1.66.2.3	21-Apr-2011	rmind	sync with head
1.66.2.2	05-Mar-2011	rmind	sync with head
1.66.2.1	30-May-2010	rmind	sync with head
1.69.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.69.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.71.2.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.73.10.1	28-Nov-2012	matt	Add LCSR definitions and NVM storage subclass.
1.73.8.1	05-Aug-2013	martin	Pullup sys/dev/pci/pcireg.h 1.74-1.82 and 1.84 via patch sys/dev/pci/pci_subr.c 1.92-1.102, 1.104-1.105 via patch Add some PCI(e) register and bit definitions in pcireg.h. Fix the definition of PCI_PCIE_SLCAP_PSN. Fix a bug that IRQ(MSI) bits in PCIe capability register is incorrectly decoded. Print more registers in "pcictl dump". Fix bug in comment. Requested by msaitoh in ticket #928
1.73.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.73.2.1	30-Oct-2012	yamt	sync with head
1.74.2.4	03-Dec-2017	jdolecek	update from HEAD
1.74.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.74.2.2	23-Jun-2013	tls	resync from head
1.74.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.84.4.1	18-May-2014	rmind	sync with head
1.85.2.1	10-Aug-2014	tls	Rebase.
1.95.2.4	11-Aug-2018	martin	Pull up the following, requested by msaitoh in ticket #1628: share/man/man4/wm.4 1.40 via patch sys/dev/mii/ihphyreg.h 1.2 sys/dev/mii/inbmphyreg.h 1.10 sys/dev/pci/if_wm.c 1.504, 1.506, 1.510-1.535, 1.539-1.540, 1.546, 1.548, 1.551-1.552, 1.558, 1.565-1.573, 1.575, 1.579, 1.582, 1.584 via patch sys/dev/pci/if_wmreg.h 1.99-1.103, 1.106-1.107 via patch sys/dev/pci/if_wmvar.h 1.34-1.39 via patch sys/dev/pci/pcidevs 1.1327 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen sys/dev/pci/pcireg.h patch Sync wm(4) up to 2018/08/08 except MSI/MSI-X and NET_MPSAFE: - remove extra "+" - Fix a bug that non-GMII devices don't send a routing message when the link status is changed. - Set WMREG_KABGTXD not in wm_init_locked() but in wm_reset(). Same as other OSes. - If a interrupt is a spurious interrupt, don't print debug message. - Don't print the Image Unique ID if an NVM is iNVM (i210 and I211). - Print sc_flags with snprintb(). - Fix a bug that a RAL was written at incorrect address when the index number is more than 16 on 82544 and newer. - The layout of RAL on PCH* are different from others. Fix it. - Flush every MTA write. Same as Linux. - Move the location of calling wm_set_filter. Same as some other OSes. - Add CSR_WRITE_FLUSH() after writing WMREG_CTRL in wm_gmii_mediachange(). - Add missing "else" in wm_nvm_release(). - Make new wm_phy_post_reset() and use this function at all location after resetting phy. - Move the location of calling wm_get_hw_control. Same as Linux. - Add I219 specific wokaround for legacy interrupt. From OpenBSD. - Move the location of calling wm_lplu_d0_disable(). - Fix latency calculation in wm_platform_pm_pch_lpt(). - Set OBFF water mark and enable OBFF on PCH_LPT and newer. - Disable D0 LPLU on 8257[12356], 82580, I350 and I21[01], too. Before this commit, above devices and non-PCIe devices accessed wrong register. - Use device_printf() instead of aprint_error_dev() for PHY read/write functions because those are used not only in device attach. - Fix a bug that wm_gmii_i82544_{read,write}reg() didn't take care of page select. PHY access from igphy() automatically did it, but accessing from wm(4) for wrokaround didn't work correctly. This change affects 8254[17], 8257[12] ICH8, ICH9 and ICH10. - Call wm_kmrn_lock_loss_workaround_ich8lan() before any PHY access in wm_linkintr_gmii(). - Register access in wm_kmrn_lock_loss_workaround_ich8lan() now works correctly. Enable this function. - Configure the LCD with the extended configuration region in NVM if it's required. - If TX is not required to flush, RX is also not required to flush in wm_flush_desc_rings(). Same as other OSes. - Remove wrong semaphore access in wm_nvm_{read,write}_{ich8,spt} to prevent hangup. A semaphore is get/put in wm_nvm_{read,write}. - Move some initialization stuff in wm_attach() before wm_reset(). Some flags and callback function are required to set correctly before wm_reset() because wm_reset() and some helper functions refer them. - Add wm_write_smbus_addr() to set SMBus address by software. - Modify wm_gmii_hv_{read,write}reg_locked() to make them access HV_SMB_ADDR correctly. - Use new nvm.{acquire,release}() for semaphore. - Our MII readreg/writereg API has not way to detect an error. kmrn_{read,write}reg() are not used for MII API, so it's not required for these functions to use the same API. So, - Change return value as error code. - Change register value from int to uint16_t. - read: pass pointer for uint16_t as an argument. - Check return value on caller side. - Check whether it's required to use MDIC workaround for 80003 or not in wm_reset(). If the workaround isn't required, don't use the workaround code in wm_gmii_i80003_{read,write}reg. - Add WM_F_WA_I210_CLSEM flag for a workaround. FreeBSD/Linux drivers say "In rare circumstances, the SW semaphore may already be held unintentionally on I21[01]". PXE boot is one of the case. - Qemu's e1000e emulation (82574L)'s SPI has only 64 words. I've never seen on real 82574 hardware with such small SPI ROM. Check sc->sc_nvm_wordsize before accessing higher address words to prevent timeout. - Check some wm_nvm_read()'s return vale. - Print NVM offset and word count when EERD polling failed. - On I219, drop TARC0 bit 28 for DMA hang workaround (from Linux). - 82583 supports jumbo frame. Fixes PR#52773 reported by Shinichi Doyashiki. - Fix typo in comment. Reported by Shinichi Doyashiki in PR#52885. - Add ASPM workaround for 8257[1234] and 82583 to prevent device timeout or hangup. Fixes PR#52818 reported by Shinichi Doyashiki. - CID-1427779: Fix uninitialized variables. - Fix a bug that wm_pll_workaround_i210() is not called when a) Chip is I211 or b) Chip is I210 and it uses iNVM (not FLASH). - Do wm_reset_mdicnfg_82580() on 82580 only. - Fix FLASH access on PCH_SPT and newer. Their FLASH access should be done by 32bit. Especially for ICH_FLASH_HSFCTL register, it's located at 0x0006, so it must be accessed via ICH_FLASH_HSFSTS(0x0004) and use shift or mask. - Make wm_nvm_valid_bank_detect_ich8lan() the same as other OSes. - If the extended configuration size in the EXTCNFSIZE register is 0, don't continue in wm_init_lcd_from_nvm(). - Add PCH_CNP support (I219 with Intel 300 series chipset). - Enable I219 support. - I354 uses an external PHY, so don't use wm_set_eee_i350(). - Fix a bug that the link can't detect in link interrupt function for non-SERDES fiber. - Fix a bug that 82542 misunderstand fiber's signal detection. - Add debug printf()s. - Update comment. - Rename functions and variables. - Add diagnostic code. - Sort registers. - Lowercase hexadecimal values. - KNF.
1.95.2.3	14-Dec-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #325): sys/dev/pci/pcireg.h: revision 1.100 sys/dev/pci/pci_subr.c: revision 1.133 Add PCIe CRS Software Visibility bit.
1.95.2.2	12-Dec-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #312): sys/dev/pci/pci_subr.c: revision 1.130 sys/dev/pci/pci_subr.c: revision 1.131 sys/dev/pci/pci_subr.c: revision 1.132 sys/dev/pci/pcireg.h: revision 1.97 sys/dev/pci/pcireg.h: revision 1.98 sys/dev/pci/pcireg.h: revision 1.99 sys/dev/pci/pci_subr.c: revision 1.127 sys/dev/pci/pci_subr.c: revision 1.128 sys/dev/pci/pci_subr.c: revision 1.129 Always print the Slot implemented bit in the PCIe Capabilities Register using with onoff(). - Add some PCI subclass and interfaces. - The interface of PCI_SUBCLASS_BRIDGE_RACEWAY is not decoded yet. - Fix typo in a message. - Add comment. - Modify comment. s/genric/generic/ Add comment. Fix typo in comment. Fix a bug that the specification revision of the Power Management function was incorrectly printed in the output of capability "list". The value is also printed in the detail output and it has no bug. - Cleanup pci_conf_print_caplist. Use table. The reason why it loops twice is that some capabilities appear multiple times (e.g. HyperTransport cap). - Print the specification revision of Power Management and AGP not in the capability list part but in the detail part. Add some HyperTransport related defines. It's required for the MSI.
1.95.2.1	12-Dec-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #311): sys/dev/pci/ppb.c: revision 1.53 sys/dev/pci/ppb.c: revision 1.54 sys/dev/pci/pcireg.h: revision 1.96 - Modify message of PCIe capability version. This field (PCIE_XCAP_VER_MASK) is not specification's version number but the capability structure's version number. To avoid confusion, print "PCI Express capability version x". - The max number of PCIe lane is not 16 but 32. Fix the bug using with macro. - Use macro instead of magic number. - Gb/s -> GT/s Rename PCIE_XCAP_VER_* macros to avoid confusion.
1.100.2.9	28-Aug-2017	skrll	Sync with HEAD
1.100.2.8	05-Feb-2017	skrll	Sync with HEAD
1.100.2.7	05-Dec-2016	skrll	Sync with HEAD
1.100.2.6	05-Oct-2016	skrll	Sync with HEAD
1.100.2.5	29-May-2016	skrll	Sync with HEAD
1.100.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.100.2.3	22-Sep-2015	skrll	Sync with HEAD
1.100.2.2	06-Jun-2015	skrll	Sync with HEAD
1.100.2.1	06-Apr-2015	skrll	Sync with HEAD
1.113.2.4	26-Apr-2017	pgoyette	Sync with HEAD
1.113.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.113.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.113.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.119.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.130.2.11	29-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1728: sys/dev/pci/pci_subr.c 1.232-1.239 via patch sys/dev/pci/pcireg.h 1.62-1.63 - Decode link control2's Compliance Preset/De-emphasis more. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Fix typo.
1.130.2.10	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1714: sys/dev/pci/pcireg.h 1.148-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.217-1.222, 1.224, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158 sys/dev/pci/ppb.c 1.74 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. - Change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Remove unused shift and mask definitions. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.130.2.9	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1388: sys/dev/pci/pcireg.h 1.147 via patch sys/dev/pci/pci_subr.c 1.212, 1.215-1.217 via patch - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Remove whitespace for consistency. - Use macro. - Whitespace fixes.
1.130.2.8	04-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1118): sys/dev/pci/pci_subr.c: revision 1.210 sys/dev/pci/pci_subr.c: revision 1.207 sys/dev/pci/pcireg.h: revision 1.143 sys/dev/pci/pci_subr.c: revision 1.208 sys/dev/pci/pcireg.h: revision 1.144 sys/dev/pci/pci_subr.c: revision 1.209 sys/dev/pci/pcireg.h: revision 1.145 sys/dev/pci/pcireg.h: revision 1.146 Decode PCI Enhanced Allocation. The register offset of the mask and pending register is depend on the 64bit address capable bit, so fix the definition of PCI MSI vector mask and pending register. This problem was not a real bug because PCI_MSI{MASK,PENDING} were not used from anywhere. The downstream port of PCIe switch is not a root port, so don't print root port related register. For example, Intel 63xxESB controller's downstream port device was printed by pcictl(8) with this bug:
1.130.2.7	30-Oct-2018	sborrill	Pull up the following revisions(s) (requested by msaitoh in ticket #1074): sys/dev/pci/pci_subr.c: revision 1.204-1.206 sys/dev/pci/pcireg.h: revision 1.141-1.142 Root Complex Event Collector Bus Number Association ECN. - If capability version is 2 (or greater), decode RCEC Associated Bus Numbers register. - Don't print TPH requester's ST Table Size if the ST table location field is not PCI_TPH_REQ_STTBLLOC_TPHREQ because the size field is only applicable for PCI_TPH_REQ_STTBLLOC_TPHREQ case. - Add comment.
1.130.2.6	23-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1028): sys/dev/pci/pci_subr.c: revision 1.203 sys/dev/pci/pcireg.h: revision 1.140 Add ATS Relaxed Ordering supported bit described in Address Translation Relaxed Ordering ECN.
1.130.2.5	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #933): sys/dev/pci/pci_subr.c: revision 1.202 sys/dev/pci/pcireg.h: revision 1.139 sys/dev/pci/ppbreg.h: revision 1.9 VGA 16 bit decode bit is not bit 3 but bit 4. -- - Print Power Management Control/status register in 32bit. - Simplify.
1.130.2.4	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #930): sys/dev/pci/pci_subr.c: revision 1.201 sys/dev/pci/pcireg.h: revision 1.138 Fix typo. s/TPL/TLP/
1.130.2.3	26-Feb-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #576): sys/dev/pci/pci_subr.c: 1.197-1.200 sys/dev/pci/pcireg.h: 1.136-1.137 sys/dev/pci/ppbreg.h: 1.8 Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified. "s/above 300W/greater than 300W/" in pci_conf_print_pcie_power(). From PCIe Base Spec 3.1a Errata 2017-12-13. Cleanup: - Don't pass a capability pointer as a argument of pci_conf_find_cap() and determine the first pointer in the pci_conf_find_cap() function. - Don't pass a capability pointer as a argument of pci_conf_find_extcap() because it's not used. - Remove unsed code. - Add PCie Link Activation ECN. - Use macro. - KNF.
1.130.2.2	21-Nov-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #362): sys/dev/pci/pcireg.h: revision 1.133 sys/dev/pci/pcireg.h: revision 1.134 sys/dev/pci/pcireg.h: revision 1.135 sys/dev/pci/pci_subr.c: revision 1.190 sys/dev/pci/pci_subr.c: revision 1.191 sys/dev/pci/pci_subr.c: revision 1.192 sys/dev/pci/pci_subr.c: revision 1.193 sys/dev/pci/pci_subr.c: revision 1.194 sys/dev/pci/pcireg.h: revision 1.132 - Official shortname of LN Requester is LNR, so change PCI_EXTCAP_LN_REQ to PCI_EXTCAP_LNR - Use macro. - Add PCI_MAPREG_ROM_ADDR_MASK macro and PCI_MAPREG_ROM_ADDR() macro. - print PCI_MAPREG_ROM_ENABLE bit. - Decode Expansion ROM Validation ECN. - Add Native PCIe Enclosure Management ECN's extended capability type. Not decoded yet. Decode IOMMU capability of PCI secure device capability. From "AMD I/O Virtualization Technology(IOMMU) Specification (#48882) Revision 3.00". IOMMU cap dump fixes: - Print Capability Register's value. - Indent output correctly. - s/cahced/cached/ - Print MSI Message number with 0x%02x Fix a bug that the TPH ST table is decoded even if it's not in the TPH Requester extended capability structure.
1.130.2.1	04-Jul-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #80): sys/dev/pci/pci_subr.c: revision 1.184 sys/dev/pci/pci_subr.c: revision 1.185 sys/dev/pci/pci_subr.c: revision 1.186 sys/dev/pci/pci_subr.c: revision 1.187 sys/dev/pci/pci_subr.c: revision 1.188 sys/dev/pci/pci_subr.c: revision 1.189 sys/dev/pci/pcireg.h: revision 1.131 Add missing return to print the Slot Power Limit Value correctly. Fix to print the following bit fields correctly. - Supported Link Speeds Vector in LCAP2 - Lower SKP OS Generation Supported Speed Vector in LCAP2 - Lower SKP OS Reception Supported Speed Vector in LCAP2 - Enable Lower SKP OS Generation Vector in LCTL3 Note that the above bitfields start from 0 and the follwing bitfields start from 1: - Maximum Link Speed in LCAP - Current Link Speed in LCSR - Target Link Speed in LCSR2 For the Target Link Speed in LCSR2, 0 is allowed for a device which supports 2.5GT/s only (and this check also works for devices which compliant to versions of the base specification prior to 3.0. Tested with BCM5709: - Target Link Speed: unknown value (0) + Target Link Speed: 2.5GT/s For Attention Indicator Control bit and Power Indicator Control bit, it's allowed to be a read only value 0 if corresponding capability register bit is 0. Fix a bug that LTR's latency in L1 PM Substates capability and Latency Tolerance Reporting capability isn't printed correctly. Fix printf format/argument.
1.137.2.6	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.137.2.5	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.137.2.4	20-Oct-2018	pgoyette	Sync with head
1.137.2.3	30-Sep-2018	pgoyette	Ssync with HEAD
1.137.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.137.2.1	21-May-2018	pgoyette	Sync with HEAD
1.138.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.138.2.1	10-Jun-2019	christos	Sync with HEAD
1.147.4.4	29-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1412: sys/dev/pci/pci_subr.c 1.232-1.239 via patch sys/dev/pci/pcireg.h 1.62-1.63 - Decode link control2's Compliance Preset/De-emphasis more. - Decode Physical Layer 16.0 GT/s extended capability. - Decode Lane Margining at the Receiver extended capability. - Print "reserved" instead of "unknown" when printing equalization preset. One of them is known to be the default value. - Fix typo.
1.147.4.3	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1384: sys/dev/pci/pcireg.h 1.152-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.222, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158, 1.163 sys/dev/pci/ppb.c 1.74 - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.147.4.2	19-Mar-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #782): sys/dev/pci/pcireg.h: revision 1.150 sys/dev/pci/pcireg.h: revision 1.151 sys/dev/pci/pci_subr.c: revision 1.220 sys/dev/pci/pci_subr.c: revision 1.221 sys/dev/pci/pcireg.h: revision 1.149 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Avoid using magic number. Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. Remove unused shift and mask definitions. Add comment.
1.147.4.1	21-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #629: sys/dev/pci/pcireg.h 1.148 sys/dev/pci/pci_subr.c 1.218-1.219 - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - Fix typo.
1.148.2.2	29-Feb-2020	ad	Sync with head.
1.148.2.1	25-Jan-2020	ad	Sync with head.
1.151.6.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.153.6.1	31-May-2021	cjep	sync with head
1.153.4.2	01-Aug-2021	thorpej	Sync with HEAD.
1.153.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.168.2.1	22-Jun-2024	martin	Pull up following revision(s) (requested by rin in ticket #723): sys/dev/pci/pcireg.h: revision 1.171 PCI_CLASS_MASK: Use unsigned to avoid undefined behavior. Found by kUBSan.
1.171.2.1	02-Aug-2025	perseant	Sync with HEAD
1.120	23-Jun-2024	riastradh	pci: Pass cookie through pci_find_device, pci_enumerate_bus, take 2. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This will allow pci_find_device callers to pass a cookie through to the match function so they can keep state or pass in extra parameters like b/d/f numbers, which will allow us to nix some horrible kludges in the Linux PCI API emulation for drm (and, perhaps, Intel wifi). This change drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway. Take 2: Make sure to handle NULL match function.
1.119	20-May-2024	riastradh	Revert "pci: Pass cookie through pci_find_device, pci_enumerate_bus." Evidently something is wrong with this, to be diagnosed and redone once the builds and tests are in better shape.
1.118	20-May-2024	riastradh	pci: Pass cookie through pci_find_device, pci_enumerate_bus. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway.
1.117	27-Feb-2022	riastradh	branches: 1.117.4; pci(9): Provide default definition of pci_get_segment, always zero. pci_machdep.h can define __HAVE_PCI_GET_SEGMENT to provide a nonzero definition.
1.116	15-Sep-2021	thorpej	Adjust the device_call() calling convention so as to provide type checking of the arguments passed to the call, using auto-generated argument structures and binding macros.
1.115	12-May-2021	thorpej	- Define a device call for PCI bus instances to fetch a direct child's device handle given the device's device/function #s (extracted from a pcitag_t). Use it to associate the handle with the child device at config_found() time. - Implement this device call for ACPI and OpenFirmware. - Enable the OpenFirmware variant for evbarm FDT, macppc, ofppc, sparc64. - Obsolete acpi_device_register(); it is no longer needed. - Obsolete setting the OpenFirmware handle in PCI devices in the sparc64 device_register(); it is no longer needed.
1.114	27-Jan-2021	thorpej	branches: 1.114.4; 1.114.6; Introduce autoconfiguration helpers based around device_compatible_entry: - pci_compatible_match(): matches against the PCI ID. - pci_compatible_match_subsys(): matches against PCI SUBSYS ID. - pci_compatible_lookup(): look up entry by PCI ID. - pci_compatible_lookup_subsys(): look up entry by PCI SUBSYS ID. - pci_compatible_lookup_id(): look up entry by an arbitrary ID using the PCI ID code conventions. - Define PCI_COMPAT_EOL as a compat data array sentinel.
1.113	01-Dec-2018	msaitoh	branches: 1.113.12; Save control registers in PCI-X, PCIe, MSI and MSI-X capability area when suspend and restore them when resume. For PCIe cababilities register, it's required to check the existence of each register to not to write the next area. This chagnge fixes a stability of suspend/resume.
1.112	19-May-2018	jakllsch	branches: 1.112.2; Refine previous change to enable PCI window decoding in Command Register upon mapping; conditionalize on a global variable, that is set to true on x86 machines booting under EFI. For now, initialize the global variable at compile time to false. This is intended to limit potential problems for other NetBSD ports, should this changeset be pulled up to netbsd-8. Related to PR #53286.
1.111	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.110	28-Feb-2018	msaitoh	branches: 1.110.2; - Add new PCI quirk PCI_QUIRK_HASEXTCNF and PCI_QUIRK_NOEXTCNF. Some devices' extended configuration area may be broken or violate spec. If an extended configuration space is strange but it really exist, use PCI_QUIRK_HASEXTCNF. If an extended configuration space is plausible to exist but it really doesn't exist, use PCI_QUIRK_NOEXTCNF. - Add PCI_PRODUCT_INTEL_XEOND_MEM_0_TTR_1(0x6fa8) and PCI_PRODUCT_INTEL_COREI76K_IMC_0(0x6f68) with PCI_QUIRK_HASEXTCNF. The document clearly states they violate spec and it support the extended configuration space.
1.109	25-Nov-2016	knakahara	branches: 1.109.8; provide all PCI MSI/MSI-X manipulation stub functions. "#ifdef __HAVE_PCI_MSI_MSIX" workaround such as nvme_pci(4) is not required any more. http://mail-index.netbsd.org/source-changes/2016/09/17/msg077799.html
1.108	11-Jul-2016	knakahara	branches: 1.108.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.107	30-Oct-2015	msaitoh	- Move PCI_INTRSTR_LEN from pcireg.h to pcivar.h. - In PCI-X cap, print 2nd bus's PCI-X mode, error protection type, Max clock frequency and Max clock period. - In SATA cap, print register location correctly. - In Virtual Channel cap, print reference clock with "ns". - In Root Complex Link Declaration, print Link Entry number.
1.106	22-Oct-2015	knakahara	add pci_intr_alloc related stubs to reduce ifdef from device drivers.
1.105	02-Oct-2015	msaitoh	PCI Extended Configuration stuff written by nonaka@: - Add PCI Extended Configuration Space support into x86. - Check register offset of pci_conf_read() in MD part. It returns (pcireg_t)-1 if it isn't accessible. - Decode Extended Capability in PCI Extended Configuration Space. Currently the following extended capabilities are decoded: - Advanced Error Reporting - Virtual Channel - Device Serial Number - Power Budgeting - Root Complex Link Declaration - Root Complex Event Collector Association - Access Control Services - Alternative Routing-ID Interpretation - Address Translation Services - Single Root IO Virtualization - Page Request - TPH Requester - Latency Tolerance Reporting - Secondary PCI Express - Process Address Space ID - LN Requester - L1 PM Substates The following extended capabilities are not decoded yet: - Root Complex Internal Link Control - Multi-Function Virtual Channel - RCRB Header - Vendor Unique - Configuration Access Correction - Multiple Root IO Virtualization - Multicast - Resizable BAR - Dynamic Power Allocation - Protocol Multiplexing - Downstream Port Containment - Precision Time Management - M-PCIe - Function Reading Status Queueing - Readiness Time Reporting - Designated Vendor-Specific
1.104	17-Aug-2015	knakahara	Add kernel code to support intrctl(8).
1.103	13-Aug-2015	msaitoh	- Don't take pci_attach_args as an argument in pci_msi[x]_count(). - Move prototypes of pci_msi[x]_count() from x86/x86/pci_machdep_common to sys/dev/pci/pcivar.h. - Move pci_msi[x]_count() from x86/pci/pci_msi_machdep.c to sys/dev/pci/pci.c
1.102	27-Apr-2015	knakahara	add x86 MD MSI/MSI-X support code.
1.101	26-Dec-2014	msaitoh	Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fix ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.100	16-Oct-2014	riastradh	branches: 1.100.2; Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.99	29-Mar-2014	christos	branches: 1.99.4; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.98	29-Jan-2012	drochner	branches: 1.98.2; 1.98.6; 1.98.10; extend the pci_aprint_devinfo slightly to cover the cases commonly used by drivers: a short name for the quiet/naive case and a string to override the "pcidevs" based name by one provided by the driver, ride on yesterday's kernel minor version bump
1.97	26-Jan-2012	drochner	put printing of the pci_devinfo into its own function (not inlined by purpose) - this is a stack hog, and with this change my uTCA amd64 system boots again a lot of similar code can be eliminated from pci device drivers this way, but before doing so (and making the new function part of the module API) I'd like to consider a modification to make it work with drivers which prefer to print names from other sources (like pciide)
1.96	21-Oct-2011	dyoung	branches: 1.96.2; 1.96.6; Tell a pci(4) instance its subordinate PCI buses using a new member in the pcibus_attach_args, pba_sub. pciN attaches to pba_bus itself. If pba_bus < pba_sub, then [pba_bus + 1, pba_sub] are subordinate to pba_bus. On i386, make mainbus0 attach pci0 with pba_sub = 255 because all buses 1 and up must be subordinate to pci0. XXX Deal with other architectures.
1.95	24-Aug-2011	dyoung	Add to pci_bus_devorder() an argument that tells the number of slots available in the devs array. Change the type of the devs array from char to uint8_t. Treat the return value of pci_bus_devorder() as the number of slots that it filled. Don't use the __PCI_BUS_DEVORDER #definition to configure the kernel but let the linker do it. Make pci_bus_devorder() available on all architectures by adding a default implementation that will DTRT on all architectures but hpcmips, the only architecture to #define __PCI_BUS_DEVORDER. On hpcmips, adapt the implementation to the new calling convention. XXX I can compile an hpcmips GENERIC kernel, but I don't have a XXX hpcmips box to test it on.
1.94	22-Jun-2011	matt	Add inline accessors for pba_pc and pa_pc so that <machine/pci_machdep.h> can use them in inline functions.
1.93	17-May-2011	dyoung	PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.92	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.91	10-Feb-2011	jmcneill	pcimmap: if the requested page is marked prefetchable in a child device's BAR, pass the BUS_SPACE_MAP_PREFETCHABLE flag down to bus_space_mmap
1.90	09-Jun-2010	mrg	branches: 1.90.2; 1.90.4; declare 'struct pci_attach_args' before it is used. fixes ia64 build.
1.89	27-May-2010	pgoyette	Extract pci_verbose interface definitions into their own header file so we don't need to include pcivar.h in the module. pcivar.h tries to pull in machine/pci_machdep.h which doesn't exist on all machine/architecture combos. Keeping track and building the module only for those that work would have been a maintenance headache; this change allows us to build the module on all systems, regardless of whether the system has pci support or not.
1.88	25-May-2010	pgoyette	Rework the pciverbose module dispatch vectors to avoid renaming the externally-visible entrypoint name. Also this avoids a potential need to bump kernel version. Requested by dyoung@ and mrg@
1.87	24-May-2010	pgoyette	Extract the vendor/product tables and related access routines into a separate kernel module. Update pci bus attach routine to load the module (if available) when we're about to start scanning the bus, and unload the module after the scan is finished. On architectures which support loading of modules by the boot loader, the 'pciverbose' module can be loaded and executed without needing to rebuild the kernel. On all architectures, using 'options PCIVERBOSE' in the kernel configuration file will create a 'builtin' module which is functionally equivalent to previous behavior. XXX Although not nearly as large as the vendor and product tables, XXX the PCI class and subclass tables might also be offloaded into XXX the module at a future time. XXX Cardbus (and possibly other) drivers should also be modified to XXX load the module before scanning/attaching devices.
1.86	28-Apr-2010	dyoung	Add data types, function prototypes, and stub implementations for pci_chipset_tag_create() and pci_chipset_tag_destroy(). On architectures that support it, an MI PCI bus driver can override the architecture's default pci(9) and pci_intr(9) implementation. Coming up next: documentation. After that: x86 implementation. Last but not least: make cbb(4) use MI PCI overrides.
1.85	12-Mar-2010	matt	branches: 1.85.2; Add placeholdr flags indicating that the bus has MSI/MSI-X support.
1.84	23-Feb-2010	dyoung	Remove unused functions pci_disable_retry() and cardbus_disable_retry().
1.83	22-Jul-2008	bjs	branches: 1.83.12; 1.83.14; 1.83.18; Add pci_mapreg_submap(): This function is pci_mapreg_map() with two additional arguments, offset and maxsize. This new functionality eases handling certain tasks within the direct rendering manager, though I hope others will also find it useful. pci_mapreg_map() is now merely a wrapper around pci_mapreg_submap(); the latter contains all of the code from the former. ok christos@
1.82	30-May-2008	ad	branches: 1.82.2; 1.82.4; pci_intr_setattr(), allows PCI interrupts to be marked MPSAFE on x86, and other platforms if the code is added. pci_intr_map(...) pci_intr_setattr(pc, ih, PCI_INTR_MPSAFE, 1); pci_intr_establish(...)
1.81	05-May-2008	dyoung	branches: 1.81.2; Restore PCI devices to a state that is closer to the state that NetBSD finds them in: save and restore power management state (D0..D3) and PCI Configuration Registers 0x0 through 0x40 during device attachment and detachment, respectively. Among other things, this will fix sip(4) detachment and re-attachment.
1.80	09-Apr-2008	dyoung	branches: 1.80.2; 1.80.4; Export some PCI autoconf routines for use by elansc(4), for example.
1.79	23-Mar-2008	cube	Split device_t and softc for pci(4).
1.78	21-Mar-2008	dyoung	pci_activate() expects for its void * argument to be a device_t, so change the type of the argument to device_t. Update each use of pci_activate(). Use device_t and accessors. Use aprint_*_dev().
1.77	09-Dec-2007	jmcneill	branches: 1.77.6; 1.77.10; Merge jmcneill-pm branch.
1.76	12-Nov-2007	joerg	branches: 1.76.2; 1.76.4; Merge pci_disable_retry function from jmcneill-pm as it is found in various drivers.
1.75	19-Oct-2007	ad	branches: 1.75.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.74	04-Mar-2007	christos	branches: 1.74.2; 1.74.14; 1.74.16; 1.74.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.73	25-Sep-2006	jmcneill	branches: 1.73.4; Create a PCI bus powerhook to handle setting PCI D# power state transitions on suspend/standby/resume.
1.72	17-Jun-2006	christos	branches: 1.72.4; 1.72.6; re-factor the pci powestate api. reviewed by gimpy
1.71	01-Mar-2006	gdamore	branches: 1.71.2; 1.71.8; 1.71.10; Add pci_find_rom() API as discussed on tech-kern.
1.70	11-Dec-2005	christos	branches: 1.70.4; 1.70.6; merge ktrace-lwp.
1.69	04-Feb-2005	perry	branches: 1.69.6; de-__P
1.68	26-Jan-2005	jmcneill	Add support functions for capturing and restoring PCI configuration registers for power management code.
1.67	13-Sep-2004	drochner	branches: 1.67.4; 1.67.6; remove macros which just hide cfdata internals (and are used at one place only)
1.66	30-Aug-2004	drochner	Phase out the use of a string as first "attach args" member to control which bustype should be attached with a specific call to config_found() (from a "mainbus" or a bus bridge). Do it for isa/eisa/mca and pci/agp for now. These buses all attach to an mi interface attribute "isabus", "eisabus" etc., and the autoconf framework now allows to specify an interface attribute on config_found() and config_search(), which limits the search of matching config data to these which attach to that specific attribute. So we basically have to call config_found_ia(..., "foobus", ...) where such a bus is attached. As a consequence, where a "mainbus" or alike also attaches other devices (eg CPUs) which do not attach to a specific attribute yet, we need at least pass an attribute name (different from "foobus") so that the foo bus is not found at these places. This made some minor changes necessary which are not obviously related to the mentioned buses.
1.65	30-Aug-2004	drochner	add centralized {eisa,isa,pci,agp,mca}busprint() functions which do what tens of the bus' parents foo{...}bridge_print()s scattered around do
1.64	17-Aug-2004	drochner	make PCI devices attachable/detachable (as far as the particular drivers allow), and allow to rescan a bus selectively (ie only the device/ function I'm looking at)
1.63	02-Aug-2004	mycroft	For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.62	29-Jul-2004	drochner	make the "generic" PCI bus enumeration code the standard case which gets used if nothing else is defined in MD headers, introduce a "PCI_MACHDEP_ENUMERATE_BUS" CPP definition which can be used by MD headers (just 1 port atm) to plug in special code
1.61	08-May-2004	christos	GC pci_{g,s}et_powerstate into pci_powerstate(). Idea from mycroft and gimpy. Nothing uses them yet.
1.60	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.59	15-Aug-2003	itojun	- check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.58	29-Jun-2003	fvdl	branches: 1.58.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.57	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.56	15-Jun-2003	fvdl	Handle 64bit DMA addresses on PCI for platforms that can (currently only enabled on amd64). Add a dmat64 field to various PCI attach structures, and pass it down where needed. Implement a simple new function called pci_dma64_available(pa) to test if 64bit DMA addresses may be used. This returns 1 iff _PCI_HAVE_DMA64 is defined in <machine/pci_machdep.h>, and there is more than 4G of memory.
1.55	25-Mar-2003	thorpej	Add PCI VPD access routines. From psi.cz!freza, PR kern/20889.
1.54	19-Mar-2003	christos	It is not appropriate for pcivar.h to include "locators.h" The two files that need it, should include it themselves.
1.53	18-Jun-2002	tshiozak	add support for the per-device power management capability. int pci_set_powerstate(pci_chipset_tag_t pc, pcitag_t tag, int newstate) set power state of the device to newstate. int pci_get_powerstate(pci_chipset_tag_t pc, pcitag_t tag) get current power state of the device. In the future, these functions will be used for ACPI support.
1.52	30-May-2002	drochner	implement a check whether a BAR is present at all at a given configuration space address and use it where the mappings of the VGA card are registered before descenting too deep into "memory" type specific code (pci_mem_find() gets noisy if it doesn't like the register)
1.51	18-May-2002	sommerfeld	branches: 1.51.2; Add "pa_rawintrpin" containing unswizzled interrupt pin to pci_attach_args.
1.50	16-May-2002	thorpej	branches: 1.50.2; * Add "pcitag_t pba_bridgetag" to pci_attach_args. This is set to NULL for root PCI busses. For busses behind a bridge, it points to a persistent copy of the bridge's pcitag_t. This can be very useful for machine-dependent PCI bus enumeration code. Implement a machine-dependent pci_enumerate_bus() for sparc64 which uses OFW device nodes to enumerate the bus. When a PCI bus that is behind a bridge is attached, pci_attach_hook() allocates a new PCI chipset tag for the new bus and sets it's "curnode" to the OFW node of the bridge. This is used as a starting point when enumerating that bus. Root busses get the OFW node of the host bridge (psycho). * Garbage-collect "ofpci" and "ofppb" from the sparc64 port.
1.49	15-May-2002	thorpej	Split the code that enumerates the PCI bus and that actually probes for a device into two functions: * pci_probe_device() actually probes/attaches the device specified by the provide pcitag_t. * pci_enumerate_bus() enumerates the bus, and calls pci_probe_device() for each device on the bus. A pci_enumerate_bus_generic() is provided which implements the old method of doing this: If something found at dev0/func0, determine number of functions and probe each one. Machine-dependent code will be able to specify the bus enumeration routine in the future.
1.48	13-Sep-2001	thorpej	Add an ioctl interface to the PCI bus. Add ioctls to read/write PCI configuration space registers, and to fetch bus info.
1.47	10-Sep-2001	fvdl	Add a pci_find_device function, to find a PCI device using a match function on all (probed) PCI buses: int pci_find_device(struct pci_attach_args pa, int (match)(struct pci_attach_args *)); The pci_attach_args structure pointed to by pa is filled in if the device is found, and 1 is returned. Otherwise 0 is returned. This function is, unfortunately, needed by the i810 agp code. It's also of use for LKMs. Also frob pci_probe_bus to take 2 extra args when used by pci_find_device.
1.46	04-May-2001	bouyer	branches: 1.46.2; 1.46.4; Add pci bus number to pci_attach_args, as suggested by Jason on tech-kern (needed for pciide)
1.45	05-Mar-2001	matt	Only include "locators.h" is _KERNEL is defined.
1.44	12-Feb-2001	mrg	branches: 1.44.2; backout the parts of the previous change (7 months ago) that were not part of the sys/vm removal, but some (fortunately disabled) work-in-progress.
1.43	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.42	10-May-2000	thorpej	branches: 1.42.4; Add support for mapping 64-bit PCI memory space. If the region is mapped in a way that is inaccessible by a 32-bit bus_addr_t, then print a message to that effect and return failure. Original patches by Bill Studenmund, with a few small changes by me.
1.41	30-Sep-1999	thorpej	branches: 1.41.2; Define flags in pci_attach_args which indicate it's okay to use the Memory Read Line, Memory Read Multiple, and Memory Write and Invalidate PCI commands.
1.40	06-May-1999	thorpej	Back out previous. Thanks to cgd for pointing out another way to do this.
1.39	06-May-1999	thorpej	Add a `bus' member to the pci_attach_args. This is not normally used, but some child drivers might need to know this information.
1.38	19-Mar-1999	cgd	branches: 1.38.4; pull pci_machdep.h in from machine/
1.37	07-Nov-1998	drochner	add support for "extended capabilities" (new in PCI spec 2.2)
1.36	15-Aug-1998	mycroft	Make copyright notices with my name consistent.
1.35	12-Jul-1998	augustss	Add USB support. Supported so far: * UHCI and OHCI host controllers on PCI * Hubs * HID devices withe special drivers for mouse and keyboard * Printers
1.34	09-Jun-1998	thorpej	Remove the PCI-ISA bridge callback mechanism; it's no longer needed.
1.33	31-May-1998	cgd	add a PCI 'quirks' mechanism, meant to be used as the mechanism of last resort when trying to communicate information about bogus behaviour of PCI devices to the MI autoconfiguration code. In general, bogus behaviour should be handled by drivers, but there are some types of bogons which can't be addressed that way. The only quirk currently defined is one which indicates that the device is multi-function even though the device's header says otherwise. (Mmm, Intel 82371FB PCI-to-ISA Bridge (PIIX); you'd think that at least Intel would have gotten it right...)
1.32	25-May-1998	mark	Include arm32/pci/pci_machdep.h if arm32 is defined.
1.31	18-May-1998	cgd	largely reimplement pci_conf_print(): * print all configuration space registers. Then, where possible, interpret them. (That is, PRESENT ALL THE DATA, then interpret it -- don't hide data behind interpretation. Also, when interpreting fields, try to print out the specific value that's being interpreted.) * handle different header types. * allow caller to specify a function which can interpret the device-dependent header and is responsible for pretty-printing it. It spews (use 'options MSGBUFSIZE=...' 8-), but when you want the data, you really want _all_ of it. Still needs some cleanup and additional code (e.g. interepretation of PCI-PCI (type 1) and PCI-Cardbus (type 2(?)) bridge headers).
1.30	15-May-1998	tsubai	Add macppc support.
1.29	14-Apr-1998	thorpej	Add pci_conf_print(), a function to dump the PCI configuration space, useful in driver debugging. From Zubin D. Dittia <zubin@clouseau.arl.wustl.edu>, PR #4249.
1.28	14-Oct-1997	sakamoto	add bebox
1.27	06-Oct-1997	thorpej	Back out last change. (Partially my fault, for not reviewing it close enough.)
1.26	03-Oct-1997	lonhyn	pci_mapreg_info() has been changed to take arguments like pci_mapreg_map()
1.25	30-Aug-1997	mycroft	Don't export pci_*_find() any more.
1.24	17-Jul-1997	jtk	branches: 1.24.2; use "locators.h" defines for indices for cf_loc[] and default values
1.23	06-Jun-1997	thorpej	Pull thorpej-bus-dma branch into mainline.
1.22	13-Apr-1997	cgd	branches: 1.22.2; create pci_mapreg_info() which simply gets mapping register information. It's used by pci_mapreg_map() and can be used directly by drivers with special needs (e.g. those being attached as console devices).
1.21	13-Apr-1997	cgd	rename pci_map_register to pci_mapreg_map. The latter name is more descriptive, and allows for a sane name for a function which just digs the info out of the mapping register but doesn't do the mapping.
1.20	13-Apr-1997	cgd	implement pci_map_register(), which gets information about a device's mapping register, maps it, and returns all of the relevant information. deprecate use of pci_{io,mem}_find(), but leave them around (for a while) for backward compatibility with third-party drivers.
1.19	10-Apr-1997	cgd	pass memory- and i/o-enabled flags down via the PCI bus and device attach arguments, so that a device can tell if its memory and I/O spaces are enabled. The flags are cleared, depending on the contents of devices CSR registers, in the machine-independent PCI bus code.
1.18	01-Dec-1996	leo	The atari has a pci_machdep.h file.
1.17	23-Nov-1996	cgd	Provide a routine so that ISA/EISA bridges can set up a callback so that their child busses can be attached after the PCI bus autoconfiguration for their parent bus is done. This works because: (1) there can be at most one ISA/EISA bridge per PCI bus, and (2) any ISA/EISA bridges must be attached to primary PCI busses (i.e. bus zero). That boils down to: there can only be one of these outstanding at a time, it is cleared when configuring PCI bus 0 before any subdevices have been found, and it is run after all subdevices of PCI bus 0 have been found. This (or something like it) is needed because there are some (legacy) PCI devices which can show up as ISA/EISA devices as well (the prime example of which are VGA controllers). If you attach ISA from a PCI-ISA/EISA bridge, and the bridge is seen before the video board is, the board can show up as an ISA device, and that can (bogusly) complicate the PCI device's attach code, or make the PCI device not be properly attached at all. This could be done with machine-dependent code, but as more ports add support for PCI (and PCI-ISA/EISA bridges) more will need it. The i386 port could (perhaps should) be converted to use it as well.
1.16	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.15	28-Mar-1996	cgd	remove inappropriate an potentially confusing comments ("unnecessary?") from the pa_device and pa_function members of the pci_attach_args structure.
1.14	27-Mar-1996	cgd	modify these to provide a new, better-specified PCI interface (soon to be documented on mailing lists; eventually in section 9 manual pages), most importantly: (1) support interrupt pin swizzling on non-i386 systems with PCI-PCI bridges (per PPB spec; done, but meaningless, on i386). (2) provide pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register. (3) provide pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (4) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.13	14-Mar-1996	cgd	(1) provide #defines for cf_loc[] entries for devices that attach to pcibus and pci. (2) remove the #ifdef i386 from pci.c, and provide a machine-dependent hook (pci_md_attach_hook()) to do any machine-dependent attachment gunk, e.g. on the i386 printing out the configuration mode (if bus 0) (3) don't pass max device number for a given bus in, use PCI_MAX_DEVICE_NUMBER, which can be defined on a per-machine basis. (defaults to 32. on i386, it's 32 if pci conf mode == 1, 16 if 2.)
1.12	08-Mar-1996	cgd	pass a bus_chipset_tag_t (defined in <machine/bus.h>) to the bus when attaching, and to the devices when attaching them. #include <machine/bus.h> to make this backward compatible with old #include requirements. Also, clean up idempotency so that isa/eisa/pci "var.h" headers are consistent (make them all idempotent).
1.11	28-Feb-1996	cgd	make PCI bus match/attach and sub-device attachment machine-independent.
1.10	22-Jan-1996	cgd	update PCIVERBOSE code in various ways: (1) remove the 'UNSUPP' keyword from the device list, because it can't be reasonably used (becuase different devices may be supported on different machines, for good reason). (2) enhance pci_devinfo so that class/subclass information is optional (so pci_devinfo can be used by drivers that match classes of devices, and want to look up the devices' names easily). (3) more known vendors and devices.
1.9	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.8	18-Jun-1995	cgd	prototype for pci_devinfo function
1.7	18-Jun-1995	cgd	add include for the alpha, alphabetize
1.6	23-May-1995	cgd	split single-subdevice lookup & attachment into a subroutine (pci_attach_subdev()). remove pciattach() function and the pcicd cfdriver struct, the former because thre are a lot of attachment actions which really are machine-dependent (perhaps even "most"), and the latter because now that both pcimatch() and pciattach() are machine-dependent it's bad style to declare them here and it gains nothing.
1.5	17-Apr-1995	cgd	clean up several ISA device interfaces: autoconfiguration, header inclusion, and interrupt configuration. more work still needs to be done, but it's getting better...
1.4	27-Jan-1995	cgd	include files from the correct places.
1.3	04-Nov-1994	mycroft	Make a wrapper match function to check the bus and device numbers, rather than insisting that every driver do it.
1.2	27-Oct-1994	cgd	new RCS ID format.
1.1	09-Aug-1994	mycroft	Add PCI autoconfiguration support.
1.22.2.1	13-May-1997	thorpej	Provide the bus dma tag to children.
1.24.2.2	15-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.24.2.1	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.38.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.41.2.2	12-Mar-2001	bouyer	Sync with HEAD.
1.41.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.42.4.1	15-May-2001	he	Pull up revision 1.46 (requested by bouyer): Add a pa_bus member to pci_attach_args, so that the PCI bus number can be used in pci device drivers.
1.44.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.44.2.3	21-Sep-2001	nathanw	Catch up to -current.
1.44.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.44.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.46.4.1	01-Oct-2001	fvdl	Catch up with -current.
1.46.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.46.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.46.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.46.2.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.50.2.4	15-Jul-2002	gehenna	catch up with -current.
1.50.2.3	20-Jun-2002	gehenna	catch up with -current.
1.50.2.2	30-May-2002	gehenna	Catch up with -current.
1.50.2.1	16-May-2002	gehenna	file pcivar.h was added on branch gehenna-devsw on 2002-05-30 14:46:31 +0000
1.51.2.1	15-Aug-2003	tron	Pull up revision 1.59 (requested by itojun in ticket #1412): - check HDRTYPE early, and ignore if it is not supported (n > 2). - defer access to interrupt configuration register, as its existence depends on HDRTYPE. - add "skip particular funtion in multifunction device" functionality to quirk table. - add GEODE/NS SC1100 quirk (now boots on soekris Net4801).
1.58.2.7	04-Feb-2005	skrll	Sync with HEAD.
1.58.2.6	21-Sep-2004	skrll	Fix the sync with head I botched.
1.58.2.5	18-Sep-2004	skrll	Sync with HEAD.
1.58.2.4	03-Sep-2004	skrll	Sync with HEAD
1.58.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.58.2.2	03-Aug-2004	skrll	Sync with HEAD
1.58.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.67.6.1	12-Feb-2005	yamt	sync with head.
1.67.4.1	29-Apr-2005	kent	sync with -current
1.69.6.7	24-Mar-2008	yamt	sync with head.
1.69.6.6	21-Jan-2008	yamt	sync with head
1.69.6.5	15-Nov-2007	yamt	sync with head.
1.69.6.4	27-Oct-2007	yamt	sync with head.
1.69.6.3	03-Sep-2007	yamt	sync with head.
1.69.6.2	30-Dec-2006	yamt	sync with head.
1.69.6.1	21-Jun-2006	yamt	sync with head.
1.70.6.1	22-Apr-2006	simonb	Sync with head.
1.70.4.1	09-Sep-2006	rpaulo	sync with head
1.71.10.1	13-Jul-2006	gdamore	Merge from HEAD.
1.71.8.1	19-Jun-2006	chap	Sync with head.
1.71.2.1	26-Jun-2006	yamt	sync with head.
1.72.6.1	22-Oct-2006	yamt	sync with head
1.72.4.1	18-Nov-2006	ad	Sync with head.
1.73.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.74.20.2	13-Nov-2007	bouyer	Sync with HEAD
1.74.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.74.16.3	23-Mar-2008	matt	sync with HEAD
1.74.16.2	09-Jan-2008	matt	sync with HEAD
1.74.16.1	06-Nov-2007	matt	sync with HEAD
1.74.14.7	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.74.14.6	20-Nov-2007	joerg	GC struct ifnet.
1.74.14.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.74.14.4	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.74.14.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.74.14.2	23-Aug-2007	joerg	Introduce pci_net_generic_power, which should be enough for most network drivers and be a good foundation for C&P for the rest. For iwi(4), don't reset the PCI retry register again, pci_conf_restore should take care of that already. For bge(4), add a NetBSD style if_stop.
1.74.14.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.74.2.1	23-Oct-2007	ad	Sync with head.
1.75.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.75.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.76.4.1	11-Dec-2007	yamt	sync with head.
1.76.2.1	26-Dec-2007	ad	Sync with head.
1.77.10.3	28-Sep-2008	mjf	Sync with HEAD.
1.77.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.77.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.77.6.1	24-Mar-2008	keiichi	sync with head.
1.80.4.4	11-Aug-2010	yamt	sync with head.
1.80.4.3	11-Mar-2010	yamt	sync with head
1.80.4.2	04-May-2009	yamt	sync with head.
1.80.4.1	16-May-2008	yamt	sync with head.
1.80.2.2	04-Jun-2008	yamt	sync with head
1.80.2.1	18-May-2008	yamt	sync with head.
1.81.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.81.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.82.4.1	19-Oct-2008	haad	Sync with HEAD.
1.82.2.1	28-Jul-2008	simonb	Sync with head.
1.83.18.1	07-Jan-2011	matt	Add/define some MSI support
1.83.14.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.83.14.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.83.12.2	15-Feb-2014	matt	Add a few OKAY flags from HEAD
1.83.12.1	05-Nov-2013	matt	Pull in support for pci_aprint_devinfo_fancy
1.85.2.5	31-May-2011	rmind	sync with head
1.85.2.4	21-Apr-2011	rmind	sync with head
1.85.2.3	05-Mar-2011	rmind	sync with head
1.85.2.2	03-Jul-2010	rmind	sync with head
1.85.2.1	30-May-2010	rmind	sync with head
1.90.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.90.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.96.6.1	18-Feb-2012	mrg	merge to -current.
1.96.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.96.2.1	17-Apr-2012	yamt	sync with head
1.98.10.1	18-May-2014	rmind	sync with head
1.98.6.2	03-Dec-2017	jdolecek	update from HEAD
1.98.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.98.2.1	16-Jan-2015	snj	Pull up following revision(s) (requested by msaitoh in ticket #1229): sys/arch/x86/pci/ichlpcib.c: revision 1.40, 1.45 sys/dev/pci/pcivar.h: revision 1.101 sys/dev/pci/pci_map.c: revision 1.32 sys/dev/ic/i82801lpcreg.h: revision 1.12 Use '\n' at the end of all aprint_error_dev() format strings. -- Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fixes ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.99.4.2	08-Jan-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #394): sys/dev/pci/pcivar.h: revision 1.101 sys/dev/pci/pci_map.c: revision 1.32 sys/dev/ic/i82801lpcreg.h: revision 1.12 sys/arch/x86/pci/ichlpcib.c: revision 1.45 Fix a bug that ichlpcib(4) maps I/O area incorrectly and then fails to attach gpio. It might also fix ACPI related problem described in PR#48960: - The LPCIB_PCI_PMBASE and LPCIB_PCI_GPIO register are alike PCI BAR but not completely compatible with it. It's ok because the registers' addresses are out of BAR0-BAR5(0x10-0x24) and are located in the device-dependent header. The PMBASE and GPIO registers define the base address and the type but not describe the size. The size is fixed to 128bytes. So use pci_mapreg_submap(). - Make pci_mapreg_submap() extern again. - Fix the calculation of the map size in pci_mapreg_submap().
1.99.4.1	17-Oct-2014	martin	Pull up following revision(s) (requested by riastradh in ticket #144): sys/dev/pci/radeonfb.c: revision 1.85 sys/dev/pci/pcivar.h: revision 1.100 sys/dev/pci/pci_map.c: revision 1.31 sys/external/bsd/drm2/include/linux/pci.h: revision 1.9 Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.100.2.6	05-Dec-2016	skrll	Sync with HEAD
1.100.2.5	05-Oct-2016	skrll	Sync with HEAD
1.100.2.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.100.2.3	22-Sep-2015	skrll	Sync with HEAD
1.100.2.2	06-Jun-2015	skrll	Sync with HEAD
1.100.2.1	06-Apr-2015	skrll	Sync with HEAD
1.108.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.109.8.2	07-Dec-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1128): sys/dev/pci/pcivar.h: revision 1.113 sys/dev/pci/pci.c: revision 1.153 Save control registers in PCI-X, PCIe, MSI and MSI-X capability area when suspend and restore them when resume. For PCIe capabilities register, it's required to check the existence of each register to not to write the next area. This change fixes a stability of suspend/resume.
1.109.8.1	07-Jun-2018	martin	Pull up following revision(s) (requested by jakllsch in ticket #832): sys/dev/pci/pcivar.h: revision 1.112 sys/dev/pci/pci_map.c: revision 1.34,1.35 sys/arch/x86/x86/efi.c: revision 1.15 Enable the appropriate memory or I/O space decode in the PCI Command/Status Register upon mapping a BAR. This should fix PR #53286. It's also possible there are other similar PRs that might be fixed by this. - Refine previous change to enable PCI window decoding in Command Register upon mapping; conditionalize on a global variable, that is set to true on x86 machines booting under EFI. For now, initialize the global variable at compile time to false. This is intended to limit potential problems for other NetBSD ports, should this changeset be pulled up to netbsd-8. Related to PR #53286.
1.110.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.110.2.2	21-May-2018	pgoyette	Sync with HEAD
1.110.2.1	22-Apr-2018	pgoyette	Sync with HEAD
1.112.2.1	10-Jun-2019	christos	Sync with HEAD
1.113.12.1	03-Apr-2021	thorpej	Sync with HEAD.
1.114.6.1	31-May-2021	cjep	sync with head
1.114.4.1	13-May-2021	thorpej	Sync with HEAD.
1.117.4.1	04-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #928): sys/external/bsd/drm2/dist/drm/drm_gem.c: revision 1.25 sys/external/bsd/drm2/dist/drm/radeon/radeon_ci_dpm.c: revision 1.7 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/priv.h: revision 1.4 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_acpi.c: revision 1.4 sys/external/bsd/drm2/dist/drm/i915/display/intel_opregion.h: revision 1.6 sys/external/bsd/drm2/dist/drm/i915/i915_drv.h: revision 1.49 sys/external/bsd/drm2/include/linux/mxm-wmi.h: revision 1.1 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/pci/nouveau_nvkm_subdev_pci_pcie.c: revision 1.4 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_base.c: revision 1.13 sys/external/bsd/common/include/linux/bitops.h: revision 1.17 sys/external/bsd/drm2/nouveau/files.nouveau: revision 1.40 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.30 sys/external/bsd/drm2/dist/drm/radeon/radeon_drv.h: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/pci/nouveau_nvkm_subdev_pci_pcie.c: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/mxm/nouveau_nvkm_subdev_mxm_base.c: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_gart.c: revision 1.12 sys/external/bsd/drm2/dist/drm/radeon/radeon_rv770.c: revision 1.3 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/disp/nouveau_nvkm_engine_disp_sorgm200.c: revision 1.3 sys/external/bsd/common/include/linux/printk.h: revision 1.14 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/instmem/nouveau_nvkm_subdev_instmem_gk20a.c: revision 1.10 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_vi.c: revision 1.4 sys/external/bsd/drm2/include/linux/acpi.h: revision 1.11 sys/external/bsd/drm2/drm/drm_cdevsw.c: revision 1.31 sys/external/bsd/drm2/dist/drm/radeon/radeon_si.c: revision 1.6 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_acpi.c: revision 1.5 sys/external/bsd/drm2/dist/drm/i915/display/intel_acpi.h: revision 1.5 sys/external/bsd/drm2/include/acpi/video.h: revision 1.3 sys/external/bsd/drm2/dist/drm/radeon/radeon_evergreen.c: revision 1.6 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_acpi.h: revision 1.4 sys/arch/sparc64/include/pci_machdep.h: revision 1.31 sys/arch/sparc64/dev/pci_machdep.c: revision 1.83 sys/external/bsd/drm2/include/linux/kref.h: revision 1.14 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_pci.c: revision 1.12 sys/external/bsd/drm2/linux/linux_dma_buf.c: revision 1.17 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/subdev/bios/nouveau_nvkm_subdev_bios_shadowacpi.c: revision 1.4 sys/external/bsd/drm2/drm/drm_module.c: revision 1.32 sys/external/bsd/drm2/dist/drm/i915/i915_gem.h: revision 1.8 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_dmabuf.c: revision 1.7 sys/external/bsd/drm2/include/linux/smp.h: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_si.c: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_device.c: revision 1.20 sys/arch/x86/x86/bus_dma.c: revision 1.91 sys/external/bsd/drm2/radeon/files.radeon: revision 1.40 sys/external/bsd/drm2/include/acpi/acpi_bus.h: revision 1.1 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_drv.h: revision 1.5 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_device.c: revision 1.21 sys/external/bsd/common/include/asm/barrier.h: revision 1.20 sys/external/bsd/drm2/include/linux/nbsd-namespace-acpi.h: revision 1.2 sys/external/bsd/common/include/asm/barrier.h: revision 1.21 sys/modules/drmkms/drmkms_pci.h: revision 1.1 sys/external/bsd/drm2/dist/drm/drm_dp_helper.c: revision 1.17 sys/external/bsd/drm2/radeon/radeon_pci.c: revision 1.23 sys/external/bsd/drm2/linux/linux_xa.c: revision 1.4 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.23 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.24 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.25 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.26 sys/dev/pci/pcivar.h: revision 1.120 sys/arch/xen/include/pci_machdep.h: revision 1.24 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.26 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.27 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.27 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.28 sys/external/bsd/drm2/dist/drm/radeon/radeon_cik.c: revision 1.8 sys/external/bsd/drm2/ttm/ttm_bo_vm.c: revision 1.28 sys/external/bsd/drm2/linux/linux_pci.c: revision 1.29 sys/external/bsd/drm2/include/linux/pci.h: revision 1.57 sys/external/bsd/drm2/include/linux/pci.h: revision 1.58 sys/external/bsd/drm2/dist/drm/radeon/radeon_acpi.c: revision 1.5 sys/external/bsd/drm2/dist/drm/nouveau/nouveau_display.c: revision 1.6 sys/external/bsd/drm2/dist/drm/amd/powerplay/hwmgr/amdgpu_hwmgr.c: revision 1.3 sys/external/bsd/drm2/dist/drm/amd/display/dc/core/amdgpu_dc_stream.c: revision 1.3 share/man/man9/bus_dma.9: revision 1.69 sys/external/bsd/drm2/drm/drm_gem_cma_helper.c: revision 1.15 sys/external/bsd/drm2/dist/drm/radeon/radeon_acpi.c: revision 1.6 sys/external/bsd/drm2/dist/drm/radeon/radeon.h: revision 1.12 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_cik.c: revision 1.7 sys/dev/acpi/acpi_mcfg.c: revision 1.29 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu_acpi.c: revision 1.6 sys/external/bsd/drm2/dist/drm/radeon/radeon_r600.c: revision 1.7 sys/external/bsd/drm2/dist/drm/radeon/radeon_bios.c: revision 1.13 sys/modules/amdgpu/Makefile: revision 1.9 sys/external/bsd/drm2/dist/drm/radeon/radeon_bios.c: revision 1.14 sys/external/bsd/common/linux/linux_tasklet.c: revision 1.12 sys/external/bsd/drm2/dist/drm/nouveau/include/nvkm/core/device.h: revision 1.10 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.23 sys/external/bsd/drm2/dist/drm/amd/amdgpu/amdgpu.h: revision 1.9 sys/external/bsd/drm2/include/linux/interval_tree.h: revision 1.14 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.26 sys/external/bsd/drm2/dist/drm/amd/powerplay/hwmgr/amdgpu_smu7_hwmgr.c: revision 1.5 sys/dev/pci/pci.c: revision 1.168 sys/external/bsd/drm2/dist/drm/i915/gem/i915_gem_mman.c: revision 1.27 sys/external/bsd/drm2/dist/drm/radeon/radeon_si_dpm.c: revision 1.9 sys/external/bsd/drm2/pci/files.drmkms_pci: revision 1.18 sys/external/bsd/drm2/linux/linux_sync_file.c: revision 1.3 sys/external/bsd/drm2/amdgpu/files.amdgpu: revision 1.31 sys/external/bsd/drm2/dist/drm/nouveau/nvkm/engine/device/nouveau_nvkm_engine_device_tegra.c: revision 1.4 sys/external/bsd/drm2/dist/drm/drm_gem.c: revision 1.24 sys/arch/xen/xen/xpci_xenbus.c: revision 1.29 drm: Eliminate __HAVE_ATOMIC_AS_MEMBAR conditionals. Discussed on tech-kern: https://mail-index.netbsd.org/tech-kern/2023/02/23/msg028729.html linux asm/barrier.h: Fix !MULTIPROCESSOR build. remove "nouveau" from a comment. noted by jmcneill. drm: KASSERT(A && B) -> KASSERT(A); KASSERT(B) comment a function that has a clear overbounds read but it isn't used. found by GCC 12. nix the NetBSD specific GEM_BUG_ON(). avoids GCC 12 warnings, and matches upstream closer. avoid uninitialised variable usage in drm_gem_cma_create_internal(). in the case nothing has returned 'error', 'nsegs' and the dma info are (potentially) uninitialised, so consider this an error. found by GCC 12. avoid a GCC 12 warning. there's a 1-element long array and a loop conditional that tries to see if indexes for it are not identical. as these indexes will always both be 0, the only valid index, the condition is always false. GCC 12 triggers a strange warning on this code that can never run (see below), so simply assert the array size is 1 and comment the rest. amdgpu_dc_stream.c:470:55: error: array subscript [0, 0] is outside array bounds of 'struct dc_writeback_info[1]' [-Werror=array-bounds] 470 \| stream->writeback_info[j] = stream->writeback_info[i]; convert a KASSERT() into an if () panic() sequence to appease GCC 12. OK riastradh@. drm: Fix conditionals around drmkms_pci and agp. Kernel should build now with all pci drm drivers stripped out but DRM_LEGACY still enabled. (Might not be very useful, but it'll build. Maybe we should also have DRM_LEGACY_PCI so those drivers can be modloaded later.) drmkms: Fix module build. avoid an unlikely array bounds issue picked up by GCC 12. nvkm_pcie_speed() can return -1, which is then used as an array index, so make this default return PCIe 1.0 speeds. drm: enable almost all PCIe functionality linux_pci.c revisions 1.24 and 1.25 implemented most of the remaining missing PCIe backends, but only enabled them for some amdgpu portions. this enables all code marked with "XXX amdgpu pcie", "XXX radeon pcie", and "XXX pcie speed". for most of it, simply removing #ifndefs __NetBSD__ to enable compliation was required, once the new "bus->max_bus_speed" member was added to struct pci_bus. add an "always fails" backend for pci_enable_atomic_ops_to_root() which seems to only be necessary for virtual GPU functionality (and could be implemented if needed.) tested on radeon 5450, 7750, R7 240 [radeon], and RX 550 [amdgpu], and nvidia 750 and 1030 [nouveau]. this still does not quite work on nvidia cards. there are two problems that remain: - the call to set the link speed is skipped because the speed is set to the default value of "-1". nvkm_pcie_set_link() will actually determine the right value for this and for some cards, calling this function if the current speed is -1 helps set the link speed. it may be that on linux other paths we don't have enabled properly would set this (there's one via debugfs, and a jetson specific one, though perhaps setting either AC or DC speed values as boot options (after hooking up these for netbsd) would currently work. - worse, cards newer than kepler - geforce 900, 1000, and newer, are all lacking the backing support to set pcie link speed. the GT 1030 card i have been testing with remains at pcie 1.0. radeon: fix and enable ACPI methods for getting ROM BIOS The hacky way of getting the BIOS mapped only works on x86. ACPI should be preferred if available. Makes BIOS reading though VFCT work on aarch64 with EDK2. (But only if EDK2 has POSTed the GPU.) XXX amdgpu should get the same treatment. drm: put_cpu() should enable preemption, not disable it again drm(4): make pr_debug equivalent to aprint_debug significantly reduces the default spam from amdgpu(4). drm: Set CONFIG_ACPI in linux/acpi.h and make it build. Leave a little ACPI-related functionality disabled for now, like getting EDID out of ACPI -- needs a bit more work to make this work, and I don't have hardware to work on that. Should help with failures of the forms: - unable to locate a BIOS ROM - bios: unable to locate usable image on various machines. radeon_acpi.c: ifdef out unused function on NetBSD. Should fix syzkaller build. drm(4): Fix st_rdev in stat. dminor->index already has the 64*type adjustment, as allocated in drm_minor_alloc. PR kern/58180 linux_sync_file: Fix missing init/fini steps. Noted by rjs@. PR kern/58210 ttm: Sync ttm_bo_uvm_fault_idle better with Linux. PR xsrc/58133 ttm: Undo mistake in previous. PR xsrc/58133 linux: Add a few more cases to pci_get_class. Should fix crash on boot with amdgpu now that the ACPI business is enabled. i915: Fix dmabuf mmap object. drm: Fix missing bounds checks in dma buf mmap. drm_gem.c: Fix sense of assertion. This is the opposite of WARN_ON. Noted by rjs@. drm_gem.c: Enable drm_gem_fence_array_add now that we emulate xa. linux_xa: Delete and replace collision in xa_store as intended. Don't free the colliding node that's still in the tree. Noted by rjs@. i915_gem_mman.c: Apply mmap types via pmap flags. This way, userland gets buffers mapped write-combining or uncached as needed. PR xsrc/58307 x86: Teach bus_dmamem_map about BUS_DMA_PREFETCHABLE. PR port-amd64/58308 bus_dma(9): Document BUS_DMA_PREFETCHABLE. Like BUS_DMA_NOCACHE. Doesn't absolve you of the need for bus_dmamap_sync, but if you later pass the vaddr to bus_dmamap_load, the DMA map might notice the mapping is write-combining and use this to make bus_dmamap_sync cheaper. PR kern/58309 nouveau_nvkm_subdev_instmem_gk20a.c: Use BUS_DMA_PREFETCHABLE. Matches Linux's pgprot_writecombine. Unclear where the appropriate bus_dmamap_sync happens, or is supposed to happen -- not using it would be wrong, but asking for a prefetchable mapping may paper over symptoms, at least! ttm: Sync more with Linux. Add the original copyright and attribution since this is now, intentionally, a modified copy of the original and not just roughly the same algorithm. ttm: Respect PGO_ALLPAGES. Not sure this is useful but it reduces XXX's and makes this match udv_fault better so it's easier to understand. ttm: Sync cacheability flag logic with Linux. ttm: Add XXX about readahead fault failures. pci: Pass cookie through pci_find_device, pci_enumerate_bus, take 2. New functions pci_find_device1 and pci_enumerate_bus1 have the cookie argument. Existing symbols pci_find_device and pci_enumerate_bus are now wrappers for the cookieless version. This will allow pci_find_device callers to pass a cookie through to the match function so they can keep state or pass in extra parameters like b/d/f numbers, which will allow us to nix some horrible kludges in the Linux PCI API emulation for drm (and, perhaps, Intel wifi). This change drops the symbol pci_probe_device, in favour of a new pci_probe_device1 with the cookie argument. But I don't think that requires a revbump because it's only called by MD pci_enumerate_bus1 implementations, which don't live in modules anyway. Take 2: Make sure to handle NULL match function. linux_pci: Nix pci enumeration kludges. Now that we can pass a cookie through, this stuff will be a little less fragile. i915: Omit needless i915_gem_object_pin/unpin_pages cycle in fault. vm_fault_cpu and vm_fault_gtt, called by i915_gem_fault, already do the pinning and unpinning internally, so there is no need for i915_gem_fault to do it. No functional change intended, except that the transient pin count will be one lower than before during the fault routine (but it will still be positive). i915: Match Linux fault routine return code actions. Omit needless EINTR interception -- this is now handled by i915_error_to_vmf_fault. Earlier revert was over a false alarm -- bisection shows the new warnings arose from linux_pci.c 1.29 here: https://mail-index.netbsd.org/source-changes/2024/06/23/msg151929.html linux_pci: Fix shifto in pci_get_class. It looks like Linux's pci_get_class also matches the interface part of the PCI class register (but not the revision part), and I hadn't noticed that in the previous shim structured differently. With GCC12 kernel ALL/amd64 triggers "'sor' may be used uninitialized". If "sublinks & 3" is zero GCC is right and sor[1] may be returned unitialized. Fix by initializing "sor" to zero to return -1 instead of uninitialized value. Ok: Taylor R Campbell <riastradh@> amdgpu: Map BAR 2, not BAR 5, on pre-bonaire chips. PR kern/58384 amdgpu: Map consecutive pages, not the same one over and over again. PR kern/58385 linux/bitops: Fix overestimate for BITS_TO_LONGS(9) Fortunately, this seems harmless except for allocating excessive buffer memory. Pointed out by nonaka@, OK riastradh@.
1.50	17-May-2023	tsutsui	Update URL of the technical manual.
1.49	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.48	22-Dec-2017	tsutsui	branches: 1.48.2; 1.48.4; Update a URL of Am53c974A technical manual.
1.47	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.46	30-Jan-2012	drochner	branches: 1.46.6; 1.46.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.45	13-Nov-2010	uebayasi	branches: 1.45.8; 1.45.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.44	28-Aug-2008	tsutsui	branches: 1.44.16; Restore Technical manual's URL.
1.43	28-Apr-2008	martin	branches: 1.43.2; 1.43.6; Remove clause 3 and 4 from TNF licenses
1.42	13-Apr-2008	tsutsui	branches: 1.42.2; 1.42.4; Split device_t/softc for MI ncr53c9x and some related devices, with various cleanup.
1.41	10-Apr-2008	cegger	use aprint_*_dev and device_xname
1.40	19-Oct-2007	ad	branches: 1.40.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.39	04-Mar-2007	christos	branches: 1.39.2; 1.39.14; 1.39.16; 1.39.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.38	16-Nov-2006	christos	branches: 1.38.4; __unused removal on arguments; approved by core.
1.37	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.36	28-Apr-2006	tsutsui	branches: 1.36.8; 1.36.10; Free one more resource on attach failure.
1.35	11-Apr-2006	tsutsui	Free resources if attach fails.
1.34	11-Dec-2005	christos	branches: 1.34.4; 1.34.6; 1.34.8; 1.34.10; 1.34.12; merge ktrace-lwp.
1.33	28-Jun-2005	thorpej	branches: 1.33.2; Use ANSI function decls and static.
1.32	21-Jan-2005	tsutsui	Unload xfer dmamap in pscsp_dma_stop().
1.31	02-Jan-2005	tsutsui	branches: 1.31.2; u_intNN_t -> uintNN_t
1.30	25-Sep-2004	tsutsui	Use ANSI function decls.
1.29	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.28	23-Nov-2003	tsutsui	Call bus_space_{read,write}_1() directly to access ncr53c9x common registers.
1.27	16-Nov-2003	tsutsui	Misc cleanup: - tweak some attach messages - establish interrupt after all dma resouces are successfully allocated - wrap an assertion with #ifdef DIAGNOSTIC - assume NCRCFG2_FE bit is always enabled - call bus_dmamap_sync for mdldmap at POSTWRITE for consistency - pass proper dmamapsize for mdldmap
1.26	19-Oct-2003	tsutsui	Utilize bus_dma(9) functions more efficiently. pcscp(4) is the first device driver I wrote about five years ago, but I still have to learn about bus_dma(9) implementation...
1.25	12-Oct-2003	tsutsui	- Update technical manual's URL. - Remove __P(). - Remove trailing whitespace.
1.24	03-May-2003	wiz	branches: 1.24.2; DMA, not dma nor Dma.
1.23	28-Dec-2002	tsutsui	sparc/tab cosmetics.
1.22	25-Nov-2002	tsutsui	- Remove code to map PCI_MAPREG_TYPE_MEM, which is not supported by 53c974. - Print PCI devinfo in the attachment message.
1.21	15-Nov-2002	tsutsui	Remove an entry for AMD PCNETS_PCI. AMD Am79C974 PCnet-SCSI Ethernet and SCSI conrtoller is a multi-function PCI device which has two device IDs of Am53C974(PCscsi-PCI) and Am79C970(PCnet-PCI), and there is no its own ID.
1.20	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.19	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.18	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.17	13-Nov-2001	lukem	add RCSID
1.16	19-Jul-2001	thorpej	Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.15	25-Apr-2001	bouyer	branches: 1.15.2; Pull up the thorpej_scsipi branch to main branch. This is a completely rewritten scsipi_xfer execution engine, and the associated changes to HBA drivers. Overview of changes & features: - All xfers are queued in the mid-layer, rather than doing so in an ad-hoc fashion in individual adapter drivers. - Adapter/channel resource management in the mid-layer, avoids even trying to start running an xfer if the adapter/channel doesn't have the resources. - Better communication between the mid-layer and the adapters. - Asynchronous event notification mechanism from adapter to mid-layer and peripherals. - Better peripheral queue management: freeze/thaw, sorted requeueing during recovery, etc. - Clean separation of peripherals, adapters, and adapter channels (no more scsipi_link). - Kernel thread for each scsipi_channel makes error recovery much easier (no more dealing with interrupt context when recovering from an error). - Mid-layer support for tagged queueing: commands can have the tag type set explicitly, tag IDs are allocated in the mid-layer (thus eliminating the need to use buggy tag ID allocation schemes in many adapter drivers). - support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command will be requeued, or a REQUEST SENSE will be sent as appropriate. Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
1.14	29-Mar-2001	petrov	use NCR_F_DMASELECT instead ncr53c9x_dmaselect
1.13	08-Mar-2001	thorpej	Add BUS_DMA_STREAMING to data transfer map loads.
1.12	28-Dec-2000	sommerfeld	branches: 1.12.2; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.11	14-Nov-2000	thorpej	NBPG -> PAGE_SIZE
1.10	05-Jun-2000	tsutsui	Adapt MI ncr53c9x changes.
1.9	02-Jun-2000	tsutsui	KNF some lines.
1.8	20-Mar-2000	tsutsui	branches: 1.8.2; Use sc_cfg4 for the Am53c974.
1.7	14-Jan-2000	tsutsui	minor cosmetic
1.6	12-Dec-1999	thorpej	Use htole32().
1.5	30-Sep-1999	thorpej	branches: 1.5.2; 1.5.8; Update for SCSPI changes.
1.4	22-Sep-1999	mhitch	Set sc_cfg3_fscsi for the AMD53C974 to match the change in ncr53c9x.c.
1.3	25-Apr-1999	thorpej	Fix a couple of problems in the PCscsi-PCI driver: * resid in pcscp_dma_intr() should also be set in the data out phase. * Don't set up DMA in the transfer pad operations. * Change URL of the PDF technical manual to the index page. * include <machine/bswap.h> on big endian machines. Fixes timeouts writing large blocks to tapes. From Izumi Tsutsui, PR 7252.
1.2	08-Jan-1999	thorpej	branches: 1.2.2; One more cosmetic change, from Izumi Tsutsui.
1.1	06-Jan-1999	thorpej	Device driver for the AMD Am53c974 PCscsi-PCI SCSI chip (i.e. the chip used on the Tekram DC-390 SCSI cards), from Izumi Tsutsui, PR#6654. Minor cosmetic changes by me.
1.2.2.2	04-Dec-1999	he	Pull up revision 1.4 (requested by mhitch): Fix the problems in detecting the clock speed on the TCDS Dual SCSI Option cards, and correctly configure the Fast SCSI interfaces. This allows installation on the 3000/500 -> 3000/900 Alphas which have the Fast SCSI interface on the on-board SCSI adapters.
1.2.2.1	26-Apr-1999	perry	branches: 1.2.2.1.2; pullup 1.2->1.3 (thorpej)
1.2.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.5.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.5.2.8	21-Apr-2001	bouyer	Sync with HEAD
1.5.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.5.2.6	22-Jan-2001	bouyer	make it compile
1.5.2.5	05-Jan-2001	bouyer	Sync with HEAD
1.5.2.4	22-Nov-2000	bouyer	Sync with HEAD.
1.5.2.3	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.2.2	20-Oct-1999	thorpej	Eliminate sleeping from adapter routines.
1.5.2.1	19-Oct-1999	thorpej	Adapt to scsipi API changes.
1.8.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.12.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.12.2.6	11-Dec-2002	thorpej	Sync with HEAD.
1.12.2.5	18-Oct-2002	nathanw	Catch up to -current.
1.12.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.12.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.12.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.12.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.15.2.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.15.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.15.2.1	03-Aug-2001	lukem	update to -current
1.24.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.24.2.6	24-Jan-2005	skrll	Sync with HEAD.
1.24.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.24.2.4	19-Oct-2004	skrll	Sync with HEAD
1.24.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.24.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.24.2.1	03-Aug-2004	skrll	Sync with HEAD
1.31.2.1	29-Apr-2005	kent	sync with -current
1.33.2.4	27-Oct-2007	yamt	sync with head.
1.33.2.3	03-Sep-2007	yamt	sync with head.
1.33.2.2	30-Dec-2006	yamt	sync with head.
1.33.2.1	21-Jun-2006	yamt	sync with head.
1.34.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.34.10.2	11-May-2006	elad	sync with head
1.34.10.1	19-Apr-2006	elad	sync with head.
1.34.8.1	24-May-2006	yamt	sync with head.
1.34.6.2	01-Jun-2006	kardel	Sync with head.
1.34.6.1	22-Apr-2006	simonb	Sync with head.
1.34.4.1	09-Sep-2006	rpaulo	sync with head
1.36.10.2	10-Dec-2006	yamt	sync with head.
1.36.10.1	22-Oct-2006	yamt	sync with head
1.36.8.1	18-Nov-2006	ad	Sync with head.
1.38.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.39.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.39.16.1	06-Nov-2007	matt	sync with HEAD
1.39.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.39.2.1	23-Oct-2007	ad	Sync with head.
1.40.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.40.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.42.4.2	04-May-2009	yamt	sync with head.
1.42.4.1	16-May-2008	yamt	sync with head.
1.42.2.1	18-May-2008	yamt	sync with head.
1.43.6.1	19-Oct-2008	haad	Sync with HEAD.
1.43.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.44.16.1	05-Mar-2011	rmind	sync with head
1.45.12.1	18-Feb-2012	mrg	merge to -current.
1.45.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.45.8.1	17-Apr-2012	yamt	sync with head
1.46.10.1	18-May-2014	rmind	sync with head
1.46.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.48.4.1	10-Jun-2019	christos	Sync with HEAD
1.48.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file pcscp.c.orig was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1	20-Nov-2000	bouyer	branches: 1.1.2; file pcscp.c.rej was initially added on branch thorpej_scsipi.
1.1.2.2	20-Nov-2000	bouyer	Remove files that should not have been added
1.1.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.1	06-Jan-1999	thorpej	branches: 1.1.140; 1.1.142; 1.1.144; Device driver for the AMD Am53c974 PCscsi-PCI SCSI chip (i.e. the chip used on the Tekram DC-390 SCSI cards), from Izumi Tsutsui, PR#6654. Minor cosmetic changes by me.
1.1.144.1	16-May-2008	yamt	sync with head.
1.1.142.1	18-May-2008	yamt	sync with head.
1.1.140.1	02-Jun-2008	mjf	Sync with HEAD.
1.35	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.34	31-Jul-2012	bouyer	branches: 1.34.2; 1.34.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.33	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.32	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.31	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.30	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.29	04-Apr-2011	dyoung	branches: 1.29.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.28	05-Nov-2010	jakllsch	branches: 1.28.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.27	19-Oct-2009	bouyer	branches: 1.27.2; 1.27.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.26	18-Mar-2008	cube	branches: 1.26.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.25	09-Feb-2007	ad	branches: 1.25.20; 1.25.36; 1.25.40; Merge newlock2 to head.
1.24	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.23	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.22	11-Dec-2005	christos	branches: 1.22.20; 1.22.22; merge ktrace-lwp.
1.21	24-May-2005	lukem	branches: 1.21.2; add __KERNEL_RCSID
1.20	27-Feb-2005	perry	nuke trailing whitespace
1.19	23-Aug-2004	bouyer	branches: 1.19.4; 1.19.6; Revert 1.12, it cause problems on promise controllers, as reported in PR kern/26568 by Paul Shupak.
1.18	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.17	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.16	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.15	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.14	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.13	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.12	02-Aug-2004	bouyer	branches: 1.12.2; If acknowledge "bogus" interrupts (that is, interrupts that have happened according to the controller-dependant registers while wdc was not waiting for an interrupt). A reset can generate spurious interrupts (or even a flow of spurious interrupts on the promise) that needs to be properly acknowledged.
1.11	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.10	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.9	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.8	22-Nov-2003	fvdl	Use the redefined Promise product ids.
1.7	15-Nov-2003	bouyer	Clear interrupts at end of attach, in case there is one pending. Avoid "bogus intr" messages when interrupts are enabled later. Patch tested by myself and Chuck Silvers.
1.6	29-Oct-2003	mycroft	Simplify a little.
1.5	25-Oct-2003	christos	Fix uninitialized variable warnings
1.4	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.3	22-Oct-2003	fvdl	Match the Promise FastTrak 378.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.12.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.12.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.12.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.12.2.2	03-Aug-2004	skrll	Sync with HEAD
1.12.2.1	02-Aug-2004	skrll	file pdcide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.19.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.19.4.1	29-Apr-2005	kent	sync with -current
1.21.2.3	24-Mar-2008	yamt	sync with head.
1.21.2.2	26-Feb-2007	yamt	sync with head.
1.21.2.1	30-Dec-2006	yamt	sync with head.
1.22.22.2	10-Dec-2006	yamt	sync with head.
1.22.22.1	22-Oct-2006	yamt	sync with head
1.22.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.22.20.1	18-Nov-2006	ad	Sync with head.
1.25.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.25.36.1	24-Mar-2008	keiichi	sync with head.
1.25.20.1	23-Mar-2008	matt	sync with HEAD
1.26.4.1	11-Mar-2010	yamt	sync with head
1.27.4.2	21-Apr-2011	rmind	sync with head
1.27.4.1	05-Mar-2011	rmind	sync with head
1.27.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.28.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.29.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.29.4.1	30-Oct-2012	yamt	sync with head
1.34.4.1	18-May-2014	rmind	sync with head
1.34.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.34.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.32	03-Sep-2021	andvar	fix typos in comments, mainly s/extention/extension/ and s/sufficent/sufficient/
1.31	03-Feb-2019	mrg	- add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.30	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.29	20-Oct-2017	jdolecek	branches: 1.29.2; 1.29.4; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.28	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.27	29-Mar-2014	christos	branches: 1.27.18; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.26	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.25	31-Jul-2012	bouyer	branches: 1.25.2; 1.25.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.24	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.23	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.22	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.21	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.20	04-Apr-2011	dyoung	branches: 1.20.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.19	05-Nov-2010	jakllsch	branches: 1.19.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.18	26-Nov-2009	njoly	branches: 1.18.2; 1.18.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.17	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.16	18-Mar-2008	cube	branches: 1.16.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.15	09-Feb-2007	ad	branches: 1.15.20; 1.15.36; 1.15.40; Merge newlock2 to head.
1.14	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.13	25-Oct-2006	bouyer	Map the SATA registers for SATA channels, and probe drives using them instead of using the old PATA way. Tested with a PDC20375 (2xSATA + 1xPATA). While there add the PDC20618-621 products (Ultra/133 controllers); untested. Yes, it's strange to support PATA-only devices in a driver called pdcsata, but that's how it is ...
1.12	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.11	03-Sep-2006	xtraeme	branches: 1.11.2; 1.11.4; Add support for PDC40519 and PDC40779... from obsd.
1.10	13-Aug-2006	dsainty	Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.9	26-Jun-2006	xtraeme	branches: 1.9.2; Add missing __KERNEL_RCSID()
1.8	26-Jun-2006	xtraeme	Add support for PDC20775 and PDC40518.
1.7	20-Jun-2006	christos	PR/33445: Brian Buhrow: Fix a number of problems with the pdcsata driver: > The driver for the family of Promise SATA controllers, > /usr/src/sys/dev/pci/pdcsata.c is not very robust when it comes to handling > transient drive errors, or interrupt hickups when the card is under load. > Worse, my experience seems to indicate, and the Linux driver confirms, > that these cards tend to fall over rather frequently during high load > operations or if drives unexpectedly reset or go to sleep. Symptoms > include interupt timeouts during heavy load, the inability to reset drives > if they go to sleep, and a failure of the card to generate interrupts at > all if the interrupt load gets too high.
1.6	26-Jan-2006	bouyer	branches: 1.6.2; 1.6.6; 1.6.14; Add support for Promise PDC2057x SATA/150 Ide controllers. From FreeBSD, tested by Douglas Wade Needham.
1.5	16-Jan-2006	bouyer	properly use ata_channel->ch_ndrive: - initialize it properly in the bus front-ends (all 2, exept in wdc_pcmcia.c for the "Sandisk CompactFlash Card" where it's set to 1) - remplace hardcoded '2' by ata_channel->ch_ndrive in MI IDE drivers. From Christos Zoulas in kern/32501.
1.4	04-Dec-2005	christos	branches: 1.4.2; PR/32238: Ryo Shimizu: Add support for Promise SATA300TX4 SATA Controller (PDC40718)
1.3	27-Feb-2005	perry	branches: 1.3.2; 1.3.4; 1.3.6; nuke trailing whitespace
1.2	28-Nov-2004	bouyer	branches: 1.2.2; 1.2.6; 1.2.8; Add RAID support, from Shin'ichiro TAYA on current-users@.
1.1	24-Nov-2004	bouyer	pdcsata(4), a driver for the Promise SATA150 (aka PDC203xx) serie of controllers. Tested with a PDC20375 (2 SATA, one PATA) controller on sparc64. Added to all kernel config file which had pdcide(4).
1.2.8.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.6.1	29-Apr-2005	kent	sync with -current
1.2.2.4	11-Dec-2005	christos	Sync with head.
1.2.2.3	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.2.2	29-Nov-2004	skrll	Sync with HEAD.
1.2.2.1	28-Nov-2004	skrll	file pdcsata.c was added on branch ktrace-lwp on 2004-11-29 07:24:17 +0000
1.3.6.4	24-Mar-2008	yamt	sync with head.
1.3.6.3	26-Feb-2007	yamt	sync with head.
1.3.6.2	30-Dec-2006	yamt	sync with head.
1.3.6.1	21-Jun-2006	yamt	sync with head.
1.3.4.3	07-Apr-2005	tron	Pull up revision 1.2 via patch (requested by bouyer in ticket #1037): Add RAID support, from Shin'ichiro TAYA on current-users@.
1.3.4.2	02-Apr-2005	he	Pull up revision 1.1 (via patch, requested by bouyer in ticket #1019): Add pdcsata(4), a driver for the Promise SATA150 (aka PDC203xx) series of controllers.
1.3.4.1	27-Feb-2005	he	file pdcsata.c was added on branch netbsd-2 on 2005-04-02 21:24:31 +0000
1.3.2.4	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #1466): sys/dev/pci/pcidevs: revision 1.798, 1.819 sys/dev/pci/pdcsata.c: revision 1.8, 1.10 Add support for PDC20775 and PDC40518. Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.3.2.3	28-Jul-2006	riz	Pull up following revision(s) (requested by simonb in ticket #1430): sys/dev/pci/pdcsata.c: revision 1.7 PR/33445: Brian Buhrow: Fix a number of problems with the pdcsata driver: The driver for the family of Promise SATA controllers, /usr/src/sys/dev/pci/pdcsata.c is not very robust when it comes to handling transient drive errors, or interrupt hickups when the card is under load. Worse, my experience seems to indicate, and the Linux driver confirms, that these cards tend to fall over rather frequently during high load operations or if drives unexpectedly reset or go to sleep. Symptoms include interupt timeouts during heavy load, the inability to reset drives if they go to sleep, and a failure of the card to generate interrupts at all if the interrupt load gets too high. --
1.3.2.2	05-Feb-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1156): sys/dev/pci/pdcsata.c: revision 1.6 Add support for Promise PDC2057x SATA/150 Ide controllers. From FreeBSD, tested by Douglas Wade Needham.
1.3.2.1	21-Jan-2006	snj	Pull up following revision(s) (requested by simonb in ticket #1126): sys/dev/pci/pcidevs: revision 1.739 sys/dev/pci/pdcsata.c: revision 1.4 PR/32238: Ryo Shimizu: Add support for Promise SATA300TX4 SATA Controller (PDC40718)
1.4.2.1	01-Feb-2006	yamt	sync with head.
1.6.14.1	13-Jul-2006	gdamore	Merge from HEAD.
1.6.6.4	14-Sep-2006	yamt	sync with head.
1.6.6.3	03-Sep-2006	yamt	sync with head.
1.6.6.2	11-Aug-2006	yamt	sync with head
1.6.6.1	26-Jun-2006	yamt	sync with head.
1.6.2.1	09-Sep-2006	rpaulo	sync with head
1.9.2.2	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #115): share/man/man4/pdcsata.4: revision 1.5 sys/dev/pci/pcidevs: revision 1.827 sys/dev/pci/pcidevs: revision 1.828 sys/dev/pci/pcidevs: revision 1.829 sys/dev/pci/pdcsata.c: revision 1.11 Add some ICH8 devices... from obsd. Add some missing devices from Promise and sort them. Add missing PDC20265 device... removed accidentally in previous. Add support for PDC40519 and PDC40779... from obsd. Mention PDC40519 and PDC40779... and bump date.
1.9.2.1	15-Aug-2006	ghen	Pull up following revision(s) (requested by xtraeme in ticket #28): sys/dev/pci/pcidevs: revision 1.819 sys/dev/pci/pcidevs.h: revision 1.819 sys/dev/pci/pcidevs_data.h: revision 1.818 (via patch) sys/dev/pci/pdcsata.c: revision 1.10 Add the Promise PDC20771, present on the Promise FastTrak TX2300 Add support for the Promise PDC20771, present on the Promise FastTrak TX2300. This support is not based on a datasheet, because a datasheet is not readily available for this chip. However, Promise have partially open sourced their driver for Linux, and all suggestions are that the PDC20771 is pretty similar to other recent SATA chips. The TX2300 has two ports, but there is unoccupied space on the board for a third PATA port. It isn't entirely obvious how many channels the PDC20771 can support. The pdc205xx_drv_probe probe is necessary to avoid probing two wd* devices for every physical device.
1.11.4.2	10-Dec-2006	yamt	sync with head.
1.11.4.1	22-Oct-2006	yamt	sync with head
1.11.2.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.11.2.1	18-Nov-2006	ad	Sync with head.
1.15.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.15.36.1	24-Mar-2008	keiichi	sync with head.
1.15.20.1	23-Mar-2008	matt	sync with HEAD
1.16.4.1	11-Mar-2010	yamt	sync with head
1.18.4.2	21-Apr-2011	rmind	sync with head
1.18.4.1	05-Mar-2011	rmind	sync with head
1.18.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.19.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.20.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.20.4.1	30-Oct-2012	yamt	sync with head
1.25.4.1	18-May-2014	rmind	sync with head
1.25.2.3	03-Dec-2017	jdolecek	update from HEAD
1.25.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.25.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.27.18.2	27-Sep-2017	jdolecek	change wdc_init_shadow_regs() to accept only struct wdc_regs, it doesn't touch anything else factor out the probe-only struct ata_channel initialization to wdcprobe(), to reduce duplication of logic in individual drivers, and to actually work now that more init is needed beyond the memset()
1.27.18.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.29.4.1	10-Jun-2019	christos	Sync with HEAD
1.29.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.67	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.66	26-May-2018	jakllsch	branches: 1.66.2; Match PCI_PRODUCT_INTEL_82Q45_IDER. Not entirely sure which chip map to use, but as this is for AMT, it's probably not like there's a real PATA interface to configure DMA for, and as the 965 AMT stuff uses the sata version of the chip map, we'll cargo cult it for now.
1.65	01-Feb-2016	msaitoh	branches: 1.65.16; Rename macro (sync with pcidevs rev. 1.1246). No functional change.
1.64	14-Nov-2012	jakllsch	branches: 1.64.12; 1.64.14; 1.64.18; enable the generic pciide detach function in piixide(4)
1.63	31-Jul-2012	bouyer	branches: 1.63.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.62	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.61	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.60	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.59	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.58	05-Mar-2012	sborrill	Add support for Intel 6 Series Serial ATA Controller
1.57	04-Apr-2011	dyoung	branches: 1.57.4; 1.57.8; 1.57.10; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.56	05-Nov-2010	jakllsch	branches: 1.56.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.55	30-Jul-2010	njoly	Adjust for Intel ICH10 devices update.
1.54	24-Feb-2010	dyoung	branches: 1.54.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.53	08-Jan-2010	dyoung	branches: 1.53.2; Expand PMF_FN_* macros.
1.52	30-Nov-2009	sborrill	Add support for Intel 3400 SATA
1.51	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.50	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.49	09-Mar-2009	christos	Match Inter 82965PM, from Anon Ymous
1.48	11-Feb-2009	markd	Add some more ICH9 SATA controller variants.
1.47	01-Oct-2008	christos	branches: 1.47.2; 1.47.4; 1.47.8; 1.47.12; ICH10 support from Brad du Plessis
1.46	18-Mar-2008	cube	branches: 1.46.4; 1.46.6; 1.46.10; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.45	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.44	15-Jan-2008	drochner	branches: 1.44.2; 1.44.6; fix some unaligned PCI config space accesses in suspend/resume functions
1.43	04-Jan-2008	joerg	Allow saving of up to 4 PCI registers in the IDE code.
1.42	09-Dec-2007	jmcneill	branches: 1.42.2; Merge jmcneill-pm branch.
1.41	09-Sep-2007	xtraeme	branches: 1.41.6; 1.41.8; 1.41.10; Another ICH8M SATA Controller found in jmcneill's laptop.
1.40	31-Aug-2007	xtraeme	Proper ICH8M IDE Controller support (missed in previous).
1.39	31-Aug-2007	xtraeme	Attach to the ICH8M IDE Controller.
1.38	30-Aug-2007	xtraeme	Attach to the ICH9 SATA devices.
1.37	10-Mar-2007	itohy	branches: 1.37.2; 1.37.10; 1.37.14; 1.37.16; I tried the test code in PR kern/35934 on PIIX4 and ICH6, both of which turned out to have the problem. Enabled (untested) workaround for all PIIX controllers. XXX Do other controllers also have this problem?
1.36	09-Feb-2007	ad	branches: 1.36.2; Merge newlock2 to head.
1.35	01-Dec-2006	cube	branches: 1.35.2; 1.35.4; Now that we have a AHCI driver (thanks a bunch, Manuel!), we can remove the nonsense quirk that switched operating mode on ICH7 and ICH8. I removed the obvious candidates for ahcisata(4), and I'll have a closer look later if there are others to be removed; ahcisata(4) will take over handling the device anyway, but there is no reason to keep AHCI devices in that list. Along the way, remove the code that tries to put the chip in Enhanced mode, it makes absolutely no sense to do that, and some BIOSes might not have prepared the BARs for that, as proven by PR#34885. If people want to use all IDE and SATA channels, they have to tell the BIOS.
1.34	24-Nov-2006	wiz	s/independant/independent/, from Zafer.
1.33	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.32	08-Nov-2006	drochner	avoid magic number
1.31	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.30	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.29	03-Sep-2006	xtraeme	branches: 1.29.2; 1.29.4; * Add support for ICH8 and ICH8M SATA/RAID controllers. * If the controller is in AHCI, ask for SATA IDE mode of operation. jsg@openbsd says: "X60/T60 Thinkpads are shipped in AHCI configuration by default, this makes them work without changing a BIOS option." Tested by eye of the beholder. From OpenBSD. Ok'ed tls.
1.28	08-Aug-2006	cube	Add support for the 631xESB/632xESB.
1.27	17-Jun-2006	jmcneill	branches: 1.27.2; Restore IDE and UDMA timings on resume for the PIIX4.
1.26	19-May-2006	markd	branches: 1.26.2; 1.26.4; Add Intel 82801GBM/GHM Serial ATA Controller.
1.25	11-Dec-2005	christos	branches: 1.25.4; 1.25.6; 1.25.8; 1.25.12; merge ktrace-lwp.
1.24	10-Aug-2005	bouyer	Not add PIIX are wired to compat mode these days. Problem reported and fix tested by Havard Eidnes.
1.23	21-Jul-2005	tron	Add support for PATA and SATA ports in the Intel ICH7 southbridge (82801G). Patch supplied by Marcello Balduccini in PR port-i386/30795.
1.22	20-Jun-2005	briggs	branches: 1.22.2; Match the 6300ESB on-board SATA/RAID controller. Set the ATA RAID flag, if appropriate, in the SATA chip map. Ensure that the interrupts are enabled in the SATA chip map.
1.21	15-Jun-2005	bouyer	Support Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern.
1.20	24-May-2005	lukem	add __KERNEL_RCSID
1.19	27-Feb-2005	perry	branches: 1.19.2; nuke trailing whitespace
1.18	26-Jan-2005	jmcneill	Add powerhooks for PIIX IDE driver.
1.17	10-Nov-2004	cube	branches: 1.17.4; 1.17.6; Add support for the ICH6 IDE controller (82801FB).
1.16	22-Oct-2004	cube	Add support for the 82801FB/FR (ICH6/6R) SATA controllers.
1.15	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.14	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.13	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.12	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.11	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.10	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.9	11-May-2004	thorpej	branches: 1.9.2; Handle the i6300ESB IDE / SATA controllers. These are apparently just ICH5.
1.8	03-Jan-2004	thorpej	branches: 1.8.2; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.7	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.6	14-Dec-2003	thorpej	Split the Intel i31244 SATA controller ("Artisea") driver out of piixide into its own driver (artsata).
1.5	06-Dec-2003	bouyer	Clone artisea_chip_map() for use by the ICH5 SATA controllers. The 31244 is really different from the ICH5, and one day the map function could be incompatibles. Suggested by thorpej@
1.4	05-Dec-2003	bouyer	Add support for the Intel 82801ER Serial ATA/Raid Controller. From Simas Mockevicius on current-users@.
1.3	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.8.2.1	28-Jul-2004	tron	branches: 1.8.2.1.2; Pull up revision 1.9 (requested by joda in ticket #721): Handle the i6300ESB IDE / SATA controllers. These are apparently just ICH5.
1.8.2.1.2.4	22-Aug-2005	riz	Pull up following revision(s) (requested by bouyer in ticket #5602): sys/dev/pci/piixide.c: revision 1.24 Not all PIIX are wired to compat mode these days. Problem reported and fix tested by Havard Eidnes.
1.8.2.1.2.3	21-Jun-2005	riz	Pull up revision 1.21 (requested by bouyer in ticket #2008): Support Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern.
1.8.2.1.2.2	19-May-2005	tron	Pull up revision 1.17 via patch (requested by riz in ticket #1531): Ad support for the ICH6 IDE controller (82801FB).
1.8.2.1.2.1	19-May-2005	tron	Pull up revision 1.16 (requested by riz in ticket #1531): Add support for the 82801FB/FR (ICH6/6R) SATA controllers.
1.9.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.9.2.8	04-Feb-2005	skrll	Sync with HEAD.
1.9.2.7	14-Nov-2004	skrll	Sync with HEAD.
1.9.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.9.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.9.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.9.2.2	03-Aug-2004	skrll	Sync with HEAD
1.9.2.1	11-May-2004	skrll	file piixide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.17.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.17.6.1	12-Feb-2005	yamt	sync with head.
1.17.4.1	29-Apr-2005	kent	sync with -current
1.19.2.5	08-Apr-2008	jdc	Pull up revisions: src/sys/dev/pci/pcidevs 1.816 src/sys/dev/pci/piixide.c 1.28 (requested by bouyer in ticket #1909). Add device IDs for Intel's 631xESB/632xESB southbridges, and start the list for the 5000X northbridge. I'll add the rest later, the spec is currently unavailable so I'll collect them from another source. Add support for the 631xESB/632xESB.
1.19.2.4	24-Aug-2005	tron	Pull up revision 1.24 (requested by bouyer in ticket #692): Not add PIIX are wired to compat mode these days. Problem reported and fix tested by Havard Eidnes.
1.19.2.3	23-Jul-2005	riz	Pull up revision 1.23 (requested by tron in ticket #609): Add support for PATA and SATA ports in the Intel ICH7 southbridge (82801G). Patch supplied by Marcello Balduccini in PR port-i386/30795.
1.19.2.2	03-Jul-2005	tron	Pull up revision 1.22 (requested by briggs in ticket #522): Match the 6300ESB on-board SATA/RAID controller. Set the ATA RAID flag, if appropriate, in the SATA chip map. Ensure that the interrupts are enabled in the SATA chip map.
1.19.2.1	21-Jun-2005	tron	Pull up revision 1.21 (requested by bouyer in ticket #479): Support Intel 82801FBM Serial ATA Controller. From Kurt Schreiner on tech-kern.
1.22.2.8	24-Mar-2008	yamt	sync with head.
1.22.2.7	17-Mar-2008	yamt	sync with head.
1.22.2.6	21-Jan-2008	yamt	sync with head
1.22.2.5	27-Oct-2007	yamt	sync with head.
1.22.2.4	03-Sep-2007	yamt	sync with head.
1.22.2.3	26-Feb-2007	yamt	sync with head.
1.22.2.2	30-Dec-2006	yamt	sync with head.
1.22.2.1	21-Jun-2006	yamt	sync with head.
1.25.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.25.8.4	14-Sep-2006	yamt	sync with head.
1.25.8.3	11-Aug-2006	yamt	sync with head
1.25.8.2	26-Jun-2006	yamt	sync with head.
1.25.8.1	24-May-2006	yamt	sync with head.
1.25.6.1	01-Jun-2006	kardel	Sync with head.
1.25.4.1	09-Sep-2006	rpaulo	sync with head
1.26.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.26.2.2	19-Jun-2006	chap	Sync with head.
1.26.2.1	19-May-2006	chap	file piixide.c was added on branch chap-midi on 2006-06-19 04:01:37 +0000
1.27.2.2	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #117): sys/dev/pci/piixide.c: revision 1.29 sys/dev/pci/pciide_piix_reg.h: revision 1.12 * Add support for ICH8 and ICH8M SATA/RAID controllers. * If the controller is in AHCI, ask for SATA IDE mode of operation. jsg@openbsd says: "X60/T60 Thinkpads are shipped in AHCI configuration by default, this makes them work without changing a BIOS option." Tested by eye of the beholder. From OpenBSD. Ok'ed tls.
1.27.2.1	11-Aug-2006	riz	Pull up following revision(s) (requested by cube in ticket #8): sys/dev/pci/piixide.c: revision 1.28 sys/dev/pci/pcidevs: revision 1.816 Add device IDs for Intel's 631xESB/632xESB southbridges, and start the list for the 5000X northbridge. I'll add the rest later, the spec is currently unavailable so I'll collect them from another source. Add support for the 631xESB/632xESB.
1.29.4.2	10-Dec-2006	yamt	sync with head.
1.29.4.1	22-Oct-2006	yamt	sync with head
1.29.2.3	06-Feb-2007	ad	Quiten noisy boot messages.
1.29.2.2	12-Jan-2007	ad	Sync with head.
1.29.2.1	18-Nov-2006	ad	Sync with head.
1.35.4.1	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.35.2.1	24-Oct-2007	pavel	Pull up following revisions (requested by msaitoh in ticket #935): sys/dev/pci/piixide.c: revision 1.38-1.41 via patch sys/dev/pci/pcidevs: revision 1.900 via patch Attach to the ICH9 SATA devices. Attach to the ICH8M IDE Controller. Proper ICH8M IDE Controller support (missed in previous). Another ICH8M SATA Controller found in jmcneill's laptop.
1.36.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.37.16.3	23-Mar-2008	matt	sync with HEAD
1.37.16.2	09-Jan-2008	matt	sync with HEAD
1.37.16.1	06-Nov-2007	matt	sync with HEAD
1.37.14.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.37.14.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.37.14.4	02-Oct-2007	joerg	Sync with HEAD.
1.37.14.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.37.14.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.37.14.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.37.10.2	10-Sep-2007	skrll	Sync with HEAD.
1.37.10.1	03-Sep-2007	skrll	Sync with HEAD.
1.37.2.1	09-Oct-2007	ad	Sync with head.
1.41.10.1	11-Dec-2007	yamt	sync with head.
1.41.8.1	26-Dec-2007	ad	Sync with head.
1.41.6.1	18-Feb-2008	mjf	Sync with HEAD.
1.42.2.2	19-Jan-2008	bouyer	Sync with HEAD
1.42.2.1	08-Jan-2008	bouyer	Sync with HEAD
1.44.6.2	05-Oct-2008	mjf	Sync with HEAD.
1.44.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.44.2.1	24-Mar-2008	keiichi	sync with head.
1.46.10.1	19-Oct-2008	haad	Sync with HEAD.
1.46.6.1	10-Oct-2008	skrll	Sync with HEAD.
1.46.4.3	11-Aug-2010	yamt	sync with head.
1.46.4.2	11-Mar-2010	yamt	sync with head
1.46.4.1	04-May-2009	yamt	sync with head.
1.47.12.1	21-Apr-2010	matt	sync to netbsd-5
1.47.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.47.4.2	17-Mar-2012	bouyer	Pull up following revision(s) (requested by sborrill in ticket #1739): sys/dev/pci/piixide.c: revision 1.58 sys/dev/pci/pcidevs: patch Add support for Intel 6 Series Serial ATA Controller
1.47.4.1	18-Dec-2009	snj	Pull up following revision(s) (requested by sborrill in ticket #1175): sys/dev/pci/piixide.c: revision 1.52 Add support for Intel 3400 SATA
1.47.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.47.2.1	03-Mar-2009	skrll	Sync with HEAD.
1.53.2.3	06-Nov-2010	uebayasi	Sync with HEAD.
1.53.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.53.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.54.2.2	21-Apr-2011	rmind	sync with head
1.54.2.1	05-Mar-2011	rmind	sync with head
1.56.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.57.10.1	07-Mar-2012	riz	Pull up following revision(s) (requested by sborrill in ticket #98): sys/dev/pci/piixide.c: revision 1.58 Add support for Intel 6 Series Serial ATA Controller
1.57.8.1	06-Mar-2012	mrg	sync to -current
1.57.4.3	16-Jan-2013	yamt	sync with (a bit old) head
1.57.4.2	30-Oct-2012	yamt	sync with head
1.57.4.1	17-Apr-2012	yamt	sync with head
1.63.2.3	03-Dec-2017	jdolecek	update from HEAD
1.63.2.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.63.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.64.18.1	18-Jan-2017	skrll	Sync with netbsd-5
1.64.14.1	19-Mar-2016	skrll	Sync with HEAD
1.64.12.1	09-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1298): sys/dev/pci/piixide.c: revision 1.65 sys/dev/pci/pcidevs: revision 1.1246 - s/82081/82801/ from OpenBSD. - 0x2829 is SATA. 0x282a is RAID. -- Rename macro (sync with pcidevs rev. 1.1246). No functional change.
1.65.16.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.65.16.1	25-Jun-2018	pgoyette	Sync with HEAD
1.66.2.1	10-Jun-2019	christos	Sync with HEAD
1.73	24-Feb-2025	jmcneill	piixpm: Cleanup log message when SMI is used
1.72	12-Jan-2023	msaitoh	branches: 1.72.6; Modify error message to match the current behavior.
1.71	10-Jan-2023	msaitoh	Use MMIO for PM register access on newer revision's devices. Newer revision's device only support memory mapped access. At least, X670E's SMBus device that the revision ID is 0x71 doesn't support I/O access. Use MMIO access for newer devices. Linux uses >= 0x51 for the check, so we do the same check. Note that X570's SMBus(rev. 0x61) supports I/O access...
1.70	09-Jan-2023	msaitoh	Modify for following MMIO support. No functional change.
1.69	09-Jan-2023	msaitoh	Simplify and modify for following MMIO support. No functional change.
1.68	09-Jan-2023	msaitoh	Modify comment. Whitespace. No functional change.
1.67	01-Apr-2022	pgoyette	branches: 1.67.4; Call iic_fini_tag() when the tag is no longer needed. Without this, a LOCKDEBUG kernel can panic when trying to re-init the tag.
1.66	12-Oct-2021	andvar	fix various typos, mainly in comments.
1.65	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.64	24-Apr-2021	thorpej	branches: 1.64.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.63	14-Jan-2020	msaitoh	branches: 1.63.8; Save/restore port number before selecting port. Linux driver says it must be required on some systems.
1.62	14-Jan-2020	msaitoh	Simplify. No functional change.
1.61	09-Jan-2020	msaitoh	Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on ECS HDC-I2. This change fixes the problem.
1.60	24-Dec-2019	thorpej	branches: 1.60.2; Make ichsmb and piixpm MP-safe: - Synchronize with the interrupt handler using a mutex. - Use a condvar to wait for completion, rather than tsleep(). - Mark our interrupt handler as such. Also, other general correctness fixes: - Loop around testing the completion condition to protect aginst spurious wakes. - The "i2c exec" function returns an error code, so actually do so.
1.59	24-Dec-2019	msaitoh	Don't force using SMBUS0SEL register. - Use it depending on USE_SMBUS0SEL bit. - If we use SMBUS0EN_LO register to select the port, update the port select bits only.
1.58	23-Dec-2019	msaitoh	- Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI or 1 on IRQ, so invert the check. - Modify comment. - Whitespace fix.
1.57	23-Dec-2019	msaitoh	Fix number of port for Hudson rev. 0x1f and newer. Same as OpenBSD and Linux.
1.56	23-Dec-2019	thorpej	piixpm_i2c_exec(): No need to check 'cold' to force I2C_F_POLL; the i2c upper layer does it for us.
1.55	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.54	13-Jul-2019	msaitoh	branches: 1.54.2; Improve SB800 and newer chipsets support: - Add newer chipset (e.g. X370/X399 and newer) support that the PCI device id is 0x790b. The register definitions are mainly taken from FreeBSD. - Rename PIIXPM_INDIRECTIO_* to SB800_INDIRECTIO_* because those are only for SB800 and newer chipsets. - SB800 also support 4 ports. - SB800's interrupt configuration bit is different from others. Use SB800_SMB_HOSTC's bit 0. - Do not bus_space_map devices which are at address 0 (it's uninitialized) in piixpm_attach(). - Add the port number to the dmesg output. - Avoid uninitiliazed use of ctl and corresponding warnings. From OpenBSD rev. 1.38
1.53	12-Jul-2019	msaitoh	Sync with OpenBSD's piixpm.c up to rev. 1.35: - Print "polling" correctly when in the attach function. Same as OpenBSD rev. 1.22-23. - Improve debug printf()'s. Part of OpenBSD 1.24. - Do not bus_space_map devices which are at address 0 (it's uninitialized). Same as OpenBSD rev. 1.25. - Add ServerWorks HT1100 device from OpenBSD. Same as OpenBSD rev. 1.32 - Use unique wait channel. From OpenBSD rev. 1.35
1.52	29-Mar-2017	msaitoh	branches: 1.52.6; 1.52.14; Fix 0x%d and 0x%u.
1.51	13-Oct-2016	jdolecek	branches: 1.51.2; provide intr xname
1.50	18-Jul-2016	pgoyette	Insert missing 'break;' statements. CID 1364142
1.49	11-Jul-2016	msaitoh	branches: 1.49.2; KNF. No functional change.
1.48	10-Jul-2016	pgoyette	The piixpm(4) driver also supports AMD's HUDSON chipset's SMBus function. From OpenBSD, verified.
1.47	10-Dec-2015	pgoyette	Initialize the mutex before calling xxx_rescan(). Otherwise if we load/attach an i2cbus at a later time (ie, load the iic module), we'll end-up re-initializing the mutex. (Caught with a LOCKDEBUG kernel)
1.46	03-May-2015	pgoyette	Teach a couple of i2cbus controllers how to rescan. This enables {,un}loading and {at,de}taching of the iic(4) driver/module at a later time. Tested piixpm on QEMU, and ichsmb on my live server.
1.45	22-Jun-2014	hannken	branches: 1.45.4; Consistently pass a "struct piixpm_softc" to piixpm_intr. Prevents a crash on hardware interrupts.
1.44	29-Mar-2014	christos	branches: 1.44.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.43	18-Mar-2014	riastradh	Merge riastradh-drm2 to HEAD.
1.42	22-Jul-2013	soren	The SB800 SMBus controller has four selectable SDA lines. Expose them as four iic busses.
1.41	18-Jul-2013	soren	Clear errors before beginning a transfer. Closes PR port-i386/46792.
1.40	14-Feb-2012	pgoyette	branches: 1.40.2; 1.40.6; 1.40.10; 1.40.18; Replace the xxx_acquire()/xxx_release() rwlocks with mutexes. There are only RW_WRITERs for these, and no RW_READERs, so no need to incur the extra overhead of allowing for both. As discussed on tech-kern. For piixpm and ichsmb, the acquire/release protocol needs to be used, even if the request is I2C_F_POLL'd (or if the device supports only polled mode). Otherwise multiple requests can be running at the same time, and they stomp on each other and create anomolous results. Part 2 addresses my PR kern/45889 3 ACKs from releng
1.39	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.38	07-Jan-2012	pgoyette	Modify the attach message for consistency with most other pci devices.
1.37	03-Oct-2011	jmcneill	branches: 1.37.2; 1.37.6; avoid printing an empty line in the sb800 case
1.36	02-Oct-2011	jmcneill	Newer ATI SB800 SMBus controllers don't report the base address in PCI config space. According to the SB800-Series Southbridges Register Reference Guide, we can still read this value from PM config space using indirect I/O.
1.35	13-Feb-2011	hannken	The ServerWorks CSB5 smbus controller often times out and hangs until reboot. Increase the delay between setting the command register and reading the status register. Reset the smbus controller if it has timed out. Tested on an IBM eServer x335.
1.34	10-Feb-2011	hannken	Make the SMbus Send/Receive functions work. If cmdlen = 0 and len = 1 use the XXX_HC_CMD_BYTE protocol, send data through XXX_SMB_HCMD or receive data from XXX_SMB_HD0. Ok: Paul Goyette
1.33	27-Jul-2010	jakllsch	branches: 1.33.2; 1.33.4; Make PIIXPM_DEBUG work with 64-bit size_t.
1.32	24-Feb-2010	dyoung	branches: 1.32.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.31	08-Jan-2010	dyoung	branches: 1.31.2; Expand PMF_FN_* macros.
1.30	03-Nov-2009	pgoyette	Set iba_type properly. The only thing that ever looks at this is the 'options I2C_SCAN' code in dev/i2c/i2c.c which is disabled by default.
1.29	18-Mar-2009	cegger	bzero -> memset
1.28	13-Feb-2009	pgoyette	Fix dmesg output: Old: piixpm0 at pci0 dev 20 function 0 piixpm0: ATI Technologies SB600/SB700/SB800 SMBus Controller (rev. 0x13) piixpm0: interrupting at SMIpiixpm0: polling New: piixpm0 at pci0 dev 20 function 0 piixpm0: ATI Technologies SB600/SB700/SB800 SMBus Controller (rev. 0x13) piixpm0: interrupting at SMI, polling Cosmetic change only, no functional changes intended.
1.27	03-Feb-2009	pgoyette	branches: 1.27.2; Update the i2c_exec() functions to handle quick_read/quick_write protocol. Although interface routines i2c_smbus_*() exist, nothing in NetBSD actually uses them yet.
1.26	12-Oct-2008	martin	branches: 1.26.2; 1.26.4; Move printing of the "polling" message at attach time out of the wrong if scope, so that it is actually printed even in non-error cases. Suggested by jmcneill.
1.25	09-Jul-2008	joerg	- device/softc split - reduce namespace pollution
1.24	10-Apr-2008	cegger	branches: 1.24.4; 1.24.6; 1.24.8; 1.24.10; use aprint_*_dev and device_xname
1.23	26-Mar-2008	jmcneill	PR#38301: Add support for SMBus controllers for ATI SB600, SB700, SB800 chipsets, from Christoph Egger.
1.22	10-Mar-2008	sketch	Fix console output.
1.21	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.20	05-Feb-2008	simonb	branches: 1.20.2; 1.20.6; Remove an extra \n from an attach message. Get's rid of a blank line during boot.
1.19	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.18	09-Dec-2007	jmcneill	branches: 1.18.2; Merge jmcneill-pm branch.
1.17	19-Oct-2007	ad	branches: 1.17.2; 1.17.4; 1.17.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	27-Aug-2007	xtraeme	branches: 1.16.2; 1.16.6; Use rwlock(9) rather than lockmgr(9).
1.15	26-Aug-2007	xtraeme	There's no need to add timecounter code inside of __HAVE_TIMECOUNTER ifdefs.
1.14	06-Aug-2007	martin	branches: 1.14.2; Some serverworks mainboards use this power management too
1.13	04-Mar-2007	christos	branches: 1.13.2; 1.13.10; 1.13.14; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.12	10-Dec-2006	uwe	branches: 1.12.2; Print pci device info instead of "Power Management Controller".
1.11	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.10	31-Oct-2006	toshii	Match ATI SB[34]00 smbus controllers.
1.9	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.8	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.7	17-Aug-2006	christos	branches: 1.7.2; 1.7.4; 1.7.6; Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.6	26-Jun-2006	drochner	use the "i2cbus" interface attribute rather than putting a string name into the i2cbus attach args
1.5	26-Jun-2006	drochner	-make sure it is a PIIX4 before accessing counter stuff (there might be usable registers on 440mx and that ATI thing too, but that needs to be checked/tested) -make sure the I/O access to the power management stuff was enabled by the BIOS before trying to map/access stuff -most PIIX4s have a bug in the timer latch which causes jitter or worse if it is read naively - check the revision -use common code in dev/ic/acpipmtimer.c
1.4	22-Jun-2006	jmcneill	Add timecounter support for Intel PIIX Power Management Controllers. Ok kardel@.
1.3	17-Jun-2006	jmcneill	branches: 1.3.2; Restore the device activity PCI configuration registers on resume.
1.2	07-May-2006	jmcneill	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12; Add __KERNEL_RCSID.
1.1	07-May-2006	jmcneill	Add support for the Intel PIIX4 power management controller, from OpenBSD.
1.2.12.1	13-Jul-2006	gdamore	Merge from HEAD.
1.2.10.2	01-Jun-2006	kardel	Sync with head.
1.2.10.1	07-May-2006	kardel	file piixpm.c was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.2.8.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.2.8.1	07-May-2006	tron	file piixpm.c was added on branch peter-altq on 2006-05-24 15:50:28 +0000
1.2.6.5	03-Sep-2006	yamt	sync with head.
1.2.6.4	11-Aug-2006	yamt	sync with head
1.2.6.3	26-Jun-2006	yamt	sync with head.
1.2.6.2	24-May-2006	yamt	sync with head.
1.2.6.1	07-May-2006	yamt	file piixpm.c was added on branch yamt-pdpolicy on 2006-05-24 10:58:01 +0000
1.2.4.1	19-Jun-2006	chap	Sync with head.
1.2.2.2	11-May-2006	elad	sync with head
1.2.2.1	07-May-2006	elad	file piixpm.c was added on branch elad-kernelauth on 2006-05-11 23:28:48 +0000
1.3.2.8	17-Mar-2008	yamt	sync with head.
1.3.2.7	11-Feb-2008	yamt	sync with head.
1.3.2.6	21-Jan-2008	yamt	sync with head
1.3.2.5	27-Oct-2007	yamt	sync with head.
1.3.2.4	03-Sep-2007	yamt	sync with head.
1.3.2.3	30-Dec-2006	yamt	sync with head.
1.3.2.2	21-Jun-2006	yamt	sync with head.
1.3.2.1	17-Jun-2006	yamt	file piixpm.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:06 +0000
1.7.6.2	10-Dec-2006	yamt	sync with head.
1.7.6.1	22-Oct-2006	yamt	sync with head
1.7.4.2	09-Sep-2006	rpaulo	sync with head
1.7.4.1	17-Aug-2006	rpaulo	file piixpm.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.7.2.2	12-Jan-2007	ad	Sync with head.
1.7.2.1	18-Nov-2006	ad	Sync with head.
1.12.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.13.14.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.13.14.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.13.14.4	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.13.14.3	05-Oct-2007	joerg	Convert piixpm(4) to PNP power management.
1.13.14.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.13.14.1	09-Aug-2007	jmcneill	Sync with HEAD.
1.13.10.2	03-Sep-2007	skrll	Sync with HEAD.
1.13.10.1	15-Aug-2007	skrll	Sync with HEAD.
1.13.2.3	23-Oct-2007	ad	Sync with head.
1.13.2.2	09-Oct-2007	ad	Sync with head.
1.13.2.1	20-Aug-2007	ad	Sync with HEAD.
1.14.2.2	06-Aug-2007	martin	Some serverworks mainboards use this power management too
1.14.2.1	06-Aug-2007	martin	file piixpm.c was added on branch matt-mips64 on 2007-08-06 22:41:23 +0000
1.16.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.2.3	23-Mar-2008	matt	sync with HEAD
1.16.2.2	09-Jan-2008	matt	sync with HEAD
1.16.2.1	06-Nov-2007	matt	sync with HEAD
1.17.6.1	11-Dec-2007	yamt	sync with head.
1.17.4.1	26-Dec-2007	ad	Sync with head.
1.17.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.18.2.1	13-Dec-2007	bouyer	Sync with HEAD
1.20.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.20.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.20.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.20.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.20.2.1	24-Mar-2008	keiichi	sync with head.
1.24.10.1	19-Oct-2008	haad	Sync with HEAD.
1.24.8.1	18-Jul-2008	simonb	Sync with head.
1.24.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.24.4.3	11-Aug-2010	yamt	sync with head.
1.24.4.2	11-Mar-2010	yamt	sync with head
1.24.4.1	04-May-2009	yamt	sync with head.
1.26.4.1	20-Nov-2010	riz	Pull up following revision(s) (requested by martin in ticket #1391): sys/dev/pci/piixpm.c: revision 1.28 Fix dmesg output: Old: piixpm0 at pci0 dev 20 function 0 piixpm0: ATI Technologies SB600/SB700/SB800 SMBus Controller (rev. 0x13) piixpm0: interrupting at SMIpiixpm0: polling New: piixpm0 at pci0 dev 20 function 0 piixpm0: ATI Technologies SB600/SB700/SB800 SMBus Controller (rev. 0x13) piixpm0: interrupting at SMI, polling Cosmetic change only, no functional changes intended.
1.26.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.26.2.1	03-Mar-2009	skrll	Sync with HEAD.
1.27.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.31.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.31.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.32.2.1	05-Mar-2011	rmind	sync with head
1.33.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.33.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.37.6.1	18-Feb-2012	mrg	merge to -current.
1.37.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.2.1	17-Apr-2012	yamt	sync with head
1.40.18.1	23-Jul-2013	riastradh	sync with HEAD
1.40.10.2	18-May-2014	rmind	sync with head
1.40.10.1	28-Aug-2013	rmind	sync with head
1.40.6.2	03-Dec-2017	jdolecek	update from HEAD
1.40.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.40.2.2	07-Aug-2014	msaitoh	Pull up following revision(s) (requested by hannken in ticket #1096): sys/dev/pci/piixpm.c: revision 1.45 Consistently pass a "struct piixpm_softc" to piixpm_intr. Prevents a crash on hardware interrupts.
1.40.2.1	20-Sep-2013	riz	Pull up following revision(s) (requested by fair in ticket #936): sys/dev/pci/piixpmreg.h: revision 1.6 sys/dev/pci/piixpm.c: revision 1.42 The SB800 SMBus controller has four selectable SDA lines. Expose them as four iic busses.
1.44.2.1	10-Aug-2014	tls	Rebase.
1.45.4.5	28-Aug-2017	skrll	Sync with HEAD
1.45.4.4	05-Dec-2016	skrll	Sync with HEAD
1.45.4.3	05-Oct-2016	skrll	Sync with HEAD
1.45.4.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.45.4.1	06-Jun-2015	skrll	Sync with HEAD
1.49.2.4	26-Apr-2017	pgoyette	Resolve a couple of conflicts
1.49.2.3	26-Apr-2017	pgoyette	Sync with HEAD
1.49.2.2	04-Nov-2016	pgoyette	Sync with HEAD
1.49.2.1	26-Jul-2016	pgoyette	Sync with HEAD
1.51.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.52.14.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.52.6.4	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1790: sys/dev/pci/piixpm.c 1.66,1.68-1.72 via patch sys/dev/pci/piixpmreg.h 1.13 Add support new AMD chipsets that do not have indirect access I/O ports.
1.52.6.3	30-Nov-2021	martin	Apply patch, requested by msaitoh in ticket #1709: sys/dev/pci/piixpm.c patch Fix a bug that I2C access panics on old AMD chipset (e.g SB600). Fixes PR kern/56525.
1.52.6.2	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1591: sys/dev/pci/piixpm.c 1.57-1.59, 1.61-1.63 via patch sys/dev/pci/piixpmreg.h 1.9-1.12 - Fix number of port for Hudson rev. 0x1f and newer. - Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI or 1 on IRQ, so invert the check. - Don't force using SMBUS0SEL register. - Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on some machines. - Save/restore port number before selecting port. - Modify comment. - Whitespace fix.
1.52.6.1	06-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1327): sys/dev/pci/piixpmreg.h: revision 1.8 sys/dev/pci/piixpm.c: revision 1.53 sys/dev/pci/piixpm.c: revision 1.54 Sync with OpenBSD's piixpm.c up to rev. 1.35: - Print "polling" correctly when in the attach function. Same as OpenBSD rev. 1.22-23. - Improve debug printf()'s. Part of OpenBSD 1.24. - Do not bus_space_map devices which are at address 0 (it's uninitialized). Same as OpenBSD rev. 1.25. - Add ServerWorks HT1100 device from OpenBSD. Same as OpenBSD rev. 1.32 - Use unique wait channel. From OpenBSD rev. 1.35 Improve SB800 and newer chipsets support: - Add newer chipset (e.g. X370/X399 and newer) support that the PCI device id is 0x790b. The register definitions are mainly taken from FreeBSD. - Rename PIIXPM_INDIRECTIO_* to SB800_INDIRECTIO_* because those are only for SB800 and newer chipsets. - SB800 also support 4 ports. - SB800's interrupt configuration bit is different from others. Use SB800_SMB_HOSTC's bit 0. - Do not bus_space_map devices which are at address 0 (it's uninitialized) in piixpm_attach(). - Add the port number to the dmesg output. - Avoid uninitiliazed use of ctl and corresponding warnings. From OpenBSD rev. 1.38
1.54.2.3	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1573: sys/dev/pci/piixpm.c 1.66,1.68-1.72 via patch sys/dev/pci/piixpmreg.h 1.13 Add support new AMD chipsets that do not have indirect access I/O ports.
1.54.2.2	30-Nov-2021	martin	Apply patch, requested by msaitoh in ticket #1378: sys/dev/pci/piixpm.c patch Fix a bug that I2C access panics on old AMD chipset (e.g SB600). Fixes PR kern/56525.
1.54.2.1	16-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1017: sys/dev/pci/piixpm.c 1.57-1.59, 1.61-1.63 via patch sys/dev/pci/piixpmreg.h 1.9-1.12 - Fix number of port for Hudson rev. 0x1f and newer. - Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI and 1 on IRQ, so invert the check. - Don't force using SMBUS0SEL register. - Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on some machines. - Save/restore port number before selecting port. - Modify comment. - Whitespace fix.
1.60.2.1	17-Jan-2020	ad	Sync with head.
1.63.8.1	24-Mar-2021	thorpej	Don't filter interface attributes in rescan functions for devices that carry only a single interface attribute. The autoconfiguration machinery already considers interface attributes when searching for possible parents.
1.64.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.67.4.1	23-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #55): sys/dev/pci/piixpm.c: revision 1.68 sys/dev/pci/piixpm.c: revision 1.69 sys/dev/pci/piixpmreg.h: revision 1.13 sys/dev/pci/piixpm.c: revision 1.70 sys/dev/pci/piixpm.c: revision 1.71 sys/dev/pci/piixpm.c: revision 1.72 Modify comment. Whitespace. No functional change. Simplify and modify for following MMIO support. No functional change. Modify for following MMIO support. No functional change. Use MMIO for PM register access on newer revision's devices. Newer revision's device only support memory mapped access. At least, X670E's SMBus device that the revision ID is 0x71 doesn't support I/O access. Use MMIO access for newer devices. Linux uses >= 0x51 for the check, so we do the same check. Note that X570's SMBus(rev. 0x61) supports I/O access... Modify error message to match the current behavior.
1.72.6.1	02-Aug-2025	perseant	Sync with HEAD
1.13	10-Jan-2023	msaitoh	Use MMIO for PM register access on newer revision's devices. Newer revision's device only support memory mapped access. At least, X670E's SMBus device that the revision ID is 0x71 doesn't support I/O access. Use MMIO access for newer devices. Linux uses >= 0x51 for the check, so we do the same check. Note that X570's SMBus(rev. 0x61) supports I/O access...
1.12	14-Jan-2020	msaitoh	branches: 1.12.24; Add AMDFCH41_SMBUS_PORTMASK.
1.11	09-Jan-2020	msaitoh	Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on ECS HDC-I2. This change fixes the problem.
1.10	24-Dec-2019	msaitoh	branches: 1.10.2; Don't force using SMBUS0SEL register. - Use it depending on USE_SMBUS0SEL bit. - If we use SMBUS0EN_LO register to select the port, update the port select bits only.
1.9	23-Dec-2019	msaitoh	- Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI or 1 on IRQ, so invert the check. - Modify comment. - Whitespace fix.
1.8	13-Jul-2019	msaitoh	branches: 1.8.2; Improve SB800 and newer chipsets support: - Add newer chipset (e.g. X370/X399 and newer) support that the PCI device id is 0x790b. The register definitions are mainly taken from FreeBSD. - Rename PIIXPM_INDIRECTIO_* to SB800_INDIRECTIO_* because those are only for SB800 and newer chipsets. - SB800 also support 4 ports. - SB800's interrupt configuration bit is different from others. Use SB800_SMB_HOSTC's bit 0. - Do not bus_space_map devices which are at address 0 (it's uninitialized) in piixpm_attach(). - Add the port number to the dmesg output. - Avoid uninitiliazed use of ctl and corresponding warnings. From OpenBSD rev. 1.38
1.7	18-Mar-2014	riastradh	branches: 1.7.22; 1.7.30; Merge riastradh-drm2 to HEAD.
1.6	22-Jul-2013	soren	The SB800 SMBus controller has four selectable SDA lines. Expose them as four iic busses.
1.5	13-Feb-2011	hannken	branches: 1.5.4; 1.5.10; 1.5.14; 1.5.18; 1.5.26; The ServerWorks CSB5 smbus controller often times out and hangs until reboot. Increase the delay between setting the command register and reading the status register. Reset the smbus controller if it has timed out. Tested on an IBM eServer x335.
1.4	26-Jun-2006	drochner	branches: 1.4.6; 1.4.84; 1.4.90; 1.4.92; oops, need to commit the register definition for PM mapping too
1.3	22-Jun-2006	jmcneill	Add timecounter support for Intel PIIX Power Management Controllers. Ok kardel@.
1.2	17-Jun-2006	jmcneill	branches: 1.2.2; Restore the device activity PCI configuration registers on resume.
1.1	07-May-2006	jmcneill	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.10; 1.1.12; Add support for the Intel PIIX4 power management controller, from OpenBSD.
1.1.12.1	13-Jul-2006	gdamore	Merge from HEAD.
1.1.10.2	01-Jun-2006	kardel	Sync with head.
1.1.10.1	07-May-2006	kardel	file piixpmreg.h was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.1.8.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.1.8.1	07-May-2006	tron	file piixpmreg.h was added on branch peter-altq on 2006-05-24 15:50:28 +0000
1.1.6.4	11-Aug-2006	yamt	sync with head
1.1.6.3	26-Jun-2006	yamt	sync with head.
1.1.6.2	24-May-2006	yamt	sync with head.
1.1.6.1	07-May-2006	yamt	file piixpmreg.h was added on branch yamt-pdpolicy on 2006-05-24 10:58:01 +0000
1.1.4.1	19-Jun-2006	chap	Sync with head.
1.1.2.2	11-May-2006	elad	sync with head
1.1.2.1	07-May-2006	elad	file piixpmreg.h was added on branch elad-kernelauth on 2006-05-11 23:28:48 +0000
1.2.2.3	30-Dec-2006	yamt	sync with head.
1.2.2.2	21-Jun-2006	yamt	sync with head.
1.2.2.1	17-Jun-2006	yamt	file piixpmreg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:06 +0000
1.4.92.1	17-Feb-2011	bouyer	Sync with HEAD
1.4.90.1	06-Jun-2011	jruoho	Sync with HEAD.
1.4.84.1	05-Mar-2011	rmind	sync with head
1.4.6.2	09-Sep-2006	rpaulo	sync with head
1.4.6.1	26-Jun-2006	rpaulo	file piixpmreg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.5.26.1	23-Jul-2013	riastradh	sync with HEAD
1.5.18.1	28-Aug-2013	rmind	sync with head
1.5.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.5.10.1	20-Sep-2013	riz	Pull up following revision(s) (requested by fair in ticket #936): sys/dev/pci/piixpmreg.h: revision 1.6 sys/dev/pci/piixpm.c: revision 1.42 The SB800 SMBus controller has four selectable SDA lines. Expose them as four iic busses.
1.5.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.30.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.7.22.3	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1790: sys/dev/pci/piixpm.c 1.66,1.68-1.72 via patch sys/dev/pci/piixpmreg.h 1.13 Add support new AMD chipsets that do not have indirect access I/O ports.
1.7.22.2	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1591: sys/dev/pci/piixpm.c 1.57-1.59, 1.61-1.63 via patch sys/dev/pci/piixpmreg.h 1.9-1.12 - Fix number of port for Hudson rev. 0x1f and newer. - Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI or 1 on IRQ, so invert the check. - Don't force using SMBUS0SEL register. - Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on some machines. - Save/restore port number before selecting port. - Modify comment. - Whitespace fix.
1.7.22.1	06-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1327): sys/dev/pci/piixpmreg.h: revision 1.8 sys/dev/pci/piixpm.c: revision 1.53 sys/dev/pci/piixpm.c: revision 1.54 Sync with OpenBSD's piixpm.c up to rev. 1.35: - Print "polling" correctly when in the attach function. Same as OpenBSD rev. 1.22-23. - Improve debug printf()'s. Part of OpenBSD 1.24. - Do not bus_space_map devices which are at address 0 (it's uninitialized). Same as OpenBSD rev. 1.25. - Add ServerWorks HT1100 device from OpenBSD. Same as OpenBSD rev. 1.32 - Use unique wait channel. From OpenBSD rev. 1.35 Improve SB800 and newer chipsets support: - Add newer chipset (e.g. X370/X399 and newer) support that the PCI device id is 0x790b. The register definitions are mainly taken from FreeBSD. - Rename PIIXPM_INDIRECTIO_* to SB800_INDIRECTIO_* because those are only for SB800 and newer chipsets. - SB800 also support 4 ports. - SB800's interrupt configuration bit is different from others. Use SB800_SMB_HOSTC's bit 0. - Do not bus_space_map devices which are at address 0 (it's uninitialized) in piixpm_attach(). - Add the port number to the dmesg output. - Avoid uninitiliazed use of ctl and corresponding warnings. From OpenBSD rev. 1.38
1.8.2.2	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1573: sys/dev/pci/piixpm.c 1.66,1.68-1.72 via patch sys/dev/pci/piixpmreg.h 1.13 Add support new AMD chipsets that do not have indirect access I/O ports.
1.8.2.1	16-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1017: sys/dev/pci/piixpm.c 1.57-1.59, 1.61-1.63 via patch sys/dev/pci/piixpmreg.h 1.9-1.12 - Fix number of port for Hudson rev. 0x1f and newer. - Read SB800_SMB_HOSTC correctly. This register is not in the PCI config space but in the I/O space. - The bit 0 of SB800_SMB_HOSTC is 0 on SMI and 1 on IRQ, so invert the check. - Don't force using SMBUS0SEL register. - Acquire/release host semaphore to share SMBus between the host and the embedded controller (IMC). Without this change, "shutdown -r" does power off and not boot on some machines. - Save/restore port number before selecting port. - Modify comment. - Whitespace fix.
1.10.2.1	17-Jan-2020	ad	Sync with head.
1.12.24.1	23-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #55): sys/dev/pci/piixpm.c: revision 1.68 sys/dev/pci/piixpm.c: revision 1.69 sys/dev/pci/piixpmreg.h: revision 1.13 sys/dev/pci/piixpm.c: revision 1.70 sys/dev/pci/piixpm.c: revision 1.71 sys/dev/pci/piixpm.c: revision 1.72 Modify comment. Whitespace. No functional change. Simplify and modify for following MMIO support. No functional change. Modify for following MMIO support. No functional change. Use MMIO for PM register access on newer revision's devices. Newer revision's device only support memory mapped access. At least, X670E's SMBus device that the revision ID is 0x71 doesn't support I/O access. Use MMIO access for newer devices. Linux uses >= 0x51 for the check, so we do the same check. Note that X570's SMBus(rev. 0x61) supports I/O access... Modify error message to match the current behavior.
1.1	17-May-2000	thorpej	branches: 1.1.6; Register definitions and preliminary API definition for the PLX 9060 family of PCI bus controllers.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	17-May-2000	bouyer	file plx9060reg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:36 +0000
1.3	11-Dec-2005	christos	merge ktrace-lwp.
1.2	04-Feb-2005	perry	de-__P
1.1	17-May-2000	thorpej	branches: 1.1.6; 1.1.30; 1.1.38; 1.1.40; Register definitions and preliminary API definition for the PLX 9060 family of PCI bus controllers.
1.1.40.1	12-Feb-2005	yamt	sync with head.
1.1.38.1	29-Apr-2005	kent	sync with -current
1.1.30.1	04-Feb-2005	skrll	Sync with HEAD.
1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.6.1	17-May-2000	bouyer	file plx9060var.h was added on branch thorpej_scsipi on 2000-11-20 11:42:36 +0000
1.37	04-Jun-2025	macallan	enable font loading
1.36	04-Jun-2025	macallan	make pipeline overrun protection in pm2fb_putchar_aa() actually work. Doh.
1.35	25-Sep-2022	thorpej	branches: 1.35.10; Remove unnecessary include of <sys/malloc.h>.
1.34	19-Dec-2021	andvar	s/forground/foreground/ in comments.
1.33	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.32	24-Apr-2021	thorpej	branches: 1.32.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.31	22-Dec-2019	thorpej	branches: 1.31.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.30	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.29	16-Dec-2016	macallan	branches: 1.29.14; 1.29.16; use hardware rendering for anti-aliased fonts
1.28	16-Sep-2015	macallan	branches: 1.28.2; prop_dictionary_get_bool() doesn't return FALSE if the property doesn't exist, so initialize things first. No more console stealing on sparc64.
1.27	17-Dec-2014	macallan	- add support for Permedia 2 - make this work on alpha from Naruaki Etomi
1.26	09-Dec-2014	macallan	begin to address PR 49229: - program aperture endianness to match host, now xf86-video-wsfb works as expected - properly restore the colour map when er-entering terminal emulation
1.25	09-Oct-2013	macallan	branches: 1.25.6; support WSDISPLAYIO_GET_FBINFO
1.24	15-Sep-2013	martin	Remove/comment out unused variables, initialize a few others.
1.23	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.22	28-Jun-2013	christos	branches: 1.22.2; this could have never worked. Cleanup bogus code. http://m00nbsd.net/ae123a9bae03f7dde5c6d654412daf5a.html
1.21	06-Oct-2012	macallan	don't crash when we're not the console
1.20	04-Oct-2012	macallan	support underlined characters while there, make sure the glyph cache doesn't exceed 2048 lines since that's the drawing engine's limit
1.19	13-Sep-2012	macallan	make video modes work that need a screen width which isn't directly supported by the rendering engine - just use the next higher supported stride
1.18	13-Sep-2012	macallan	add mode setting support While there, don't pretend to support the Permedia2 non-V - I don't have the hardware and it has never been tested in the first place. The only hardware this has been tested on is a TechSource Raptor GFX 8P / Sun PGX32 which happens to be a Permedia2V with Sun firmware.
1.17	12-Sep-2012	macallan	support anti-aliased fonts in 8 bit while there, fix pm2fb_bitblt() packed mode, for real this time TODO: - use packed mode when uploading glyphs - let the hardware do the alpha blending if possible - mode setting
1.16	05-Sep-2012	macallan	branches: 1.16.2; fix horizontal scrolling in 8bit, while there use an R3G3B2 palette
1.15	05-Sep-2012	macallan	some preparations for mode setting support and such: - DDC2 support, so far only used for WSDISPLAYIO_GET_EDID - fix 8 bit support, use packed mode for some extra speed - pm2 and pm2v have different DACs, deal with it where appropriate
1.14	23-May-2012	macallan	first step to fix PR 46461 - don't jump through a NULL pointer when we're not the console but the firmware set up the graphics hardware anyway
1.13	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.12	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.11	11-Jan-2012	macallan	use rasops_init(0, 0)
1.10	24-Nov-2011	macallan	branches: 1.10.2; fix debug build
1.9	24-Nov-2011	macallan	#include "opt_pm2fb.h"
1.8	24-Nov-2011	macallan	support DDC2, so far only for WSDISPLAYIO_GET_EDID
1.7	22-Jan-2011	cegger	branches: 1.7.4; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.6	16-Dec-2010	cegger	branches: 1.6.2; 1.6.4; ioctl: KNF switch-case, remove useless brackets
1.5	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.4	04-May-2010	macallan	use alternate font if available ( for autogenerated box drawing characters )
1.3	16-Mar-2010	macallan	Implement character drawing by hardware so we don't need to map the framebuffer anymore. Also do some cleanup, remove no longer relevant debugging code etc.
1.2	28-Oct-2009	macallan	branches: 1.2.2; 1.2.4; 1.2.6; fix occasionally wrong colours in rectangle fills
1.1	28-Oct-2009	macallan	A driver for the Permedia 2 graphics processor, so far it's been tested only on sparc64 with a Sun PGX32 / TechSource Raptor GFX 8P So far it accelerates block copies, rectangle fills and the cursor, it supports the usual wsdisplay / vcons stuff. TODO: - character drawing by hardware so we don't need to map the framebuffer - DDC2 support - mode setting This driver still relies on the firmware to set up graphics modes etc. so it won't work on x86 ( or non-OpenFirmware for that matter ) unless you use the VESA BIOS to set up graphics and pass the right properties ( width, height etc. ) to the driver. For the same reason it will work only if it's the OpenFirmware console.
1.2.6.2	05-Mar-2011	rmind	sync with head
1.2.6.1	30-May-2010	rmind	sync with head
1.2.4.3	11-Aug-2010	yamt	sync with head.
1.2.4.2	11-Mar-2010	yamt	sync with head
1.2.4.1	28-Oct-2009	yamt	file pm2fb.c was added on branch yamt-nfs-mp on 2010-03-11 15:03:59 +0000
1.2.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.2.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.6.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.6.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.7.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.4.2	30-Oct-2012	yamt	sync with head
1.7.4.1	17-Apr-2012	yamt	sync with head
1.10.2.3	02-Jun-2012	mrg	sync to latest -current.
1.10.2.2	05-Apr-2012	mrg	sync to latest -current.
1.10.2.1	18-Feb-2012	mrg	merge to -current.
1.16.2.3	03-Dec-2017	jdolecek	update from HEAD
1.16.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.16.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.22.2.2	18-May-2014	rmind	sync with head
1.22.2.1	28-Aug-2013	rmind	sync with head
1.25.6.3	05-Feb-2017	skrll	Sync with HEAD
1.25.6.2	22-Sep-2015	skrll	Sync with HEAD
1.25.6.1	06-Apr-2015	skrll	Sync with HEAD
1.28.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.29.16.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.29.16.1	10-Jun-2019	christos	Sync with HEAD
1.29.14.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.31.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.32.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.35.10.1	02-Aug-2025	perseant	Sync with HEAD
1.12	16-Dec-2016	macallan	flesh out alpha operations properly
1.11	16-Dec-2016	macallan	add bit definitions for dither and alpha blending registers
1.10	17-Dec-2014	macallan	branches: 1.10.2; - add support for Permedia 2 - make this work on alpha from Naruaki Etomi
1.9	14-Dec-2014	macallan	fix bit flip in PM2_DAC_COLOR_MODE register from Naruaki Etomi
1.8	12-Sep-2012	macallan	branches: 1.8.14; support anti-aliased fonts in 8 bit while there, fix pm2fb_bitblt() packed mode, for real this time TODO: - use packed mode when uploading glyphs - let the hardware do the alpha blending if possible - mode setting
1.7	05-Sep-2012	macallan	branches: 1.7.2; add a few more registers
1.6	02-Feb-2012	macallan	add a bunch of register definitions not used yet but I better commit them now than have he tell me that I forgot them when pm2fb.c starts using them
1.5	24-Nov-2011	macallan	branches: 1.5.2; add MONITOR_DATA register
1.4	16-Mar-2010	macallan	branches: 1.4.8; Implement character drawing by hardware so we don't need to map the framebuffer anymore. Also do some cleanup, remove no longer relevant debugging code etc.
1.3	30-Oct-2009	christos	branches: 1.3.2; 1.3.4; 1.3.6; spell the guys name properly.
1.2	28-Oct-2009	macallan	fix occasionally wrong colours in rectangle fills
1.1	28-Oct-2009	macallan	A driver for the Permedia 2 graphics processor, so far it's been tested only on sparc64 with a Sun PGX32 / TechSource Raptor GFX 8P So far it accelerates block copies, rectangle fills and the cursor, it supports the usual wsdisplay / vcons stuff. TODO: - character drawing by hardware so we don't need to map the framebuffer - DDC2 support - mode setting This driver still relies on the firmware to set up graphics modes etc. so it won't work on x86 ( or non-OpenFirmware for that matter ) unless you use the VESA BIOS to set up graphics and pass the right properties ( width, height etc. ) to the driver. For the same reason it will work only if it's the OpenFirmware console.
1.3.6.1	30-May-2010	rmind	sync with head
1.3.4.3	11-Aug-2010	yamt	sync with head.
1.3.4.2	11-Mar-2010	yamt	sync with head
1.3.4.1	30-Oct-2009	yamt	file pm2reg.h was added on branch yamt-nfs-mp on 2010-03-11 15:03:59 +0000
1.3.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.4.8.2	30-Oct-2012	yamt	sync with head
1.4.8.1	17-Apr-2012	yamt	sync with head
1.5.2.1	18-Feb-2012	mrg	merge to -current.
1.7.2.2	03-Dec-2017	jdolecek	update from HEAD
1.7.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.8.14.2	05-Feb-2017	skrll	Sync with HEAD
1.8.14.1	06-Apr-2015	skrll	Sync with HEAD
1.10.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.14	16-Jun-2025	macallan	set & support RI_FULLCLEAR now we get a clean console when X exits
1.13	16-Jun-2025	macallan	defflag PM3FB_DEBUG wipe glyphcache when re-entering terminal mode
1.12	27-May-2025	macallan	add support for anti-aliased fonts while there, enable support for font loading
1.11	25-May-2025	macallan	fix tpyo when calculating PM3_HORIZ_TOTAL - now mode setting works properly, no more semi-random garbage at the right border
1.10	14-Aug-2024	macallan	initialize a screen even if we're not the console
1.9	25-Sep-2022	thorpej	branches: 1.9.10; Remove unnecessary include of <sys/malloc.h>.
1.8	25-Aug-2021	andvar	Add missing RCSID and __KERNEL_RCSID().
1.7	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.6	24-Apr-2021	thorpej	branches: 1.6.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.5	22-Dec-2019	thorpej	branches: 1.5.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.4	06-Mar-2017	macallan	branches: 1.4.12; 1.4.16; print out the actual stride value
1.3	07-Dec-2016	macallan	branches: 1.3.2; 1.3.4; - don't flush the engine after every blitter operation - make sure we don't overrun the pipeline in pm3fb_init()
1.2	07-Dec-2016	macallan	fix downward scrolling, set BUS_SPACE_MAP_PREFETCHABLE
1.1	20-Nov-2016	macallan	branches: 1.1.2; add driver for Permedia 3 based graphics devices from Naruaki Etomi, via PR 49518
1.1.2.4	28-Aug-2017	skrll	Sync with HEAD
1.1.2.3	05-Feb-2017	skrll	Sync with HEAD
1.1.2.2	05-Dec-2016	skrll	Sync with HEAD
1.1.2.1	20-Nov-2016	skrll	file pm3fb.c was added on branch nick-nhusb on 2016-12-05 10:55:16 +0000
1.3.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.3.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.3.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.3.2.1	07-Dec-2016	pgoyette	file pm3fb.c was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.4.16.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.4.12.2	03-Dec-2017	jdolecek	update from HEAD
1.4.12.1	06-Mar-2017	jdolecek	file pm3fb.c was added on branch tls-maxphys on 2017-12-03 11:37:28 +0000
1.5.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.6.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.9.10.1	02-Aug-2025	perseant	Sync with HEAD
1.4	27-May-2025	macallan	add PM3_SOURCE_DATA register
1.3	25-Aug-2021	andvar	branches: 1.3.10; Add missing RCSID and __KERNEL_RCSID().
1.2	17-Jun-2017	macallan	branches: 1.2.4; moar registers
1.1	20-Nov-2016	macallan	branches: 1.1.2; 1.1.4; add driver for Permedia 3 based graphics devices from Naruaki Etomi, via PR 49518
1.1.4.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.1	20-Nov-2016	pgoyette	file pm3reg.h was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.1.2.3	28-Aug-2017	skrll	Sync with HEAD
1.1.2.2	05-Dec-2016	skrll	Sync with HEAD
1.1.2.1	20-Nov-2016	skrll	file pm3reg.h was added on branch nick-nhusb on 2016-12-05 10:55:16 +0000
1.2.4.2	03-Dec-2017	jdolecek	update from HEAD
1.2.4.1	17-Jun-2017	jdolecek	file pm3reg.h was added on branch tls-maxphys on 2017-12-03 11:37:28 +0000
1.3.10.1	02-Aug-2025	perseant	Sync with HEAD
1.76	29-Aug-2024	andvar	improve grammar in the comment.
1.75	26-Nov-2023	rin	branches: 1.75.2; ppb(4): Print out PCIe Gen5 link speed correctly
1.74	10-Oct-2021	msaitoh	branches: 1.74.4; Use PCI-SIG official acronyms: - RP stands for Root Port. - RC stands for Root Complex. - RCIEP stands for Root Complex Integrated End Point.
1.73	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.72	11-May-2021	thorpej	branches: 1.72.4; Forward the bridge's devhandle to the downstream bus.
1.71	24-Apr-2021	thorpej	branches: 1.71.2; 1.71.4; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.70	12-Jul-2020	rin	branches: 1.70.4; Remove duplicate include of opt_ppb.h. No binary changes.
1.69	09-Jul-2019	msaitoh	branches: 1.69.2; Identify 16GT/s.
1.68	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.67	29-Jan-2019	msaitoh	If the secondary bus is configured and the bus mastering is not enabled, enable it. Suggested by thorpej@.
1.66	28-Jan-2019	msaitoh	Explicitly enable bus masterling in case BIOS, UEFI or firmware don't enable it. Might fix PR kern/53811.
1.65	27-Jan-2019	pgoyette	Merge the [pgoyette-compat] branch
1.64	07-Dec-2018	msaitoh	- defflag PPB_USEINTR - Print "interrupting at "
1.63	10-May-2017	msaitoh	branches: 1.63.2; 1.63.8; 1.63.10; Fix typos.
1.62	09-May-2017	msaitoh	- Fix a bug that a device which has no PCIe capability incorrectly accessess the PCI config area in ppbdetach(). - Don't add event counters if slot interrupt isn't used.
1.61	27-Apr-2017	msaitoh	Make ppb(4) interrupt support stable: - Disable all interrupts in the beginning of attach. Without this, interrupt storm occurs while cold == 1 on some environment. - Disable command complete interrput for a while to prevent hangup on some enviroment. I'm sorry, I don't know what this bit is :-\| - Check all status bits and return 0 if an interrupt is not for me. It's required for INTx. Tested on XEN3_DOM0 because it doesn't support MSI yet. - Return 1 when a interrupt is processed.
1.60	26-Apr-2017	msaitoh	branches: 1.60.2; Disable ppb(4)'s interrupt for a while. It causes hangup on some environment. Define PPB_USEINTR if you'd like to use interrupt.
1.59	26-Apr-2017	msaitoh	Disable and disestablish interrupt in ppbdetach().
1.58	24-Apr-2017	chs	in ppbdetach(), try to detach the children before tearing down our own state, to avoid trouble if detaching the children fails.
1.57	18-Apr-2017	msaitoh	Enable PCIe's interrupt as much as possilbe in ppb(4) to detect and count status change event. HotPlug function itself have not implemented yet. - Interrupt and each event are counted by evcnt(9). Example: ppb0 Interrupt 0 0 intr ppb0 Attention Button Pressed 0 0 misc ppb0 Power Fault Detected 0 0 misc ppb0 MRL Sensor Changed 0 0 misc ppb0 Presence Detect Changed 0 0 misc ppb0 Command Completed 0 0 misc ppb0 Data Link Layer State Changed 0 0 misc - Print message if ppb_printevent is not zero. The default vaule is 0. The output messages: Attention Button Pressed Power Fault Detected MRL Sensor Changed Presence Detect Changed Command Completed Data Link Layer State Changed - Remove workaround code to disable interrupt (ppb.c rev. 1.35). Tested with Dell Latitude 2120 without if_bge.c rev. 1.304's workaround. dmesg when bge's device timeout occured: ppb3: Presence Detect Changed ppb3: Data Link Layer State Changed ppb3: Presence Detect Changed vmstat -e \|grep ppb ppb3 Interrupt 2 0 intr ppb3 Presence Detect Changed 2 0 misc ppb3 Data Link Layer State Changed 1 0 misc
1.56	05-Apr-2017	msaitoh	Move struct ppb_softc into ppbvar.h.
1.55	16-Nov-2015	msaitoh	branches: 1.55.2; 1.55.4; Define PCIE_XCAP_{VER,TYPE}(x) and use them.
1.54	24-Sep-2014	msaitoh	branches: 1.54.2; Rename PCIE_XCAP_VER_* macros to avoid confusion.
1.53	24-Sep-2014	msaitoh	- Modify message of PCIe capability version. This field (PCIE_XCAP_VER_MASK) is not specification's version number but the capability structure's version number. To avoid confusion, print "PCI Express capability version x". - The max number of PCIe lane is not 16 but 32. Fix the bug using with macro. - Use macro instead of magic number. - Gb/s -> GT/s
1.52	21-Apr-2013	msaitoh	branches: 1.52.10; Delete "PCI_" from PCIX and PICE capability registers.
1.51	06-Mar-2013	yamt	ppb: fix link speed print
1.50	20-Oct-2012	matt	Print out negotiated link width and speed for PCIe (merged from matt-nb5-mips64).
1.49	29-Jan-2012	drochner	branches: 1.49.6; extend the pci_aprint_devinfo slightly to cover the cases commonly used by drivers: a short name for the quiet/naive case and a string to override the "pcidevs" based name by one provided by the driver, ride on yesterday's kernel minor version bump
1.48	26-Jan-2012	drochner	put printing of the pci_devinfo into its own function (not inlined by purpose) - this is a stack hog, and with this change my uTCA amd64 system boots again a lot of similar code can be eliminated from pci device drivers this way, but before doing so (and making the new function part of the module API) I'd like to consider a modification to make it work with drivers which prefer to print names from other sources (like pciide)
1.47	21-Oct-2011	dyoung	branches: 1.47.2; 1.47.6; Tell a pci(4) instance its subordinate PCI buses using a new member in the pcibus_attach_args, pba_sub. pciN attaches to pba_bus itself. If pba_bus < pba_sub, then [pba_bus + 1, pba_sub] are subordinate to pba_bus. On i386, make mainbus0 attach pci0 with pba_sub = 255 because all buses 1 and up must be subordinate to pci0. XXX Deal with other architectures.
1.46	17-Aug-2011	dyoung	Redefine PCI_MSI_* and PCI_PCIE_* constants in terms of bits(3). Use named constants and more conventional variable names in pci_msi_establish() and pci_msi_disestablish(). Fix a couple of bugs: pci_msi_establish() returned a pointer to the struct intrhand instead of to the struct msi_hdl as it was intended to, and pci_msi_disestablish() did not free(9) the msi_hdl.
1.45	10-Jan-2011	cegger	add missing break
1.44	10-Jan-2011	jmcneill	ppb_fix_pcix changes: - rename to ppb_fix_pcie - support version PCI-E 2.0 - print version and device/port type information - use constants from pcireg.h instead of magic numbers changes: ppb2 at pci0 dev 21 function 0: vendor 0x15ad product 0x07a0 (rev. 0x01) ppb2: unsupported PCI Express version to: ppb2 at pci0 dev 21 function 0: vendor 0x15ad product 0x07a0 (rev. 0x01) ppb2: PCI Express 2.0 <Root Port of PCI-E Root Complex>
1.43	11-Dec-2010	matt	On powerpc, recognize PCI Express RC root bridges.
1.42	24-Feb-2010	dyoung	branches: 1.42.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.41	08-Jan-2010	dyoung	branches: 1.41.2; Expand PMF_FN_* macros.
1.40	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.39	03-May-2008	cegger	branches: 1.39.8; 1.39.10; 1.39.14; 1.39.18; 1.39.22; device_t / softc split. Tested on amd64. "looks good" spz
1.38	03-May-2008	cegger	unuspported -> unsupported
1.37	22-Feb-2008	dyoung	branches: 1.37.2; 1.37.4; Add methods for detaching self and for detaching children. Use device_t and accessors. Use aprint_*_dev().
1.36	09-Dec-2007	jmcneill	branches: 1.36.6; 1.36.10; Merge jmcneill-pm branch.
1.35	16-Oct-2007	joerg	branches: 1.35.6; 1.35.8; Merge 1.34.22.7 from jmcneill-pm: Next attempt at trying to fix the irregular interrupt storms on my Thinkpad: when we find a PCI Express device, check the list of notification events and if any are sets, clear them. We can't handle them ATM anyway.
1.34	16-Nov-2006	christos	branches: 1.34.8; 1.34.22; 1.34.24; 1.34.26; __unused removal on arguments; approved by core.
1.33	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.32	11-Dec-2005	christos	branches: 1.32.20; 1.32.22; merge ktrace-lwp.
1.31	28-Jun-2005	thorpej	branches: 1.31.2; Use ANSI function decls and static.
1.30	04-Feb-2005	perry	de-__P
1.29	30-Aug-2004	drochner	branches: 1.29.4; 1.29.6; Phase out the use of a string as first "attach args" member to control which bustype should be attached with a specific call to config_found() (from a "mainbus" or a bus bridge). Do it for isa/eisa/mca and pci/agp for now. These buses all attach to an mi interface attribute "isabus", "eisabus" etc., and the autoconf framework now allows to specify an interface attribute on config_found() and config_search(), which limits the search of matching config data to these which attach to that specific attribute. So we basically have to call config_found_ia(..., "foobus", ...) where such a bus is attached. As a consequence, where a "mainbus" or alike also attaches other devices (eg CPUs) which do not attach to a specific attribute yet, we need at least pass an attribute name (different from "foobus") so that the foo bus is not found at these places. This made some minor changes necessary which are not obviously related to the mentioned buses.
1.28	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.27	09-Dec-2003	briggs	Use aprint_*()
1.26	15-Jun-2003	fvdl	branches: 1.26.2; Handle 64bit DMA addresses on PCI for platforms that can (currently only enabled on amd64). Add a dmat64 field to various PCI attach structures, and pass it down where needed. Implement a simple new function called pci_dma64_available(pa) to test if 64bit DMA addresses may be used. This returns 1 iff _PCI_HAVE_DMA64 is defined in <machine/pci_machdep.h>, and there is more than 4G of memory.
1.25	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.24	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.23	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.22	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.21	16-May-2002	thorpej	branches: 1.21.2; * Add "pcitag_t pba_bridgetag" to pci_attach_args. This is set to NULL for root PCI busses. For busses behind a bridge, it points to a persistent copy of the bridge's pcitag_t. This can be very useful for machine-dependent PCI bus enumeration code. Implement a machine-dependent pci_enumerate_bus() for sparc64 which uses OFW device nodes to enumerate the bus. When a PCI bus that is behind a bridge is attached, pci_attach_hook() allocates a new PCI chipset tag for the new bus and sets it's "curnode" to the OFW node of the bridge. This is used as a starting point when enumerating that bus. Root busses get the OFW node of the host bridge (psycho). * Garbage-collect "ofpci" and "ofppb" from the sparc64 port.
1.20	13-Nov-2001	lukem	add RCSID
1.19	04-Nov-1999	thorpej	branches: 1.19.6; 1.19.8; Don't pass rd/mult capability though a PCI-PCI bridge. The bridge would have to break it up into mutliple rd/line's anyhow, so why bother letting the device issue it in the first place.
1.18	08-Jun-1998	thorpej	branches: 1.18.14; 1.18.16; 1.18.20; Nuke __BROKEN_INDIRECT_CONFIG.
1.17	04-Mar-1998	cgd	clean up slightly, correct a few comments
1.16	12-Jan-1998	thorpej	branches: 1.16.2; Update for config changes.
1.15	30-Aug-1997	mycroft	branches: 1.15.2; 1.15.4; Pass down bus_dma_tag_t's as appropriate (per Jason's bus_dma code).
1.14	30-Aug-1997	mycroft	Pass the I/O and memory enable flags through the bridge.
1.13	05-Dec-1996	cgd	update these so they compile whether or not __BROKEN_INDIRECT_CONFIG is defined.
1.12	21-Oct-1996	thorpej	New bus.h implementation/interface: - No more distinction between i/o-mapped and memory-mapped devices. It's all "bus space" now, and space tags differentiate the space with finer grain than the bus chipset tag. - Add memory barrier methods. - Implement space alloc/free methods. - Implement region read/write methods (like memcpy to/from bus space). This interface provides a better abstraction for dealing with machine-independent chipset drivers.
1.11	13-Oct-1996	christos	backout kprintf changes
1.10	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.9	27-Aug-1996	cgd	change cfprint_t type definition to take a const char , rather than a char , because that's what was really intended, and because if the print function modifies the string, various things could become unhappy (so the string should _not_ be modified).
1.8	03-May-1996	christos	remove unused variables
1.7	27-Mar-1996	cgd	modify these to provide a new, better-specified PCI interface (soon to be documented on mailing lists; eventually in section 9 manual pages), most importantly: (1) support interrupt pin swizzling on non-i386 systems with PCI-PCI bridges (per PPB spec; done, but meaningless, on i386). (2) provide pci_{io,mem}_find(), to determine what I/O or memory space is described by a given PCI configuration space mapping register. (3) provide pci_intr_map(), pci_intr_string(), and pci_intr_{,dis}establish() to manipulate and print info about PCI interrupts. (4) make pci functions take as an argument a machine-dependent cookie, to allow more flexibility in implementation.
1.6	17-Mar-1996	cgd	spacing nit
1.5	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.4	14-Mar-1996	cgd	These devices don't actually need softc's that are supersets of 'struct device.' No point in adding that complexity + space if they're not needed, so axe them.
1.3	14-Mar-1996	cgd	(1) provide #defines for cf_loc[] entries for devices that attach to pcibus and pci. (2) remove the #ifdef i386 from pci.c, and provide a machine-dependent hook (pci_md_attach_hook()) to do any machine-dependent attachment gunk, e.g. on the i386 printing out the configuration mode (if bus 0) (3) don't pass max device number for a given bus in, use PCI_MAX_DEVICE_NUMBER, which can be defined on a per-machine basis. (defaults to 32. on i386, it's 32 if pci conf mode == 1, 16 if 2.)
1.2	04-Mar-1996	cgd	fix aux argument to config_found() for secondary bus.
1.1	28-Feb-1996	cgd	Preliminary support for PCI-PCI bridges. Recognize a PCI-PCI bridge and attach the secondary pci bus as a 'pci' device. Note that this support is incomplete and will not yet work for ports other than that i386. (The i386 can rely on the PCI interrupt 'line' information to determine interrupt mapping, which is not necessarily possible on other systems.)
1.15.4.2	01-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.15.4.1	30-Aug-1997	thorpej	file ppb.c was added on branch marc-pcmcia on 1997-09-01 20:29:55 +0000
1.15.2.2	30-Aug-1997	mycroft	Pass down bus_dma_tag_t's as appropriate (per Jason's bus_dma code).
1.15.2.1	30-Aug-1997	mycroft	file ppb.c was added on branch thorpej-bus-dma on 1997-08-30 06:53:59 +0000
1.16.2.2	12-Jan-1998	thorpej	Update for config changes.
1.16.2.1	12-Jan-1998	thorpej	file ppb.c was added on branch bouyer-scsipi on 1998-01-12 09:40:13 +0000
1.18.20.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.18.16.1	15-Nov-1999	fvdl	Sync with -current
1.18.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.19.8.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.19.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.19.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.6.4	03-Jan-2003	thorpej	Sync with HEAD.
1.19.6.3	18-Oct-2002	nathanw	Catch up to -current.
1.19.6.2	20-Jun-2002	nathanw	Catch up to -current.
1.19.6.1	14-Nov-2001	nathanw	Catch up to -current.
1.21.2.2	16-May-2002	thorpej	* Add "pcitag_t pba_bridgetag" to pci_attach_args. This is set to NULL for root PCI busses. For busses behind a bridge, it points to a persistent copy of the bridge's pcitag_t. This can be very useful for machine-dependent PCI bus enumeration code. Implement a machine-dependent pci_enumerate_bus() for sparc64 which uses OFW device nodes to enumerate the bus. When a PCI bus that is behind a bridge is attached, pci_attach_hook() allocates a new PCI chipset tag for the new bus and sets it's "curnode" to the OFW node of the bridge. This is used as a starting point when enumerating that bus. Root busses get the OFW node of the host bridge (psycho). * Garbage-collect "ofpci" and "ofppb" from the sparc64 port.
1.21.2.1	16-May-2002	thorpej	file ppb.c was added on branch gehenna-devsw on 2002-05-16 01:01:31 +0000
1.26.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.26.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.26.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.26.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.26.2.2	03-Sep-2004	skrll	Sync with HEAD
1.26.2.1	03-Aug-2004	skrll	Sync with HEAD
1.29.6.1	12-Feb-2005	yamt	sync with head.
1.29.4.1	29-Apr-2005	kent	sync with -current
1.31.2.4	27-Feb-2008	yamt	sync with head.
1.31.2.3	21-Jan-2008	yamt	sync with head
1.31.2.2	27-Oct-2007	yamt	sync with head.
1.31.2.1	30-Dec-2006	yamt	sync with head.
1.32.22.2	10-Dec-2006	yamt	sync with head.
1.32.22.1	22-Oct-2006	yamt	sync with head
1.32.20.1	18-Nov-2006	ad	Sync with head.
1.34.26.1	18-Oct-2007	yamt	sync with head.
1.34.24.3	23-Mar-2008	matt	sync with HEAD
1.34.24.2	09-Jan-2008	matt	sync with HEAD
1.34.24.1	06-Nov-2007	matt	sync with HEAD
1.34.22.10	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.34.22.9	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.34.22.8	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.34.22.7	08-Oct-2007	joerg	Next attempt at trying to fix the irregular interrupt storms on my Thinkpad: when we find a PCI Express device, check the list of notification events and if any are sets, clear them. We can't handle them ATM anyway.
1.34.22.6	04-Oct-2007	joerg	Remove SCI hack as the ACPI EC issues are fixed and those were the likely source of this problem.
1.34.22.5	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.34.22.4	04-Sep-2007	joerg	Disable PCI-E SCI on all ICH6+ chipsets.
1.34.22.3	21-Aug-2007	joerg	Disable SCI on the ICH6 PCI Express Ports -- if the BIOS enabled it and such an event happened, the resulting interrupt storm would at best keep the CPU over 40% busy and otherwise hang the box. This is a bandaid until proper PCI Express support exists.
1.34.22.2	08-Aug-2007	jmcneill	Improve ppb power management support.
1.34.22.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.34.8.1	23-Oct-2007	ad	Sync with head.
1.35.8.1	11-Dec-2007	yamt	sync with head.
1.35.6.1	26-Dec-2007	ad	Sync with head.
1.36.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.36.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.36.6.1	24-Mar-2008	keiichi	sync with head.
1.37.4.3	11-Mar-2010	yamt	sync with head
1.37.4.2	04-May-2009	yamt	sync with head.
1.37.4.1	16-May-2008	yamt	sync with head.
1.37.2.1	18-May-2008	yamt	sync with head.
1.39.22.1	07-Jan-2011	matt	On powerpc systems, match powerpc processors which are really PCIe PPBs.
1.39.18.4	15-Feb-2014	matt	sync pcireg.h with HEAD. update if_wm.c and ppb.c accordingly.
1.39.18.3	24-Dec-2011	matt	Pull latest from -HEAD. Add printing bus lanes and speed.
1.39.18.2	28-Jan-2010	matt	Change ppb_fix_pcix to ppb_fix_pcie since pcix is usually PCI-X while pcie is PCI-Express.
1.39.18.1	21-Jan-2010	matt	Support the SB1250 LDT HB.
1.39.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.39.10.1	11-May-2013	riz	Applied patch (requested by msaitoh in ticket #1844): sys/dev/pci/pcireg.h 1.69 sys/dev/pci/ppb.c 1.44-1.45 Support PCI Express 2.0. Print version and device/port type information [msaitoh, ticket #1844]
1.39.8.1	28-Apr-2009	skrll	Sync with HEAD.
1.41.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.42.2.1	05-Mar-2011	rmind	sync with head
1.47.6.1	18-Feb-2012	mrg	merge to -current.
1.47.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.47.2.2	30-Oct-2012	yamt	sync with head
1.47.2.1	17-Apr-2012	yamt	sync with head
1.49.6.3	03-Dec-2017	jdolecek	update from HEAD
1.49.6.2	23-Jun-2013	tls	resync from head
1.49.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.52.10.1	12-Dec-2014	martin	Pull up following revision(s) (requested by msaitoh in ticket #311): sys/dev/pci/ppb.c: revision 1.53 sys/dev/pci/ppb.c: revision 1.54 sys/dev/pci/pcireg.h: revision 1.96 - Modify message of PCIe capability version. This field (PCIE_XCAP_VER_MASK) is not specification's version number but the capability structure's version number. To avoid confusion, print "PCI Express capability version x". - The max number of PCIe lane is not 16 but 32. Fix the bug using with macro. - Use macro instead of magic number. - Gb/s -> GT/s Rename PCIE_XCAP_VER_* macros to avoid confusion.
1.54.2.2	28-Aug-2017	skrll	Sync with HEAD
1.54.2.1	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.55.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.55.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.60.2.2	11-May-2017	pgoyette	Sync with HEAD
1.60.2.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.63.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.63.10.1	10-Jun-2019	christos	Sync with HEAD
1.63.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.63.2.3	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1714: sys/dev/pci/pcireg.h 1.148-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.217-1.222, 1.224, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158 sys/dev/pci/ppb.c 1.74 - Print Bridge Config Retry Enable bit and Retimer Presence Detect Supported bit. - Add PCIe 4.0 stuff a little: - 10-bit Tag Requester/Completer. - Add Data link Feature extended capability. - Add Physical Layer 16.0 GT/s extended capability. Not decode yet. - Change pci_conf_print() to allocate memory for the regs dynamically instead of on-stack. - Print some DPC register values not with %04x but with %08x because those are 32bit. - Fix a bug that the virtual channel extended configuration's arbitration phase register can't be decoded correctly. - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Remove unused shift and mask definitions. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.63.2.2	17-Jul-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1295): sys/dev/pci/pci_subr.c: revision 1.213 sys/dev/pci/ppb.c: revision 1.69 Identify 16GT/s.
1.63.2.1	01-Feb-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1181): sys/dev/pci/ppb.c: revision 1.66 sys/dev/pci/ppb.c: revision 1.67 Explicitly enable bus masterling in case BIOS, UEFI or firmware don't enable it. Might fix PR kern/53811. - If the secondary bus is configured and the bus mastering is not enabled, enable it. Suggested by thorpej@.
1.69.2.1	03-Dec-2021	martin	Pull up the following revisions, requested by msaitoh in ticket #1384: sys/dev/pci/pcireg.h 1.152-1.154, 1.156-1.161 sys/dev/pci/pci_subr.c 1.222, 1.227-1.232 via patch sys/dev/pci/nvme_pci.c 1.31 sys/dev/pci/pci.c 1.158, 1.163 sys/dev/pci/ppb.c 1.74 - When parsing Enhanced Allocation entries, use the correct calculation for finding the next entry. - Add 32.0GT/s to the list of pcie speeds (PCIe 5.x.). - Add Some PCI config information: - Lane Margining at the Receiver - NVME admin interface - UFSHCI - InfiniBand - Host fabric - HDA 1.0 with vendor ext - USB4 HCI - MIPI I3C - Cellular controller/modem (+ Ethernet) - Change PCI_VENDOR_MASK and PCI_PRODUCT_MASK to unsigned values, to prevent sign extension of product ID when shifted up into place in PCI_ID_CODE(). Fixes PR kern/56176. - Add LCAP & LCAP2 definitions. - Use PCI-SIG official acronyms for some macros. - Fix typo in some messages. - Fix typo in comments. - Whitespace fixes.
1.70.4.1	23-Mar-2021	thorpej	Convert config_found_ia() call sites where the device only carries a single interface attribute to bare config_found() calls.
1.71.4.1	31-May-2021	cjep	sync with head
1.71.2.1	13-May-2021	thorpej	Sync with HEAD.
1.72.4.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.74.4.1	06-Feb-2024	martin	Pull up following revision(s) (requested by rin in ticket #588): sys/dev/pci/ppb.c: revision 1.75 ppb(4): Print out PCIe Gen5 link speed correctly
1.75.2.1	02-Aug-2025	perseant	Sync with HEAD
1.10	01-Mar-2019	msaitoh	- Almost all ppbreg.h's definitions are also in pcireg.h. Remove duplicated definitions from ppbreg.h and move some definitions from ppbreg.h to pcireg.h. - Change fast back-to-back "capable" to "enable" in pci_subr.c. - Print Primary Discard Timer, Secondary Discard Timer, Discard Timer Status and Discard Timer SERR# Enable bit in pci_subr.c. - PCI_BRIDGE_PREFETCHBASE32_REG and PCI_BRIDGE_PREFETCHLIMIT32_REG are "upper" 32bit registers, rename to *UP32_REG to avoid confusion. - Use macro.
1.9	02-Jul-2018	msaitoh	VGA 16 bit decode bit is not bit 3 but bit 4.
1.8	18-Dec-2017	msaitoh	branches: 1.8.2; 1.8.4; Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified.
1.7	10-May-2017	msaitoh	branches: 1.7.2; Fix typos.
1.6	11-Dec-2005	christos	branches: 1.6.120; 1.6.140; 1.6.154; merge ktrace-lwp.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	08-Nov-2001	thorpej	branches: 1.4.16; 1.4.24; 1.4.26; Fix the Bridge Control Register bit definitions, add ones that appeared in PCI 2.2.
1.3	06-Jul-2001	mcr	branches: 1.3.2; 1.3.6; added bridge secondary bus reset macros.
1.2	14-Mar-1996	cgd	branches: 1.2.42; (1) provide #defines for cf_loc[] entries for devices that attach to pcibus and pci. (2) remove the #ifdef i386 from pci.c, and provide a machine-dependent hook (pci_md_attach_hook()) to do any machine-dependent attachment gunk, e.g. on the i386 printing out the configuration mode (if bus 0) (3) don't pass max device number for a given bus in, use PCI_MAX_DEVICE_NUMBER, which can be defined on a per-machine basis. (defaults to 32. on i386, it's 32 if pci conf mode == 1, 16 if 2.)
1.1	28-Feb-1996	cgd	Preliminary support for PCI-PCI bridges. Recognize a PCI-PCI bridge and attach the secondary pci bus as a 'pci' device. Note that this support is incomplete and will not yet work for ports other than that i386. (The i386 can rely on the PCI interrupt 'line' information to determine interrupt mapping, which is not necessarily possible on other systems.)
1.2.42.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.42.1	24-Aug-2001	nathanw	Catch up with -current.
1.3.6.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.3.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.26.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.24.1	29-Apr-2005	kent	sync with -current
1.4.16.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.154.1	11-May-2017	pgoyette	Sync with HEAD
1.6.140.1	28-Aug-2017	skrll	Sync with HEAD
1.6.120.1	03-Dec-2017	jdolecek	update from HEAD
1.7.2.2	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #933): sys/dev/pci/pci_subr.c: revision 1.202 sys/dev/pci/pcireg.h: revision 1.139 sys/dev/pci/ppbreg.h: revision 1.9 VGA 16 bit decode bit is not bit 3 but bit 4. -- - Print Power Management Control/status register in 32bit. - Simplify.
1.7.2.1	26-Feb-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #576): sys/dev/pci/pci_subr.c: 1.197-1.200 sys/dev/pci/pcireg.h: 1.136-1.137 sys/dev/pci/ppbreg.h: 1.8 Add VGA 16bit decode bit into the PCI bridge control register. This bit is defined in PCI-to-PCI Bridge Architecture Specification Revision 1.2. This bit has meaning if the VGA enable bit or the VGA Palette Snoop Enable bit is set. NOTE: sys/arch/x86/pci/pci_ranges.c::mmio_range_extend_by_vga_enable() and/or some other functions should be modified. "s/above 300W/greater than 300W/" in pci_conf_print_pcie_power(). From PCIe Base Spec 3.1a Errata 2017-12-13. Cleanup: - Don't pass a capability pointer as a argument of pci_conf_find_cap() and determine the first pointer in the pci_conf_find_cap() function. - Don't pass a capability pointer as a argument of pci_conf_find_extcap() because it's not used. - Remove unsed code. - Add PCie Link Activation ECN. - Use macro. - KNF.
1.8.4.1	10-Jun-2019	christos	Sync with HEAD
1.8.2.1	28-Jul-2018	pgoyette	Sync with HEAD
1.4	16-Apr-2020	rin	Revert previous for now: http://mail-index.netbsd.org/source-changes/2020/04/16/msg116278.html The reasoning turned out to be wrong; __KERNEL_RCSID() in header files does not overwrite RCSID in main source files. The real problem is that it inserts its RCSID into every object files. However, it can be still useful even if heavily duplicated.
1.3	16-Apr-2020	rin	Stop using __KERNEL_RCSID() in header files; it confuses ident(1) by overwriting RCSID in main source files. XXX The first argument of __KERNEL_RCSID() is neglected for ELF. If we wish to have RCSID of header files in kernel binary, we need something like __FBSDID() macro in FreeBSD.
1.2	18-Apr-2017	msaitoh	branches: 1.2.2; 1.2.4; 1.2.12; 1.2.16; 1.2.32; Enable PCIe's interrupt as much as possilbe in ppb(4) to detect and count status change event. HotPlug function itself have not implemented yet. - Interrupt and each event are counted by evcnt(9). Example: ppb0 Interrupt 0 0 intr ppb0 Attention Button Pressed 0 0 misc ppb0 Power Fault Detected 0 0 misc ppb0 MRL Sensor Changed 0 0 misc ppb0 Presence Detect Changed 0 0 misc ppb0 Command Completed 0 0 misc ppb0 Data Link Layer State Changed 0 0 misc - Print message if ppb_printevent is not zero. The default vaule is 0. The output messages: Attention Button Pressed Power Fault Detected MRL Sensor Changed Presence Detect Changed Command Completed Data Link Layer State Changed - Remove workaround code to disable interrupt (ppb.c rev. 1.35). Tested with Dell Latitude 2120 without if_bge.c rev. 1.304's workaround. dmesg when bge's device timeout occured: ppb3: Presence Detect Changed ppb3: Data Link Layer State Changed ppb3: Presence Detect Changed vmstat -e \|grep ppb ppb3 Interrupt 2 0 intr ppb3 Presence Detect Changed 2 0 misc ppb3 Data Link Layer State Changed 1 0 misc
1.1	05-Apr-2017	msaitoh	Move struct ppb_softc into ppbvar.h.
1.2.32.1	20-Apr-2020	bouyer	Sync with HEAD
1.2.16.2	03-Dec-2017	jdolecek	update from HEAD
1.2.16.1	18-Apr-2017	jdolecek	file ppbvar.h was added on branch tls-maxphys on 2017-12-03 11:37:28 +0000
1.2.12.2	28-Aug-2017	skrll	Sync with HEAD
1.2.12.1	18-Apr-2017	skrll	file ppbvar.h was added on branch nick-nhusb on 2017-08-28 17:52:25 +0000
1.2.4.2	26-Apr-2017	pgoyette	Sync with HEAD
1.2.4.1	18-Apr-2017	pgoyette	file ppbvar.h was added on branch pgoyette-localcount on 2017-04-26 02:53:22 +0000
1.2.2.2	21-Apr-2017	bouyer	Sync with HEAD
1.2.2.1	18-Apr-2017	bouyer	file ppbvar.h was added on branch bouyer-socketcan on 2017-04-21 16:53:51 +0000
1.42	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.41	24-Apr-2021	thorpej	branches: 1.41.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.40	30-Nov-2018	jmcneill	branches: 1.40.14; Add support for polling com devices when no interrupt is available.
1.39	07-Jul-2016	msaitoh	branches: 1.39.16; 1.39.18; KNF. Remove extra spaces. No functional change.
1.38	04-May-2015	ryo	PR/49819: Roberto E. Vargas Caballero: Add support for SystemBase SB16C1050 PCI serial card
1.37	07-Feb-2014	msaitoh	branches: 1.37.6; s/mesage/message/
1.36	23-Jul-2013	soren	Restrict the hack for offset BAR's to x86. A cleaner fix requires a rewrite of puc.c, but not today.
1.35	22-Jul-2013	soren	Oops.
1.34	22-Jul-2013	martin	Unbreak the build - soren, please review!
1.33	22-Jul-2013	soren	Allow console on com_puc without a compile-time option so that PC servers can become headless after the first reboot (sadly, e.g. Intel AMT presents as a com_puc, but doesn't appear in the BIOS serial port table, so you need a keyboard and monitor to install and set the installboot parameters first). Fix com_puc console on devices with offset BAR's.
1.32	28-May-2011	ryo	branches: 1.32.4; 1.32.14; 1.32.18; 1.32.26; Add support for SYSTEMBASE SB16C105x 4-port/8-port serial PCI cards. - Added initialize code for SB16C105x to puc.c, but It is better to add a member (config_function)() to struct puc_device_description and use it - It seems SB16C1054 rev 0x91* has different BAR layout, but not supported yet.
1.31	09-Jul-2008	joerg	branches: 1.31.16; 1.31.22; - device/softc split
1.30	10-Apr-2008	cegger	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.10; use aprint_*_dev and device_xname
1.29	16-Nov-2006	christos	branches: 1.29.48; __unused removal on arguments; approved by core.
1.28	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.27	11-Dec-2005	christos	branches: 1.27.20; 1.27.22; merge ktrace-lwp.
1.26	26-Aug-2005	drochner	kill some more simple submatch() functions, use config_stdsubmatch()
1.25	25-Aug-2005	drochner	replace the "locdesc_t" structure carrying the number of locators explicitely by a plain integer array the length in now known to all relevant parties, so this avoids duplication of information, and we can allocate that thing in drivers without hacks
1.24	28-Jun-2005	thorpej	branches: 1.24.2; Use ANSI function decls and static.
1.23	27-Feb-2005	perry	nuke trailing whitespace
1.22	04-Feb-2005	perry	de-__P
1.21	13-Sep-2004	drochner	branches: 1.21.4; 1.21.6; a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.20	03-Feb-2004	fredb	Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.19	25-Jan-2004	jdolecek	pass the DMA tags to child devices, too
1.18	19-Mar-2003	christos	branches: 1.18.2; It is not appropriate for pcivar.h to include "locators.h" The two files that need it, should include it themselves.
1.17	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.16	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.15	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.14	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.13	27-Sep-2002	thorpej	Introduce a new routine, config_match(), which invokes the cfattach->ca_match function in behalf of the caller. Use it rather than invoking cfattach->ca_match directly.
1.12	13-Nov-2001	lukem	add RCSID
1.11	02-Mar-2001	thorpej	branches: 1.11.2; 1.11.4; In the self-describing bogus code block, at least use plausible subclass IDs.
1.10	03-Jan-2001	bouyer	Add a 'flag' member to struct puc_attach_args and struct port. For PUC_PORT_TYPE_COM, use it to store the clock frequency (with 8 lower bits to 0, used for real flags if needed). Update all descriptions to set flags to 0 for LPT or COM_FREQ for COM. Add support for the VScom PCI-800H 8 port serial adapter (which uses a 14.7456 Mhz crystal instead of the standart 1.8432Mhz :) XXX now that we can pass other frequency than COM_FREQ, the VScom PCI-800 entry could probably be updated to DTRT - does anyone have one ?
1.9	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.8	29-Jul-2000	jhawk	Whitespace fix (space rather than tab after #ifdef)
1.7	29-Jul-2000	jlam	Make this compile without PUCCN defined.
1.6	28-Jul-2000	castor	Guard the console code in puc.c with the defopt PUCCN so we can use the puc driver without having the com drivers.
1.5	25-Jul-2000	jeffs	Add code to allow the PCI com serial ports to be used as the system console. This is not enabled by default, and is turned on with options PUCCN. Done by castor@netbsd.org.
1.4	17-Apr-2000	cgd	branches: 1.4.4; #if 0 the (way bogus) PCI_CLASS_COMMUNICATIONS/PCI_SUBCLASS_BRIDGE_PCI (!!!) match pointed out (again?) by Johan Danielsson, after i repeatedly told him not to do something similar in com_cardbus.
1.3	06-Feb-1999	cgd	branches: 1.3.8; add code to cope with multiple ports per BAR, as are found, for instance, on the 8 port card Simon Gerraty has. In general, cards which have this lots of ports also have a separate interrupt status register, but this change is just to talk to the various ports independently. It works, but it's not optimal. (XXX still need a good name for the card in the comments, and to update the manual page.)
1.2	21-Dec-1998	drochner	use a symbolic definition for the PCI subsystem ID
1.1	26-Jun-1998	cgd	add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.3.8.3	12-Mar-2001	bouyer	Sync with HEAD.
1.3.8.2	05-Jan-2001	bouyer	Sync with HEAD
1.3.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.4.2	22-Mar-2001	he	Apply patch (requested by he): The function puc_find_description() is now external, so make it so.
1.4.4.1	22-Mar-2001	he	Pull up revisions 1.10-1.11 (requested by sommerfeld): Add support for NetMos NM9835, Titan PCI-800H, Lava 8-port, Actiontec 56K PCI Master. This also adds a ``flags'' member to two structs in this driver.
1.11.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.11.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.11.2.3	03-Jan-2003	thorpej	Sync with HEAD.
1.11.2.2	18-Oct-2002	nathanw	Catch up to -current.
1.11.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.18.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.18.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.18.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.18.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.18.2.1	03-Aug-2004	skrll	Sync with HEAD
1.21.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.6.1	12-Feb-2005	yamt	sync with head.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.24.2.2	30-Dec-2006	yamt	sync with head.
1.24.2.1	21-Jun-2006	yamt	sync with head.
1.27.22.2	10-Dec-2006	yamt	sync with head.
1.27.22.1	22-Oct-2006	yamt	sync with head
1.27.20.1	18-Nov-2006	ad	Sync with head.
1.29.48.2	28-Sep-2008	mjf	Sync with HEAD.
1.29.48.1	02-Jun-2008	mjf	Sync with HEAD.
1.30.10.1	19-Oct-2008	haad	Sync with HEAD.
1.30.8.1	18-Jul-2008	simonb	Sync with head.
1.30.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.30.4.1	04-May-2009	yamt	sync with head.
1.31.22.1	06-Jun-2011	jruoho	Sync with HEAD.
1.31.16.1	31-May-2011	rmind	sync with head
1.32.26.1	23-Jul-2013	riastradh	sync with HEAD
1.32.18.2	18-May-2014	rmind	sync with head
1.32.18.1	28-Aug-2013	rmind	sync with head
1.32.14.2	03-Dec-2017	jdolecek	update from HEAD
1.32.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.32.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.37.6.2	09-Jul-2016	skrll	Sync with HEAD
1.37.6.1	06-Jun-2015	skrll	Sync with HEAD
1.39.18.1	10-Jun-2019	christos	Sync with HEAD
1.39.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.40.14.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.40.14.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.41.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.14	05-Mar-2014	msaitoh	No functional change: - Rename function for consistency. - Add prototypes. - KNF.
1.13	26-Jan-2014	msaitoh	PUCCN improvements: - Fix a bug that the puc cn mechanism doesn't use the UART's frequency in pucdata.c's table. - Add a new option PUC_CNAUTO. If this option is set, consinit() in x86/x86/consinit.c checks puc com device to use it as console. Without this option, the behavior is the same as before. - Add a new config parameter PUC_CNBUS. The old code scans bus #0 only. If PUC_CNBUS is set, the specified number's bus will be scanned. - Rename comcnprobe() to puc_cnprobe() to make it clear. - Rename comcninit() to puc_cninit() to make it clear. - Add code for a device that a device's com register is MMIO (#if0 ed).
1.12	23-Jan-2014	msaitoh	Add comments and KNF. No functional change.
1.11	22-Jul-2013	soren	Allow console on com_puc without a compile-time option so that PC servers can become headless after the first reboot (sadly, e.g. Intel AMT presents as a com_puc, but doesn't appear in the BIOS serial port table, so you need a keyboard and monitor to install and set the installboot parameters first). Fix com_puc console on devices with offset BAR's.
1.10	28-Apr-2010	dyoung	branches: 1.10.8; 1.10.18; 1.10.22; 1.10.30; On x86, change the bus_space_tag_t to a pointer to a struct bus_space_tag. For now, bus_space_tag's only member is bst_type, the type of space, which is either X86_BUS_SPACE_IO or X86_BUS_SPACE_MEM. In the future, new bus_space_tag members will refer to override-functions installed by a new function, bus_space_tag_create(9). Add pointers to constant struct bus_space_tag, x86_bus_space_io and x86_bus_space_mem. Use them to replace most uses of X86_BUS_SPACE_IO and X86_BUS_SPACE_MEM. Add an x86-specific bus_space_is_equal(9) implementation that compares the two tags' bst_type.
1.9	13-Jan-2007	cube	branches: 1.9.44; 1.9.66; 1.9.68; Work around false-positives for gcc's detection of use of uninitialised variables.
1.8	11-Dec-2005	christos	branches: 1.8.20; merge ktrace-lwp.
1.7	27-Feb-2005	perry	branches: 1.7.4; nuke trailing whitespace
1.6	14-Jun-2003	thorpej	branches: 1.6.2; 1.6.10; 1.6.12; Also pass a type argument to comcnattach() and com_kgdb_attach(). comspeed() (and thus cominit()) may need this information.
1.5	29-Mar-2003	matt	Make this compile again.
1.4	13-Nov-2001	lukem	add RCSID
1.3	30-May-2001	lukem	branches: 1.3.2; add missing #include "opt_kgdb.h"
1.2	09-Dec-2000	imp	branches: 1.2.2; Change $Id: to $NetBSD$
1.1	25-Jul-2000	jeffs	branches: 1.1.2; Add code to allow the PCI com serial ports to be used as the system console. This is not enabled by default, and is turned on with options PUCCN. Done by castor@netbsd.org.
1.1.2.3	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.1.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	25-Jul-2000	bouyer	file puccn.c was added on branch thorpej_scsipi on 2000-11-20 11:42:36 +0000
1.2.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.2.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.3.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.6.10.1	29-Apr-2005	kent	sync with -current
1.6.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.4.1	26-Feb-2007	yamt	sync with head.
1.8.20.1	01-Feb-2007	ad	Sync with head.
1.9.68.1	30-May-2010	rmind	sync with head
1.9.66.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.9.44.1	11-Aug-2010	yamt	sync with head.
1.10.30.1	23-Jul-2013	riastradh	sync with HEAD
1.10.22.2	18-May-2014	rmind	sync with head
1.10.22.1	28-Aug-2013	rmind	sync with head
1.10.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.10.8.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7	05-Mar-2014	msaitoh	No functional change: - Rename function for consistency. - Add prototypes. - KNF.
1.6	26-Jan-2014	msaitoh	PUCCN improvements: - Fix a bug that the puc cn mechanism doesn't use the UART's frequency in pucdata.c's table. - Add a new option PUC_CNAUTO. If this option is set, consinit() in x86/x86/consinit.c checks puc com device to use it as console. Without this option, the behavior is the same as before. - Add a new config parameter PUC_CNBUS. The old code scans bus #0 only. If PUC_CNBUS is set, the specified number's bus will be scanned. - Rename comcnprobe() to puc_cnprobe() to make it clear. - Rename comcninit() to puc_cninit() to make it clear. - Add code for a device that a device's com register is MMIO (#if0 ed).
1.5	22-Jul-2013	soren	Allow console on com_puc without a compile-time option so that PC servers can become headless after the first reboot (sadly, e.g. Intel AMT presents as a com_puc, but doesn't appear in the BIOS serial port table, so you need a keyboard and monitor to install and set the installboot parameters first). Fix com_puc console on devices with offset BAR's.
1.4	11-Dec-2005	christos	branches: 1.4.110; 1.4.120; 1.4.126; 1.4.134; merge ktrace-lwp.
1.3	08-Jul-2003	itojun	function prototype must not have variable name
1.2	09-Dec-2000	imp	branches: 1.2.24; Change $Id: to $NetBSD$
1.1	25-Jul-2000	jeffs	branches: 1.1.2; Add code to allow the PCI com serial ports to be used as the system console. This is not enabled by default, and is turned on with options PUCCN. Done by castor@netbsd.org.
1.1.2.3	13-Dec-2000	bouyer	Sync with HEAD (for UBC fixes).
1.1.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.2.1	25-Jul-2000	bouyer	file puccn.h was added on branch thorpej_scsipi on 2000-11-20 11:42:36 +0000
1.2.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.24.1	03-Aug-2004	skrll	Sync with HEAD
1.4.134.1	23-Jul-2013	riastradh	sync with HEAD
1.4.126.2	18-May-2014	rmind	sync with head
1.4.126.1	28-Aug-2013	rmind	sync with head
1.4.120.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.4.110.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.119	03-May-2025	rin	puc(4): Add support to EXAR XR17D358 Tested on M.2 card model.
1.118	12-Dec-2024	msaitoh	puc(4): Add some Oxford Semiconductor devices support from FreeBSD.
1.117	11-Nov-2024	msaitoh	Add many Brainboxes devices. Repoted in PR/kern 58824 by Cameron Williams.
1.116	23-Jun-2024	jakllsch	branches: 1.116.2; Add Brainboxes puc(4) devices From Cameron Williams in PR 58358.
1.115	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.114	29-Jan-2023	thorpej	Add and entry for the EXAR XR17V354 4-port UART. From Håkan Engvall. PR kern/57202
1.113	01-May-2022	mrg	branches: 1.113.4; attach the other pciid for intel 100 series KT serial. seen on a thinkpad 460s.
1.112	12-Oct-2021	msaitoh	Add Intel 400 Series PCH-V devices.
1.111	11-Oct-2021	msaitoh	Add some Intel KT devices.
1.110	17-Jul-2021	riastradh	puc(4): Add Intel Core 5G (mobile) KT.
1.109	02-Feb-2021	bouyer	branches: 1.109.4; Also match Oxford Semiconductor Exsys EX-41098 PCI UARTs
1.108	13-Jun-2020	ryo	branches: 1.108.2; add entry for ASIX AX99100 PCIe 4port serial card
1.107	05-Apr-2020	ryo	add entry for NetMos NM9900 Octa UART notice: On the boards I've obtained, the number on the cable stamp doesn't match the actual unit number of the attached device. Cable's 0,1,2,..,7 corresponds to comN+0,2,4,6,1,3,5,7 :-(
1.106	24-Mar-2020	ryo	add entry for NetMos NM9900 Quad UART. The NM9900 seems to have variations such as 4UART and 2UART+1Printer, but I tested on 4UART board.
1.105	14-Nov-2019	hauke	The 16C1054 and 16C1058 serial multi-port controllers need a clock multiplier of 8, just like the 16C1050 controller. Verified with an ExSys EX-41388. ryo@ checked back with the hardware his original commit was based on, and confirmed the change. XXX Pull-up to netbsd-{7,8,9}
1.104	02-May-2019	jdolecek	branches: 1.104.2; add NetMos NM9912 entries PR kern/54149 by Mouse
1.103	29-Nov-2018	jmcneill	Add Amazon.com EC2 virtual 16650-compatible PCI serial device
1.102	21-May-2018	jnemeth	branches: 1.102.2; add Oxford Semiconductor OXPCIe954 PCIe 4-port serial card
1.101	13-Apr-2018	msaitoh	Add 300 series chipset support.
1.100	27-Dec-2017	christos	branches: 1.100.2; PR/52868: Petar Bogdanovic: Add support for Manhattan 158220 card
1.99	15-Jan-2017	msaitoh	branches: 1.99.8; Add Intel 200 series devices.
1.98	01-Dec-2015	msaitoh	branches: 1.98.2; Add Intel 100 Series chipset's devices.
1.97	23-Aug-2015	jakllsch	whitespace
1.96	23-Aug-2015	jakllsch	Support PCI_PRODUCT_INTEL_82Q45_KT_1 in addition to PCI_PRODUCT_INTEL_82Q45_KT.
1.95	04-May-2015	ryo	PR/49819: Roberto E. Vargas Caballero: Add support for SystemBase SB16C1050 PCI serial card
1.94	18-Mar-2015	msaitoh	Add 9 Series support.
1.93	06-Feb-2014	msaitoh	branches: 1.93.4; 1.93.6; USE PCI_BAR[0-5]. Part of PR#48016.
1.92	26-Jan-2014	msaitoh	Add some OXPCIe952 devices.
1.91	23-Jan-2014	msaitoh	Add entry for Intel Quark X1000 UART.
1.90	20-Jan-2014	msaitoh	Fix a bug that EG20T's UARTs from #1 to #3 never match. Tested with my MinnowBoard.
1.89	25-Dec-2013	msaitoh	Add DIGI Neo 8-port serial(PCIe) and OXFORD OXPCIe952 Parallel.
1.88	25-Dec-2013	msaitoh	Add support S1200's UART. Part of PR#48016.
1.87	07-Nov-2013	msaitoh	Add some Mobile 4th generation Intel Core Processor devices (Lynx Point-LP).
1.86	04-Sep-2013	christos	PR/48180: Christian Groessler: Support for Oxford Semi OXuPCI952 dual serial PCI card
1.85	17-Jul-2013	soren	Extend support for Nanjing QinHeng Electronics puc(4) devices.
1.84	03-Jun-2013	msaitoh	branches: 1.84.2; 1.84.4; Add Intel 8 Series KT.
1.83	19-Dec-2012	msaitoh	space change.
1.82	18-Oct-2012	msaitoh	- Indent. - Fix some product names or device names.
1.81	25-Sep-2012	gson	Add Lava Computers SSERIAL-PCI single port serial PCI card.
1.80	15-Aug-2012	msaitoh	branches: 1.80.2; Add EXSYS EX-41098-2 UARTs support. The clock frequency is different from other card(s). Fixes PR#46362 reported by Wolfgang Stukenbrock.
1.79	07-Aug-2012	msaitoh	Add some Intel's KT (Serial over LAN) devices.
1.78	05-Aug-2012	msaitoh	Use PCI_VENDOR_* and PCI_PRODUCT_*. Fixes PR#46303.
1.77	27-Jul-2012	matt	Add EG20T UART entries. Closes PR/45567.
1.76	09-Aug-2011	dyoung	branches: 1.76.2; 1.76.8; Add support for the Oxford Semiconductor OXPCIe952, a 2-port serial chip with a PCI Express interface, in legacy mode. Cursorily tested with a StarTech PEX2S952LP. See the manual for your adapter to see how to put it into legacy mode. There is also a "native" mode that uses a different clock multiplier and (I think) provides a richer feature set, but also requires a bit of additional initialization (see FreeBSD's puc(4)). The native-mode PCI functions use different PCI Device IDs than the legacy-mode functions.
1.75	28-May-2011	ryo	Add support for SYSTEMBASE SB16C105x 4-port/8-port serial PCI cards. - Added initialize code for SB16C105x to puc.c, but It is better to add a member (config_function)() to struct puc_device_description and use it - It seems SB16C1054 rev 0x91* has different BAR layout, but not supported yet.
1.74	14-May-2011	martin	PR kern/44956: add support for Qinheng PCI cards to the puc driver
1.73	02-May-2011	manu	Add support for new SUNIX PCI seriel and/or parallel adapters, with new SUNIX vendor Id 0x1fd4 and product Id 0x1999. Constants were taken from Linux open source driver distributed by SUNIX. The following models should be supported now: 5008 1 parallel port 5016 16 serial ports, only 8 supported right now (PUC_MAX_PORTS limit) 5027 1 serial port 5037 2 serial ports 5056 4 serial ports 5066 8 serial ports, this model is the only one that was actually tested. 5069 1 serial port and 1 parallel port 5079 2 serial ports and 1 parallel port 5099 4 serial ports and 1 parallel port
1.72	02-May-2011	manu	Add new PCI vendor Id for Sunix, and the corresponding device Id.
1.71	11-Mar-2011	he	Add two NetMos NM9865 1-port and 2-port controllers, as found in 2x and 1x on a 4-port card sold as Sunsway/ST Lab I-430.
1.70	20-Feb-2011	msaitoh	- Add support for Moschip NM9904 and NM9922. These devices use PCI multi function, so only one PUC_PORT_TYPE_COM entry is used in these devices. - Use macro.
1.69	01-Aug-2010	msaitoh	branches: 1.69.2; 1.69.4; Add the following cards' support from FreeBSD: - Comtrol RocketPort 550/4, 550/8 and 550/16 series - Lava Computers Quatro-PCI - some NetMos NM9865 cards - VScom PCI-200Li
1.68	13-Jul-2010	msaitoh	Add three card support from Tsubai. - Advantech PCI-1604UP (2S) - InnoSys Keyspan SX Pro (4S) - I-O DATA RSA-PCI2 (2S)
1.67	11-Jul-2010	mrg	use PCI_PRODUCT_NETMOS_NM9901 instead of 0x9901 magic number.
1.66	08-Jul-2010	msaitoh	- Fix the frequency for Advantech PCI1620 port 4-8. - Fix the mask for IBM SurePos 300 - s/Titan/VScom/ - Add the following cards: - Avlab PCI 2 Serial - Avlab Low Profile PCI 4 Serial (another revision?) - Moxa CP-102/PCI. - SUNIX 4036 2S - Syba Tech Ltd. PCI-4S - Syba Tech Ltd. PCI-4S2P-550-ECP - PR#37299 - VScom PCI-010L - VScom PCI-100L - VScom PCI-110L - VScom PCI-200L - VScom PCI-210L - VScom PCI-400L - VScom PCI-011H - VScom PCI-100H
1.65	08-Jul-2010	msaitoh	- Fix an entry for CP-168EL-A/PCIe reported by Ryo SHIMIZU. - Add an entry for NetMos NM9901 UART. Tested my own NM9901 card.
1.64	08-Jul-2010	msaitoh	- Fix a table for Advantech PCI-1620 - Add entry for PCI-1610 and PCI-1612
1.63	08-Jul-2010	msaitoh	Use real company name and product name.
1.62	08-Jul-2010	msaitoh	- revert pucdata.c rev. 1.60 and add a new entry which checks subvendor code. - Add the following cards' support (include fixing for PR#37566 and PR#39288): Advantech PCI-1620A-AE (8-port serial) Decision Computer Inc PCCOM PCI 2 Port (2 port serial) Decision Computer Inc PCCOM PCI 4 Port (4 port serial) Decision Computer Inc PCCOM PCI 8 Port (8 port serial) Exsys EX-41098 (4 port serial) IO-DATA RSA-PCI2 (8 port serial) Moxa Technologies SmartIO C168EL-A/PCI (8-port serial) Moxa Technologies SmartIO C168EL/PCIe (8-port serial) Moxa Technologies SmartIO C168U/PCI (8-port serial) Moxa Technologies SmartIO CP104-EL/PCIe (4-port serial) Multi-Tech ISI5634PCI/4 (4 port serial) NetMos NM9835 UART (1 port serial) NetMos NM9865 4 UART (4 port serial) NetMos NM9865 UART (1 port serial) NetMos NM9901 PCIe (single parallel) Oxford OX16PCI958 (8 port serial) SIIG Cyber 4 PCI 16550 (4-port serial) SIIG Cyber 4S PCI 16C650 (20x family) (4-port serial)
1.61	08-Jul-2010	msaitoh	Sort by PCI_VENDOR_*. No functional change.
1.60	02-Jul-2010	msaitoh	Fix frequency for OX16PCI954.
1.59	19-Jan-2010	mbalmer	branches: 1.59.2; 1.59.4; Add the IBM 481033H SCC
1.58	18-Nov-2009	enami	Add an entry for AVlab's low profile PCI 4 port serial card, which is sold in japan from kuroutoshikou as SERIAL4P-LPPCI2.
1.57	29-Aug-2009	tsutsui	Add support for Oxford Semiconductor OXmPCI952 2 port UARTs. Also update URL of Oxford UART chips.
1.56	08-Dec-2008	dyoung	Distinguish the NetMos NM9845 Quad UART chip from the NetMos NM9845 Quad UART + One Parallel chip.
1.55	06-Sep-2008	rmind	branches: 1.55.2; 1.55.4; 1.55.10; PR/38012: David Dudley: Support for B&B Electronics MIPort serial cards.
1.54	03-Apr-2008	riz	branches: 1.54.4; 1.54.6; 1.54.10; Support for Digi International 4 and 8 port boards, from OpenBSD. XXX need to convert the rest of the PCI vendor/product IDs to macros
1.53	22-Feb-2008	tsutsui	Add an entry for I-O DATA RSA-PCI 2-port serial board. (works only on 16550A compatible mode)
1.52	01-Nov-2007	jnemeth	branches: 1.52.12; 1.52.16; PR/37259 - Daniel Neri -- typo in EXAR XR17C158 description
1.51	28-Jan-2007	chs	branches: 1.51.18; 1.51.20; 1.51.24; add some SIIG and Moxa cards.
1.50	09-Dec-2006	manu	Add support for 4 and 8 ports modem boards from Perle systems
1.49	13-Nov-2006	bouyer	branches: 1.49.2; Add NetMos 9855 Quad UART and 1284 Printer port
1.48	31-Oct-2006	manu	Add support for Perle systems PCI-RAS 4 modem ports See http://www.perle.com/products/Multimodem-Cards.shtml for product details
1.47	09-Oct-2006	cube	Add an entry for the 2 UARTs version of the Moschip NM9835.
1.46	03-Sep-2006	christos	branches: 1.46.2; 1.46.4; add missing initializers
1.45	11-Jun-2006	bouyer	Add entry for "Moxa Technologies, SmartIO CP104-V2/PCI". From Gert Doering in kern/33566.
1.44	09-Mar-2006	hamajima	branches: 1.44.4; Add support for EXAR XR17D15x serial. (I tested only XR17D154.)
1.43	06-Mar-2006	martin	branches: 1.43.2; Add support for Moxa CP104 serial card, from Olivier Cherrier
1.42	24-Jan-2006	christos	branches: 1.42.2; 1.42.4; 1.42.6; Add support for more puc chips from Pavel Cahyna. This might NetMos 9845 card which need the COM_FREQ * 4 hack, but this can be re-added by a more specific entry (matching subsystem IDs).
1.41	11-Dec-2005	christos	branches: 1.41.2; merge ktrace-lwp.
1.40	27-Feb-2005	perry	branches: 1.40.2; 1.40.4; nuke trailing whitespace
1.39	16-Oct-2004	bouyer	branches: 1.39.4; 1.39.6; Add the dual-serial VScom PCI-200 adapter. Tested by Jukka Salmi.
1.38	13-Oct-2004	drochner	support ADDI-DATA APCI-7800 8-port serial adapter
1.37	04-Jul-2004	mycroft	Support for the VS-Com PCI-210H 2s/1p card, from PR 18478.
1.36	03-Feb-2004	fredb	branches: 1.36.2; Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.35	25-Jan-2004	jdolecek	fix register offset of second port for NetMos NM9815 Dual 1284 Printer port card
1.34	06-Dec-2003	jdolecek	add entry for another NetMos card - NetMos NM9815 Dual 1284 Printer port
1.33	28-Sep-2003	jdolecek	add entry for Lava Quattro-PCI (4 port serial) contributed in PR kern/22977 by Heison Chak
1.32	29-Oct-2002	chs	branches: 1.32.6; add Boca Research 4- and 8-port cards. from openbsd.
1.31	02-Aug-2002	bouyer	Add Siig 20x 8 ports and and Oxford x952 cards. From Paul Shupak in kern/17790.
1.30	05-Jul-2002	enami	Add an entry for Avlab Technology, Inc. Low Profile PCI 4 Serial (in Japan, sold by Kuroutoshikou as SERIAL4P-LPPCI.)
1.29	02-Jun-2002	seb	Support Titan PCI-800L, a 8 serial puc card, from PR 17140. Reviewed/approved by veego@ and martin@.
1.28	24-May-2002	lukem	Support NetMos NM9845 Quad UART and 1284 Printer port. All 4 serial ports and parallel port seem to work.
1.27	13-May-2002	hannken	branches: 1.27.2; 1.27.4; Add support for NetMos NM9805 1284 Printer port. Tested by David Kwok <dkwok@iware.com.au>.
1.26	11-Apr-2002	hannken	Add config for the SUNIX line of pci serial/parallel cards. Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
1.25	16-Dec-2001	thorpej	Fix a minor typo.
1.24	16-Dec-2001	thorpej	Add the serial port on the Middle Digital, Inc. PCI Weasel.
1.23	26-Nov-2001	jonathan	* Add entry for Dolphin Peripherals 4036 dual-serial card. (Note: PCI ID conflicts with Dolphin Peripherals 4025; a 4025 may attach two serial units, but I cannot find one to test).
1.22	18-Nov-2001	christos	Add Titan 200H (from Wojciech Puchar)
1.21	13-Nov-2001	lukem	add RCSID
1.20	06-Aug-2001	soren	Add support for the PCI-400 card and correct the clock rate for all the VScom/Titan cards.
1.19	04-Jul-2001	thorpej	branches: 1.19.2; Add another Lava serial board.
1.18	13-Mar-2001	bouyer	Looks like "VSCom PCI-800H" should be in fact "Titan PCI-800H".
1.17	04-Mar-2001	msaitoh	branches: 1.17.2; add support for NetMos NM9835 Dual UART and 1284 Printer port.
1.16	14-Feb-2001	christos	make this compile again.
1.15	14-Feb-2001	christos	PR/12204: Andreas Wrede: Add support for Lava 8-port serial card.
1.14	04-Jan-2001	bouyer	Ops, add COM_FREQ to a few PUC_PORT_TYPE_COM I missed. Thanks to Bernd Ernesti for pointing this out.
1.13	04-Jan-2001	veego	Use COM_FREQ for COM part of the last commited entry. Why are there other drivers here which have PUC_PORT_TYPE_COM but don't use COM_FREQ or even don't have the extra field?
1.12	04-Jan-2001	veego	Add support for boards with an Oxford Semiconductor chip. This one was an I/O Flex PCI I/O Card Model-223 with 4 serial and 1 parallel ports. XXX We need to move the PCI IDs to pcidevs.
1.11	03-Jan-2001	bouyer	Add a 'flag' member to struct puc_attach_args and struct port. For PUC_PORT_TYPE_COM, use it to store the clock frequency (with 8 lower bits to 0, used for real flags if needed). Update all descriptions to set flags to 0 for LPT or COM_FREQ for COM. Add support for the VScom PCI-800H 8 port serial adapter (which uses a 14.7456 Mhz crystal instead of the standart 1.8432Mhz :) XXX now that we can pass other frequency than COM_FREQ, the VScom PCI-800 entry could probably be updated to DTRT - does anyone have one ?
1.10	06-Sep-2000	bouyer	Add definition for "Actiontec 56K PCI Master", from Bill Hudspeth.
1.9	08-Feb-2000	bouyer	branches: 1.9.4; Supports "USR 3CP5609 PCI 16550 Modem", from Kent Polk.
1.8	04-Dec-1999	castor	Add support for Lava Computers MFG 2sp-pci and quattro cards.
1.7	16-Nov-1999	jun	Commit patch kern/8778 Subject: NEC PCI modem support
1.6	03-Jul-1999	cgd	branches: 1.6.2; 1.6.8; 'Dolphin' -> 'Dolphin Peripherals' where appropriate
1.5	03-Jul-1999	cgd	Add support for 36 more SIIG PCI serial/parallel cards (for a total of 37). This means that all SIIG PCI serial/parallel cards (described on their web site as "I/O boards" as of July 2, 1999) should now be supported.
1.4	06-Feb-1999	cgd	branches: 1.4.2; 1.4.4; 1.4.6; put in better name/description for Simon G's board
1.3	06-Feb-1999	cgd	add code to cope with multiple ports per BAR, as are found, for instance, on the 8 port card Simon Gerraty has. In general, cards which have this lots of ports also have a separate interrupt status register, but this change is just to talk to the various ports independently. It works, but it's not optimal. (XXX still need a good name for the card in the comments, and to update the manual page.)
1.2	12-Jul-1998	cgd	I went to Fry's and noticed a few more cards that probably will be supportable with this driver, but currently aren't supported. Add them.
1.1	26-Jun-1998	cgd	add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.4.6.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.4.4.1	02-Aug-1999	thorpej	Update from trunk.
1.4.2.2	10-Feb-2000	he	Pull up revision 1.9 (via patch, requested by bouyer): Add support for US Robotics 3CP5609 PCI 16550 Modem.
1.4.2.1	06-Jul-1999	perry	pullup 1.4->1.6 (cgd)
1.6.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.6.2.4	27-Mar-2001	bouyer	Sync with HEAD.
1.6.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.6.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.6.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.9.4.3	04-Sep-2002	itojun	pullup sys/dev/pci/pucdata.c 1.31 via patch (bouyer) Add Siig 20x 8 ports and and Oxford x952 cards. From Paul Shupak in kern/17790.
1.9.4.2	22-Mar-2001	he	Pull up revisions 1.11-1.18 (requested by sommerfeld): Add support for NetMos NM9835, Titan PCI-800H, Lava 8-port, Actiontec 56K PCI Master. This also adds a ``flags'' member to two structs in this driver.
1.9.4.1	06-Sep-2000	bouyer	Pull up 1.9 -> 1.10, approved by thorpej: Add a new modem: Actiontec 56K PCI Master
1.17.2.9	11-Nov-2002	nathanw	Catch up to -current
1.17.2.8	13-Aug-2002	nathanw	Catch up to -current.
1.17.2.7	01-Aug-2002	nathanw	Catch up to -current.
1.17.2.6	20-Jun-2002	nathanw	Catch up to -current.
1.17.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.17.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.17.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.2	24-Aug-2001	nathanw	Catch up with -current.
1.17.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.19.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.19.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.19.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.19.2.1	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.27.4.3	02-Aug-2002	lukem	Pull up revision 1.31 (requested by bouyer in ticket #611): Add Siig 20x 8 ports and and Oxford x952 cards. From Paul Shupak in kern/17790.
1.27.4.2	11-Jun-2002	lukem	Pull up revision 1.29 (requested by seb in ticket #249): Support Titan PCI-800L, a 8 serial puc card, from PR 17140. Reviewed/approved by veego@ and martin@.
1.27.4.1	11-Jun-2002	lukem	Pull up revision 1.28 (requested by lukem in ticket #167): Support NetMos NM9845 Quad UART and 1284 Printer port. All 4 serial ports and parallel port seem to work.
1.27.2.4	29-Aug-2002	gehenna	catch up with -current.
1.27.2.3	15-Jul-2002	gehenna	catch up with -current.
1.27.2.2	20-Jun-2002	gehenna	catch up with -current.
1.27.2.1	30-May-2002	gehenna	Catch up with -current.
1.32.6.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.32.6.4	19-Oct-2004	skrll	Sync with HEAD
1.32.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.32.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.32.6.1	03-Aug-2004	skrll	Sync with HEAD
1.36.2.1	04-Jul-2004	he	branches: 1.36.2.1.2; Pull up revision 1.37 (requested by mycroft in ticket #597): Support for the VS-Com PCI-210H 2s/1p card. Fixes PR#18478.
1.36.2.1.2.2	21-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #10749): sys/dev/pci/pucdata.c: revision 1.49 sys/dev/pci/pcidevs: revision 1.849 sys/dev/pci/pcidevs: revision 1.850 Add NetMos 9855 Quad UART and 1284 Printer port Fix typo in previous (hi bouyer).
1.36.2.1.2.1	06-Aug-2006	tron	Apply patch (requested by bouyer in ticket #10643): Add support for Moxa SmartIO CP104 and CP104-V2/PCI and EXAR XR17D15x.
1.39.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.39.4.1	29-Apr-2005	kent	sync with -current
1.40.4.5	27-Feb-2008	yamt	sync with head.
1.40.4.4	15-Nov-2007	yamt	sync with head.
1.40.4.3	26-Feb-2007	yamt	sync with head.
1.40.4.2	30-Dec-2006	yamt	sync with head.
1.40.4.1	21-Jun-2006	yamt	sync with head.
1.40.2.4	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1584): sys/dev/pci/pucdata.c: revision 1.49 sys/dev/pci/pcidevs: revision 1.849 sys/dev/pci/pcidevs: revision 1.850 Add NetMos 9855 Quad UART and 1284 Printer port Fix typo in previous (hi bouyer).
1.40.2.3	15-Jun-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #1373): sys/dev/pci/pucdata.c: revision 1.45 Add entry for "Moxa Technologies, SmartIO CP104-V2/PCI". From Gert Doering in kern/33566.
1.40.2.2	15-Jun-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #1373): sys/dev/pci/pucdata.c: revision 1.44 Add support for EXAR XR17D15x serial. (I tested only XR17D154.)
1.40.2.1	15-Jun-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #1373): share/man/man4/puc.4: revision 1.19 sys/dev/pci/pucdata.c: revision 1.43 Add support for Moxa CP104 serial card, from Olivier Cherrier
1.41.2.1	01-Feb-2006	yamt	sync with head.
1.42.6.3	03-Sep-2006	yamt	sync with head.
1.42.6.2	26-Jun-2006	yamt	sync with head.
1.42.6.1	13-Mar-2006	yamt	sync with head.
1.42.4.1	22-Apr-2006	simonb	Sync with head.
1.42.2.1	09-Sep-2006	rpaulo	sync with head
1.43.2.1	19-Apr-2006	elad	sync with head.
1.44.4.1	19-Jun-2006	chap	Sync with head.
1.46.4.2	10-Dec-2006	yamt	sync with head.
1.46.4.1	22-Oct-2006	yamt	sync with head
1.46.2.3	01-Feb-2007	ad	Sync with head.
1.46.2.2	12-Jan-2007	ad	Sync with head.
1.46.2.1	18-Nov-2006	ad	Sync with head.
1.49.2.3	09-Mar-2008	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #1097): sys/dev/pci/pucdata.c: revision 1.53 sys/dev/pci/pcidevs: revision 1.930 Add I-O DATA RSA-PCI 2 port serial board. Add an entry for I-O DATA RSA-PCI 2-port serial board. (works only on 16550A compatible mode)
1.49.2.2	04-Nov-2007	xtraeme	Pull up following revision(s) (requested by jnemeth in ticket #968): sys/dev/pci/pucdata.c: revision 1.52 PR/37259 - Daniel Neri -- typo in EXAR XR17C158 description
1.49.2.1	18-Feb-2007	tron	branches: 1.49.2.1.2; Pull up following revision(s) (requested by chs in ticket #424): sys/dev/pci/pucdata.c: revision 1.51 add some SIIG and Moxa cards.
1.49.2.1.2.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.49.2.1.2.1	06-Jan-2008	wrstuden	Catch up to netbsd-4.0 release.
1.51.24.1	13-Nov-2007	bouyer	Sync with HEAD
1.51.20.2	23-Mar-2008	matt	sync with HEAD
1.51.20.1	06-Nov-2007	matt	sync with HEAD
1.51.18.1	04-Nov-2007	jmcneill	Sync with HEAD.
1.52.16.4	17-Jan-2009	mjf	Sync with HEAD.
1.52.16.3	28-Sep-2008	mjf	Sync with HEAD.
1.52.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.52.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.52.12.1	24-Mar-2008	keiichi	sync with head.
1.54.10.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.54.10.1	19-Oct-2008	haad	Sync with HEAD.
1.54.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.54.4.4	11-Aug-2010	yamt	sync with head.
1.54.4.3	11-Mar-2010	yamt	sync with head
1.54.4.2	16-Sep-2009	yamt	sync with head
1.54.4.1	04-May-2009	yamt	sync with head.
1.55.10.1	21-Apr-2010	matt	sync to netbsd-5
1.55.4.2	23-Jan-2010	bouyer	Pull up following revision(s) (requested by mbalmer in ticket #1255): sys/dev/pci/pucdata.c: revision 1.59 Add the IBM 481033H SCC
1.55.4.1	16-Oct-2009	snj	Pull up following revision(s) (requested by tsutsui in ticket #1070): sys/dev/pci/pucdata.c: revision 1.57 Add support for Oxford Semiconductor OXmPCI952 2 port UARTs. Also update URL of Oxford UART chips.
1.55.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.59.4.4	31-May-2011	rmind	sync with head
1.59.4.3	21-Apr-2011	rmind	sync with head
1.59.4.2	05-Mar-2011	rmind	sync with head
1.59.4.1	03-Jul-2010	rmind	sync with head
1.59.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.69.4.1	05-Mar-2011	bouyer	Sync with HEAD
1.69.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.76.8.1	31-Oct-2012	riz	sys/dev/pci/pucdata.c 1.77-1.82 sys/dev/pci/pcidevs 1.1125, 1.1129 via patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen share/man/man4/puc.4 1.36-1.38 Add entries for puc(4) devices into pcidevs. Add some Intel devices from document (Intel 7 Series / C216 Chipset Family Platform Controller Hub (PCH) Family Specification Update) into pcidevs. Modify some devices in pcidevs to clarify. Add Intel chipset internal serial over lan devices. Closes PR/45567. Use PCI_VENDOR_* and PCI_PRODUCT_*. Fixes PR#46303. Add EXSYS EX-41098-2 UARTs support. The clock frequency is different from other card(s). Fixes PR#46362 reported by Wolfgang Stukenbrock. Add Lava Computers SSERIAL-PCI single port serial PCI card. Update the manual.
1.76.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.76.2.2	23-Jan-2013	yamt	sync with head
1.76.2.1	30-Oct-2012	yamt	sync with head
1.80.2.5	03-Dec-2017	jdolecek	update from HEAD
1.80.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.80.2.3	23-Jun-2013	tls	resync from head
1.80.2.2	25-Feb-2013	tls	resync with head
1.80.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.84.4.1	23-Jul-2013	riastradh	sync with HEAD
1.84.2.2	18-May-2014	rmind	sync with head
1.84.2.1	28-Aug-2013	rmind	sync with head
1.93.6.5	05-Feb-2017	skrll	Sync with HEAD
1.93.6.4	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.93.6.3	22-Sep-2015	skrll	Sync with HEAD
1.93.6.2	06-Jun-2015	skrll	Sync with HEAD
1.93.6.1	06-Apr-2015	skrll	Sync with HEAD
1.93.4.4	14-Nov-2019	martin	Pull up following revision(s) (requested by hauke in ticket #1712): sys/dev/pci/pucdata.c: revision 1.105 The 16C1054 and 16C1058 serial multi-port controllers need a clock multiplier of 8, just like the 16C1050 controller. Verified with an ExSys EX-41388. ryo@ checked back with the hardware his original commit was based on, and confirmed the change. XXX Pull-up to netbsd-{7,8,9}
1.93.4.3	03-Jan-2018	snj	Pull up following revision(s) (requested by maya in ticket #1547): sys/dev/pci/pucdata.c: revision 1.99-1.100 sys/dev/pci/pcidevs: revision 1.1278 Add Intel 200 series chipset devices from "Table 2-2. PCH-H Device and Revision ID Table, Intel 200 Series Chipset Family Platform Controller Hub(PCI) Datasheet Volume 1 of 2 (335192-001)". -- Add Intel 200 series devices. -- PR/52868: Petar Bogdanovic: Add support for Manhattan 158220 card
1.93.4.2	08-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1292): sys/dev/pci/pucdata.c: revisions 1.95-1.98 share/man/man4/puc.4: revisions 1.39-40 Add three devices to puc(4): - Add support for SystemBase SB16C1050 PCI serial card. PR#49819 reported by Roberto E. Vargas Caballero. - Add another Intel Q45 KT. - Add Intel 100 Series Chipset KT.
1.93.4.1	30-Apr-2015	snj	branches: 1.93.4.1.4; Pull up following revision(s) (requested by msaitoh in ticket #725): sys/arch/x86/pci/ichlpcib.c: revisions 1.47, 1.48 sys/dev/pci/ichsmb.c: revisions 1.39, 1.40, 1.41 sys/dev/pci/pucdata.c: revision 1.94 Add 9 Series support. -- Add Intel C61x and X99 devices. -- attach Mobile 5th Gen. Core SMBus
1.93.4.1.4.1	18-Jan-2017	skrll	Sync with netbsd-5
1.98.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.99.8.6	03-Dec-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1712: sys/dev/pci/pucdata.c 1.103, 1.106-1.112 share/man/man4/puc.4 1.42-1.43 - Add Amazon.com EC2 virtual 16650-compatible PCI serial device. - Add NetMos NM9900 Quad and Octal serial card. - Add ASIX AX99100 PCIe 4port serial card. - Add Oxford Semiconductor Exsys EX-41098 PCI serial card.
1.99.8.5	14-Nov-2019	martin	Pull up following revision(s) (requested by hauke in ticket #1440): sys/dev/pci/pucdata.c: revision 1.105 The 16C1054 and 16C1058 serial multi-port controllers need a clock multiplier of 8, just like the 16C1050 controller. Verified with an ExSys EX-41388. ryo@ checked back with the hardware his original commit was based on, and confirmed the change. XXX Pull-up to netbsd-{7,8,9}
1.99.8.4	05-May-2019	martin	Pull up following revision(s) (requested by jdolecek in ticket #1255): sys/dev/pci/pucdata.c: revision 1.104 add NetMos NM9912 entries PR kern/54149 by Mouse
1.99.8.3	07-Jun-2018	martin	Pull up following revision(s) (requested by jnemeth in ticket #846): sys/dev/pci/pcidevs: revision 1.1334 sys/dev/pci/pucdata.c: revision 1.102 add Oxford Semiconductor OXPCIe954 PCIe 4-port serial card
1.99.8.2	16-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #763): sys/dev/pci/ichsmb.c: revision 1.57 sys/dev/pci/pucdata.c: revision 1.101 Add 300 series chipset support.
1.99.8.1	13-Jan-2018	snj	Pull up following revision(s) (requested by maya in ticket #472): sys/dev/pci/pucdata.c: revision 1.100 PR/52868: Petar Bogdanovic: Add support for Manhattan 158220 card
1.100.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.100.2.2	25-Jun-2018	pgoyette	Sync with HEAD
1.100.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.102.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.102.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.102.2.1	10-Jun-2019	christos	Sync with HEAD
1.104.2.4	24-Aug-2024	martin	Pull up following revision(s) (requested by jakllsch in ticket #1880): sys/dev/pci/pucdata.c: revision 1.116 Add Brainboxes puc(4) devices From Cameron Williams in PR 58358.
1.104.2.3	30-Jan-2023	martin	Pull up following revision(s) (requested by thorpej in ticket #1580): share/man/man4/puc.4: revision 1.44 sys/dev/pci/pucdata.c: revision 1.114 Add and entry for the EXAR XR17V354 4-port UART. From Hakan Engvall. PR kern/57202
1.104.2.2	03-Dec-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1382: sys/dev/pci/pucdata.c 1.106-1.112 share/man/man4/puc.4 1.42-1.43 - Add NetMos NM9900 Quad and Octal serial card. - Add ASIX AX99100 PCIe 4port serial card. - Add Oxford Semiconductor Exsys EX-41098 PCI serial card.
1.104.2.1	14-Nov-2019	martin	Pull up following revision(s) (requested by hauke in ticket #426): sys/dev/pci/pucdata.c: revision 1.105 The 16C1054 and 16C1058 serial multi-port controllers need a clock multiplier of 8, just like the 16C1050 controller. Verified with an ExSys EX-41388. ryo@ checked back with the hardware his original commit was based on, and confirmed the change. XXX Pull-up to netbsd-{7,8,9}
1.108.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.109.4.1	01-Aug-2021	thorpej	Sync with HEAD.
1.113.4.3	06-Dec-2024	snj	Pull up following revision(s) (requested by msaitoh in ticket #1025): share/man/man4/puc.4: revision 1.46 sys/dev/pci/pucdata.c: revision 1.117 Add many Brainboxes devices. Reported in PR/kern 58824 by Cameron Williams.
1.113.4.2	24-Aug-2024	martin	Pull up following revision(s) (requested by jakllsch in ticket #806): sys/dev/pci/pucdata.c: revision 1.116 Add Brainboxes puc(4) devices From Cameron Williams in PR 58358.
1.113.4.1	30-Jan-2023	martin	Pull up following revision(s) (requested by thorpej in ticket #61): share/man/man4/puc.4: revision 1.44 sys/dev/pci/pucdata.c: revision 1.114 Add and entry for the EXAR XR17V354 4-port UART. From Hakan Engvall. PR kern/57202
1.116.2.1	02-Aug-2025	perseant	Sync with HEAD
1.12	30-Nov-2018	jmcneill	Add support for polling com devices when no interrupt is available.
1.11	18-Mar-2014	riastradh	branches: 1.11.28; 1.11.30; Merge riastradh-drm2 to HEAD.
1.10	17-Jul-2013	soren	Extend support for Nanjing QinHeng Electronics puc(4) devices.
1.9	01-Aug-2010	msaitoh	branches: 1.9.8; 1.9.18; 1.9.22; 1.9.30; Add the following cards' support from FreeBSD: - Comtrol RocketPort 550/4, 550/8 and 550/16 series - Lava Computers Quatro-PCI - some NetMos NM9865 cards - VScom PCI-200Li
1.8	11-Dec-2005	christos	branches: 1.8.74; 1.8.96; 1.8.98; merge ktrace-lwp.
1.7	04-Feb-2005	perry	de-__P
1.6	03-Feb-2004	fredb	branches: 1.6.8; 1.6.10; Enable the 8X clock on the SIIG Cyberserial serial and combination PCI cards at attachment time, in order to support bit rates greater than 115K, as discussed on tech-kern.
1.5	25-Jan-2004	jdolecek	pass the DMA tags to child devices, too
1.4	03-Jan-2001	bouyer	branches: 1.4.24; Add a 'flag' member to struct puc_attach_args and struct port. For PUC_PORT_TYPE_COM, use it to store the clock frequency (with 8 lower bits to 0, used for real flags if needed). Update all descriptions to set flags to 0 for LPT or COM_FREQ for COM. Add support for the VScom PCI-800H 8 port serial adapter (which uses a 14.7456 Mhz crystal instead of the standart 1.8432Mhz :) XXX now that we can pass other frequency than COM_FREQ, the VScom PCI-800 entry could probably be updated to DTRT - does anyone have one ?
1.3	25-Jul-2000	jeffs	Add code to allow the PCI com serial ports to be used as the system console. This is not enabled by default, and is turned on with options PUCCN. Done by castor@netbsd.org.
1.2	06-Feb-1999	cgd	branches: 1.2.8; 1.2.18; add code to cope with multiple ports per BAR, as are found, for instance, on the 8 port card Simon Gerraty has. In general, cards which have this lots of ports also have a separate interrupt status register, but this change is just to talk to the various ports independently. It works, but it's not optimal. (XXX still need a good name for the card in the comments, and to update the manual page.)
1.1	26-Jun-1998	cgd	add a driver for "PCI 'universal' communications" cards, that is, PCI cards which contain 'standard' com- and lpt-type ports. Some of these present as PCI simple-communications/serial or simple-communications/parallel devices, but many do not. (Additionally, there is no document that I can find that describes the "specific well-konwn register-level" description of how the 'standard' devices' config space headers shold work.) Eventually, some of the devices driven by this code should become simple pci attachments for the 'lpt' and 'com' drivers, but that requires solid documentation.
1.2.18.1	22-Mar-2001	he	Pull up revisions 1.3-1.4 (requested by sommerfeld): Add support for NetMos NM9835, Titan PCI-800H, Lava 8-port, Actiontec 56K PCI Master. This also adds a ``flags'' member to two structs in this driver.
1.2.8.2	05-Jan-2001	bouyer	Sync with HEAD
1.2.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.24.4	04-Feb-2005	skrll	Sync with HEAD.
1.4.24.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.24.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.24.1	03-Aug-2004	skrll	Sync with HEAD
1.6.10.1	12-Feb-2005	yamt	sync with head.
1.6.8.1	29-Apr-2005	kent	sync with -current
1.8.98.1	05-Mar-2011	rmind	sync with head
1.8.96.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.8.74.1	11-Aug-2010	yamt	sync with head.
1.9.30.1	23-Jul-2013	riastradh	sync with HEAD
1.9.22.1	28-Aug-2013	rmind	sync with head
1.9.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.8.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.11.30.1	10-Jun-2019	christos	Sync with HEAD
1.11.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.5	06-Sep-2025	riastradh	pvscsi(4): Zero rings before using them. bus_dmamem_alloc(9) doesn't guarantee zeroing, as far as I can tell, and who knows what might happen if the ring header and contents have anything nonzero initially. Insert an initial preread/prewrite sync between zeroing and first use, and, out of paranoia, a final postread/postwrite sync between last use and unload/free. Prompted by (but not really related to): PR kern/59618: occasional virtio block device lock ups/hangs
1.4	06-Sep-2025	riastradh	pvscsi(4): Use bus_dmamap_sync, not membar_, for DMA. membar_ may be a noop if we're booting on a single _virtual_ CPU, but the barriers are still needed in case the host behind pvscsi(4) is running on another _physical_ CPU. Prompted by (and related to but not the same issue as): PR kern/59618: occasional virtio block device lock ups/hangs
1.3	06-Sep-2025	riastradh	pvscsi(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs (pvscsi(4) is not virtio(4) but the same fundamental issue arises.)
1.2	05-Aug-2025	skrll	branches: 1.2.2; 1.2.4; Add the barrier that a comment describes.
1.1	04-Aug-2025	skrll	Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots.
1.2.4.2	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #1148): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.2.4.1	05-Aug-2025	martin	file pvscsi.c was added on branch netbsd-10 on 2025-08-08 14:21:39 +0000
1.2.2.4	19-Oct-2025	martin	Pull up following revision(s) (requested by riastradh in ticket #64): sys/dev/pci/pvscsi.c: revision 1.4 sys/dev/pci/pvscsi.c: revision 1.5 pvscsi(4): Use bus_dmamap_sync, not membar_, for DMA. membar_ may be a noop if we're booting on a single _virtual_ CPU, but the barriers are still needed in case the host behind pvscsi(4) is running on another _physical_ CPU. Prompted by (and related to but not the same issue as): PR kern/59618: occasional virtio block device lock ups/hangs pvscsi(4): Zero rings before using them. bus_dmamem_alloc(9) doesn't guarantee zeroing, as far as I can tell, and who knows what might happen if the ring header and contents have anything nonzero initially. Insert an initial preread/prewrite sync between zeroing and first use, and, out of paranoia, a final postread/postwrite sync between last use and unload/free. Prompted by (but not really related to): PR kern/59618: occasional virtio block device lock ups/hangs
1.2.2.3	19-Oct-2025	martin	Pull up following revision(s) (requested by riastradh in ticket #63): sys/dev/pci/pvscsi.c: revision 1.3 pvscsi(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs (pvscsi(4) is not virtio(4) but the same fundamental issue arises.)
1.2.2.2	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #6): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.2.2.1	05-Aug-2025	martin	file pvscsi.c was added on branch netbsd-11 on 2025-08-08 14:19:54 +0000
1.1	04-Aug-2025	skrll	branches: 1.1.2; 1.1.4; Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots.
1.1.4.2	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #1148): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.1.4.1	04-Aug-2025	martin	file pvscsi.h was added on branch netbsd-10 on 2025-08-08 14:21:39 +0000
1.1.2.2	08-Aug-2025	martin	Pull up following revision(s) (requested by skrll in ticket #6): sys/arch/evbarm/conf/GENERIC64: revision 1.227 sys/dev/pci/files.pci: revision 1.451 sys/dev/pci/pvscsi.c: revision 1.1 sys/dev/pci/pvscsi.c: revision 1.2 sys/dev/pci/pvscsi.h: revision 1.1 sys/arch/amd64/conf/GENERIC: revision 1.617 Add a port of the FreeBSD driver for the VMware PV SCSI controller found in VMware's ESXi product. Tested on ESXi/Arm such that an installation from ISO succeeds and boots. Add the barrier that a comment describes.
1.1.2.1	04-Aug-2025	martin	file pvscsi.h was added on branch netbsd-11 on 2025-08-08 14:19:54 +0000
1.10	20-Jan-2017	maya	Check pmf_device_register return value. NFC Appeases static analyzers.
1.9	14-Jul-2016	msaitoh	branches: 1.9.2; 1.9.4; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.8	23-Apr-2015	pgoyette	Update module dependencies for all the existing modules that depend on sysmon components.
1.7	30-Aug-2011	mbalmer	branches: 1.7.12; 1.7.30; Unconditionally include <sys/modules.h>.
1.6	29-Aug-2011	mbalmer	Register as built-in module when not compiled as a module. Reminded by jmcneill.
1.5	29-Aug-2011	jmcneill	have pci register itself as a module, and make PCI driver modules depend on it
1.4	26-Aug-2011	mbalmer	Add a pmf resume function to put the hardware in a defined state after resume. Problem found by jmcneill@.
1.3	26-Aug-2011	mbalmer	Remove one of the copyrights, it is the same person. Remove variable name in prototype.
1.2	20-Aug-2011	mbalmer	Remove a commented out include and clarify a comment why the timeout period is ignored.
1.1	11-Aug-2011	mbalmer	pwdog(4) is a device driver for QUANCOM Electronic PWDOG1 PCI attached watchdog timers.
1.7.30.3	05-Feb-2017	skrll	Sync with HEAD
1.7.30.2	05-Oct-2016	skrll	Sync with HEAD
1.7.30.1	06-Jun-2015	skrll	Sync with HEAD
1.7.12.1	03-Dec-2017	jdolecek	update from HEAD
1.9.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.9.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.45	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.44	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.43	24-Apr-2021	thorpej	branches: 1.43.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.42	21-May-2020	macallan	branches: 1.42.4; prune unneeded includes
1.41	21-Jun-2017	macallan	allow font loading etc.
1.40	20-Jan-2017	maya	Check pmf_device_register return value. NFC Appeases static analyzers.
1.39	16-Sep-2015	macallan	branches: 1.39.2; 1.39.4; prop_dictionary_get_bool() doesn't return FALSE if the property doesn't exist, so initialize things first. No more console stealing on sparc64.
1.38	09-Oct-2013	macallan	branches: 1.38.6; support WSDISPLAYIO_GET_FBINFO
1.37	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.36	04-Oct-2012	macallan	branches: 1.36.2; don't call allocattr() for the default attribute
1.35	23-May-2012	macallan	branches: 1.35.2; first step to fix PR 46461 - don't jump through a NULL pointer when we're not the console but the firmware set up the graphics hardware anyway
1.34	19-Apr-2012	macallan	glyphcache_init() works much better with its parameters in the right order
1.33	19-Apr-2012	macallan	use rasops_get_cmap()
1.32	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.31	08-Mar-2012	macallan	when doing host blits, let the fifo drain every now and then. Apparently even a relatively slow CPU can overload a relatively fast Rage 128 in some circumstances. This cures weird artifacts seen when scrolling lots and lots of text very fast. Why this only showed up with glyph cacheing which should reduce the frequency of host blits isn't quite clear though.
1.30	16-Feb-2012	macallan	remove debug spam
1.29	16-Feb-2012	macallan	cache glyphs in video memory when using anti-aliased fonts on r128fb and radeonfb
1.28	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.27	11-Jan-2012	macallan	use rasops_init(0, 0)
1.26	04-Jan-2012	macallan	use WSFONT_GLYPH()
1.25	04-Jan-2012	macallan	generate a better r3g3b2 colour map which actually produces white for 0xff
1.24	04-Jan-2012	macallan	split putchar into separate methods for bitmap and alpha fonts, use FONT_IS_ALPHA()
1.23	28-Dec-2011	macallan	support anti-aliased fonts in 8 bit r3g3b2 colour
1.22	29-Jun-2011	macallan	branches: 1.22.2; 1.22.6; support WSDISPLAYIO_GET_EDID
1.21	15-Feb-2011	macallan	don't muck with the LVDS_BLON bit
1.20	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.19	06-Jan-2011	macallan	branches: 1.19.2; 1.19.4; clear the screen before redrawing on reentering text mode
1.18	25-Dec-2010	macallan	there is no R128_LITTLE_ENDIAN_EN flag, just lack of R128_HOST_BIG_ENDIAN_EN Someone(tm) needs to test this on little endian hw
1.17	16-Dec-2010	cegger	ioctl: KNF switch-case, remove useless brackets
1.16	15-Nov-2010	macallan	explicitly set and clear the backlight on bit, just in case someone messes with it while we're not looking
1.15	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.14	07-Oct-2010	macallan	use WSDISPLAYIO_PARAM_BACKLIGHT as a switch and WSDISPLAYIO_PARAM_BRIGHTNESS for backlight level instead of following radeonfb TODO: adapt radeonfb and powerd scripts
1.13	30-Sep-2010	macallan	use the drawing engine to draw characters so we don't have to sync after each command and don't need to map the framebuffer anymore
1.12	14-Sep-2010	macallan	support backlight control via wsconsctl
1.11	09-Sep-2010	macallan	add backlight control via PMF TODO: support wsconsctl
1.10	01-Oct-2009	jmmv	branches: 1.10.2; 1.10.4; Recognize PCI_PRODUCT_ATI_RAGEGLPCI. Works on a PowerMac G3 B&W running NetBSD/macppc.
1.9	20-Aug-2009	macallan	use vcons_replay_msgbuf() when appropriate
1.8	06-May-2009	elad	Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.7	03-Jan-2009	yamt	branches: 1.7.2; remove extra semicolons.
1.6	25-Aug-2008	jmcneill	branches: 1.6.2; Match on Rage Mobility M3 (AGP)
1.5	30-May-2008	macallan	branches: 1.5.4; remove test for VGA subclass, also match PCI_PRODUCT_ATI_RAGE3AGP4XT this makes r128fb work on 2nd generation iMacs From Brett Slager
1.4	08-May-2008	macallan	nuke clause 3
1.3	27-Feb-2008	macallan	branches: 1.3.2; 1.3.4; 1.3.6; get rid of the last #ifdef macppc and use PCI_MAGIC_IO_RANGE to select at which offset we mmap PCI IO space
1.2	23-Nov-2007	macallan	branches: 1.2.10; 1.2.12; 1.2.16; make this compile on sparc64
1.1	07-Nov-2007	macallan	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; add preliminary driver for ATI Rage 128 graphics controllers So far it supports some acceleration ( copies and rectangle fills ) but no video mode programming at all. Known problems: - tested only on macppc - matches only one r128 chip out of ca. 30 - character drawing is unaccelerated - no attempt is made to restore the console when X exits
1.1.8.3	08-Dec-2007	mjf	Sync with HEAD.
1.1.8.2	19-Nov-2007	mjf	Sync with HEAD.
1.1.8.1	07-Nov-2007	mjf	file r128fb.c was added on branch mjf-devfs on 2007-11-19 00:48:18 +0000
1.1.6.4	17-Mar-2008	yamt	sync with head.
1.1.6.3	07-Dec-2007	yamt	sync with head
1.1.6.2	15-Nov-2007	yamt	sync with head.
1.1.6.1	07-Nov-2007	yamt	file r128fb.c was added on branch yamt-lazymbuf on 2007-11-15 11:44:27 +0000
1.1.4.2	13-Nov-2007	bouyer	Sync with HEAD
1.1.4.1	07-Nov-2007	bouyer	file r128fb.c was added on branch bouyer-xenamd64 on 2007-11-13 16:01:33 +0000
1.1.2.3	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.1.2.2	11-Nov-2007	joerg	Sync with HEAD.
1.1.2.1	07-Nov-2007	joerg	file r128fb.c was added on branch jmcneill-pm on 2007-11-11 16:47:43 +0000
1.2.16.4	17-Jan-2009	mjf	Sync with HEAD.
1.2.16.3	28-Sep-2008	mjf	Sync with HEAD.
1.2.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.2.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.2.12.1	24-Mar-2008	keiichi	sync with head.
1.2.10.3	23-Mar-2008	matt	sync with HEAD
1.2.10.2	09-Jan-2008	matt	sync with HEAD
1.2.10.1	23-Nov-2007	matt	file r128fb.c was added on branch matt-armv6 on 2008-01-09 01:54:01 +0000
1.3.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.3.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.3.4.6	09-Oct-2010	yamt	sync with head
1.3.4.5	11-Mar-2010	yamt	sync with head
1.3.4.4	16-Sep-2009	yamt	sync with head
1.3.4.3	16-May-2009	yamt	sync with head
1.3.4.2	04-May-2009	yamt	sync with head.
1.3.4.1	16-May-2008	yamt	sync with head.
1.3.2.2	04-Jun-2008	yamt	sync with head
1.3.2.1	18-May-2008	yamt	sync with head.
1.5.4.1	19-Oct-2008	haad	Sync with HEAD.
1.6.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.7.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.10.4.1	05-Mar-2011	rmind	sync with head
1.10.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.19.4.2	17-Feb-2011	bouyer	Sync with HEAD
1.19.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.19.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.22.6.5	02-Jun-2012	mrg	sync to latest -current.
1.22.6.4	29-Apr-2012	mrg	sync to latest -current.
1.22.6.3	05-Apr-2012	mrg	sync to latest -current.
1.22.6.2	11-Mar-2012	mrg	sync to latest -current
1.22.6.1	18-Feb-2012	mrg	merge to -current.
1.22.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.22.2.3	30-Oct-2012	yamt	sync with head
1.22.2.2	23-May-2012	yamt	sync with head.
1.22.2.1	17-Apr-2012	yamt	sync with head
1.35.2.3	03-Dec-2017	jdolecek	update from HEAD
1.35.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.35.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.36.2.2	18-May-2014	rmind	sync with head
1.36.2.1	28-Aug-2013	rmind	sync with head
1.38.6.3	28-Aug-2017	skrll	Sync with HEAD
1.38.6.2	05-Feb-2017	skrll	Sync with HEAD
1.38.6.1	22-Sep-2015	skrll	Sync with HEAD
1.39.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.39.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.42.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.43.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	06-Jan-2012	macallan	Forgot to commit this along with r128fb.c. Again.
1.4	28-Dec-2011	macallan	add a bunch of register definitions
1.3	30-Sep-2010	macallan	branches: 1.3.8; 1.3.12; add a couple more register definitions, mostly from xf86-video-r128
1.2	09-Sep-2010	macallan	add backlight control via PMF TODO: support wsconsctl
1.1	07-Nov-2007	macallan	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; 1.1.18; 1.1.28; 1.1.48; 1.1.50; add preliminary driver for ATI Rage 128 graphics controllers So far it supports some acceleration ( copies and rectangle fills ) but no video mode programming at all. Known problems: - tested only on macppc - matches only one r128 chip out of ca. 30 - character drawing is unaccelerated - no attempt is made to restore the console when X exits
1.1.50.1	05-Mar-2011	rmind	sync with head
1.1.48.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.1.28.1	09-Oct-2010	yamt	sync with head
1.1.18.2	09-Jan-2008	matt	sync with HEAD
1.1.18.1	07-Nov-2007	matt	file r128fbreg.h was added on branch matt-armv6 on 2008-01-09 01:54:02 +0000
1.1.8.2	19-Nov-2007	mjf	Sync with HEAD.
1.1.8.1	07-Nov-2007	mjf	file r128fbreg.h was added on branch mjf-devfs on 2007-11-19 00:48:19 +0000
1.1.6.2	15-Nov-2007	yamt	sync with head.
1.1.6.1	07-Nov-2007	yamt	file r128fbreg.h was added on branch yamt-lazymbuf on 2007-11-15 11:44:27 +0000
1.1.4.2	13-Nov-2007	bouyer	Sync with HEAD
1.1.4.1	07-Nov-2007	bouyer	file r128fbreg.h was added on branch bouyer-xenamd64 on 2007-11-13 16:01:33 +0000
1.1.2.2	11-Nov-2007	joerg	Sync with HEAD.
1.1.2.1	07-Nov-2007	joerg	file r128fbreg.h was added on branch jmcneill-pm on 2007-11-11 16:47:44 +0000
1.3.12.1	18-Feb-2012	mrg	merge to -current.
1.3.8.1	17-Apr-2012	yamt	sync with head
1.117	08-Aug-2021	tnn	radeonfb: use BUS_SPACE_MAP_PREFETCHABLE for the framebuffer mapping Makes the driver work on cia(4) based alpha machines, which can only do dense access to linear mappings with prefetch. ok macallan@
1.116	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.115	21-Jun-2021	christos	branches: 1.115.2; fix proplib deprecation
1.114	24-Apr-2021	thorpej	branches: 1.114.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.113	14-Mar-2021	rin	branches: 1.113.2; Initialize dp->rd_cmap_{red,green,blue} in radeonfb_init_palette() for 8bpp, so that color map can be obtained by WSDISPLAYIO_GETCMAP ioctl(2). Now, mlterm-fb from pkgsrc/x11/mlterm works just fine on radeonfb(4).
1.112	30-Oct-2020	macallan	branches: 1.112.2; support screen blanking of r5xx somewhat crude but works
1.111	11-Oct-2020	jdc	Handle EDID:N entries, where N is the port number, so that we can set up the display on port1 when port0 has no EDID information. Try harder not to return NULL from radeonfb_modelookup() because we use the result without checking later. While here, adjust RADEONFB_DEBUG output: don't print RADEON_PIXCLKS_CNTL (fixes a crash on sparc64 because it's not aligned) don't ignore the bottom 200 lines of the display (for no apparent reason))
1.110	28-Sep-2020	macallan	first step towards r5xx support: - no CRTC programming yet, we just use whatever mode the (open)firmware hands us - palette control, fb location etc. all work - hardware acceleration for character drawing, scrolling etc. works - hardware cursor works ... tested & still works on various radeons in various macppc machines
1.109	05-Jul-2020	martin	Adapt to proplib api changes
1.108	11-Jun-2020	macallan	only read the backlight level register on mobility chips
1.107	11-Jun-2020	macallan	reduce stack usage in radeonfb_pickres() and radeonfb_set_cursor()
1.106	06-Feb-2020	macallan	sent PMF events when (un)blanking so things like lmu can turn their lights on or off as well
1.105	15-Aug-2019	rin	branches: 1.105.2; Support bitmap fonts with stride 4 (i.e., width 17 to 32).
1.104	27-Mar-2019	macallan	branches: 1.104.4; fix the fix(es) - don't try to access a BIOS we didn't find. Now this works on sparc64 again.
1.103	22-Mar-2019	martin	Even if you try very hard, some cards just do not have a BIOS.
1.102	21-Mar-2019	macallan	don't malloc(0) spotted by mrg@
1.101	20-Mar-2019	macallan	add code to read disabled ROMs, adapted from xf86-video-radeon With this radeonfb does The Right Thing(tm) on my 2xDVI mac radeon with decidedly non-standard output wiring. ( apparently at least some mac radeons have a hidden x86 BIOS with valid connector tables )
1.100	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.99	28-Jun-2018	macallan	branches: 1.99.2; restrict a magic number 'fix' inherited from xf86-video-radeon to actual rv100 chips instead of applying it to anything older than r300. Now DVI output works properly on 1st generation Mac Minis. tested by christos
1.98	16-Jun-2018	macallan	explicitly turn off tiling translation - now software rendered characters look right again on r3xx hardware
1.97	15-Jun-2018	macallan	cosmetics, no functional change
1.96	14-Jun-2018	macallan	make this work on rv350 chips found in last generation *Books Needs testing on non-Apple hardware.
1.95	06-Jun-2018	maya	Remove duplicate ;
1.94	24-Jan-2018	riastradh	branches: 1.94.2; Fix integer overflows noted by Silvio Cesare of InfoSect. Someone^TM should name these idioms so we can eliminate this class of copypasta bug.
1.93	11-Oct-2017	macallan	add missing break;s
1.92	25-Aug-2017	macallan	Radeon PLLs can't have odd post dividers when used with digital outputs - add a flag to track this in relevant functions. Now DVI output works properly on my Mini 2 Needs testing on Mini 1 and other hardware...
1.91	23-Aug-2017	macallan	Mac Mini WIP: - plug the external TMDS into the right CRTC when in use - use the right output types for Mini configs now VGA works properly ( and not just by accident ) on my Mini 2. DVI is still garbled but at least produces output.
1.90	11-Aug-2017	macallan	set display output parameters according to properties Now VGA output works as intended on my Mini 2. DVI is still garbled ( but visible ) and Mini 1 is untested.
1.89	02-Jun-2017	macallan	enable font loading / screen resizing
1.88	05-Nov-2014	macallan	branches: 1.88.2; - unmap registers in WSDISPLAYIO_MODE_MAPPED only if a drm attached - don't try to touch backlight registers when they're not mapped
1.87	21-Oct-2014	macallan	remove outdated comments
1.86	21-Oct-2014	macallan	get rid of RADEONFB_MMAP_BARS which hasn't been useful for years
1.85	16-Oct-2014	riastradh	Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.84	22-Jul-2014	riastradh	branches: 1.84.2; Do not printf(foobar()). Bad bad bad bad bad!
1.83	20-Feb-2014	joerg	branches: 1.83.2; Mark radeonfb_unclip as unused.
1.82	22-Jan-2014	macallan	implement WSDISPLAYIO_{GET\|PUT}CMAP ioctl()s while there: - unmap registers and vram only in WSDISPLAYIO_MODE_MAPPED ( so DRM can map them ) but not in WSDISPLAYIO_MODE_DUMBFB ( so wsfb can do things that require hardware access, like palette and cursor ioctl()s ) - reject ioctl()s that need hardware access if registers are unmapped ( so we don't crash instead ) - use the same bit order in cursor sprite images as the other drivers do with all this, Xorg with wsfb works properly on top of radeonfb tested on i386 and macppc ( fingers crossed for alpha ;) )
1.81	14-Jan-2014	macallan	cap mappable framebuffer size at 64MB before dividing it up for each output now we can actually access the 2nd fb on boards with more than 64MB VRAM
1.80	14-Jan-2014	macallan	- move cursor area to the end of each display's chunk of fb RAM - use memory between display and cursor as glyph cache instead of blindly trying to use 8MB should fix PR 48492
1.79	09-Oct-2013	macallan	support WSDISPLAYIO_GET_FBINFO
1.78	15-Sep-2013	martin	Remove an unused variable
1.77	15-Sep-2013	martin	Comment out (but leave for documentation purposes) a variable.
1.76	28-Mar-2013	macallan	branches: 1.76.4; clear screen when leaving X so we don't leave garbage in the borders
1.75	25-Mar-2013	macallan	change hsync start to match xf86-video-radeon so CRT modes set up by radeonfb and Xorg no longer differ by a few pixels horizontally
1.74	07-Jan-2013	macallan	add PCI_PRODUCT_ATI_RADEON_RV370_5B63 - confirmed working by Dieter Roelants
1.73	01-Jan-2013	macallan	add a compile option to override the fallback to software drawing of bitmap fonts on R3xx ( options RADEONFB_ALWAYS_ACCEL_PUTCHAR ) since this works just fine on my RV350. Next step: don't map VRAM if we don't need it.
1.72	31-Dec-2012	macallan	fix some typos & comments
1.71	31-Dec-2012	macallan	R4xx support is untested at best, so don't pretend to support them.
1.70	31-Dec-2012	macallan	clean up some more confusion between (ws)displays and CRTCs. Now palette initialization no longer stomps over the port table, which gives us a fighting chance to intentionally enable the right outputs. How on earth did this ever work?
1.69	30-Dec-2012	macallan	Make R3xx actually work by wiring outputs, DACs and CRTCs sanely, add a bunch of R3xx-specific workarounds. Tested in xvr-100, iBook G4 and an RV350.
1.68	20-Dec-2012	macallan	don't access registers when they're not mapped now X works on i386 with radeonfb
1.67	20-Dec-2012	macallan	make hardware accelerated drawing work on little endian hardware tested on i386 and macppc
1.66	20-Dec-2012	macallan	first step to make this work (again) on x86 - make sure that if we use the 2nd CRTC it's actually enabled
1.65	04-Dec-2012	macallan	make sure to use the same stride as the drawing engine when setting up video modes. Now things work with horizontal resolutions that aren't multiples of 64 ( like 1440x900 )
1.64	04-Oct-2012	macallan	support underlined characters
1.63	27-Jun-2012	macallan	branches: 1.63.2; don't spam the EDID block to the console unless options RADEONFB_DEBUG
1.62	22-Apr-2012	uebayasi	Consistent declaration.
1.61	12-Apr-2012	macallan	- set colour channel info only in 32bit colour - fix comment ( don't claim to run in some byte-swapped mode when we don't )
1.60	26-Mar-2012	macallan	correctly draw alpha fonts with an uneven number of pixels per characters radeon is weird how it handles leftovers
1.59	15-Mar-2012	macallan	adjust the backlight control interface to match the other drivers ( genfb, r128fb, voyagerfb etc. ) so the lid_switch script does the right thing noticed by riz, I thought I fixed that a long time ago
1.58	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.57	28-Feb-2012	macallan	let the fifo drain periodically - we need this even in 8 bit on some hardware, like rv100 paired with a semi-fast CPU
1.56	28-Feb-2012	macallan	add methods to map/unmap video memory and registers when entering/leaving X so DRI has a fighting chance of working
1.55	28-Feb-2012	macallan	support anti-aliased fonts in 8 bit as well
1.54	16-Feb-2012	macallan	cache glyphs in video memory when using anti-aliased fonts on r128fb and radeonfb
1.53	30-Jan-2012	drochner	branches: 1.53.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.52	11-Jan-2012	macallan	use rasops_init(0, 0)
1.51	05-Jan-2012	macallan	turns out we can feed some radeons faster than they can process image uploads, so make periodic stops to let the fifo drain a bit now this works on my Blade 2500 with a 32MB XVR-100
1.50	04-Jan-2012	macallan	use WSFONT_GLYPH()
1.49	03-Jan-2012	macallan	support anti-aliased fonts in 32bit colour enable with options RADEONFB_DEPTH_32
1.48	29-Dec-2011	macallan	rework putchar(): - get rid of engine stalls when using the blitter to draw characters - add a wrapper for non-accelerated putchar() so we only wait for the engine when we actually want to scribble into video memory - rework accelerated putchar(), should work on R3xx now but needs testing
1.47	29-Dec-2011	macallan	finish device_t-ification
1.46	18-Aug-2011	christos	branches: 1.46.2; 1.46.6; fix gcc-4.5 warnings
1.45	21-Jul-2011	njoly	Fix RADEONFB_DEBUG build.
1.44	11-May-2011	dyoung	Constify pci_attach_args.
1.43	15-Apr-2011	dyoung	Refer to bits in the PCI CSR with using PCI_COMMAND_* instead of PCI_FLAGS_*: change PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED to PCI_COMMAND_IO_ENABLE and PCI_COMMAND_MEM_ENABLE, respectively.
1.42	06-Feb-2011	jmcneill	- add support for using compressed images as splash images - retire SPLASHSCREEN_PROGRESS and SPLASHSCREEN_IMAGE options
1.41	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.40	16-Dec-2010	cegger	branches: 1.40.2; 1.40.4; ioctl: KNF switch-case, remove useless brackets
1.39	03-Nov-2010	macallan	pull in rasops8 for radeonfb, while there defflag a bunch of options and make them all use the same prefix
1.38	02-Nov-2010	macallan	when drawing characters by software make sure the engine is idle before accessing video memory
1.37	24-Aug-2010	macallan	re-initialize the DAC when re-entering terminal emulation mode
1.36	17-Aug-2010	macallan	- fix 8 bit support and run the console in 8 bit for a little extra speed - fix colour oddities seen occasionally on sparc64 - try to use firmware DAC wiring until we figure out what's wrong TODO: - find out why radeonfb gets the DAC wiring ass-backwards on my xvr-100 while Xorg's radeon driver gets it right - find out why reverse characters aren't drawn in reverse tested on macppc and sparc64
1.35	04-May-2010	macallan	use alternate font if available ( for autogenerated box drawing characters )
1.34	23-Jul-2009	macallan	branches: 1.34.2; 1.34.4; quick hack to allow brightness control via PMF
1.33	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.32	06-May-2009	elad	Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.31	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.30	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.29	01-Jun-2008	macallan	branches: 1.29.12; adjust virtual resolution to preferred mode if available From Marco Trillo
1.28	05-May-2008	jmcneill	branches: 1.28.2; Change autoconf message 'direct rendering on dev unsupported' to the more accurate 'drm at dev not configured'.
1.27	10-Apr-2008	cegger	branches: 1.27.2; 1.27.4; use aprint_*_dev and device_xname
1.26	12-Mar-2008	phx	Make console drivers support PCI_IOC_CFGREAD/WRITE ioctls. This is useful for XFree86, which can alternatively find the graphics card through /dev/ttyE0. Approved by macallan.
1.25	28-Feb-2008	macallan	switch to PCI_MAGIC_IO_RANGE
1.24	01-Dec-2007	ad	branches: 1.24.10; 1.24.14; Back out previous.
1.23	01-Dec-2007	ad	Noisy printfs.
1.22	26-Nov-2007	bjs	Add PCI_PRODUCT_ATI_RADEON_R430_554F to the R420 chip class. This will not work until (at the very least) we map the memory properly (MMIO/VRAM BARs are 64-bit)--some of the values in the configuration space may be different. I will discuss this with others, and maybe it will work someday.
1.21	19-Oct-2007	ad	branches: 1.21.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.20	03-Sep-2007	macallan	branches: 1.20.4; back out the reference divider probing - doesn't work right on R1xx
1.19	02-Sep-2007	macallan	always try to probe the reference divider, use default only if that fails
1.18	18-Aug-2007	macallan	branches: 1.18.2; use bool with prop_dictionary_get_bool
1.17	03-Aug-2007	macallan	branches: 1.17.2; 1.17.4; unbork fallout from newlock2 merge: - reinitializing video after X exits works again - backlight levels should be right again
1.16	03-Aug-2007	macallan	make IO mapping actually work
1.15	09-Jul-2007	ad	branches: 1.15.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.14	21-Mar-2007	macallan	try to attach a drm
1.13	21-Mar-2007	macallan	use EDID data passed as property when present This should allow radeonfb to do The Right Thing on Macs that use non- standard video timings and have built-in monitors that don't provide data via DDC since those usually have an EDID block in the OF tree.
1.12	04-Mar-2007	christos	branches: 1.12.2; 1.12.4; 1.12.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.11	09-Feb-2007	ad	branches: 1.11.2; Merge newlock2 to head.
1.10	09-Feb-2007	macallan	don't hardcode inverted backlight levels needed on some macppc models
1.9	03-Jan-2007	macallan	add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.8	13-Dec-2006	macallan	branches: 1.8.2; various minor fixes: - avoid virtual resolution < physical resolution (Marco Trillo) - don't use hardware-accelerated putchar() on R3xx - clear screen when attaching wsdisplay
1.7	28-Nov-2006	christos	branches: 1.7.2; PR/35146: George Michaelson: minor fix to let radeonfb.c compile
1.6	07-Oct-2006	gdamore	Fix an incorrect calculation for the destination X direction when doing blitting. Thanks to David Redman (Tadpole) for noticing it. This probably escaped notice before, since we never do overlapping blits (in the X direction), but this fix may prevent problems if someone ever does use it for that.
1.5	29-Aug-2006	macallan	branches: 1.5.2; 1.5.4; 1.5.6; 1.5.8; - use RADEON_DEFAULT_MODE when no EDID data available - let only root mmap PCI resources through wscons - allow IO access on macppc
1.4	19-Aug-2006	macallan	fix a typo so we actually write sane values into the clipping registers
1.3	19-Aug-2006	macallan	fix some typos, add mmap() support for X needs options RADEONFB_MMAP_BARS for now
1.2	19-Aug-2006	macallan	make this work on macppc. fixes and caveats: - will switch to 32bit colour, 8bit support needs some more work - added support for fonts with widths that aren't multiples of 8 - for now the driver will always try to become system console - mode switching works but is ugly - all the acceleration bits work - X should work with wsfb, mmap() needs some more work - it still needs a hack to allow wsdisplay_cnattach to be called twice
1.1	16-Aug-2006	gdamore	Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.5.8.3	18-Dec-2006	yamt	sync with head.
1.5.8.2	10-Dec-2006	yamt	sync with head.
1.5.8.1	22-Oct-2006	yamt	sync with head
1.5.6.2	09-Sep-2006	rpaulo	sync with head
1.5.6.1	29-Aug-2006	rpaulo	file radeonfb.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.5.4.2	12-Jan-2007	ad	Sync with head.
1.5.4.1	18-Nov-2006	ad	Sync with head.
1.5.2.2	03-Sep-2006	yamt	sync with head.
1.5.2.1	29-Aug-2006	yamt	file radeonfb.c was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.7.2.3	12-May-2007	pavel	Pull up following revision(s) (requested by macallan in ticket #615): sys/dev/pci/radeonfb.c: revision 1.13 use EDID data passed as property when present This should allow radeonfb to do The Right Thing on Macs that use non- standard video timings and have built-in monitors that don't provide data via DDC since those usually have an EDID block in the OF tree.
1.7.2.2	04-Mar-2007	bouyer	Pull up following revision(s) (requested by macallan in ticket #477): sys/dev/pci/radeonfb.c: revision 1.10 sys/dev/pci/radeonfbvar.h: revision 1.4 don't hardcode inverted backlight levels needed on some macppc models
1.7.2.1	04-Mar-2007	bouyer	Pull up following revision(s) (requested by macallan in ticket #472): sys/dev/pci/radeonfb.c: revision 1.9 sys/dev/pci/radeonfbreg.h: revision 1.3 sys/dev/pci/radeonfbvar.h: revision 1.3 add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.8.2.7	17-Mar-2008	yamt	sync with head.
1.8.2.6	07-Dec-2007	yamt	sync with head
1.8.2.5	27-Oct-2007	yamt	sync with head.
1.8.2.4	03-Sep-2007	yamt	sync with head.
1.8.2.3	26-Feb-2007	yamt	sync with head.
1.8.2.2	30-Dec-2006	yamt	sync with head.
1.8.2.1	13-Dec-2006	yamt	file radeonfb.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.11.2.2	24-Mar-2007	yamt	sync with head.
1.11.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.12.6.1	29-Mar-2007	reinoud	Pullup to -current
1.12.4.1	11-Jul-2007	mjf	Sync with head.
1.12.2.4	23-Oct-2007	ad	Sync with head.
1.12.2.3	20-Aug-2007	ad	Sync with HEAD.
1.12.2.2	01-Jul-2007	ad	Adapt to callout API change.
1.12.2.1	10-Apr-2007	ad	Sync with head.
1.15.2.2	03-Sep-2007	skrll	Sync with HEAD.
1.15.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.17.4.2	03-Aug-2007	macallan	unbork fallout from newlock2 merge: - reinitializing video after X exits works again - backlight levels should be right again
1.17.4.1	03-Aug-2007	macallan	file radeonfb.c was added on branch matt-mips64 on 2007-08-03 05:40:48 +0000
1.17.2.4	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.17.2.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.17.2.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.17.2.1	03-Aug-2007	jmcneill	file radeonfb.c was added on branch jmcneill-pm on 2007-09-03 16:48:24 +0000
1.18.2.3	23-Mar-2008	matt	sync with HEAD
1.18.2.2	09-Jan-2008	matt	sync with HEAD
1.18.2.1	06-Nov-2007	matt	sync with HEAD
1.20.4.1	25-Oct-2007	bouyer	Sync with HEAD.
1.21.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.24.14.2	02-Jun-2008	mjf	Sync with HEAD.
1.24.14.1	03-Apr-2008	mjf	Sync with HEAD.
1.24.10.1	24-Mar-2008	keiichi	sync with head.
1.27.4.6	09-Oct-2010	yamt	sync with head
1.27.4.5	11-Aug-2010	yamt	sync with head.
1.27.4.4	19-Aug-2009	yamt	sync with head.
1.27.4.3	16-May-2009	yamt	sync with head
1.27.4.2	04-May-2009	yamt	sync with head.
1.27.4.1	16-May-2008	yamt	sync with head.
1.27.2.2	04-Jun-2008	yamt	sync with head
1.27.2.1	18-May-2008	yamt	sync with head.
1.28.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.29.12.2	23-Jul-2009	jym	Sync with HEAD.
1.29.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.34.4.4	31-May-2011	rmind	sync with head
1.34.4.3	21-Apr-2011	rmind	sync with head
1.34.4.2	05-Mar-2011	rmind	sync with head
1.34.4.1	30-May-2010	rmind	sync with head
1.34.2.3	06-Nov-2010	uebayasi	Sync with HEAD.
1.34.2.2	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.34.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.40.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.40.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.46.6.6	29-Apr-2012	mrg	sync to latest -current.
1.46.6.5	05-Apr-2012	mrg	sync to latest -current.
1.46.6.4	06-Mar-2012	mrg	sync to -current
1.46.6.3	06-Mar-2012	mrg	sync to -current
1.46.6.2	04-Mar-2012	mrg	sync to latest -current.
1.46.6.1	18-Feb-2012	mrg	merge to -current.
1.46.2.6	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.46.2.5	23-Jan-2013	yamt	sync with head
1.46.2.4	16-Jan-2013	yamt	sync with (a bit old) head
1.46.2.3	30-Oct-2012	yamt	sync with head
1.46.2.2	23-May-2012	yamt	sync with head.
1.46.2.1	17-Apr-2012	yamt	sync with head
1.53.2.2	29-Sep-2014	msaitoh	Pull up following revision(s) (requested by riastradh in ticket #1120): sys/dev/pci/radeonfb.c: revision 1.84 sys/dev/wsfb/genfb.c: revision 1.54 Do not printf(foobar()). Bad bad bad bad bad!
1.53.2.1	21-Mar-2012	riz	Pull up following revision(s) (requested by macallan in ticket #132): sys/dev/pci/radeonfb.c: revision 1.59 sys/dev/pci/radeonfbvar.h: revision 1.16 adjust the backlight control interface to match the other drivers ( genfb, r128fb, voyagerfb etc. ) so the lid_switch script does the right thing noticed by riz, I thought I fixed that a long time ago
1.63.2.5	03-Dec-2017	jdolecek	update from HEAD
1.63.2.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.63.2.3	23-Jun-2013	tls	resync from head
1.63.2.2	25-Feb-2013	tls	resync with head
1.63.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.76.4.1	18-May-2014	rmind	sync with head
1.83.2.1	10-Aug-2014	tls	Rebase.
1.84.2.1	17-Oct-2014	martin	Pull up following revision(s) (requested by riastradh in ticket #144): sys/dev/pci/radeonfb.c: revision 1.85 sys/dev/pci/pcivar.h: revision 1.100 sys/dev/pci/pci_map.c: revision 1.31 sys/external/bsd/drm2/include/linux/pci.h: revision 1.9 Generalize pci_find_rom and use it to locate x86 video ROM in drm2. - Make pci_find_rom take the ROM `BAR' size as a parameter, instead of using pci_find_mem with the ROM `BAR' to detect the size. - Use it to find the x86 video ROM in [0xc0000, 0xe0000) in drm2, when nothing else reports that location. - Adapt the one other caller in radeonfb, which already has the maximum ROM size handy (romsz). XXX pullup to netbsd-7
1.88.2.1	28-Aug-2017	skrll	Sync with HEAD
1.94.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.94.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.94.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.99.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.99.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.99.2.1	10-Jun-2019	christos	Sync with HEAD
1.104.4.2	17-Jul-2020	martin	Pull up following revision(s) (requested by macallan in ticket #1019): sys/dev/pci/radeonfbvar.h: revision 1.21 sys/dev/pci/radeonfb.c: revision 1.107 reduce stack usage in radeonfb_pickres() and radeonfb_set_cursor() forgot to commit a header change, again...
1.104.4.1	18-Aug-2019	martin	Pull up following revision(s) (requested by rin in ticket #86): sys/dev/pci/radeonfb.c: revision 1.105 Support bitmap fonts with stride 4 (i.e., width 17 to 32).
1.105.2.1	29-Feb-2020	ad	Sync with head.
1.112.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.113.2.3	24-Apr-2021	thorpej	Make sure to explcitly specify the "wsemuldisplaydev" interface attribute when attaching the wscons display instance, because these devices can also attach children using the "drm" interface attribute.
1.113.2.2	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.113.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.114.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.115.2.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.7	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.6	18-Sep-2020	macallan	minor adaptatioons to new radeonfbreg.h
1.5	08-Feb-2018	dholland	Typos.
1.4	03-Nov-2010	macallan	pull in rasops8 for radeonfb, while there defflag a bunch of options and make them all use the same prefix
1.3	03-Jan-2009	yamt	branches: 1.3.4; 1.3.6; remove extra semicolons.
1.2	19-Oct-2007	ad	branches: 1.2.16; 1.2.20; 1.2.28; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.1	16-Aug-2006	gdamore	branches: 1.1.2; 1.1.6; 1.1.12; 1.1.18; 1.1.32; 1.1.34; 1.1.38; Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.1.38.1	25-Oct-2007	bouyer	Sync with HEAD.
1.1.34.1	06-Nov-2007	matt	sync with HEAD
1.1.32.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.1.18.1	23-Oct-2007	ad	Sync with head.
1.1.12.3	27-Oct-2007	yamt	sync with head.
1.1.12.2	30-Dec-2006	yamt	sync with head.
1.1.12.1	16-Aug-2006	yamt	file radeonfb_bios.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.1.6.2	09-Sep-2006	rpaulo	sync with head
1.1.6.1	16-Aug-2006	rpaulo	file radeonfb_bios.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.2.2	03-Sep-2006	yamt	sync with head.
1.1.2.1	16-Aug-2006	yamt	file radeonfb_bios.c was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.2.28.1	19-Jan-2009	skrll	Sync with HEAD.
1.2.20.1	04-May-2009	yamt	sync with head.
1.2.16.1	17-Jan-2009	mjf	Sync with HEAD.
1.3.6.1	05-Mar-2011	rmind	sync with head
1.3.4.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.4	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.3	22-Dec-2019	thorpej	Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.2	19-Oct-2007	ad	branches: 1.2.108; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.1	16-Aug-2006	gdamore	branches: 1.1.2; 1.1.6; 1.1.12; 1.1.18; 1.1.32; 1.1.34; 1.1.38; Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.1.38.1	25-Oct-2007	bouyer	Sync with HEAD.
1.1.34.1	06-Nov-2007	matt	sync with HEAD
1.1.32.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.1.18.1	23-Oct-2007	ad	Sync with head.
1.1.12.3	27-Oct-2007	yamt	sync with head.
1.1.12.2	30-Dec-2006	yamt	sync with head.
1.1.12.1	16-Aug-2006	yamt	file radeonfb_i2c.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.1.6.2	09-Sep-2006	rpaulo	sync with head
1.1.6.1	16-Aug-2006	rpaulo	file radeonfb_i2c.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.2.2	03-Sep-2006	yamt	sync with head.
1.1.2.1	16-Aug-2006	yamt	file radeonfb_i2c.c was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.2.108.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.9	28-Sep-2020	macallan	moar registers
1.8	18-Sep-2020	macallan	update register definitions from xsrc/external/mit/xf86-video-ati/ merge in enough of the local changes to make radeonfb compile tested on various macppc boxes
1.7	15-Mar-2019	macallan	moar registers
1.6	23-Aug-2017	macallan	branches: 1.6.4; consistency...
1.5	30-Dec-2012	macallan	branches: 1.5.14; Make R3xx actually work by wiring outputs, DACs and CRTCs sanely, add a bunch of R3xx-specific workarounds. Tested in xvr-100, iBook G4 and an RV350.
1.4	30-Dec-2011	macallan	branches: 1.4.6; add some register definitions
1.3	03-Jan-2007	macallan	branches: 1.3.80; 1.3.84; add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.2	29-Aug-2006	macallan	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.10; 1.2.12; those should have been committed along with radeonfb.c doh
1.1	16-Aug-2006	gdamore	Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.2.12.3	26-Feb-2007	yamt	sync with head.
1.2.12.2	30-Dec-2006	yamt	sync with head.
1.2.12.1	29-Aug-2006	yamt	file radeonfbreg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.2.10.1	04-Mar-2007	bouyer	Pull up following revision(s) (requested by macallan in ticket #472): sys/dev/pci/radeonfb.c: revision 1.9 sys/dev/pci/radeonfbreg.h: revision 1.3 sys/dev/pci/radeonfbvar.h: revision 1.3 add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.2.6.2	09-Sep-2006	rpaulo	sync with head
1.2.6.1	29-Aug-2006	rpaulo	file radeonfbreg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.2.4.1	12-Jan-2007	ad	Sync with head.
1.2.2.2	03-Sep-2006	yamt	sync with head.
1.2.2.1	29-Aug-2006	yamt	file radeonfbreg.h was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.3.84.1	18-Feb-2012	mrg	merge to -current.
1.3.80.2	23-Jan-2013	yamt	sync with head
1.3.80.1	17-Apr-2012	yamt	sync with head
1.4.6.2	03-Dec-2017	jdolecek	update from HEAD
1.4.6.1	25-Feb-2013	tls	resync with head
1.5.14.1	28-Aug-2017	skrll	Sync with HEAD
1.6.4.1	10-Jun-2019	christos	Sync with HEAD
1.23	23-Jul-2023	andvar	s/probaby/probably/ in comments.
1.22	28-Sep-2020	macallan	first step towards r5xx support: - no CRTC programming yet, we just use whatever mode the (open)firmware hands us - palette control, fb location etc. all work - hardware acceleration for character drawing, scrolling etc. works - hardware cursor works ... tested & still works on various radeons in various macppc machines
1.21	11-Jun-2020	macallan	forgot to commit a header change, again...
1.20	05-Nov-2014	macallan	branches: 1.20.26; - unmap registers in WSDISPLAYIO_MODE_MAPPED only if a drm attached - don't try to touch backlight registers when they're not mapped
1.19	22-Jan-2014	macallan	implement WSDISPLAYIO_{GET\|PUT}CMAP ioctl()s while there: - unmap registers and vram only in WSDISPLAYIO_MODE_MAPPED ( so DRM can map them ) but not in WSDISPLAYIO_MODE_DUMBFB ( so wsfb can do things that require hardware access, like palette and cursor ioctl()s ) - reject ioctl()s that need hardware access if registers are unmapped ( so we don't crash instead ) - use the same bit order in cursor sprite images as the other drivers do with all this, Xorg with wsfb works properly on top of radeonfb tested on i386 and macppc ( fingers crossed for alpha ;) )
1.18	28-Mar-2013	macallan	branches: 1.18.4; clear screen when leaving X so we don't leave garbage in the borders
1.17	20-Dec-2012	macallan	don't access registers when they're not mapped now X works on i386 with radeonfb
1.16	15-Mar-2012	macallan	branches: 1.16.2; adjust the backlight control interface to match the other drivers ( genfb, r128fb, voyagerfb etc. ) so the lid_switch script does the right thing noticed by riz, I thought I fixed that a long time ago
1.15	28-Feb-2012	macallan	support anti-aliased fonts in 8 bit as well
1.14	16-Feb-2012	macallan	cache glyphs in video memory when using anti-aliased fonts on r128fb and radeonfb
1.13	30-Jan-2012	drochner	branches: 1.13.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.12	30-Dec-2011	macallan	this should have been committed along with the previous changes in radeonfb.c
1.11	06-Feb-2011	jmcneill	branches: 1.11.4; 1.11.8; - add support for using compressed images as splash images - retire SPLASHSCREEN_PROGRESS and SPLASHSCREEN_IMAGE options
1.10	03-Nov-2010	macallan	branches: 1.10.2; 1.10.4; pull in rasops8 for radeonfb, while there defflag a bunch of options and make them all use the same prefix
1.9	02-Nov-2010	macallan	when drawing characters by software make sure the engine is idle before accessing video memory
1.8	17-Aug-2010	macallan	- fix 8 bit support and run the console in 8 bit for a little extra speed - fix colour oddities seen occasionally on sparc64 - try to use firmware DAC wiring until we figure out what's wrong TODO: - find out why radeonfb gets the DAC wiring ass-backwards on my xvr-100 while Xorg's radeon driver gets it right - find out why reverse characters aren't drawn in reverse tested on macppc and sparc64
1.7	28-Jul-2009	macallan	branches: 1.7.2; 1.7.4; forgot to commit that along with radeonfb.c
1.6	03-Aug-2007	macallan	branches: 1.6.2; 1.6.28; 1.6.36; unbork fallout from newlock2 merge: - reinitializing video after X exits works again - backlight levels should be right again
1.5	09-Feb-2007	ad	branches: 1.5.6; 1.5.14; Merge newlock2 to head.
1.4	09-Feb-2007	macallan	don't hardcode inverted backlight levels needed on some macppc models
1.3	03-Jan-2007	macallan	add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.2	29-Aug-2006	macallan	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.10; 1.2.12; those should have been committed along with radeonfb.c doh
1.1	16-Aug-2006	gdamore	Initial import of radeonfb. Works to some degree on at least some hardware. Most of the testing was done on MIPS hardware -- it probably needs work before it will be useful with x86 hardware, and it is probably incompatible with the X11 server. "ATI Technologies Inc. ("ATI") has not assisted in the creation of, and does not endorse, this software. ATI will not be responsible or liable for any actual or alleged damage or loss caused by or in connection with the use of or reliance on this software." Enjoy!
1.2.12.4	03-Sep-2007	yamt	sync with head.
1.2.12.3	26-Feb-2007	yamt	sync with head.
1.2.12.2	30-Dec-2006	yamt	sync with head.
1.2.12.1	29-Aug-2006	yamt	file radeonfbvar.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.2.10.2	04-Mar-2007	bouyer	Pull up following revision(s) (requested by macallan in ticket #477): sys/dev/pci/radeonfb.c: revision 1.10 sys/dev/pci/radeonfbvar.h: revision 1.4 don't hardcode inverted backlight levels needed on some macppc models
1.2.10.1	04-Mar-2007	bouyer	Pull up following revision(s) (requested by macallan in ticket #472): sys/dev/pci/radeonfb.c: revision 1.9 sys/dev/pci/radeonfbreg.h: revision 1.3 sys/dev/pci/radeonfbvar.h: revision 1.3 add preliminary support for setting backlight levels via ioctl() from Johan Wallen
1.2.6.2	09-Sep-2006	rpaulo	sync with head
1.2.6.1	29-Aug-2006	rpaulo	file radeonfbvar.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.2.4.1	12-Jan-2007	ad	Sync with head.
1.2.2.2	03-Sep-2006	yamt	sync with head.
1.2.2.1	29-Aug-2006	yamt	file radeonfbvar.h was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.5.14.1	15-Aug-2007	skrll	Sync with HEAD.
1.5.6.1	20-Aug-2007	ad	Sync with HEAD.
1.6.36.2	03-Aug-2007	macallan	unbork fallout from newlock2 merge: - reinitializing video after X exits works again - backlight levels should be right again
1.6.36.1	03-Aug-2007	macallan	file radeonfbvar.h was added on branch matt-mips64 on 2007-08-03 05:40:48 +0000
1.6.28.2	09-Oct-2010	yamt	sync with head
1.6.28.1	19-Aug-2009	yamt	sync with head.
1.6.2.2	03-Aug-2007	macallan	unbork fallout from newlock2 merge: - reinitializing video after X exits works again - backlight levels should be right again
1.6.2.1	03-Aug-2007	macallan	file radeonfbvar.h was added on branch jmcneill-pm on 2007-08-03 05:40:48 +0000
1.7.4.1	05-Mar-2011	rmind	sync with head
1.7.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.7.2.1	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.10.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.10.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.11.8.5	05-Apr-2012	mrg	sync to latest -current.
1.11.8.4	06-Mar-2012	mrg	sync to -current
1.11.8.3	06-Mar-2012	mrg	sync to -current
1.11.8.2	04-Mar-2012	mrg	sync to latest -current.
1.11.8.1	18-Feb-2012	mrg	merge to -current.
1.11.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.11.4.2	23-Jan-2013	yamt	sync with head
1.11.4.1	17-Apr-2012	yamt	sync with head
1.13.2.1	21-Mar-2012	riz	Pull up following revision(s) (requested by macallan in ticket #132): sys/dev/pci/radeonfb.c: revision 1.59 sys/dev/pci/radeonfbvar.h: revision 1.16 adjust the backlight control interface to match the other drivers ( genfb, r128fb, voyagerfb etc. ) so the lid_switch script does the right thing noticed by riz, I thought I fixed that a long time ago
1.16.2.4	03-Dec-2017	jdolecek	update from HEAD
1.16.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.16.2.2	23-Jun-2013	tls	resync from head
1.16.2.1	25-Feb-2013	tls	resync with head
1.18.4.1	18-May-2014	rmind	sync with head
1.20.26.1	17-Jul-2020	martin	Pull up following revision(s) (requested by macallan in ticket #1019): sys/dev/pci/radeonfbvar.h: revision 1.21 sys/dev/pci/radeonfb.c: revision 1.107 reduce stack usage in radeonfb_pickres() and radeonfb_set_cursor() forgot to commit a header change, again...
1.27	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.26	31-Jul-2012	bouyer	branches: 1.26.2; 1.26.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.25	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.24	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.23	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.22	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.21	04-Apr-2011	dyoung	branches: 1.21.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.20	05-Nov-2010	jakllsch	branches: 1.20.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.19	14-Mar-2009	dsl	branches: 1.19.2; 1.19.4; Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.18	18-Mar-2008	cube	branches: 1.18.4; 1.18.12; 1.18.18; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.17	24-Oct-2007	xtraeme	branches: 1.17.12; 1.17.16; - Add support for the ServerWorks HT-1000 IDE Controller in rccide. - Add support for another ServerWorks HT-1000 SATA Controller. From OpenBSD.
1.16	09-Feb-2007	ad	branches: 1.16.18; 1.16.20; 1.16.24; Merge newlock2 to head.
1.15	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.14	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.13	11-Dec-2005	christos	branches: 1.13.20; 1.13.22; merge ktrace-lwp.
1.12	21-Aug-2004	thorpej	branches: 1.12.12; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.11	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.10	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.9	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.8	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.7	03-Jan-2004	thorpej	branches: 1.7.4; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.6	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.5	14-Dec-2003	thorpej	Liberally apply "static", and wrap a line.
1.4	02-Dec-2003	bouyer	Use a different interrupt routine for the CSB6. The CSB6 doens't assert IDEDMA_CTL_INTR for non-DMA commands. It would be nice to have a way to know if the interrupt was for us or not at this point, but for now let wdcintr() deal with it. Problem reported and patch tested by Martti Kuparinen.
1.3	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.2	07-Nov-2003	enami	Sync with sys/dev/pci/pcidevs change (ProductID 0x0213 was renamed to CSB6_RAID from CSB6_IDE). And exclude the device of RAID subclass (i.e., match IDE subclass only).
1.1	04-Nov-2003	mycroft	Re-add ServerWorks IDE support, as "rccide". Note: this is untested, as I no longer have a ServerWorks-based motherboard.
1.7.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.7.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.7.4.2	03-Aug-2004	skrll	Sync with HEAD
1.7.4.1	03-Jan-2004	skrll	file rccide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.12.12.4	24-Mar-2008	yamt	sync with head.
1.12.12.3	27-Oct-2007	yamt	sync with head.
1.12.12.2	26-Feb-2007	yamt	sync with head.
1.12.12.1	30-Dec-2006	yamt	sync with head.
1.13.22.2	10-Dec-2006	yamt	sync with head.
1.13.22.1	22-Oct-2006	yamt	sync with head
1.13.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.13.20.1	18-Nov-2006	ad	Sync with head.
1.16.24.1	13-Nov-2007	bouyer	Sync with HEAD
1.16.20.2	23-Mar-2008	matt	sync with HEAD
1.16.20.1	06-Nov-2007	matt	sync with HEAD
1.16.18.1	28-Oct-2007	joerg	Sync with HEAD.
1.17.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.17.12.1	24-Mar-2008	keiichi	sync with head.
1.18.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.18.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.18.4.1	04-May-2009	yamt	sync with head.
1.19.4.2	21-Apr-2011	rmind	sync with head
1.19.4.1	05-Mar-2011	rmind	sync with head
1.19.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.20.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.21.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.21.4.1	30-Oct-2012	yamt	sync with head
1.26.4.1	18-May-2014	rmind	sync with head
1.26.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.26.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.8	08-Jul-2014	msaitoh	Add support for RDC R1012 IDE controller. Tested on EduCake(Vortex86EX) with sdcard: \| read \| write \| ------+---------+-------+ pciide\| 999519\| 244574\| rdcide\| 16011238\| 939895\| ------+---------+-------+ (bytes/sec) rd: dd if=/dev/rwd0a of=/dev/null bs=1m count=100 wr: dd if=/dev/zero of=/var/tmp/hoge bs=1m count=100
1.7	31-Jul-2012	bouyer	branches: 1.7.2; 1.7.12; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.6	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.5	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.4	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.3	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.2	04-Apr-2011	dyoung	branches: 1.2.2; 1.2.6; 1.2.8; Make rdcide_chip_map()'s pci_attach_args argument const.
1.1	04-Apr-2011	bouyer	Add rdcide(4), a driver for the IDE controller found in RDC's vortex86/PMX-1000 system-on-chip.
1.2.8.1	30-Oct-2012	yamt	sync with head
1.2.6.2	06-Jun-2011	jruoho	Sync with HEAD.
1.2.6.1	04-Apr-2011	jruoho	file rdcide.c was added on branch jruoho-x86intr on 2011-06-06 09:08:26 +0000
1.2.2.2	21-Apr-2011	rmind	sync with head
1.2.2.1	04-Apr-2011	rmind	file rdcide.c was added on branch rmind-uvmplock on 2011-04-21 01:42:00 +0000
1.7.12.1	10-Aug-2014	tls	Rebase.
1.7.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.7.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.1	04-Apr-2011	bouyer	branches: 1.1.2; 1.1.6; Add rdcide(4), a driver for the IDE controller found in RDC's vortex86/PMX-1000 system-on-chip.
1.1.6.2	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.1	04-Apr-2011	jruoho	file rdcide_reg.h was added on branch jruoho-x86intr on 2011-06-06 09:08:26 +0000
1.1.2.2	21-Apr-2011	rmind	sync with head
1.1.2.1	04-Apr-2011	rmind	file rdcide_reg.h was added on branch rmind-uvmplock on 2011-04-21 01:42:00 +0000
1.11	18-Jan-2025	mrg	match RTS5249. fix a comment.
1.10	25-Sep-2022	thorpej	branches: 1.10.10; Remove unnecessary include of <sys/malloc.h>.
1.9	27-Apr-2020	jmcneill	Add support for Realtek RTS522A
1.8	09-Dec-2018	jdolecek	branches: 1.8.4; use pci_intr_establish_xname() everywhere
1.7	24-Apr-2018	maya	branches: 1.7.2; rtsx(4): Add support for RT525A, from openbsd. ok rkujawa
1.6	07-Jul-2016	msaitoh	branches: 1.6.10; 1.6.16; KNF. Remove extra spaces. No functional change.
1.5	06-Nov-2015	nonaka	Always use pci_intr_alloc(9)/pci_intr_release(9).
1.4	05-Sep-2015	nonaka	Use MSI/MSI-X instead of legacy INTx, if available.
1.3	29-Oct-2014	nonaka	branches: 1.3.2; Added RTS5227/RTL8402/RTL8411/RTL8411B support.
1.2	29-Mar-2014	christos	branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.1	19-Mar-2014	nonaka	Add a driver for Realtek RTS5209/RTS5229 Card Reader. Ported from OpenBSD.
1.2.10.3	03-Dec-2017	jdolecek	update from HEAD
1.2.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.10.1	29-Mar-2014	tls	file rtsx_pci.c was added on branch tls-maxphys on 2014-08-20 00:03:48 +0000
1.2.8.1	17-Jan-2015	martin	Pull up following revision(s) (requested by nonaka in ticket #432): sys/dev/ic/rtsxreg.h: revision 1.2 sys/dev/ic/rtsxvar.h: revision 1.2 sys/dev/pci/rtsx_pci.c: revision 1.3 sys/dev/ic/rtsx.c: revision 1.2 share/man/man4/rtsx.4: revision 1.3-1.4 Added RTS5227/RTL8402/RTL8411/RTL8411B support.
1.2.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.6.1	29-Mar-2014	yamt	file rtsx_pci.c was added on branch yamt-pagecache on 2014-05-22 11:40:33 +0000
1.2.4.2	18-May-2014	rmind	sync with head
1.2.4.1	29-Mar-2014	rmind	file rtsx_pci.c was added on branch rmind-smpnet on 2014-05-18 17:45:44 +0000
1.3.2.3	09-Jul-2016	skrll	Sync with HEAD
1.3.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.3.2.1	22-Sep-2015	skrll	Sync with HEAD
1.6.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.6.16.1	02-May-2018	pgoyette	Synch with HEAD
1.6.10.1	14-Jun-2018	martin	Pull up following revision(s) (requested by maya in ticket #878): share/man/man4/rtsx.4: revision 1.5 sys/dev/pci/pcidevs: revision 1.1331 sys/dev/ic/rtsxreg.h: revision 1.3 sys/dev/ic/rtsxvar.h: revision 1.3 sys/dev/pci/rtsx_pci.c: revision 1.7 sys/dev/ic/rtsx.c: revision 1.3 add realtek RT525A PCI-E card reader found in my Dell XPS 15 9550. rtsx(4): Add support for RT525A, from openbsd. ok rkujawa Denote RTS525A, bump date
1.7.2.1	10-Jun-2019	christos	Sync with HEAD
1.8.4.1	28-Apr-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #864): share/man/man4/rtsx.4: revision 1.6 sys/dev/pci/pcidevs: revision 1.1410 sys/dev/pci/rtsx_pci.c: revision 1.9 Add an eclectic mix of GeForce GTX 680, RTS522A PCI-E Card Reader, intel wifi 9560 + whitespace Add support for Realtek RTS522A
1.10.10.1	02-Aug-2025	perseant	Sync with HEAD
1.58	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.57	21-Oct-2019	christos	follow the channel locking protocol during probe (like the other drivers)
1.56	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.55	20-Oct-2017	jdolecek	branches: 1.55.2; 1.55.4; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.54	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.53	23-Aug-2016	msaitoh	branches: 1.53.4; Remove PCI_PRODUCT_CMDTECH_240's entry because CMDTECH_AAR_1210SA has the same ID.
1.52	29-Mar-2014	christos	branches: 1.52.4; 1.52.6; 1.52.12; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.51	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.50	15-Sep-2013	martin	Comment out unused variables
1.49	27-Oct-2012	chs	branches: 1.49.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.48	31-Jul-2012	bouyer	branches: 1.48.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.47	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.46	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.45	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.44	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.43	04-Apr-2011	dyoung	branches: 1.43.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.42	05-Nov-2010	jakllsch	branches: 1.42.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.41	08-Jun-2010	mrg	attach on Sil240 and IXP 300 chips as well, as if they are 3112 variants. should fix PR#32577, id's found in linux driver.
1.40	23-Mar-2010	mrg	only apply the satalink 3112 rev 0.1 and earlier to actual 3112 based cards, not eg, 3512 cards. should help performance for 3512 cards, derived from discussions in PR#42980: satalink DMA fails under amd64.
1.39	26-Nov-2009	njoly	branches: 1.39.2; 1.39.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.38	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.37	10-Apr-2008	cegger	branches: 1.37.2; 1.37.4; use aprint_*_dev and device_xname
1.36	18-Mar-2008	cube	Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.35	27-Dec-2007	ws	branches: 1.35.2; 1.35.6; Change cacheline fixup for 3114's second set of channels, too. Fix provided by Masanori Kanaoka in PR 37519. Fixed differently to avoid code duplication.
1.34	09-Feb-2007	ad	branches: 1.34.20; 1.34.26; 1.34.32; Merge newlock2 to head.
1.33	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.32	25-Oct-2006	bouyer	Use common wdc_sataprobe() where possible; explain why we can't in satalink.c
1.31	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.30	26-Jun-2006	xtraeme	branches: 1.30.4; 1.30.6; Add missing __KERNEL_RCSID()
1.29	29-Mar-2006	thorpej	branches: 1.29.4; Use device_cfdata().
1.28	16-Jan-2006	bouyer	branches: 1.28.2; 1.28.4; 1.28.6; 1.28.8; 1.28.10; properly use ata_channel->ch_ndrive: - initialize it properly in the bus front-ends (all 2, exept in wdc_pcmcia.c for the "Sandisk CompactFlash Card" where it's set to 1) - remplace hardcoded '2' by ata_channel->ch_ndrive in MI IDE drivers. From Christos Zoulas in kern/32501.
1.27	24-Dec-2005	perry	branches: 1.27.2; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.26	11-Dec-2005	christos	merge ktrace-lwp.
1.25	01-Jun-2005	ragge	branches: 1.25.2; The read/read multiple cacheline problem may also appear on the 3114 controller, so apply the cacheline fix to it as well.
1.24	27-Feb-2005	perry	branches: 1.24.2; nuke trailing whitespace
1.23	05-Sep-2004	skd	branches: 1.23.4; 1.23.6; Fill out correct cap variable so that DMA works again.
1.22	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.21	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.20	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.19	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.18	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.17	19-Jul-2004	msaitoh	branches: 1.17.2; Add Adaptec AAR-1210SA serial ATA RAID. RAID is not supported by this driver.
1.16	23-Jun-2004	briggs	When the Silicon Image 3112 retries a PCI memory read command, it may retry it as a memory read multiple command under some circumstances. This can totally confuse some PCI controllers, so ensure that it will never do this by making sure that the Read Threshold (FIFO Read Request Control) field of the FIFO Valid Byte Count and Control registers for both channels (BA5 offset 0x40 and 0x44) are set to be at least as large as the cacheline size register (the unit of measure for these registers is 32 bytes).
1.15	28-May-2004	thorpej	Use the common sata_speed() routine.
1.14	25-May-2004	thorpej	Add the notion of "shadow registers" to the wdc driver. These shadow registers are registers that overlap with others on many controllers, but which may actually be distinct on some controllers. Right now, the two shadows are: - wd_status (usually overlaps wd_command) - wd_features (usually overlaps wd_error) Add a new helper function, wdc_init_shadow_regs(), used to initialize the shadow register handles on controllers where they do actually overlap. Partially from Jordan Rhody @ Wasabi Systems, Inc.
1.13	07-May-2004	sekiya	Add support for Silicon Image 3512 chip to satalink driver -- treat it as a 3112 variant (without RAID support), which does the right thing with the controller on my K8NNXP.
1.12	07-Jan-2004	thorpej	branches: 1.12.2; Initialize some importantant members of the drive info structure. (I have no idea how this worked for me before, but I swear that it did...) kern/24015.
1.11	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.10	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.9	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.8	31-Dec-2003	martin	Include <dev/ata/atareg.h>, so that WDSD_IBM is defined.
1.7	30-Dec-2003	thorpej	Move the drive probing code out of atabusconfig() and into a new wdc_drvprobe() function. wdc_drvprobe() is used if the controller does not specify a custom one prior to calling wdcattach(). The WDC_CAPABILITY_DRVPROBE bit is gone.
1.6	20-Dec-2003	thorpej	Fix a typo.
1.5	20-Dec-2003	thorpej	Add support for the Silicon Image 3114 SATALink 4-port SATA controller.
1.4	19-Dec-2003	thorpej	Use a register map to find the offsets of various per-channel registers in the BA5 space.
1.3	15-Dec-2003	thorpej	Remove the notice that was in cmdide.c before SATALink support was split out. All of the code in this file was written by me.
1.2	15-Dec-2003	thorpej	Use the SControl and SStatus registers to query the SATA PHY for device detection.
1.1	13-Dec-2003	thorpej	Split out the Silicon Image SATALink support into its own driver, "satalink".
1.12.2.1	11-May-2004	tron	Pull up revision 1.13 (requested by sekiya in ticket #284): Add support for Silicon Image 3512 chip to satalink driver -- treat it as a 3112 variant (without RAID support), which does the right thing with the controller on my K8NNXP.
1.17.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.17.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.17.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.17.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.17.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.17.2.2	03-Aug-2004	skrll	Sync with HEAD
1.17.2.1	19-Jul-2004	skrll	file satalink.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.23.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.23.4.1	29-Apr-2005	kent	sync with -current
1.24.2.1	01-Jun-2005	tron	Pull up revision 1.25 (requested by ragge in ticket #364): The read/read multiple cacheline problem may also appear on the 3114 controller, so apply the cacheline fix to it as well.
1.25.2.5	24-Mar-2008	yamt	sync with head.
1.25.2.4	21-Jan-2008	yamt	sync with head
1.25.2.3	26-Feb-2007	yamt	sync with head.
1.25.2.2	30-Dec-2006	yamt	sync with head.
1.25.2.1	21-Jun-2006	yamt	sync with head.
1.27.2.1	01-Feb-2006	yamt	sync with head.
1.28.10.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.28.8.1	19-Apr-2006	elad	sync with head.
1.28.6.2	11-Aug-2006	yamt	sync with head
1.28.6.1	01-Apr-2006	yamt	sync with head.
1.28.4.1	22-Apr-2006	simonb	Sync with head.
1.28.2.1	09-Sep-2006	rpaulo	sync with head
1.29.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.30.6.2	10-Dec-2006	yamt	sync with head.
1.30.6.1	22-Oct-2006	yamt	sync with head
1.30.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.30.4.1	18-Nov-2006	ad	Sync with head.
1.34.32.1	02-Jan-2008	bouyer	Sync with HEAD
1.34.26.1	18-Feb-2008	mjf	Sync with HEAD.
1.34.20.2	23-Mar-2008	matt	sync with HEAD
1.34.20.1	09-Jan-2008	matt	sync with HEAD
1.35.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.35.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.35.2.1	24-Mar-2008	keiichi	sync with head.
1.37.4.3	11-Aug-2010	yamt	sync with head.
1.37.4.2	11-Mar-2010	yamt	sync with head
1.37.4.1	16-May-2008	yamt	sync with head.
1.37.2.1	18-May-2008	yamt	sync with head.
1.39.4.4	21-Apr-2011	rmind	sync with head
1.39.4.3	05-Mar-2011	rmind	sync with head
1.39.4.2	03-Jul-2010	rmind	sync with head
1.39.4.1	30-May-2010	rmind	sync with head
1.39.2.3	06-Nov-2010	uebayasi	Sync with HEAD.
1.39.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.39.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.42.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.43.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.43.4.1	30-Oct-2012	yamt	sync with head
1.48.2.3	03-Dec-2017	jdolecek	update from HEAD
1.48.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.48.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.49.2.1	18-May-2014	rmind	sync with head
1.52.12.1	18-Jan-2017	skrll	Sync with netbsd-5
1.52.6.1	05-Oct-2016	skrll	Sync with HEAD
1.52.4.1	08-Dec-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1290): sys/dev/pci/satalink.c: revision 1.53 Remove PCI_PRODUCT_CMDTECH_240's entry because CMDTECH_AAR_1210SA has the same ID.
1.53.4.2	27-Sep-2017	jdolecek	change wdc_init_shadow_regs() to accept only struct wdc_regs, it doesn't touch anything else factor out the probe-only struct ata_channel initialization to wdcprobe(), to reduce duplication of logic in individual drivers, and to actually work now that more init is needed beyond the memset()
1.53.4.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.55.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.55.4.1	10-Jun-2019	christos	Sync with HEAD
1.55.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.8	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.7	31-Jul-2012	bouyer	branches: 1.7.2; 1.7.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.6	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.5	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.4	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.3	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.2	04-Apr-2011	dyoung	branches: 1.2.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.1	06-Nov-2010	jakllsch	branches: 1.1.2; 1.1.6; Add Intel SCH IDE controller driver. From OpenBSD via Tomokazu HARADA in PR#42310.
1.1.6.3	21-Apr-2011	rmind	sync with head
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	06-Nov-2010	rmind	file schide.c was added on branch rmind-uvmplock on 2011-03-05 20:53:57 +0000
1.1.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.2.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.1	30-Oct-2012	yamt	sync with head
1.7.4.1	18-May-2014	rmind	sync with head
1.7.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.7.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.21	05-Jan-2023	msaitoh	Add quirk setting for some Intel eMMC devices. On some Intel eMMC controllers, the driver reports "autoconfiguration error: couldn't enable card: 60" even though they really have eMMC device. This change fixes the problem on some machines. It might be required more quirks for newer devices (or HS400 support). At least, this change fixes the problem on GIGABYTE MA10-ST0.
1.20	04-Jan-2023	msaitoh	Fix typo. No functional change.
1.19	14-Oct-2022	jmcneill	branches: 1.19.2; Disable ADMA2 on Ricoh SDHCI controllers. PR# 57015
1.18	10-Nov-2021	msaitoh	Support 64bit BAR.
1.17	01-Jan-2020	mlelstv	Add support for single power writes, and enable for Ricoh 5u823.
1.16	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.15	28-Sep-2019	mlelstv	Fix typo, remove debug printf.
1.14	27-Apr-2017	msaitoh	branches: 1.14.2; 1.14.10; 1.14.16; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.13	17-Feb-2017	nonaka	branches: 1.13.4; sdhc(4): hardware reset support for Intel eMMC controller
1.12	09-Aug-2015	mlelstv	branches: 1.12.2; 1.12.4; Replace the LOWER_FREQ quirk with the SLOW_SDR50 quirk for the Ricoh 5U823 controller. This stabilizes SD cards at standard speed and high speed and also UHS-I SDR50.
1.11	29-Mar-2014	christos	branches: 1.11.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.10	20-Dec-2012	jakllsch	branches: 1.10.2; Add detachment support for sdhc_pci. Tested on an ExpressCard JMicron JMB38[89] with as-of-yet-uncommitted changes.
1.9	29-Jul-2012	matt	branches: 1.9.2; Add FORCE_DMA quirk for 5U822 (already in lace for 5C822).
1.8	12-Jul-2012	jakllsch	Ricoh R5C822 has working DMA, but does not advertise support for it, use force.
1.7	02-Mar-2012	nonaka	Added Ricoh 5U822/5U823 SD/MMC Controller support.
1.6	02-Feb-2012	nonaka	Use "opt_sdmmc.h" for SDHC_DEBUG.
1.5	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.4	02-Feb-2011	jakllsch	branches: 1.4.4; 1.4.8; Like some other TI SDHCI parts, PCI_PRODUCT_TI_PCIXX12SD also needs the SD card moved to the SDHCI controller.
1.3	02-Oct-2009	uebayasi	branches: 1.3.2; 1.3.6; 1.3.8; 1.3.10; 1.3.12; Pass not (struct sdhc_pci_softc ) but (struct sdhc_softc ) to sdhc_intr(). No functional change because struct sdhc_pci_softc has struct sdhc_softc as the 1st member.
1.2	29-Jun-2009	hubertf	Use PCI IDs from pcidevs, instead of storing them in here. Compile tested and OK'd by NONAKA Kimihiro.
1.1	21-Apr-2009	nonaka	branches: 1.1.2; 1.1.4; 1.1.6; Added SD/MMC support from OpenBSD. tested on i386, amd64 at current-users ML by pgoyette@. tested on zaurus by myself.
1.1.6.3	23-Jul-2009	jym	Sync with HEAD.
1.1.6.2	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.1.6.1	21-Apr-2009	jym	file sdhc_pci.c was added on branch jym-xensuspend on 2009-05-13 17:20:29 +0000
1.1.4.4	11-Mar-2010	yamt	sync with head
1.1.4.3	18-Jul-2009	yamt	sync with head.
1.1.4.2	04-May-2009	yamt	sync with head.
1.1.4.1	21-Apr-2009	yamt	file sdhc_pci.c was added on branch yamt-nfs-mp on 2009-05-04 08:13:01 +0000
1.1.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.1.2.1	21-Apr-2009	skrll	file sdhc_pci.c was added on branch nick-hppapmap on 2009-04-28 07:35:59 +0000
1.3.12.1	08-Feb-2011	bouyer	Sync with HEAD
1.3.10.1	06-Jun-2011	jruoho	Sync with HEAD.
1.3.8.2	21-Apr-2010	matt	sync to netbsd-5
1.3.8.1	02-Oct-2009	matt	file sdhc_pci.c was added on branch matt-nb5-mips64 on 2010-04-21 00:27:49 +0000
1.3.6.1	05-Mar-2011	rmind	sync with head
1.3.2.2	28-Oct-2009	bouyer	Pull up the following revisions, requested by sborrill in ticket #1114: share/man/man4/sdhc.4 1.1-1.2 sys/dev/pci/sdhc_pci.c 1.1-1.3 distrib/sets/lists/man/mi patch share/man/man4/Makefile patch sys/arch/amd64/conf/GENERIC patch sys/arch/i386/conf/ALL patch sys/arch/i386/conf/GENERIC patch sys/dev/pci/files.pci patch sys/dev/pci/pcidevs patch sys/dev/pci/pcidevs.h regen sys/dev/pci/pcidevs_data.h regen Add sdhc(4), a driver for SD controllers following the SD Host Controller Standard Simplified Specification.
1.3.2.1	02-Oct-2009	bouyer	file sdhc_pci.c was added on branch netbsd-5 on 2009-10-28 09:03:42 +0000
1.4.8.4	06-Mar-2012	mrg	sync to -current
1.4.8.3	06-Mar-2012	mrg	sync to -current
1.4.8.2	04-Mar-2012	mrg	sync to latest -current.
1.4.8.1	18-Feb-2012	mrg	merge to -current.
1.4.4.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.4.4.3	23-Jan-2013	yamt	sync with head
1.4.4.2	30-Oct-2012	yamt	sync with head
1.4.4.1	17-Apr-2012	yamt	sync with head
1.9.2.3	03-Dec-2017	jdolecek	update from HEAD
1.9.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.2.1	25-Feb-2013	tls	resync with head
1.10.2.1	18-May-2014	rmind	sync with head
1.11.6.2	28-Aug-2017	skrll	Sync with HEAD
1.11.6.1	22-Sep-2015	skrll	Sync with HEAD
1.12.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.12.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.13.4.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.14.16.3	22-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1721): sys/dev/pci/sdhc_pci.c: revision 1.21 Add quirk setting for some Intel eMMC devices. On some Intel eMMC controllers, the driver reports "autoconfiguration error: couldn't enable card: 60" even though they really have eMMC device. This change fixes the problem on some machines. It might be required more quirks for newer devices (or HS400 support). At least, this change fixes the problem on GIGABYTE MA10-ST0.
1.14.16.2	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1566): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1469 sys/dev/pci/pcidevs: revision 1.1470 sys/dev/pci/pcidevs: revision 1.1471 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Add some AMD 17h/9xh devices from OpenBSD. Add AMD F17/Axh devices. Add some Xeon Scalable devices from OpenBSD. Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.14.16.1	03-Dec-2021	martin	Pull up the followin revisions (all via patch), requested by msaitoh in ticket #1383: sys/dev/sdmmc/sdhc.c 1.110, 1.112 sys/dev/sdmmc/sdmmc_mem.c 1.74 sys/dev/pci/sdhc_pci.c 1.18 - Support 64bit BAR. - Use unsigned to avoid undefined behavior in hwrite[12]() and sdmmc_mem_sd_switch(). - Fix typo in comment.
1.14.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.14.10.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.14.2.3	22-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1889): sys/dev/pci/sdhc_pci.c: revision 1.21 Add quirk setting for some Intel eMMC devices. On some Intel eMMC controllers, the driver reports "autoconfiguration error: couldn't enable card: 60" even though they really have eMMC device. This change fixes the problem on some machines. It might be required more quirks for newer devices (or HS400 support). At least, this change fixes the problem on GIGABYTE MA10-ST0.
1.14.2.2	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1783): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1469 sys/dev/pci/pcidevs: revision 1.1470 sys/dev/pci/pcidevs: revision 1.1471 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Add some AMD 17h/9xh devices from OpenBSD. Add AMD F17/Axh devices. Add some Xeon Scalable devices from OpenBSD. Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.14.2.1	03-Dec-2021	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1713: sys/dev/sdmmc/sdhc.c 1.110, 1.112 sys/dev/sdmmc/sdmmc_mem.c 1.74 sys/dev/pci/sdhc_pci.c 1.18 - Support 64bit BAR. - Use unsigned to avoid undefined behavior in hwrite[12]() and sdmmc_mem_sd_switch(). - Fix typo in comment.
1.19.2.2	22-Aug-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #336): sys/dev/pci/sdhc_pci.c: revision 1.21 Add quirk setting for some Intel eMMC devices. On some Intel eMMC controllers, the driver reports "autoconfiguration error: couldn't enable card: 60" even though they really have eMMC device. This change fixes the problem on some machines. It might be required more quirks for newer devices (or HS400 support). At least, this change fixes the problem on GIGABYTE MA10-ST0.
1.19.2.1	18-Jan-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #50): sys/dev/pci/sdhc_pci.c: revision 1.20 sys/dev/pci/pcidevs: revision 1.1472 sys/dev/pci/pcidevs: revision 1.1473 sys/dev/pci/pcidevs: revision 1.1474 sys/dev/pci/pcidevs: revision 1.1475 sys/dev/pci/pcidevs: revision 1.1476 Update Intel Elkhart Lake devices. Add Elkhart Lake Shared SRAM. Fix typo. s/SSC/SCC/. Fix typo. No functional change. add Aquantia (Marvell) AQC113 ethernet devices and the variants Add AMD Family 19h/1xh devices.
1.23	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.22	31-Dec-2021	andvar	s/interrups/interrupts/
1.21	05-May-2021	jdolecek	disable MSI for SiI3124 - interrupts don't seem to work on this old board when MSI is enabled, maybe because it's behind a PCI bridge PR kern/55115 by John D. Baker
1.20	25-Oct-2018	jdolecek	branches: 1.20.4; 1.20.16; release intr if pci_intr_establish_xname() fails
1.19	24-Oct-2018	jdolecek	detach the controller itself on shutdown; adjust to not detach already detached atabus/channel
1.18	22-Oct-2018	jdolecek	first disestablish interrupt, then release
1.17	22-Oct-2018	jdolecek	enable MSI support where available, blatantly copied from jmcneill's msk(4)
1.16	21-Jun-2017	jdolecek	branches: 1.16.4; 1.16.6; switch to pci_intr_establish_xname()
1.15	17-Jun-2017	jdolecek	match Adaptec 1220SA; that card actually uses Silicon Image 3132, and works well with the siisata(4) driver
1.14	14-Jul-2016	msaitoh	branches: 1.14.6; 1.14.10; KNF. No functional change.
1.13	29-Mar-2014	christos	branches: 1.13.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.12	17-Aug-2013	jakllsch	Revert previous; the SiI3114 is not even BAR/register compatible with siisata(4).
1.11	20-Oct-2012	matt	branches: 1.11.2; Recognize 3114.
1.10	30-Jan-2012	drochner	branches: 1.10.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.9	13-Nov-2010	uebayasi	branches: 1.9.8; 1.9.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.8	24-Feb-2010	dyoung	branches: 1.8.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.7	30-Jan-2010	jakllsch	branches: 1.7.2; Sprinkle __KERNEL_RCSID() into siisata(4). Slightly adjust some comment styling.
1.6	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.5	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.4	04-Jul-2009	jakllsch	Remove unused elements from siisata_softc. Clean up siisata pci attachment. noteable changes: - Use aprint_*_dev where appropriate. - Condense board match table and use C99 initializers. - Use aprint_verbose for extra info. Also, add 2009 to my copyright, as well as remove my local CVS tags.
1.3	17-Jun-2009	jakllsch	A few changes for siisata(4): - Support detachment. From KIYOHARA Takashi. - Add PCI detachment functionality (albeit not very interesting when the bus can not yet be rescanned). - Rework interrupt handlers to reduce near-duplicate code. Borrowed from ahcisata(4). - Attempt to make polled I/O work. Untested. - Fix formatting of some messages. - For always-polled commands, disable interrupt at slot level rather than port level. - Instead of busy-waiting indefinitely for completion of some commands move on after 31 seconds. Use cv_timedwait(9) instead of DELAY(9). - Use abstracted SATA FIS code. - Enable use of disks that don't respond with the standard signature.
1.2	16-Dec-2008	jakllsch	branches: 1.2.2; 1.2.4; In case the PCI command bits aren't already set, do so before performing an access that requires them to be. From KIYOHARA Takashi.
1.1	23-May-2008	jnemeth	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.12; 1.1.14; Import siisata(4) by Jonathan A. Kollasch. The siisata driver supports the Silicon Image SteelVine family of SATA-II controllers, interfacing the hardware with the ata(4) and atapi(4) sub- systems. The following controllers are supported by the siisata driver: Silicon Image SiI3124 4-port PCI/PCI-X Silicon Image SiI3132 2-port PCI-Express x1 Silicon Image SiI3531 1-port PCI-Express x1 SATA Native Command Queueing is not yet supported. Device hot swapping is not yet supported. Silicon Image's Software RAID is not yet supported by the ataraid(4) driver. Approved by: core (christos), releng (bouyer)
1.1.14.3	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ic/siisatavar.h: revision 1.4 sys/dev/pci/siisata_pci.c: revision 1.4 Remove unused elements from siisata_softc. Clean up siisata pci attachment. noteable changes: - Use aprint_*_dev where appropriate. - Condense board match table and use C99 initializers. - Use aprint_verbose for extra info. Also, add 2009 to my copyright, as well as remove my local CVS tags.
1.1.14.2	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ic/siisata.c: revision 1.3 sys/dev/ic/siisatareg.h: revision 1.3 sys/dev/ic/siisatavar.h: revision 1.3 sys/dev/pci/files.pci: revision 1.315 sys/dev/pci/siisata_pci.c: revision 1.3 A few changes for siisata(4): - Support detachment. From KIYOHARA Takashi. - Add PCI detachment functionality (albeit not very interesting when the bus can not yet be rescanned). - Rework interrupt handlers to reduce near-duplicate code. Borrowed from ahcisata(4). - Attempt to make polled I/O work. Untested. - Fix formatting of some messages. - For always-polled commands, disable interrupt at slot level rather than port level. - Instead of busy-waiting indefinitely for completion of some commands move on after 31 seconds. Use cv_timedwait(9) instead of DELAY(9). - Use abstracted SATA FIS code. - Enable use of disks that don't respond with the standard signature.
1.1.14.1	09-Jan-2009	snj	branches: 1.1.14.1.4; Pull up following revision(s) (requested by jakllsch in ticket #244): sys/dev/pci/siisata_pci.c: revision 1.2 In case the PCI command bits aren't already set, do so before performing an access that requires them to be. From KIYOHARA Takashi.
1.1.14.1.4.2	29-Apr-2011	matt	Add 3114 entry
1.1.14.1.4.1	21-Apr-2010	matt	sync to netbsd-5
1.1.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.1.8.2	23-Jun-2008	wrstuden	Add files to branch that were added on -current. After this, all that's left of update is to merge some changes that had conflicts.
1.1.8.1	23-May-2008	wrstuden	file siisata_pci.c was added on branch wrstuden-revivesa on 2008-06-23 05:02:13 +0000
1.1.4.2	04-Jun-2008	yamt	sync with head
1.1.4.1	23-May-2008	yamt	file siisata_pci.c was added on branch yamt-pf42 on 2008-06-04 02:05:14 +0000
1.1.2.3	17-Jan-2009	mjf	Sync with HEAD.
1.1.2.2	02-Jun-2008	mjf	Sync with HEAD.
1.1.2.1	23-May-2008	mjf	file siisata_pci.c was added on branch mjf-devfs2 on 2008-06-02 13:23:43 +0000
1.2.4.5	11-Mar-2010	yamt	sync with head
1.2.4.4	18-Jul-2009	yamt	sync with head.
1.2.4.3	20-Jun-2009	yamt	sync with head
1.2.4.2	04-May-2009	yamt	sync with head.
1.2.4.1	16-Dec-2008	yamt	file siisata_pci.c was added on branch yamt-nfs-mp on 2009-05-04 08:13:01 +0000
1.2.2.1	23-Jul-2009	jym	Sync with HEAD.
1.7.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.8.2.1	05-Mar-2011	rmind	sync with head
1.9.12.1	18-Feb-2012	mrg	merge to -current.
1.9.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.9.8.2	30-Oct-2012	yamt	sync with head
1.9.8.1	17-Apr-2012	yamt	sync with head
1.10.6.3	03-Dec-2017	jdolecek	update from HEAD
1.10.6.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.10.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.11.2.2	18-May-2014	rmind	sync with head
1.11.2.1	28-Aug-2013	rmind	sync with head
1.13.6.2	28-Aug-2017	skrll	Sync with HEAD
1.13.6.1	05-Oct-2016	skrll	Sync with HEAD
1.14.10.2	25-Jun-2017	snj	Pull up following revision(s) (requested by jdolecek in ticket #56): sys/dev/pci/mvsata_pci.c: revision 1.9 sys/dev/pci/siisata_pci.c: revision 1.16 switch to pci_intr_establish_xname()
1.14.10.1	25-Jun-2017	snj	Pull up following revision(s) (requested by jdolecek in ticket #55): share/man/man4/siisata.4: revision 1.6 sys/dev/pci/pcidevs: revision 1.1291 sys/dev/pci/siisata_pci.c: revision 1.15 add entry for Adaptec 1220SA -- match Adaptec 1220SA; that card actually uses Silicon Image 3132, and works well with the siisata(4) driver -- add Adaptec 1220SA
1.14.6.2	17-Jun-2017	jdolecek	revert, bad tree
1.14.6.1	17-Jun-2017	jdolecek	match Adaptec 1220SA; that card actually uses Silicon Image 3132, and works well with the siisata(4) driver
1.16.6.1	10-Jun-2019	christos	Sync with HEAD
1.16.4.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.20.16.1	13-May-2021	thorpej	Sync with HEAD.
1.20.4.1	09-May-2021	martin	Pull up following revision(s) (requested by dolecek in ticket #1270): sys/dev/pci/siisata_pci.c: revision 1.21 disable MSI for SiI3124 - interrupts don't seem to work on this old board when MSI is enabled, maybe because it's behind a PCI bridge PR kern/55115 by John D. Baker
1.27	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.26	19-Oct-2009	bouyer	branches: 1.26.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.25	15-May-2009	tsutsui	Split device_t/softc. No crash on: esiop0 at pci0 dev 11 function 0: Symbios Logic 53c875 (ultra-wide scsi) siop0 at pci0 dev 9 function 0: Symbios Logic 53c810a (fast scsi) siop0 at gsc0 hpa 0xf0830000 path 2/0/7 irq 3 ipl 6: NCR53C720 rev 2
1.24	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.23	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.22	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.21	16-Nov-2006	christos	branches: 1.21.52; 1.21.68; __unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	11-Dec-2005	christos	branches: 1.19.20; 1.19.22; merge ktrace-lwp.
1.18	28-Jun-2005	thorpej	branches: 1.18.2; Use ANSI function decls and static.
1.17	27-Feb-2005	perry	nuke trailing whitespace
1.16	04-Feb-2005	perry	de-__P
1.15	09-Apr-2003	thorpej	branches: 1.15.2; 1.15.10; 1.15.12; Use PAGE_SIZE rather than NBPG.
1.14	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.13	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.12	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.11	23-Apr-2002	bouyer	More copyright fixes, pointed out by Thomas. Thanks !
1.10	20-Apr-2002	bouyer	Move a few things around, so that ic/siop_common.c and pci/siop_pci_common.c can be compiled without including siopvar.h.
1.9	13-Nov-2001	lukem	add RCSID
1.8	15-May-2000	bouyer	branches: 1.8.6; 1.8.8; 1.8.10; - split siop_pci in attachements vs helper functions, for comming esiop - add a reset callback to enable PCI-specific features. This improve bandwith by a factor of 2 on my alpha with a 875 ! - sync copyrigth notice
1.7	10-May-2000	thorpej	Support 64-bit PCI memory mapping registers, as sometimes are used on the Symbios 53c896 controller. Also, prefer memory space over I/O space.
1.6	09-May-2000	bouyer	Ops, abort if we can't map the registers ! Pointed out by Matthias Drochner.
1.5	06-May-2000	soren	Print the usual interrupt line.
1.4	02-May-2000	bouyer	- do sync/wide negotiation - use a circular queue for the start slots, so that order has better chances to be preserved.
1.3	02-May-2000	nathanw	In siop_lookup_product(), actually check the PCI vendor ID as well as the product ID. Prevents false matches of other devices, such as the DEC 21050 ppb.
1.2	27-Apr-2000	bouyer	Fix copyrigth notice.
1.1	21-Apr-2000	bouyer	Snapshot of work in progress: new driver for the NCR 53c8xx SCSI controller (the name 'siop' is still being discussed, may change). Only basic disconnect/reselect for now, no sync/wide negotiation. Tested with 810A, 875 and 895 on i386 only. The bus-independant part should also be able to handle the 53c720 and 53c770. A new driver with enhanced script should appear for the 825/875/895 'soon'.
1.8.10.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.8.10.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.8.10.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.8.3	18-Oct-2002	nathanw	Catch up to -current.
1.8.8.2	20-Jun-2002	nathanw	Catch up to -current.
1.8.8.1	14-Nov-2001	nathanw	Catch up to -current.
1.8.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.6.1	15-May-2000	bouyer	file siop_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:37 +0000
1.15.12.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.15.12.1	12-Feb-2005	yamt	sync with head.
1.15.10.1	29-Apr-2005	kent	sync with -current
1.15.2.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.15.2.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.15.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.18.2.1	30-Dec-2006	yamt	sync with head.
1.19.22.2	10-Dec-2006	yamt	sync with head.
1.19.22.1	22-Oct-2006	yamt	sync with head
1.19.20.1	18-Nov-2006	ad	Sync with head.
1.21.68.2	23-Jul-2009	jym	Sync with HEAD.
1.21.68.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.21.52.2	11-Mar-2010	yamt	sync with head
1.21.52.1	16-May-2009	yamt	sync with head
1.26.4.1	05-Mar-2011	rmind	sync with head
1.39	04-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "use_pciclock".
1.38	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.37	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.36	17-Oct-2016	flxd	branches: 1.36.14; 1.36.16; Add 53C810AP for FreeAXP's emulation of a DEC KZPAA adapter.
1.35	29-Mar-2014	christos	branches: 1.35.6; 1.35.10; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.34	13-Nov-2010	uebayasi	branches: 1.34.8; 1.34.18; 1.34.22; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.33	26-Nov-2009	njoly	branches: 1.33.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.32	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.31	16-May-2009	tsutsui	Misc cleanup: - some KNF - u_intNN_t -> uintNN_t - wrap long lines and fix indent - remove return statements at the end of void functions - use __arraycount() XXX: many inconsistent aprint_error_dev(9) vs printf(9)
1.30	15-May-2009	tsutsui	Split device_t/softc. No crash on: esiop0 at pci0 dev 11 function 0: Symbios Logic 53c875 (ultra-wide scsi) siop0 at pci0 dev 9 function 0: Symbios Logic 53c810a (fast scsi) siop0 at gsc0 hpa 0xf0830000 path 2/0/7 irq 3 ipl 6: NCR53C720 rev 2
1.29	13-Feb-2009	bouyer	Make sure PCI_COMMAND_MASTER_ENABLE is set in PCI_COMMAND_STATUS_REG.
1.28	11-Jun-2008	kiyohara	branches: 1.28.4; 1.28.10; Support use PCI Clock(SF_CHIP_USEPCIC). We can set the value "use_pciclock" by prop_dictionary_set_bool().
1.27	10-Apr-2008	cegger	branches: 1.27.2; 1.27.4; 1.27.6; 1.27.8; use aprint_*_dev and device_xname
1.26	11-Dec-2005	christos	branches: 1.26.70; merge ktrace-lwp.
1.25	28-Jun-2005	thorpej	Use ANSI function decls and static.
1.24	27-Feb-2005	perry	nuke trailing whitespace
1.23	04-Feb-2005	perry	de-__P
1.22	10-Mar-2004	bouyer	branches: 1.22.8; 1.22.10; The 1010-66 always generate AIP values on the bus, regardless of the SCNTL4 settings. Disable AIP completely, as this confuse pre-Ultra160 drives. From FreeBSD.
1.21	25-Oct-2003	christos	return not break.
1.20	25-Oct-2003	christos	Fix uninitialized variable warnings
1.19	09-Apr-2003	thorpej	branches: 1.19.2; Use PAGE_SIZE rather than NBPG.
1.18	31-Jan-2003	thorpej	Use aprint_*().
1.17	04-May-2002	bouyer	branches: 1.17.4; It seems that the 1010-33 has a bug: it sometimes generate spurious SCSI gross errors for narrow transfers after a reselect. FreeBSD sym driver has a comment about this, but their workaround (disable SCSI gross error reports) doesn't work for me. Instead dissallow disconnect if the target is not wide (FreeBSD doens't allow disconnect until the target has been fully probed, which is why they may not have noticed my problem).
1.16	26-Apr-2002	thorpej	53c1010 chips are ultra3, not ultra2.
1.15	23-Apr-2002	bouyer	More copyright fixes, pointed out by Thomas. Thanks !
1.14	23-Apr-2002	bouyer	- factor out parts of (e)siop_attach() to siop_common_attach() - Add support for DT transfers (aka Ultra/160) in esiop Note that DT transfers are not enabled for 53c1010-33 rev 0 yet; if I trust FreeBSD it has a bug which prevent them to do DT properly. From the same source there may be issues with some revs of 53c1010-66.
1.13	23-Apr-2002	bouyer	Enable software LED control based on LED0 feature, not #define. For now, set the LED0 feature if SIOP_SYMLED is defined in siop_pci_common.c. From Jason R Thorpe.
1.12	23-Apr-2002	bouyer	Set SF_CHIP_LEDC for chips that support it. Based on patch sent in private mail by Jason R Thorpe.
1.11	20-Apr-2002	bouyer	Move a few things around, so that ic/siop_common.c and pci/siop_pci_common.c can be compiled without including siopvar.h.
1.10	08-Feb-2002	briggs	Dup 53c1010-33 entry for 53c1010-66 (PCI_PRODUCT_SYMBIOS_1010_2).
1.9	13-Nov-2001	lukem	add RCSID
1.8	15-Mar-2001	bouyer	branches: 1.8.2; 1010 has clock quadrupler but without PPL, so use SF_CHIP_DBLR.
1.7	12-Mar-2001	bouyer	Preliminary support for SYM53C1010-33 (at 80MB/s only for now), based on patches sent by Matthias Drochner <M.Drochner@fz-juelich.de> and Kenneth R Westerback <kwesterback@home.com>
1.6	10-Jan-2001	thorpej	branches: 1.6.2; Add the 53c1510d and 53c895a. 1510d confirmed working by Mark Davies <mark@MCS.VUW.AC.NZ>.
1.5	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.4	29-Nov-2000	matt	Fix c896 support. If the first mem bar is 64bit, then the second one will be too. And offset for the 2nd being after the 1st by a 64-bit bar, not 32-bit.
1.3	23-Oct-2000	bouyer	branches: 1.3.2; Add ram size to struct properties, and pass it to the bus independant part.
1.2	25-May-2000	bouyer	branches: 1.2.4; Separate the sheduler from the main script, allocate another DMA-safe memory page for the sheduler. Put the main script in the on-chip RAM when available. Avoid a null-pointer dereference when DSA is invalid.
1.1	15-May-2000	bouyer	- split siop_pci in attachements vs helper functions, for comming esiop - add a reset callback to enable PCI-specific features. This improve bandwith by a factor of 2 on my alpha with a 875 ! - sync copyrigth notice
1.2.4.3	29-Jan-2002	he	Pull up revisions 1.6-1.8 (requested by bouyer): Add support for the 53c1010-33, and bring documentation up to date.
1.2.4.2	20-Mar-2001	he	Pull up revision 1.4 (requested by matt): Fix c896 support related to PCI configuration and handling of 64-bit BARs.
1.2.4.1	15-Dec-2000	he	Pull up revision 1.3 (requested by bouyer): Speed improvements to the siop driver, and add tagged queueing support. As a side effect, better handling of some not so common phase or message sequences.
1.3.2.7	27-Mar-2001	bouyer	Sync with HEAD.
1.3.2.6	12-Mar-2001	bouyer	Sync with HEAD.
1.3.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.3.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.3.2.3	08-Dec-2000	bouyer	Sync with HEAD.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	23-Oct-2000	bouyer	file siop_pci_common.c was added on branch thorpej_scsipi on 2000-11-20 11:42:37 +0000
1.6.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.6.2.3	28-Feb-2002	nathanw	Catch up to -current.
1.6.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.6.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.8.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.8.2.2	16-Mar-2002	jdolecek	Catch up with -current.
1.8.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.17.4.1	17-Mar-2005	tron	Pull up revision 1.22 (requested by bouyer in ticket #1748): The 1010-66 always generate AIP values on the bus, regardless of the SCNTL4 settings. Disable AIP completely, as this confuse pre-Ultra160 drives. From FreeBSD.
1.19.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.19.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.19.2.4	04-Feb-2005	skrll	Sync with HEAD.
1.19.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.19.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.19.2.1	03-Aug-2004	skrll	Sync with HEAD
1.22.10.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.22.10.1	12-Feb-2005	yamt	sync with head.
1.22.8.1	29-Apr-2005	kent	sync with -current
1.26.70.2	29-Jun-2008	mjf	Sync with HEAD.
1.26.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.27.8.1	18-Jun-2008	simonb	Sync with head.
1.27.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.27.4.3	11-Mar-2010	yamt	sync with head
1.27.4.2	16-May-2009	yamt	sync with head
1.27.4.1	04-May-2009	yamt	sync with head.
1.27.2.1	17-Jun-2008	yamt	sync with head.
1.28.10.2	23-Jul-2009	jym	Sync with HEAD.
1.28.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.28.4.1	03-Mar-2009	skrll	Sync with HEAD.
1.33.4.1	05-Mar-2011	rmind	sync with head
1.34.22.1	18-May-2014	rmind	sync with head
1.34.18.2	03-Dec-2017	jdolecek	update from HEAD
1.34.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.34.8.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.35.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.35.6.1	05-Dec-2016	skrll	Sync with HEAD
1.36.16.1	10-Jun-2019	christos	Sync with HEAD
1.36.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.9	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.8	16-May-2009	tsutsui	Misc cleanup: - some KNF - u_intNN_t -> uintNN_t - wrap long lines and fix indent - remove return statements at the end of void functions - use __arraycount() XXX: many inconsistent aprint_error_dev(9) vs printf(9)
1.7	11-Dec-2005	christos	branches: 1.7.74; 1.7.90; merge ktrace-lwp.
1.6	27-Feb-2005	perry	nuke trailing whitespace
1.5	04-Feb-2005	perry	de-__P
1.4	23-Apr-2002	bouyer	branches: 1.4.10; 1.4.18; 1.4.20; More copyright fixes, pointed out by Thomas. Thanks !
1.3	20-Apr-2002	bouyer	Move a few things around, so that ic/siop_common.c and pci/siop_pci_common.c can be compiled without including siopvar.h.
1.2	23-Oct-2000	bouyer	branches: 1.2.2; 1.2.4; 1.2.6; Add ram size to struct properties, and pass it to the bus independant part.
1.1	15-May-2000	bouyer	branches: 1.1.4; - split siop_pci in attachements vs helper functions, for comming esiop - add a reset callback to enable PCI-specific features. This improve bandwith by a factor of 2 on my alpha with a 875 ! - sync copyrigth notice
1.1.4.1	15-Dec-2000	he	Pull up revision 1.2 (requested by bouyer): Speed improvements to the siop driver, and add tagged queueing support. As a side effect, better handling of some not so common phase or message sequences.
1.2.6.1	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.4.1	20-Jun-2002	nathanw	Catch up to -current.
1.2.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.1	23-Oct-2000	bouyer	file siop_pci_common.h was added on branch thorpej_scsipi on 2000-11-20 11:42:37 +0000
1.4.20.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.20.1	12-Feb-2005	yamt	sync with head.
1.4.18.1	29-Apr-2005	kent	sync with -current
1.4.10.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.10.1	04-Feb-2005	skrll	Sync with HEAD.
1.7.90.1	23-Jul-2009	jym	Sync with HEAD.
1.7.74.2	11-Mar-2010	yamt	sync with head
1.7.74.1	16-May-2009	yamt	sync with head
1.9	08-Aug-2023	mrg	ensure all data passed to copyout() has been initialised. found by GCC 12.
1.8	21-Sep-2021	christos	use KAUTH_MACHDEP_UNMANAGEDMEM instead of KAUTH_GENERIC_ISSUSER like all the other framebuffers.
1.7	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.6	24-Apr-2021	thorpej	branches: 1.6.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.5	26-Jan-2014	bouyer	branches: 1.5.46; Complete WSDISPLAYIO_SMODE/WSDISPLAYIO_GMODE support. In sisfb_mmap(), allow mapping the framebuffer and mmio and io registers when in WSDISPLAYIO_MODE_MAPPED mode. This can be used by X11.
1.4	24-Jan-2014	bouyer	Snapshot of work in progress: Add support xorg wsfb driver. Tested on evbmips/loongson
1.3	18-Jan-2014	bouyer	Request a 80x25 tty emulation, which is more likely to work than 80x30 on a 640x400 VGA.
1.2	30-Jan-2012	drochner	branches: 1.2.2; 1.2.6; 1.2.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.1	27-Aug-2011	bouyer	branches: 1.1.2; 1.1.6; Add a basic driver for the SIS315 Pro frame buffer found on some loogson2-based systems, from OpenBSD. At this time it can only map the framebuffer and attaches wscons, using whatever resolution was configured by the firmware. It doens't support any accelaration or mode change.
1.1.6.1	18-Feb-2012	mrg	merge to -current.
1.1.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.2.1	17-Apr-2012	yamt	sync with head
1.2.10.1	18-May-2014	rmind	sync with head
1.2.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.2.1	07-Feb-2014	sborrill	Pull up the following revisions(s) (requested by bouyer in ticket #1019): sys/dev/pci/sisfb.c: revision 1.3 sys/arch/evbmips/conf/LOONGSON: revision 1.22 Make sysinst much more useable when used with the VGA display on LOONGSON.
1.5.46.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.6.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	27-Aug-2011	bouyer	Add a basic driver for the SIS315 Pro frame buffer found on some loogson2-based systems, from OpenBSD. At this time it can only map the framebuffer and attaches wscons, using whatever resolution was configured by the firmware. It doens't support any accelaration or mode change.
1.38	05-Feb-2019	mrg	add missing break; after "100NEW" configuration. don't also set "133OLD", since the 100NEW cards don't support 133.
1.37	23-Jun-2018	maxv	branches: 1.37.2; constify
1.36	31-May-2014	christos	branches: 1.36.20; 1.36.26; PR/48858: Roy Bixler: siside0:0:1:0: lost interrupt, submitter claims this makes the cdrom work.
1.35	24-May-2014	abs	Add in missing space after "96X UDMA%d". Changes "96X UDMA6746" to "96X UDMA6 746"
1.34	07-Oct-2013	jakllsch	branches: 1.34.2; Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.33	31-Jul-2012	bouyer	branches: 1.33.2; 1.33.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.32	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.31	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.30	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.29	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.28	24-May-2011	joerg	branches: 1.28.4; 1.28.10; Properly use format string
1.27	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.26	05-Nov-2010	jakllsch	branches: 1.26.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.25	19-Oct-2009	bouyer	branches: 1.25.2; 1.25.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.24	12-Mar-2009	bsh	Turn off interrupt disable bit in command status reg. This is necessary to make S-ATA drives work on some motherboards with SiS964 including my ASUS P4S800D.
1.23	18-Mar-2008	cube	branches: 1.23.4; 1.23.12; 1.23.18; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.22	09-Feb-2007	ad	branches: 1.22.20; 1.22.36; 1.22.40; Merge newlock2 to head.
1.21	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.20	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.19	26-Jun-2006	xtraeme	branches: 1.19.4; 1.19.6; Add support for SIS 660/760 IDE Controllers and SIS 181/182 SATA Controllers (from obsd).
1.18	11-Dec-2005	christos	branches: 1.18.4; 1.18.8; 1.18.16; merge ktrace-lwp.
1.17	16-Jun-2005	bouyer	branches: 1.17.2; Add quirk for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.16	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.15	24-May-2005	lukem	add __KERNEL_RCSID
1.14	27-Feb-2005	perry	branches: 1.14.2; nuke trailing whitespace
1.13	21-Aug-2004	thorpej	branches: 1.13.4; 1.13.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.12	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.11	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.10	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.9	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.8	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.7	21-Jul-2004	mycroft	branches: 1.7.2; Add the SiS 741, per PR 26394.
1.6	22-Apr-2004	skd	Added initial support for SIS 180 SATA controller. Patterned after VIA 8237 support.
1.5	03-Jan-2004	thorpej	branches: 1.5.2; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.4	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.3	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.5.2.1	23-Jul-2004	he	branches: 1.5.2.1.2; Pull up revision 1.7 (requested by mycroft in ticket #691): Add support for SiS 741. Fixes PR#26394.
1.5.2.1.2.1	05-Jul-2005	riz	Pull up revision 1.17 via patch (requested by bouyer in ticket #2032): Add quick for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.7.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.7.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.7.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.7.2.2	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	21-Jul-2004	skrll	file siside.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.13.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.13.4.1	29-Apr-2005	kent	sync with -current
1.14.2.1	26-Jun-2005	tron	Pull up revision 1.17 (requested by bouyer in ticket #498): Add quirk for SIS 96x that masquerade as a SIS503. Based on patch sent by Robert Fuller in kern/30431, information found in the linux kernel. While here, add SIS 965 to the table (from FreeBSD).
1.17.2.3	24-Mar-2008	yamt	sync with head.
1.17.2.2	26-Feb-2007	yamt	sync with head.
1.17.2.1	30-Dec-2006	yamt	sync with head.
1.18.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.18.8.1	11-Aug-2006	yamt	sync with head
1.18.4.1	09-Sep-2006	rpaulo	sync with head
1.19.6.2	10-Dec-2006	yamt	sync with head.
1.19.6.1	22-Oct-2006	yamt	sync with head
1.19.4.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.19.4.1	18-Nov-2006	ad	Sync with head.
1.22.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.22.36.1	24-Mar-2008	keiichi	sync with head.
1.22.20.1	23-Mar-2008	matt	sync with HEAD
1.23.18.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.23.12.1	28-Apr-2009	skrll	Sync with HEAD.
1.23.4.2	11-Mar-2010	yamt	sync with head
1.23.4.1	04-May-2009	yamt	sync with head.
1.25.4.3	31-May-2011	rmind	sync with head
1.25.4.2	21-Apr-2011	rmind	sync with head
1.25.4.1	05-Mar-2011	rmind	sync with head
1.25.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.26.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.28.10.1	03-Jun-2014	msaitoh	Pull up following revision(s) (requested by abs in ticket #1069): sys/dev/pci/siside.c: revision 1.35 Add in missing space after "96X UDMA%d". Changes "96X UDMA6746" to "96X UDMA6 746"
1.28.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.28.4.1	30-Oct-2012	yamt	sync with head
1.33.4.1	18-May-2014	rmind	sync with head
1.33.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.33.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.34.2.1	10-Aug-2014	tls	Rebase.
1.36.26.1	25-Jun-2018	pgoyette	Sync with HEAD
1.36.20.1	01-Aug-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1314): sys/dev/pci/siside.c: revision 1.38 add missing break; after "100NEW" configuration. don't also set "133OLD", since the 100NEW cards don't support 133.
1.37.2.1	10-Jun-2019	christos	Sync with HEAD
1.29	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.28	31-Jul-2012	bouyer	branches: 1.28.2; 1.28.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.27	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.26	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.25	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.24	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.23	08-Apr-2012	jakllsch	As suggested in the previous commit, the overwrite of the pci_attach_args is a bug. Although I have not tested on real hardware, Linux's pata_sl82c105.c only looks at the revision of the PCI bridge.
1.22	04-Apr-2011	dyoung	branches: 1.22.4; 1.22.8; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.21	05-Nov-2010	jakllsch	branches: 1.21.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.20	28-Apr-2008	martin	branches: 1.20.20; 1.20.22; Remove clause 3 and 4 from TNF licenses
1.19	18-Mar-2008	cube	branches: 1.19.2; 1.19.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.18	26-Apr-2007	garbled	branches: 1.18.8; 1.18.24; 1.18.28; Remove the PCIIDE_SLIDE_SETIRQA hack. Recent revalations have uncovered the fact that it's useless.
1.17	20-Mar-2007	garbled	Put a small hack in slide.c disabled by default. PCIIDE_SLIDE_SETIRQA. This sets IDE_IRQA in the IDE Control/Status Register. This is needed on prep machines with Motorola RAVEN host bridges, as they wire up the ide device oddly.
1.16	09-Feb-2007	ad	branches: 1.16.2; 1.16.6; 1.16.8; 1.16.10; Merge newlock2 to head.
1.15	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.14	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.13	03-Sep-2006	christos	branches: 1.13.2; 1.13.4; add missing initializer
1.12	11-Dec-2005	christos	branches: 1.12.4; 1.12.8; merge ktrace-lwp.
1.11	24-May-2005	lukem	branches: 1.11.2; add __KERNEL_RCSID
1.10	27-Feb-2005	perry	nuke trailing whitespace
1.9	21-Aug-2004	thorpej	branches: 1.9.4; 1.9.6; atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.8	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.7	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.6	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.5	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.4	03-Jan-2004	thorpej	branches: 1.4.4; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.3	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.4.4.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.4.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.4.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.4.4.2	03-Aug-2004	skrll	Sync with HEAD
1.4.4.1	03-Jan-2004	skrll	file slide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.9.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.4.1	29-Apr-2005	kent	sync with -current
1.11.2.3	24-Mar-2008	yamt	sync with head.
1.11.2.2	26-Feb-2007	yamt	sync with head.
1.11.2.1	30-Dec-2006	yamt	sync with head.
1.12.8.1	03-Sep-2006	yamt	sync with head.
1.12.4.1	09-Sep-2006	rpaulo	sync with head
1.13.4.2	10-Dec-2006	yamt	sync with head.
1.13.4.1	22-Oct-2006	yamt	sync with head
1.13.2.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.13.2.1	18-Nov-2006	ad	Sync with head.
1.16.10.1	29-Mar-2007	reinoud	Pullup to -current
1.16.8.1	11-Jul-2007	mjf	Sync with head.
1.16.6.2	27-May-2007	ad	Sync with head.
1.16.6.1	10-Apr-2007	ad	Sync with head.
1.16.2.2	07-May-2007	yamt	sync with head.
1.16.2.1	24-Mar-2007	yamt	sync with head.
1.18.28.2	02-Jun-2008	mjf	Sync with HEAD.
1.18.28.1	03-Apr-2008	mjf	Sync with HEAD.
1.18.24.1	24-Mar-2008	keiichi	sync with head.
1.18.8.1	23-Mar-2008	matt	sync with HEAD
1.19.4.1	16-May-2008	yamt	sync with head.
1.19.2.1	18-May-2008	yamt	sync with head.
1.20.22.2	21-Apr-2011	rmind	sync with head
1.20.22.1	05-Mar-2011	rmind	sync with head
1.20.20.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.21.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.22.8.1	29-Apr-2012	mrg	sync to latest -current.
1.22.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.22.4.2	30-Oct-2012	yamt	sync with head
1.22.4.1	17-Apr-2012	yamt	sync with head
1.28.4.1	18-May-2014	rmind	sync with head
1.28.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.28.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.2	05-Oct-2017	msaitoh	Use pcireg.h's macro. No functional change.
1.1	09-Nov-2010	skrll	branches: 1.1.6; 1.1.20; Update sti(4) from OpenBSD bringing across support for sti @ pci. Thanks to Adam Hoka for testing.
1.1.20.1	03-Dec-2017	jdolecek	update from HEAD
1.1.6.2	05-Mar-2011	rmind	sync with head
1.1.6.1	09-Nov-2010	rmind	file sti_pci.c was added on branch rmind-uvmplock on 2011-03-05 20:53:58 +0000
1.27	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.26	31-Jul-2012	bouyer	branches: 1.26.2; 1.26.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.25	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.24	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.23	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.22	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.21	04-Apr-2011	dyoung	branches: 1.21.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.20	05-Nov-2010	jakllsch	branches: 1.20.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.19	26-May-2008	nisimura	branches: 1.19.18; 1.19.20; - assign 2 clause TNF license to the files cited as my ownership. - abandon and remove my copyright notice from the three files were copied in order to adapt HW I don't have; pvr.c, pm.c and stic.c
1.18	18-Mar-2008	cube	branches: 1.18.2; 1.18.4; 1.18.6; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.17	09-Feb-2007	ad	branches: 1.17.20; 1.17.36; 1.17.40; Merge newlock2 to head.
1.16	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.15	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.14	11-Dec-2005	christos	branches: 1.14.20; 1.14.22; merge ktrace-lwp.
1.13	24-May-2005	lukem	branches: 1.13.2; add __KERNEL_RCSID
1.12	27-Feb-2005	perry	nuke trailing whitespace
1.11	21-Aug-2004	nisimura	branches: 1.11.4; 1.11.6; Redo the previous adjustment work.
1.10	21-Aug-2004	nisimura	Remove fluctuation of written name in alphabetical representation of copyright holder.
1.9	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.8	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.7	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.6	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.5	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.4	03-Jan-2004	thorpej	branches: 1.4.2; 1.4.4; More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.3	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.2	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.1	31-Oct-2003	nisimura	Added stpcide(4) driver for STMicroelectronics STPC IDE controllers.
1.4.4.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.4.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.4.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.4.4	18-Sep-2004	skrll	Sync with HEAD.
1.4.4.3	25-Aug-2004	skrll	Sync with HEAD.
1.4.4.2	03-Aug-2004	skrll	Sync with HEAD
1.4.4.1	03-Jan-2004	skrll	file stpcide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.4.2.2	22-Aug-2004	tron	Pull up revision 1.11 (requested by nisimura in ticket #779): Redo the previous adjustment work.
1.4.2.1	22-Aug-2004	tron	Pull up revision 1.10 (requested by nisimura in ticket #779): Remove fluctuation of written name in alphabetical representation of copyright holder.
1.11.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.11.4.1	29-Apr-2005	kent	sync with -current
1.13.2.3	24-Mar-2008	yamt	sync with head.
1.13.2.2	26-Feb-2007	yamt	sync with head.
1.13.2.1	30-Dec-2006	yamt	sync with head.
1.14.22.2	10-Dec-2006	yamt	sync with head.
1.14.22.1	22-Oct-2006	yamt	sync with head
1.14.20.2	06-Feb-2007	ad	Quiten noisy boot messages.
1.14.20.1	18-Nov-2006	ad	Sync with head.
1.17.40.2	02-Jun-2008	mjf	Sync with HEAD.
1.17.40.1	03-Apr-2008	mjf	Sync with HEAD.
1.17.36.1	24-Mar-2008	keiichi	sync with head.
1.17.20.1	23-Mar-2008	matt	sync with HEAD
1.18.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.18.4.1	04-May-2009	yamt	sync with head.
1.18.2.1	04-Jun-2008	yamt	sync with head
1.19.20.2	21-Apr-2011	rmind	sync with head
1.19.20.1	05-Mar-2011	rmind	sync with head
1.19.18.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.20.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.21.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.21.4.1	30-Oct-2012	yamt	sync with head
1.26.4.1	18-May-2014	rmind	sync with head
1.26.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.26.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.63	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.62	08-Jan-2024	chs	mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.61	07-Aug-2021	thorpej	branches: 1.61.6; Merge thorpej-cfargs2.
1.60	24-Apr-2021	thorpej	branches: 1.60.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.59	06-Feb-2021	isaki	branches: 1.59.2; Add a missing mutex_spin_exit().
1.58	19-Apr-2020	isaki	branches: 1.58.2; Fix round_blocksize not to return 0.
1.57	28-Oct-2019	joerg	branches: 1.57.6; Fix operand priority of bit op vs ternary operator.
1.56	08-Jun-2019	isaki	branches: 1.56.2; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.55	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.54	16-Mar-2019	isaki	branches: 1.54.2; Use C99 style struct initializer to audio_hw_if.
1.53	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.52	01-Jun-2017	chs	branches: 1.52.8; 1.52.10; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.51	14-Jul-2016	msaitoh	- Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.50	29-Mar-2014	christos	branches: 1.50.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.49	18-Mar-2014	riastradh	Merge riastradh-drm2 to HEAD.
1.48	17-Jul-2013	soren	A few "isa" files are so tightly integrated into the x86 platform code as to not really be part of the optional isa bus autoconf machinery. Allows configuring a kernel like so: include "arch/amd64/conf/GENERIC" no isa
1.47	27-Oct-2012	chs	branches: 1.47.2; 1.47.8; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.46	24-Nov-2011	mrg	branches: 1.46.8; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.45	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.44	26-Nov-2009	njoly	branches: 1.44.12; 1.44.14; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.43	18-Aug-2009	drochner	comparision with a string literal by "==" can't work (found by gcc-4.4.1), replace with strcmp()
1.42	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.41	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.40	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.39	28-Apr-2008	martin	branches: 1.39.12; 1.39.14; Remove clause 3 and 4 from TNF licenses
1.38	10-Apr-2008	cegger	branches: 1.38.2; 1.38.4; use aprint_*_dev and device_xname
1.37	19-Oct-2007	ad	branches: 1.37.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.36	04-Mar-2007	christos	branches: 1.36.2; 1.36.14; 1.36.16; 1.36.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.35	16-Nov-2006	christos	branches: 1.35.4; 1.35.6; __unused removal on arguments; approved by core.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	03-Sep-2006	christos	branches: 1.33.2; 1.33.4; add missing initializer
1.32	08-Mar-2006	lukem	Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.31	11-Dec-2005	christos	branches: 1.31.4; 1.31.6; 1.31.8; 1.31.10; merge ktrace-lwp.
1.30	28-Jun-2005	thorpej	branches: 1.30.2; Use ANSI function decls and static.
1.29	30-Apr-2005	hannken	sv_set_params(): avoid null pointer dereference when `setmode' is 0. Reviewed by: Lennart Augustsson <augustss@netbsd.org>
1.28	27-Feb-2005	perry	branches: 1.28.2; nuke trailing whitespace
1.27	15-Jan-2005	kent	branches: 1.27.2; 1.27.4; ansify and KNF
1.26	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.25	29-Oct-2004	yamt	branches: 1.25.2; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.24	09-Jul-2004	mycroft	Clean up. Attempt to make these safer against stray interrupts.
1.23	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.22	03-May-2003	wiz	branches: 1.22.2; DMA, not dma nor Dma.
1.21	22-Feb-2003	tsutsui	{HZ,hz,hZ} -> Hz
1.20	20-Feb-2003	hannken	The SonicVibes DMA is broken and only works on 24-bit addresses. As long as bus_dmamem_alloc_range() is missing we use the ISA dma tag on i386. The iorange for the dma registers now gets allocated near the used range. Chances are high, that this range is valid on this PCI bus. This one is really ugly :-) Approved by: Lennart Augustsson <augustss@netbsd.org>
1.19	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.18	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.17	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.16	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.15	13-Nov-2001	lukem	add RCSID
1.14	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.13	19-Jul-2001	kleink	branches: 1.13.2; Hint BUS_DMA_READ and BUS_DMA_WRITE.
1.12	28-Dec-2000	sommerfeld	branches: 1.12.2; 1.12.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.11	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.10	01-Nov-1999	augustss	branches: 1.10.6; Make little endian and big endian version of all functions that deal with 16 bit samples. Let the drivers decide which ones to use.
1.9	10-Jul-1999	kleink	branches: 1.9.2; 1.9.4; 1.9.6; Fix a buglet in the driver's DMA buffer list ipmlementation, causing the next- in-list (wrong) descriptor structure of the buffer in question to be freed.
1.8	14-Apr-1999	jonathan	initialize 'p' for lame gcc-2.7.2.2 flow analysis.
1.7	22-Mar-1999	mycroft	branches: 1.7.2; Make 48KHz output work.
1.6	19-Feb-1999	mycroft	This is an utterly revolting hack to work around the previously mentioned lossage. On the Alpha, we force the buffers to be allocated through the SGMAP so that the PCI bus addresses are low enough. At least it's only one line of code...
1.5	18-Feb-1999	mycroft	Pass the correct address (PCI bus address, not PA) to the card. Also fix some typos and add more debugging printf()s. This would work on the Alpha, but the card I have appears to have the upper address bits chopped off, and the ring buffer gets mapped using the DGMAP, which uses the upper bits. Boom.
1.4	18-Feb-1999	mycroft	spaces -> tabs
1.3	18-Feb-1999	mycroft	A few things: * Use the trigger interface. * Permit different encodings for record and playback. * Set AUDIO_PROP_INDEPENDENT. * Fix the mmap(2) hole again. * Use 16-bit mode for a-law and u-law playback.
1.2	17-Feb-1999	mycroft	Pass the direction to the allocm and round_buffersize methods. Some drivers need this to properly allocate DMAable memory.
1.1	10-Dec-1998	augustss	Add a driver for the S3 SonicVibes chip. From OpenBSD and Constantine Paul Sapuntzakis (csapuntz@cvs.openbsd.org) with some changes by me. XXX The driver still needs work.
1.7.2.1	14-Apr-1999	jonathan	branches: 1.7.2.1.2; Pull up 1.8 delta (initialize 'p' for lame gcc-2.7.2.2 flow analysis) to 1.4.
1.7.2.1.2.2	02-Aug-1999	thorpej	Update from trunk.
1.7.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.9.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.9.4.1	15-Nov-1999	fvdl	Sync with -current
1.9.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.9.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.10.6.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.12.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.12.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.12.4.1	03-Aug-2001	lukem	update to -current
1.12.2.4	18-Oct-2002	nathanw	Catch up to -current.
1.12.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.12.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.12.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.13.2.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.22.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.22.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.22.2.5	17-Jan-2005	skrll	Sync with HEAD.
1.22.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.22.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.22.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.22.2.1	03-Aug-2004	skrll	Sync with HEAD
1.25.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.25.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.27.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.27.2.1	29-Apr-2005	kent	sync with -current
1.28.2.1	01-May-2005	tron	Pull up revision 1.29 (requested by hannken in ticket #222): sv_set_params(): avoid null pointer dereference when `setmode' is 0. Reviewed by: Lennart Augustsson <augustss@netbsd.org>
1.30.2.4	27-Oct-2007	yamt	sync with head.
1.30.2.3	03-Sep-2007	yamt	sync with head.
1.30.2.2	30-Dec-2006	yamt	sync with head.
1.30.2.1	21-Jun-2006	yamt	sync with head.
1.31.10.1	19-Apr-2006	elad	sync with head.
1.31.8.2	03-Sep-2006	yamt	sync with head.
1.31.8.1	13-Mar-2006	yamt	sync with head.
1.31.6.1	22-Apr-2006	simonb	Sync with head.
1.31.4.1	09-Sep-2006	rpaulo	sync with head
1.33.4.2	10-Dec-2006	yamt	sync with head.
1.33.4.1	22-Oct-2006	yamt	sync with head
1.33.2.1	18-Nov-2006	ad	Sync with head.
1.35.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.35.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.36.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.36.16.1	06-Nov-2007	matt	sync with HEAD
1.36.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.36.2.1	23-Oct-2007	ad	Sync with head.
1.37.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.38.4.4	11-Mar-2010	yamt	sync with head
1.38.4.3	19-Aug-2009	yamt	sync with head.
1.38.4.2	16-May-2009	yamt	sync with head
1.38.4.1	16-May-2008	yamt	sync with head.
1.38.2.1	18-May-2008	yamt	sync with head.
1.39.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.39.12.2	12-Dec-2008	ad	Checkpoint work in progress.
1.39.12.1	11-Dec-2008	ad	Checkpoint work in progress.
1.44.14.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.44.14.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.44.12.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.44.12.2	30-Oct-2012	yamt	sync with head
1.44.12.1	17-Apr-2012	yamt	sync with head
1.46.8.3	03-Dec-2017	jdolecek	update from HEAD
1.46.8.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.46.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.47.8.1	23-Jul-2013	riastradh	sync with HEAD
1.47.2.2	18-May-2014	rmind	sync with head
1.47.2.1	28-Aug-2013	rmind	sync with head
1.50.6.2	28-Aug-2017	skrll	Sync with HEAD
1.50.6.1	05-Oct-2016	skrll	Sync with HEAD
1.52.10.3	21-Apr-2020	martin	Sync with HEAD
1.52.10.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.52.10.1	10-Jun-2019	christos	Sync with HEAD
1.52.8.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.54.2.3	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.54.2.2	29-Apr-2019	isaki	Adapt to audio2. - Drop INDEPENDENT property. Both play and rec seems to share the sample rate (however, they require different calculations?). - XXX It's better to modify frequency list more strictly. - Remove obsoleted and empty methods.
1.54.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.56.2.1	27-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #492): sys/dev/pci/sv.c: revision 1.57 sys/dev/pci/eso.c: revision 1.72 sys/dev/ieee1394/fwohci.c: revision 1.144 Fix operand priority of bit op vs ternary operator.
1.57.6.1	20-Apr-2020	bouyer	Sync with HEAD
1.58.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.59.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.60.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.61.6.1	14-Jan-2024	martin	Pull up following revision(s) (requested by chs in ticket #542): sys/dev/pci/yds.c: revision 1.69 sys/dev/pci/cmpci.c: revision 1.61 sys/dev/pci/sv.c: revision 1.62 sys/dev/pci/fms.c: revision 1.50 mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	27-Feb-2005	perry	nuke trailing whitespace
1.2	15-Jan-2005	kent	branches: 1.2.2; 1.2.4; ansify and KNF
1.1	10-Dec-1998	augustss	branches: 1.1.42; Add a driver for the S3 SonicVibes chip. From OpenBSD and Constantine Paul Sapuntzakis (csapuntz@cvs.openbsd.org) with some changes by me. XXX The driver still needs work.
1.1.42.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.42.1	17-Jan-2005	skrll	Sync with HEAD.
1.2.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.2.1	29-Apr-2005	kent	sync with -current
1.8	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.7	23-Nov-2011	jmcneill	branches: 1.7.8; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.6	11-Dec-2005	christos	branches: 1.6.28; 1.6.88; 1.6.110; 1.6.112; merge ktrace-lwp.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	03-May-2003	wiz	branches: 1.4.2; 1.4.10; 1.4.12; DMA, not dma nor Dma.
1.3	18-Feb-1999	mycroft	spaces -> tabs
1.2	18-Feb-1999	mycroft	A few things: * Use the trigger interface. * Permit different encodings for record and playback. * Set AUDIO_PROP_INDEPENDENT. * Fix the mmap(2) hole again. * Use 16-bit mode for a-law and u-law playback.
1.1	10-Dec-1998	augustss	Add a driver for the S3 SonicVibes chip. From OpenBSD and Constantine Paul Sapuntzakis (csapuntz@cvs.openbsd.org) with some changes by me. XXX The driver still needs work.
1.4.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.10.1	29-Apr-2005	kent	sync with -current
1.4.2.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.112.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.6.110.2	30-Oct-2012	yamt	sync with head
1.6.110.1	17-Apr-2012	yamt	sync with head
1.6.88.1	11-Dec-2008	ad	Checkpoint work in progress.
1.6.28.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.7.8.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.24	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.23	06-Jun-2018	kamil	branches: 1.23.2; Enhance the syntax in order to be acceptable by Clang Replace "((crv == 0))" with "(crv == 0)", as the intention is to check the value, whether it is equal to 0. This fixes kernel=GENERIC build under MKLLVM=yes. Sponsored by <The NetBSD Foundation>
1.22	01-Jun-2018	macallan	set ATACH_DMA_BEFORE_CMD
1.21	01-Jun-2018	macallan	another G5-specific fix - do a 32bit read of the status register before checking for channel interrupts. No more interrupt storms. Adapted from FreeBSD
1.20	07-Oct-2017	jdolecek	branches: 1.20.2; Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.19	29-Mar-2014	christos	branches: 1.19.18; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.18	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.17	28-Apr-2013	macallan	branches: 1.17.4; make this work on PowerMac G5 from Phileas Fogg
1.16	31-Jul-2012	bouyer	branches: 1.16.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.15	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.14	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.13	04-Apr-2011	dyoung	branches: 1.13.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.12	05-Nov-2010	jakllsch	branches: 1.12.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.11	26-Nov-2009	njoly	branches: 1.11.2; 1.11.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.10	18-Mar-2008	cube	branches: 1.10.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.9	05-Feb-2008	simonb	branches: 1.9.2; 1.9.6; Print the "bus-master DMA support present" message out the same other a few other SATA drivers do. Nicer output than just a line with ": DMA" on it...
1.8	24-Oct-2007	xtraeme	branches: 1.8.2; - Add support for the ServerWorks HT-1000 IDE Controller in rccide. - Add support for another ServerWorks HT-1000 SATA Controller. From OpenBSD.
1.7	30-Nov-2006	hannken	branches: 1.7.22; 1.7.24; 1.7.28; Enable WDC_CAPABILITY_WIDEREGS. With this change LBA48 operations no longer wrap at 128GB. Tested on an "ServerWorks HT-1000 SATA Controller (rev. 0x00)". Approved by: Manuel Bouyer <bouyer@netbsd.org>
1.6	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.5	25-Oct-2006	bouyer	Use common wdc_sataprobe() where possible; explain why we can't in satalink.c
1.4	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.3	29-Mar-2006	thorpej	branches: 1.3.2; 1.3.4; 1.3.10; 1.3.14; 1.3.16; 1.3.18; Use device_cfdata().
1.2	07-Mar-2006	bouyer	branches: 1.2.2; 1.2.4; Also support Serverwork Frodo4, Frodo8 and HT-1000 in svwsata. From OpenBSD.
1.1	06-Mar-2006	bouyer	branches: 1.1.2; Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.1.2.1	19-Apr-2006	elad	sync with head.
1.2.4.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.2.2.3	01-Apr-2006	yamt	sync with head.
1.2.2.2	13-Mar-2006	yamt	sync with head.
1.2.2.1	07-Mar-2006	yamt	file svwsata.c was added on branch yamt-pdpolicy on 2006-03-13 09:07:27 +0000
1.3.18.2	10-Dec-2006	yamt	sync with head.
1.3.18.1	22-Oct-2006	yamt	sync with head
1.3.16.2	09-Sep-2006	rpaulo	sync with head
1.3.16.1	29-Mar-2006	rpaulo	file svwsata.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.3.14.2	12-Jan-2007	ad	Sync with head.
1.3.14.1	18-Nov-2006	ad	Sync with head.
1.3.10.6	24-Mar-2008	yamt	sync with head.
1.3.10.5	11-Feb-2008	yamt	sync with head.
1.3.10.4	27-Oct-2007	yamt	sync with head.
1.3.10.3	30-Dec-2006	yamt	sync with head.
1.3.10.2	21-Jun-2006	yamt	sync with head.
1.3.10.1	29-Mar-2006	yamt	file svwsata.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:06 +0000
1.3.4.3	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): sys/dev/pci/svwsata.c: revision 1.2 share/man/man4/svwsata.4: revision 1.2 Also support Serverwork Frodo4, Frodo8 and HT-1000 in svwsata. From OpenBSD.
1.3.4.2	04-May-2006	tron	Pull up following revision(s) (requested by jonathan in ticket #1298): share/man/man4/pciide.4: revision 1.60 share/man/man4/Makefile: revision 1.376 via patch sys/dev/pci/svwsata.c: revision 1.1 sys/dev/pci/files.pci: revision 1.243 sys/arch/i386/conf/INSTALL: revision 1.281 sys/dev/pci/pcidevs: revision 1.771 via patch sys/dev/pci/pciidevar.h: revision 1.33 sys/arch/i386/conf/XEN0: revision 1.34 sys/arch/i386/conf/GENERIC: revision 1.732 sys/dev/pci/pciide_svwsata_reg.h: revision 1.1 Add svwsata(4), a driver for Serverworks K2 SATA controllers. From OpenBSD via Joerg Sonnenberger.
1.3.4.1	29-Mar-2006	tron	file svwsata.c was added on branch netbsd-3 on 2006-05-04 12:12:07 +0000
1.3.2.2	22-Apr-2006	simonb	Sync with head.
1.3.2.1	29-Mar-2006	simonb	file svwsata.c was added on branch simonb-timecounters on 2006-04-22 11:39:16 +0000
1.7.28.1	13-Nov-2007	bouyer	Sync with HEAD
1.7.24.2	23-Mar-2008	matt	sync with HEAD
1.7.24.1	06-Nov-2007	matt	sync with HEAD
1.7.22.1	28-Oct-2007	joerg	Sync with HEAD.
1.8.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.9.2.1	24-Mar-2008	keiichi	sync with head.
1.10.4.1	11-Mar-2010	yamt	sync with head
1.11.4.2	21-Apr-2011	rmind	sync with head
1.11.4.1	05-Mar-2011	rmind	sync with head
1.11.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.12.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.13.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.13.4.1	30-Oct-2012	yamt	sync with head
1.16.2.4	03-Dec-2017	jdolecek	update from HEAD
1.16.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.16.2.2	23-Jun-2013	tls	resync from head
1.16.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.17.4.1	18-May-2014	rmind	sync with head
1.19.18.1	27-Sep-2017	jdolecek	change wdc_init_shadow_regs() to accept only struct wdc_regs, it doesn't touch anything else factor out the probe-only struct ata_channel initialization to wdcprobe(), to reduce duplication of logic in individual drivers, and to actually work now that more init is needed beyond the memset()
1.20.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.20.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.23.2.1	10-Jun-2019	christos	Sync with HEAD
1.11	01-Aug-2023	andvar	s/diable/disable/ in comments.
1.10	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.9	24-Apr-2021	thorpej	branches: 1.9.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.8	28-Feb-2014	matt	branches: 1.8.46; Use %zu for size_t
1.7	10-Nov-2013	mrg	- avoid an uninitialised variable warning - fix a series of comparisions that duplicated one test and did not test one variable (found by gcc set-but-unused.) ok rkujawa
1.6	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.5	31-Jan-2013	rkujawa	branches: 1.5.2; - Map memory space with BUS_SPACE_MAP_LINEAR - Support WSDISPLAYIO_GET_FBINFO - Improve debugging
1.4	29-Jul-2012	rkujawa	branches: 1.4.2; 1.4.4; Add wsdisplay ioctls and support for mmap'ing of linear frame buffer.
1.3	20-Jul-2012	rkujawa	Preliminary blitter support for Voodoo2.
1.2	20-Jul-2012	rkujawa	Add 3Dfx Voodoo Graphics (aka Voodoo 1) support.
1.1	18-Jul-2012	rkujawa	Add 3Dfx Voodoo2 driver. Still needs some cleanup and prettyfying, but hey it works.
1.4.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.4.4.2	30-Oct-2012	yamt	sync with head
1.4.4.1	29-Jul-2012	yamt	file tdvfb.c was added on branch yamt-pagecache on 2012-10-30 17:21:54 +0000
1.4.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.4.2.1	25-Feb-2013	tls	resync with head
1.5.2.2	18-May-2014	rmind	sync with head
1.5.2.1	28-Aug-2013	rmind	sync with head
1.8.46.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.9.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.3	20-Jul-2012	rkujawa	branches: 1.3.4; Preliminary blitter support for Voodoo2.
1.2	20-Jul-2012	rkujawa	Add 3Dfx Voodoo Graphics (aka Voodoo 1) support.
1.1	18-Jul-2012	rkujawa	Add 3Dfx Voodoo2 driver. Still needs some cleanup and prettyfying, but hey it works.
1.3.4.2	30-Oct-2012	yamt	sync with head
1.3.4.1	20-Jul-2012	yamt	file tdvfbreg.h was added on branch yamt-pagecache on 2012-10-30 17:21:54 +0000
1.3	29-Jul-2012	rkujawa	branches: 1.3.4; Add wsdisplay ioctls and support for mmap'ing of linear frame buffer.
1.2	20-Jul-2012	rkujawa	Add 3Dfx Voodoo Graphics (aka Voodoo 1) support.
1.1	18-Jul-2012	rkujawa	Add 3Dfx Voodoo2 driver. Still needs some cleanup and prettyfying, but hey it works.
1.3.4.2	30-Oct-2012	yamt	sync with head
1.3.4.1	29-Jul-2012	yamt	file tdvfbvar.h was added on branch yamt-pagecache on 2012-10-30 17:21:54 +0000
1.93	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.92	16-Oct-2022	tsutsui	Use a consistent pointer variable. No functional change.
1.91	17-Sep-2022	tsutsui	Use C99 designated initializers.
1.90	17-Sep-2022	tsutsui	Make local functions static.
1.89	19-Aug-2021	andvar	s/memry/memory+s/softare/software/+s/grapics/graphics+s/ouput/output
1.88	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.87	24-Apr-2021	thorpej	branches: 1.87.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.86	09-Dec-2018	jdolecek	branches: 1.86.14; use pci_intr_establish_xname() everywhere
1.85	29-Mar-2014	christos	branches: 1.85.28; 1.85.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.84	04-Nov-2013	christos	remove unused code
1.83	11-Jan-2012	macallan	branches: 1.83.6; 1.83.10; add extra parameter for wsfont_find()
1.82	25-Jan-2011	mrg	branches: 1.82.4; 1.82.8; fix the previous. there's no "vc" variable, nor does the local "dc" have a "softc" member. use sc->sc_dev instead. compile tested only, but that's better than the previous revision.
1.81	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.80	13-Nov-2010	uebayasi	branches: 1.80.2; 1.80.4; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.79	15-May-2010	tsutsui	Clear RI_NO_AUTO in actual attach functions during autoconf(9) for console devices that set the flag during consinit().
1.78	15-May-2010	tsutsui	Set RI_NO_AUTO in rip->ri_flg if tga_init() is invoked from cnattach.
1.77	16-May-2009	tsutsui	branches: 1.77.2; 1.77.4; KNF and misc cosmetics.
1.76	16-May-2009	tsutsui	Split device_t/softc. Tested on arc: tga0 at pci0 dev 3 function 0: DC21030 step C, board type T8-02
1.75	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.74	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.73	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.72	16-Mar-2009	dsl	ANSIfy functions with function-pointer arguments
1.71	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.70	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.69	07-Jan-2009	ahoka	branches: 1.69.2; Implement the required ioctls for running Xorg through /dev/ttyE0. Reviewed by macallan@
1.68	10-Apr-2008	cegger	branches: 1.68.4; 1.68.12; use aprint_*_dev and device_xname
1.67	19-Oct-2007	ad	branches: 1.67.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.66	04-Mar-2007	yamt	branches: 1.66.2; 1.66.14; 1.66.16; 1.66.20; fix fallout from caddr_t changes.
1.65	04-Mar-2007	christos	Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.64	13-Jan-2007	cube	branches: 1.64.2; Complete initializers so those files compile.
1.63	26-Apr-2006	rpaulo	branches: 1.63.8; In tgamatch(), ifdef tag test under __alpha__ \|\| arc.
1.62	12-Apr-2006	jmmv	Add an extra cookie to the ioctl and mmap wsdisplay accessops that points to the screen on which they are being called. The driver cannot guess this by itself but it is needed to implement, at least, the getwschar and putwschar functions in the correct place. There are no functional changes yet. Tested on i386 (vga, vga_raster, machfb, vesafb), macppc and sparc64. Suggested and reviewed by macallan@.
1.61	11-Dec-2005	christos	branches: 1.61.4; 1.61.6; 1.61.8; 1.61.10; 1.61.12; merge ktrace-lwp.
1.60	27-Feb-2005	perry	branches: 1.60.4; nuke trailing whitespace
1.59	04-Feb-2005	perry	de-__P
1.58	29-Jun-2004	kleink	branches: 1.58.4; 1.58.6; Remove redundant tga_getconf() and tga_identify() prototypes.
1.57	13-Nov-2003	chs	eliminate uvm_useracc() in favor of checking the return value of copyin() or copyout(). uvm_useracc() tells us whether the mapping permissions allow access to the desired part of an address space, and many callers assume that this is the same as knowing whether an attempt to access that part of the address space will succeed. however, access to user space can fail for reasons other than insufficient permission, most notably that paging in any non-resident data can fail due to i/o errors. most of the callers of uvm_useracc() make the above incorrect assumption. the rest are all misguided optimizations, which optimize for the case where an operation will fail. we'd rather optimize for operations succeeding, in which case we should just attempt the access and handle failures due to insufficient permissions the same way we handle i/o errors. since there appear to be no good uses of uvm_useracc(), we'll just remove it.
1.56	29-Oct-2003	mycroft	Don't bother with one of the loops if the data to move is too narrow.
1.55	29-Oct-2003	mycroft	Calculate lastx and lastleft outside the loops...
1.54	27-Oct-2003	chs	uninitialized variables
1.53	21-Oct-2003	tsutsui	vaddr_t is not pointer, so don't compare it against NULL.
1.52	29-Jun-2003	fvdl	branches: 1.52.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.51	29-Jun-2003	simonb	Fix 'struct lwp *' lossage.
1.50	29-Oct-2002	tsutsui	Remove no longer needed #ifdef __alpha__ and #ifdef __mips__.
1.49	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.48	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.47	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.46	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.45	16-Sep-2002	mycroft	Force GPSR to 0 before a blit, so my text doesn't get shifted to the right...
1.44	16-Sep-2002	mycroft	Copy forwards even when moving to the right, if the regions do not overlap.
1.43	16-Sep-2002	mycroft	Simplify the blitting code slightly, and fix fenceposts.
1.42	04-Jul-2002	junyoung	alloc_attr -> allocattr Approved by Matthias Drochner.
1.41	17-Mar-2002	atatat	branches: 1.41.4; 1.41.6; Convert ioctl code to use EPASSTHROUGH instead of -1 or ENOTTY for indicating an unhandled "command". ERESTART is -1, which can lead to confusion. ERESTART has been moved to -3 and EPASSTHROUGH has been placed at -4. No ioctl code should now return -1 anywhere. The ioctl() system call is now properly restartable.
1.40	13-Mar-2002	ad	Reorganise the wsfont stuff slightly so that multiple display adapters with different bit/byte order requirements can co-exist happily.
1.39	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.38	12-Dec-2001	elric	Added support for the PowerStorm 4d20, initial support that is. I plan on cleaning a few things up over the next few days, but this appears to be working for me.
1.37	13-Nov-2001	lukem	add RCSID
1.36	04-Sep-2001	thorpej	Use bus_space_mmap().
1.35	05-Aug-2001	jdolecek	use unsigned variable types as appropriate
1.34	16-Jul-2001	elric	So, the PowerStorm 4d20 a.k.a. 32bit TGA2 w/ IBM RGB561 RAMDAC was causing the kernel to panic since it is recognised as a TGA and the TGA driver doesn't [yet] know what to do with it. This patch fixes that by: o making tgamatch() try to actually figure out what kind of TGA card is there, rather than simply relying on the vendor/product ids. o creating a tga_cnmatch() so that the console code in arch/alpha/pci/pci_machdep.c can cause the same to occur. o breaking up some of tga_getdevconfig() into a few different functions to re-use code that would have been duplicated. o changed arch/alpha/pci/pci_machdep.c so that it calls out to tga_cnmatch() if DEVICE_IS_TGA() matches before it decides to attach the console as a TGA. Addresses PR: port-alpha/12923
1.33	07-Jul-2001	thorpej	branches: 1.33.2; bzero -> memset
1.32	24-Jun-2001	elric	This was able to crash my new alpha consistently when I started X.
1.31	11-Feb-2001	nathanw	branches: 1.31.2; Make tga_rop_vtov deal with copies of <256 byte granularity, necessary for proper console text handling (especially in-line insertion) on 8-bit displays. From Christian Groessler <cpg@aladdin.de>. Tested on 32-bit TGA by me.
1.30	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.29	17-Dec-2000	thorpej	Add back the code that clears "odd" in VCHR. This shouldn't be necessary, but we haven't made X cope with the wierd size yet.
1.28	30-Jun-2000	mjacob	need uvm_extern.h for a function definition
1.27	28-Jun-2000	mrg	remove include of <vm/vm.h>. <vm/vm.h> -> <uvm/uvm_extern.h>
1.26	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.25	14-Jun-2000	soda	branches: 1.25.2; yet another change for NEC RISCstation 2250 of arc port.
1.24	09-Jun-2000	soda	changes for NEC RISCstation 2250 of arc port.
1.23	20-Apr-2000	nathanw	branches: 1.23.2; Replace rcons console code with rasops console code. Console text should now be supported on all TGA cards; 8- and 32-bit TGA and 8-bit TGA2 have been tested. Implement accelerated character drawing, scrolling, and clearing. Stop clearing "odd" in VHCR; it's unnecessary and hurts performance. Use bus_space_vaddr() instead of a local hack.
1.22	02-Apr-2000	nathanw	Add code to read and write bt463-style ramdacs; pick ramdac interface functions based on type returned from tga_getconf(). Adapt to change in ramdac interface. Fix race condition in tga_sched_update(). Make tga_sched_update() work before interrupts are enabled.
1.21	12-Mar-2000	nathanw	Bus_space-ify TGA register accesses. (raster operations are still using pointer-based access)
1.20	09-Mar-2000	drochner	put "#ifdef __alpha__" around "alpha_mb()" calls (I know the driver is useless without console initialisation, but it should at least compile.)
1.19	05-Mar-2000	elric	Moved things around a bit so that TGA2 can be used as a console. Addresses PR: port-alpha/9542
1.18	05-Mar-2000	elric	I got the mode sense bits of the monitor wrong they needed a twiddle (~).
1.17	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.16	25-Jan-2000	drochner	use BUS_SPACE_MAP_PREFETCHABLE instead BUS_SPACE_MAP_CACHEABLE where the PCI BAR bit is referred to
1.15	06-Dec-1999	drochner	update for changed struct wsdisplay_accessops:show_screen signature. no functional changes
1.14	28-Apr-1999	ross	branches: 1.14.2; 1.14.8; Accelerate scrolling. Originally by Roland Dowdeswell, revised by {drocher,ross}@netbsd.org.
1.13	24-Mar-1999	mrg	branches: 1.13.2; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.12	11-Jan-1999	thorpej	Fix a few glitches in the nwscons tga driver, PR #6655, Brett Slager.
1.11	11-Jan-1999	drochner	remove dummy "load_font" entries
1.10	19-Nov-1998	mrg	fix problems in many d_mmap routines: - returned EOPNOTSUPP rather than -1. - no check for negative offset. many of these fix potential security problems in these drivers. XXX XXX XXX the d_mmap cdev routine should be changed to have a prototype like: paddr_t (*d_mmap) __P((dev_t, off_t, int)); by someone!
1.9	02-Sep-1998	drochner	Since the RAMDAC drivers use the !=0-ness of fields for diagnostic purposes, we have to zero-initialize the tga_devconfig structure.
1.8	18-Aug-1998	thorpej	Fix a couple of obvious bugs in the hardware cursor handling code, and use the RAMDAC callbacks to get/set the cursor colormap.
1.7	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.6	26-Jun-1998	drochner	branches: 1.6.2; adapt to wsdisplay interface change
1.5	20-Jun-1998	drochner	adapt to wscons changes (character output call)
1.4	14-May-1998	drochner	Simple screen attribute handling for wscons. The graphics device driver passes a "default attribute" for normal text output to the wscons framework. If the emulation module needs more attributes (for different "renditions") it can allocate them via a callback. For now, only the "sun" emulation makes use of it.
1.3	29-Apr-1998	thorpej	Use the symbolic name VVR_BLANK rather than a magic constant.
1.2	16-Apr-1998	drochner	count allocated screens correctly
1.1	15-Apr-1998	drochner	The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.6.2.2	08-Aug-1998	eeh	Revert cdevsw mmap routines to return int.
1.6.2.1	07-Aug-1998	drochner	make it compile in paddr_t world
1.13.2.1	29-Apr-1999	perry	branches: 1.13.2.1.2; pullup 1.13->1.14 (ross): scrolling fixes.
1.13.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.14.8.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.14.2.3	12-Mar-2001	bouyer	Sync with HEAD.
1.14.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.14.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.23.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.25.2.4	16-Aug-2001	tv	Pullup [jdolecek]: sys/arch/amiga/dev/grf_cl.c 1.26 sys/arch/amiga/dev/grfioctl.h 1.14 sys/arch/hpcmips/dev/plumvideo.c 1.20 sys/arch/macppc/dev/ofb.c 1.25 sys/arch/sparc/dev/cgtwo.c 1.35 sys/arch/sparc/include/fbio.h patch sys/arch/sparc64/include/fbio.h patch sys/arch/sun3/dev/cg2.c 1.14 sys/arch/sun3/include/fbio.h patch sys/dev/pci/tga.c 1.35 sys/dev/tc/cfb.c 1.28 sys/dev/tc/mfb.c 1.27 sys/dev/tc/sfb.c 1.46 sys/dev/tc/sfbplus.c 1.10 via patch sys/dev/tc/tfb.c 1.30 sys/dev/tc/xcfb.c 1.23 sys/net/if_ppp.c 1.71 Use unsigned variable types to make bounds checking more correct.
1.25.2.3	25-Jun-2001	he	Pull up revision 1.32 (requested by elric): Fix problem that some systems would panic when XalphaNetBSD is started.
1.25.2.2	26-Dec-2000	jhawk	Pull up revision 1.29 (requested by thorpej): Restore code to clear "odd" in VCHR. This seems to be required for some TGA cards to run X.
1.25.2.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.31.2.10	11-Nov-2002	nathanw	Catch up to -current
1.31.2.9	18-Oct-2002	nathanw	Catch up to -current.
1.31.2.8	17-Sep-2002	nathanw	Catch up to -current.
1.31.2.7	01-Aug-2002	nathanw	Catch up to -current.
1.31.2.6	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.31.2.5	28-Feb-2002	nathanw	Catch up to -current.
1.31.2.4	08-Jan-2002	nathanw	Catch up to -current.
1.31.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.31.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.31.2.1	24-Aug-2001	nathanw	Catch up with -current.
1.33.2.9	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.33.2.8	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.33.2.7	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.33.2.6	16-Mar-2002	jdolecek	Catch up with -current.
1.33.2.5	11-Feb-2002	jdolecek	Sync w/ -current.
1.33.2.4	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.33.2.3	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.33.2.2	25-Aug-2001	thorpej	Merge Aug 24 -current into the kqueue branch.
1.33.2.1	03-Aug-2001	lukem	update to -current
1.41.6.1	01-Dec-2002	he	Pull up revisions 1.43-1.45 (requested by mycroft in ticket #837): Simplify blitting code slightly, fix fenceposts. Copy forward when moving to the right if no overlap. Force GPSR to 0 before a blit, so text doesn't shift right. Fixes a few problems causing occasional corruption with TGA/TGA2 consoles.
1.41.4.1	15-Jul-2002	gehenna	catch up with -current.
1.52.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.52.2.5	04-Feb-2005	skrll	Sync with HEAD.
1.52.2.4	21-Nov-2004	skrll	Adapt to branch. Alpha kernels now compile.
1.52.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.52.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.52.2.1	03-Aug-2004	skrll	Sync with HEAD
1.58.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.58.6.1	12-Feb-2005	yamt	sync with head.
1.58.4.1	29-Apr-2005	kent	sync with -current
1.60.4.4	27-Oct-2007	yamt	sync with head.
1.60.4.3	03-Sep-2007	yamt	sync with head.
1.60.4.2	26-Feb-2007	yamt	sync with head.
1.60.4.1	21-Jun-2006	yamt	sync with head.
1.61.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.61.10.2	11-May-2006	elad	sync with head
1.61.10.1	19-Apr-2006	elad	sync with head.
1.61.8.1	24-May-2006	yamt	sync with head.
1.61.6.2	01-Jun-2006	kardel	Sync with head.
1.61.6.1	22-Apr-2006	simonb	Sync with head.
1.61.4.1	09-Sep-2006	rpaulo	sync with head
1.63.8.1	01-Feb-2007	ad	Sync with head.
1.64.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.66.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.66.16.1	06-Nov-2007	matt	sync with HEAD
1.66.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.66.2.1	23-Oct-2007	ad	Sync with head.
1.67.16.2	17-Jan-2009	mjf	Sync with HEAD.
1.67.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.68.12.2	28-Apr-2009	skrll	Sync with HEAD.
1.68.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.68.4.4	11-Aug-2010	yamt	sync with head.
1.68.4.3	20-Jun-2009	yamt	sync with head
1.68.4.2	16-May-2009	yamt	sync with head
1.68.4.1	04-May-2009	yamt	sync with head.
1.69.2.2	23-Jul-2009	jym	Sync with HEAD.
1.69.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.77.4.2	05-Mar-2011	rmind	sync with head
1.77.4.1	30-May-2010	rmind	sync with head
1.77.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.80.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.80.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.82.8.1	18-Feb-2012	mrg	merge to -current.
1.82.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.82.4.1	17-Apr-2012	yamt	sync with head
1.83.10.1	18-May-2014	rmind	sync with head
1.83.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.85.30.1	10-Jun-2019	christos	Sync with HEAD
1.85.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.86.14.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.87.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.6	02-Apr-2000	nathanw	Move bt463 code to dev/ic and adapt to the new ramdac interface. Set default window type to 24-bit truecolor. Correctly set pseudocolor mode.
1.5	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.4	30-Apr-1999	nathanw	branches: 1.4.2; Rototillage; make this actually useful. To wit: - Initialize more of the Bt463's registers, instead of leaving them in undefined states. Notably, the window type table is set up with 8-plane pseudocolor and 24-plane truecolor modes. - Bus-space-ify, mostly. Could use some more cleanup, but not until the rest of the tga stuff is converted, too. - Do the TGA/RAMDAC communication dance more carefully. - Explain a lot more of what's going on in comments.
1.3	24-Mar-1999	mrg	branches: 1.3.4; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.2	18-Aug-1998	thorpej	Initial support for the Brooktree Bt463 RAMDAC.
1.1	15-Apr-1998	drochner	The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.3.4.1	21-Jun-1999	thorpej	Sync w/ -current.
1.4.2.1	20-Nov-2000	bouyer	Remove files that are no longer on the trunck, and commit Makefile which I forgot in the batch of commits.
1.5	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.4	24-Mar-1999	mrg	branches: 1.4.8; completely remove Mach VM support. all that is left is the all the header files as UVM still uses (most of) these.
1.3	18-Aug-1998	thorpej	When updating the cursor colormap, actually store the data in our copy of the cursor colormap, not the regular colormap!
1.2	18-Aug-1998	thorpej	Add set_curcmap and get_curcmap RAMDAC methods, for use by the TGA hardware cursor code. These are unneeded on the Bt485, since the RAMDAC handles the hardware cursor. They are needed for the Bt463, however, since the RAMDAC only provides cursor color.
1.1	15-Apr-1998	drochner	The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.4.8.1	20-Nov-2000	bouyer	Remove files that are no longer on the trunck, and commit Makefile which I forgot in the batch of commits.
1.7	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.6	12-Dec-2001	elric	branches: 1.6.118; 1.6.128; 1.6.134; Added support for the PowerStorm 4d20, initial support that is. I plan on cleaning a few things up over the next few days, but this appears to be working for me.
1.5	13-Nov-2001	lukem	add RCSID
1.4	02-Apr-2000	nathanw	branches: 1.4.6; 1.4.8; Restore ramdac-type field to TGA configuration structure.
1.3	12-Mar-2000	nathanw	Bus_space-ify TGA register accesses. (raster operations are still using pointer-based access)
1.2	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.1	15-Apr-1998	drochner	branches: 1.1.14; The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.1.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.6.2	08-Jan-2002	nathanw	Catch up to -current.
1.4.6.1	14-Nov-2001	nathanw	Catch up to -current.
1.6.134.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.6.128.1	28-Apr-2009	skrll	Sync with HEAD.
1.6.118.1	04-May-2009	yamt	sync with head.
1.7	03-Jul-2022	andvar	fix various typos in comments, mainly s/pallete/palette/.
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	12-Mar-2000	nathanw	branches: 1.4.28; 1.4.36; 1.4.38; Bus_space-ify TGA register accesses. (raster operations are still using pointer-based access)
1.3	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.2	29-Apr-1998	thorpej	branches: 1.2.14; Define the contents of the Video Valid register on the TGA and TGA2.
1.1	15-Apr-1998	drochner	The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.2.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.38.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.36.1	29-Apr-2005	kent	sync with -current
1.4.28.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.18	18-May-2009	he	Complete the device_t / softc split by changing the sc_dev member from "struct device" to "device_t". Commit overlooked when tga.c was changed?
1.17	07-Jan-2009	ahoka	branches: 1.17.2; Implement the required ioctls for running Xorg through /dev/ttyE0. Reviewed by macallan@
1.16	11-Dec-2005	christos	branches: 1.16.70; 1.16.74; 1.16.84; merge ktrace-lwp.
1.15	01-Jun-2005	drochner	add "const" where needed
1.14	27-Feb-2005	perry	nuke trailing whitespace
1.13	04-Feb-2005	perry	de-__P
1.12	12-Dec-2001	elric	branches: 1.12.16; 1.12.24; 1.12.26; Added support for the PowerStorm 4d20, initial support that is. I plan on cleaning a few things up over the next few days, but this appears to be working for me.
1.11	04-Sep-2001	thorpej	Use bus_space_mmap().
1.10	16-Jul-2001	elric	So, the PowerStorm 4d20 a.k.a. 32bit TGA2 w/ IBM RGB561 RAMDAC was causing the kernel to panic since it is recognised as a TGA and the TGA driver doesn't [yet] know what to do with it. This patch fixes that by: o making tgamatch() try to actually figure out what kind of TGA card is there, rather than simply relying on the vendor/product ids. o creating a tga_cnmatch() so that the console code in arch/alpha/pci/pci_machdep.c can cause the same to occur. o breaking up some of tga_getdevconfig() into a few different functions to re-use code that would have been duplicated. o changed arch/alpha/pci/pci_machdep.c so that it calls out to tga_cnmatch() if DEVICE_IS_TGA() matches before it decides to attach the console as a TGA. Addresses PR: port-alpha/12923
1.9	20-Apr-2000	nathanw	branches: 1.9.6; 1.9.8; Replace rcons console code with rasops console code. Console text should now be supported on all TGA cards; 8- and 32-bit TGA and 8-bit TGA2 have been tested. Implement accelerated character drawing, scrolling, and clearing. Stop clearing "odd" in VHCR; it's unnecessary and hurts performance. Use bus_space_vaddr() instead of a local hack.
1.8	02-Apr-2000	nathanw	Add code to read and write bt463-style ramdacs; pick ramdac interface functions based on type returned from tga_getconf(). Adapt to change in ramdac interface. Fix race condition in tga_sched_update(). Make tga_sched_update() work before interrupts are enabled.
1.7	12-Mar-2000	nathanw	Bus_space-ify TGA register accesses. (raster operations are still using pointer-based access)
1.6	05-Mar-2000	elric	Moved things around a bit so that TGA2 can be used as a console. Addresses PR: port-alpha/9542
1.5	04-Mar-2000	elric	Added basic TGA2 support to the TGA driver. Abstracted the RAMDAC code out into ../ic/bt485.c. Disabled tga_bt463.c, which we don't currently support.
1.4	18-Aug-1998	thorpej	branches: 1.4.12; Need an entry point to check cursor colormap entries for validity before actually setting them.
1.3	18-Aug-1998	thorpej	Add set_curcmap and get_curcmap RAMDAC methods, for use by the TGA hardware cursor code. These are unneeded on the Bt485, since the RAMDAC handles the hardware cursor. They are needed for the Bt463, however, since the RAMDAC only provides cursor color.
1.2	13-Aug-1998	eeh	Merge paddr_t changes into the main branch.
1.1	15-Apr-1998	drochner	branches: 1.1.2; The "mi" TGA driver. It's still not of much use on anything but alpha because it depends on initialization by the firmware. Further, an mi mmap() is needed...
1.1.2.1	30-Jul-1998	eeh	Split vm_offset_t and vm_size_t into paddr_t, psize_t, vaddr_t, and vsize_t.
1.4.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.9.8.3	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.8.2	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.9.8.1	03-Aug-2001	lukem	update to -current
1.9.6.3	08-Jan-2002	nathanw	Catch up to -current.
1.9.6.2	21-Sep-2001	nathanw	Catch up to -current.
1.9.6.1	24-Aug-2001	nathanw	Catch up with -current.
1.12.26.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.12.26.1	12-Feb-2005	yamt	sync with head.
1.12.24.1	29-Apr-2005	kent	sync with -current
1.12.16.3	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.12.16.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.16.1	04-Feb-2005	skrll	Sync with HEAD.
1.16.84.1	19-Jan-2009	skrll	Sync with HEAD.
1.16.74.2	20-Jun-2009	yamt	sync with head
1.16.74.1	04-May-2009	yamt	sync with head.
1.16.70.1	17-Jan-2009	mjf	Sync with HEAD.
1.17.2.1	23-Jul-2009	jym	Sync with HEAD.
1.5	30-Jun-2001	thorpej	Move the Tigon/Tigon2 firmware, version 12.4.13+wpaul, to a more appropriate location.
1.4	28-Mar-2001	msaitoh	add NetBSD RCS id
1.3	18-Jan-2001	jdolecek	branches: 1.3.2; constify
1.2	17-Nov-2000	bouyer	Update to firmware 12.4.13+wpaul, from FreeBSD
1.1	01-Sep-1999	drochner	branches: 1.1.2; 1.1.4; 1.1.14; quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver. todo: -IPv6 -clean up jumbo buffer allocation - NetBSD provides an opaque argument to the free function, thus doesn't need the hack done here -deal correctly with the mapping of the shared memory
1.1.14.1	31-Dec-2000	jhawk	Pull up revision 1.2 (requested by bouyer): For the ti driver, add hardware 802.1Q support, update firmware to 12.4.13+wpaul, and support the Netgear GA620T copper gigabit card, and the Alteon ACEnic 100baseSX and 1000baseT gigabit cards. Various small changes from FreeBSD.
1.1.4.3	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.4.1	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.2	05-Oct-1999	he	Pull up revision 1.1 (new, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.1.2.1	01-Sep-1999	he	file ti_fw.h was added on branch netbsd-1-4 on 1999-10-05 22:07:31 +0000
1.3.2.2	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.5	30-Jun-2001	thorpej	Move the Tigon/Tigon2 firmware, version 12.4.13+wpaul, to a more appropriate location.
1.4	28-Mar-2001	msaitoh	add NetBSD RCS id again
1.3	18-Jan-2001	jdolecek	branches: 1.3.2; constify
1.2	17-Nov-2000	bouyer	Update to firmware 12.4.13+wpaul, from FreeBSD
1.1	01-Sep-1999	drochner	branches: 1.1.2; 1.1.4; 1.1.14; quick 'n dirty port of FreeBSD's Alteon Gigabit Ethernet driver. todo: -IPv6 -clean up jumbo buffer allocation - NetBSD provides an opaque argument to the free function, thus doesn't need the hack done here -deal correctly with the mapping of the shared memory
1.1.14.3	28-Mar-2001	msaitoh	back out previous change
1.1.14.2	28-Mar-2001	msaitoh	add NetBSD RCS id again
1.1.14.1	31-Dec-2000	jhawk	Pull up revision 1.2 (requested by bouyer): For the ti driver, add hardware 802.1Q support, update firmware to 12.4.13+wpaul, and support the Netgear GA620T copper gigabit card, and the Alteon ACEnic 100baseSX and 1000baseT gigabit cards. Various small changes from FreeBSD.
1.1.4.3	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.2	11-Feb-2001	bouyer	Sync with HEAD.
1.1.4.1	22-Nov-2000	bouyer	Sync with HEAD.
1.1.2.2	05-Oct-1999	he	Pull up revision 1.1 (new, requested by drochner): Back-port of driver for Alteon Gigabit Ethernet driver, originally from FreeBSD.
1.1.2.1	01-Sep-1999	he	file ti_fw2.h was added on branch netbsd-1-4 on 1999-10-05 22:08:13 +0000
1.3.2.2	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.11	07-Oct-2013	jakllsch	Drop trailing whitespace.
1.10	07-Oct-2013	jakllsch	Use the universal pciide_detach() for detach function of all remaining pciide-core drivers, except pciide at pnpbios which isn't actually PCI.
1.9	31-Jul-2012	bouyer	branches: 1.9.2; 1.9.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.8	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.7	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.6	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.5	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.4	04-Apr-2011	dyoung	branches: 1.4.4; Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.3	05-Nov-2010	jakllsch	branches: 1.3.2; Pave the way for detachment of pciide(4)-family ATA controllers.
1.2	20-Sep-2009	christos	branches: 1.2.2; 1.2.4; 1.2.6; remove dup decl.
1.1	20-Sep-2009	christos	PR/42100: Dave J. Barnes: Support for old toshiba PICCOLO IDE controllers.
1.2.6.2	21-Apr-2011	rmind	sync with head
1.2.6.1	05-Mar-2011	rmind	sync with head
1.2.4.2	11-Mar-2010	yamt	sync with head
1.2.4.1	20-Sep-2009	yamt	file toshide.c was added on branch yamt-nfs-mp on 2010-03-11 15:03:59 +0000
1.2.2.1	06-Nov-2010	uebayasi	Sync with HEAD.
1.3.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.4.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.4.4.1	30-Oct-2012	yamt	sync with head
1.9.4.1	18-May-2014	rmind	sync with head
1.9.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.44	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.43	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.42	24-Apr-2021	thorpej	branches: 1.42.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.41	10-Nov-2019	chs	branches: 1.41.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.40	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.39	09-Feb-2017	macallan	branches: 1.39.12; 1.39.14; fix tpyo
1.38	09-Feb-2017	macallan	make fallback to PIO actually work properly
1.37	07-Jul-2016	msaitoh	branches: 1.37.2; 1.37.4; KNF. Remove extra spaces. No functional change.
1.36	29-Mar-2014	christos	branches: 1.36.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.35	24-Nov-2013	bouyer	XS_NOERROR \| XS_RESET doens't make sense, XS_RESET is what is wanted here. Pointed out by Edgar Fu�
1.34	15-Sep-2013	martin	Remove unused variables
1.33	10-May-2012	macallan	branches: 1.33.2; 1.33.4; use MMIO if available
1.32	13-Nov-2010	uebayasi	branches: 1.32.8; 1.32.12; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.31	27-Jul-2010	jakllsch	printf() size_t with 'z' format modifier in debug code.
1.30	14-May-2008	tsutsui	branches: 1.30.18; 1.30.20; Normalize my licenses.
1.29	12-Apr-2008	tsutsui	branches: 1.29.2; 1.29.4; 1.29.6; Split device_t/softc, with misc cosmetics.
1.28	10-Apr-2008	cegger	use aprint_*_dev and device_xname
1.27	19-Oct-2007	ad	branches: 1.27.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.26	04-Mar-2007	christos	branches: 1.26.2; 1.26.14; 1.26.16; 1.26.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.25	16-Nov-2006	christos	branches: 1.25.4; __unused removal on arguments; approved by core.
1.24	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.23	22-May-2006	christos	branches: 1.23.6; 1.23.8; add void casts.
1.22	11-Apr-2006	rpaulo	branches: 1.22.2; Coverity ID 2440: in trm_timeout(), check if srb and xs are NULL before trying to dereference them.
1.21	11-Dec-2005	christos	branches: 1.21.4; 1.21.6; 1.21.8; 1.21.10; 1.21.12; merge ktrace-lwp.
1.20	30-May-2005	christos	branches: 1.20.2; - const poisoning - avoid variable shadowing.
1.19	16-Apr-2005	bouyer	Move epprom settings checks to the right place, so that sync/wide nego is started based on the eeprom values and adapter capabilities. Avoid the driver to negotiate wide transfers with a wide drive connected to a non-wide capable controller. Problem reported and fix tested by Dan LaBell on netbsd-help.
1.18	27-Feb-2005	perry	branches: 1.18.2; nuke trailing whitespace
1.17	21-Feb-2005	thorpej	Part 1 of a cleanup pass over the SCSI subsystem. The aim is to name everything "scsi_*", since we really are talking about the SCSI command set, ATAPI transport not withstanding. Improve the names of many structures, and prepend "SCSI_" onto all SCSI command opcodes. Place items described by the SCSI Primary Commands document into scsi_spc.h.
1.16	02-Jan-2005	tsutsui	branches: 1.16.2; 1.16.4; u_intNN_t -> uintNN_t
1.15	25-Sep-2004	tsutsui	Use ANSI function decls.
1.14	30-Oct-2003	simonb	branches: 1.14.4; Remove some assigned-to but otherwise unused variables.
1.13	02-Oct-2002	thorpej	branches: 1.13.6; Add trailing ; to CFATTACH_DECL.
1.12	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.11	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.10	05-Apr-2002	bouyer	branches: 1.10.4; Implement mstohz() as discussed on tech-kern, and use it in SCSI drivers to convert xs->timeout to callout() parameter.
1.9	05-Apr-2002	mjacob	Fix: ...trm.c:578: warning: too many arguments for format problem.
1.8	13-Mar-2002	wiz	Fix some typos in comments.
1.7	06-Mar-2002	tsutsui	Major overhaul of trm(4) driver: - Change structures for each device to make them more suitable with our scsipi mid-layer. - Use ADAPTER_REQ_SET_XFER_MODE callback. - Cleanup misc functions/structures/style. XXX Tagged queuieng support is disabled for now. XXX Maybe we should have common library of the scsi protocol engine XXX to share it among all other drivers..
1.6	16-Dec-2001	tsutsui	branches: 1.6.2; malloc() + memset() -> malloc() with M_ZERO flag
1.5	18-Nov-2001	tsutsui	Misc cleanups: - correct trm_clock_period[] value - pass BUS_DMA_{READ,WRITE,STREAMING} to bus_dmamap_load() as appropriate - make sure to call bus_dmamap_unload() even after request sense - saved srb params on request sense do not seem to be used, so remove them from struct trm_srb - handle target status more properly in trm_srb_done() and no need to handle error retry in lower driver layer - fix some comments XXX Maybe need more cleanup around TQING for MI SCSI callback..
1.4	13-Nov-2001	lukem	branches: 1.4.2; add RCSID
1.3	11-Nov-2001	tsutsui	branches: 1.3.2; - Fix some bogus comments. - Add a missed splx(s) in error case.
1.2	04-Nov-2001	tsutsui	Fix a couple of bugs: - make sure to allocate memory for sc_tempsrb properly. - correct links of SRB queue in trm_link_srb(). Now trm driver works on arc.
1.1	03-Nov-2001	tsutsui	Add a driver for the Tekram DC-395U/UW/F and DC-315/U SCSI host adapters, which have the Tekram TRM-S1040 ASIC. This driver is written by Rui-Xiang Guo <rxg@ms25.url.com.tw>, and a number of cosmetic changes by me. Tested on i386 by the author, and on macppc and sparc64 by me. XXX On arc, kernel got panic in ltsleep() called from scsipi_execute_xs(), XXX but I'm not sure what is wrong...
1.3.2.2	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.3.2.1	11-Nov-2001	thorpej	file trm.c was added on branch thorpej-mips-cache on 2001-11-12 21:18:21 +0000
1.4.2.6	18-Oct-2002	nathanw	Catch up to -current.
1.4.2.5	17-Apr-2002	nathanw	Catch up to -current.
1.4.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.4.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.4.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.1	13-Nov-2001	nathanw	file trm.c was added on branch nathanw_sa on 2001-11-14 19:15:31 +0000
1.6.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.6.2.4	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.6.2.3	16-Mar-2002	jdolecek	Catch up with -current.
1.6.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.6.2.1	16-Dec-2001	thorpej	file trm.c was added on branch kqueue on 2002-01-10 19:57:07 +0000
1.10.4.1	16-Apr-2005	tron	Pull up revision 1.19 (requested by bouyer in ticket #5751): Move epprom settings checks to the right place, so that sync/wide nego is started based on the eeprom values and adapter capabilities. Avoid the driver to negotiate wide transfers with a wide drive connected to a non-wide capable controller. Problem reported and fix tested by Dan LaBell on netbsd-help.
1.13.6.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.13.6.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.13.6.5	17-Jan-2005	skrll	Sync with HEAD.
1.13.6.4	19-Oct-2004	skrll	Sync with HEAD
1.13.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.6.1	03-Aug-2004	skrll	Sync with HEAD
1.14.4.1	16-Apr-2005	tron	Pull up revision 1.19 (requested by bouyer in ticket #1429): Move epprom settings checks to the right place, so that sync/wide nego is started based on the eeprom values and adapter capabilities. Avoid the driver to negotiate wide transfers with a wide drive connected to a non-wide capable controller. Problem reported and fix tested by Dan LaBell on netbsd-help.
1.16.4.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.16.2.1	29-Apr-2005	kent	sync with -current
1.18.2.1	16-Apr-2005	tron	Pull up revision 1.19 (requested by bouyer in ticket #159): Move epprom settings checks to the right place, so that sync/wide nego is started based on the eeprom values and adapter capabilities. Avoid the driver to negotiate wide transfers with a wide drive connected to a non-wide capable controller. Problem reported and fix tested by Dan LaBell on netbsd-help.
1.20.2.4	27-Oct-2007	yamt	sync with head.
1.20.2.3	03-Sep-2007	yamt	sync with head.
1.20.2.2	30-Dec-2006	yamt	sync with head.
1.20.2.1	21-Jun-2006	yamt	sync with head.
1.21.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.21.10.1	19-Apr-2006	elad	sync with head.
1.21.8.1	24-May-2006	yamt	sync with head.
1.21.6.2	01-Jun-2006	kardel	Sync with head.
1.21.6.1	22-Apr-2006	simonb	Sync with head.
1.21.4.1	09-Sep-2006	rpaulo	sync with head
1.22.2.1	19-Jun-2006	chap	Sync with head.
1.23.8.2	10-Dec-2006	yamt	sync with head.
1.23.8.1	22-Oct-2006	yamt	sync with head
1.23.6.1	18-Nov-2006	ad	Sync with head.
1.25.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.26.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.26.16.1	06-Nov-2007	matt	sync with HEAD
1.26.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.26.2.1	23-Oct-2007	ad	Sync with head.
1.27.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.29.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.29.4.2	11-Aug-2010	yamt	sync with head.
1.29.4.1	16-May-2008	yamt	sync with head.
1.29.2.1	18-May-2008	yamt	sync with head.
1.30.20.1	05-Mar-2011	rmind	sync with head
1.30.18.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.32.12.1	02-Jun-2012	mrg	sync to latest -current.
1.32.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.32.8.1	23-May-2012	yamt	sync with head.
1.33.4.1	18-May-2014	rmind	sync with head
1.33.2.2	03-Dec-2017	jdolecek	update from HEAD
1.33.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.36.6.2	28-Aug-2017	skrll	Sync with HEAD
1.36.6.1	09-Jul-2016	skrll	Sync with HEAD
1.37.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.37.2.1	20-Mar-2017	pgoyette	Sync with HEAD
1.39.14.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.39.14.1	10-Jun-2019	christos	Sync with HEAD
1.39.12.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.41.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.42.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	10-May-2012	macallan	use MMIO if available
1.1	03-Nov-2001	tsutsui	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.160; 1.1.164; Add a driver for the Tekram DC-395U/UW/F and DC-315/U SCSI host adapters, which have the Tekram TRM-S1040 ASIC. This driver is written by Rui-Xiang Guo <rxg@ms25.url.com.tw>, and a number of cosmetic changes by me. Tested on i386 by the author, and on macppc and sparc64 by me. XXX On arc, kernel got panic in ltsleep() called from scsipi_execute_xs(), XXX but I'm not sure what is wrong...
1.1.164.1	02-Jun-2012	mrg	sync to latest -current.
1.1.160.1	23-May-2012	yamt	sync with head.
1.1.6.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.6.1	03-Nov-2001	thorpej	file trmreg.h was added on branch kqueue on 2002-01-10 19:57:07 +0000
1.1.4.2	14-Nov-2001	nathanw	Catch up to -current.
1.1.4.1	03-Nov-2001	nathanw	file trmreg.h was added on branch nathanw_sa on 2001-11-14 19:15:31 +0000
1.1.2.2	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.1.2.1	03-Nov-2001	thorpej	file trmreg.h was added on branch thorpej-mips-cache on 2001-11-12 21:18:22 +0000
1.61	10-Dec-2021	andvar	s/occured/occurred/ in comments, log messages and man pages.
1.60	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.59	24-Apr-2021	thorpej	branches: 1.59.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.58	10-Nov-2019	chs	branches: 1.58.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.57	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.56	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.55	27-Sep-2016	pgoyette	branches: 1.55.6; 1.55.14; 1.55.16; Add a missing ')' in the call to bswap64(). Should fix compile error of twa module for sparc and sparc64
1.54	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.53	07-Jul-2016	msaitoh	branches: 1.53.2; KNF. Remove extra spaces. No functional change.
1.52	27-Sep-2014	christos	branches: 1.52.2; fix size in memset() parameter. Thanks max.
1.51	21-Sep-2014	christos	fix memset size inconsistency
1.50	25-Jul-2014	dholland	branches: 1.50.2; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.49	29-Mar-2014	christos	branches: 1.49.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.48	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.47	13-Mar-2014	hannken	Add __diagused.
1.46	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.45	17-Oct-2013	christos	- remove unused variables - move debugging code inside debugging sections
1.44	14-Sep-2013	joerg	Comment out twa_error_table, but keep it around for documentation purposes.
1.43	02-Nov-2012	chs	branches: 1.43.2; fix autoconf output (don't print from the match routine).
1.42	28-Jul-2012	matt	branches: 1.42.2; Make this compile with i386/conf/ALL
1.41	27-Jul-2012	jakllsch	Convert twa(4) to device_t (CFATTACH_DECL_NEW).
1.40	30-Jan-2012	drochner	branches: 1.40.2; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.39	27-Jan-2012	para	extending vmem(9) to be able to allocated resources for it's own needs. simplifying uvm_map handling (no special kernel entries anymore no relocking) make malloc(9) a thin wrapper around kmem(9) (with private interface for interrupt safety reasons) releng@ acknowledged
1.38	31-Aug-2011	plunky	branches: 1.38.2; 1.38.6; NULL does not need a cast
1.37	24-May-2011	joerg	Explicitly get rid of constness.
1.36	22-Nov-2010	dholland	branches: 1.36.2; Extend the workaround for the 3ware 9650 full-queue bug to a full-fledged quirk, and use it for the 9690 as well where it's also needed. Patch by me, problem hunted down by Jiri Novotny and Zdenek Slavet and reported in PR 44002.
1.35	14-Nov-2010	uebayasi	Avoid pointless use of a sysctl definition, CTL_NAMES. Clean up whitespace.
1.34	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.33	18-Aug-2009	drochner	branches: 1.33.4; put some strange text within an "#if 0"'d section into a comment additionally to appease picky parsers
1.32	31-Jul-2009	bouyer	Disable completely bogus DIAGNOSTIC check.
1.31	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.30	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.29	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.28	27-Jul-2008	joerg	branches: 1.28.4; 1.28.8; 1.28.10; 1.28.12; Unbreak twa after the last two commits.
1.27	25-Jun-2008	gmcgarry	branches: 1.27.2; Revert unintentional whitespace changes. From simonb.
1.26	24-Jun-2008	gmcgarry	Ragge says that 6.5.4 clause 2 says: "Unless the type name specifies a void type, the type name shall specify qualified or unqualified scalar type and the operand shall have scalar type." Therefore, don't cast to a union.
1.25	08-Jun-2008	joerg	branches: 1.25.2; Mention 64bit support once during boot and not every time the controller is reset.
1.24	08-Jun-2008	tsutsui	Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.23	10-May-2008	joerg	Remove #if 0'd code for flashing firmware. It was never active and even the FreeBSD version of the driver removed it.
1.22	08-May-2008	joerg	Add support for 3ware 9650 and 9690 based on patches contributed by Wasabi Systems. Fix bus_dma usage to fix write operations in the light of bounce buffers. Use the 64bit bus_dma tag for 9550+ to avoid bouncing buffers.
1.21	07-May-2008	joerg	branches: 1.21.2; First part of twa(4) improvements: - more/better diagnostic messages - dynamically allocate units in preparation for 9650+ support - better distribute queue slots over disks according to size - drain large request queue on 9550 - keep track of resets, will be needed for 9650. This changes have been contributed by Wasabi Systems.
1.20	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.19	10-Apr-2008	cegger	branches: 1.19.2; 1.19.4; use aprint_*_dev and device_xname
1.18	19-Oct-2007	ad	branches: 1.18.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.17	04-Mar-2007	christos	branches: 1.17.2; 1.17.14; 1.17.16; 1.17.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.16	16-Nov-2006	christos	branches: 1.16.2; 1.16.4; 1.16.12; __unused removal on arguments; approved by core.
1.15	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.14	11-Oct-2006	jmcneill	Size is the third argument to memset, not the second.
1.13	23-Sep-2006	manu	Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done.
1.12	03-Sep-2006	christos	branches: 1.12.2; 1.12.4; 1.12.6; - add missing initializer - inline comes first
1.11	26-Aug-2006	christos	PR/34292: Aaron Hodgen: twa driver incorrectly reports drive size/description
1.10	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.9	30-Jul-2006	bouyer	branches: 1.9.2; Add some missing splvm()/splx() around uvm_km_* I noticed.
1.8	29-Jul-2006	wrstuden	Updated driver, contributed by Jordan Rhody of Wasabi Systems. Main fix is to improve reliability under heavy load and frequent status monitoring.
1.7	11-Jul-2006	simonb	Use c99 uintN_t.
1.6	11-Jul-2006	simonb	Sprinkle some KNF. Still a bit more needed...
1.5	10-Jul-2006	simonb	Remove some incorrect licencing goop.
1.4	10-Jul-2006	simonb	Update for timecounters (s/time.tv_sec/time_second/).
1.3	25-May-2006	wrstuden	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; 1.3.10; Add $NetBSD: ids, which work better than $netbsd: ones.
1.2	25-May-2006	wrstuden	Style changes from simonb at wasabisystems dot com: Use config_stdsubmatch() instead of our own submatch function. Use "__inline" in header files and "inline" in .c files. Make some local function static.
1.1	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.3.10.4	03-Sep-2006	yamt	sync with head.
1.3.10.3	11-Aug-2006	yamt	sync with head
1.3.10.2	26-Jun-2006	yamt	sync with head.
1.3.10.1	25-May-2006	yamt	file twa.c was added on branch yamt-pdpolicy on 2006-06-26 12:51:23 +0000
1.3.8.5	27-Oct-2007	yamt	sync with head.
1.3.8.4	03-Sep-2007	yamt	sync with head.
1.3.8.3	30-Dec-2006	yamt	sync with head.
1.3.8.2	21-Jun-2006	yamt	sync with head.
1.3.8.1	25-May-2006	yamt	file twa.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.3.6.2	19-Jun-2006	chap	Sync with head.
1.3.6.1	25-May-2006	chap	file twa.c was added on branch chap-midi on 2006-06-19 04:01:37 +0000
1.3.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.3.2.2	01-Jun-2006	kardel	Sync with head.
1.3.2.1	25-May-2006	kardel	file twa.c was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.9.2.3	11-Nov-2006	bouyer	Pull up following revision(s) (requested by manu in ticket #1528): etc/MAKEDEV.tmpl: revision 1.66 via patch sys/conf/majors: revision 1.30 via patch sys/dev/pci/twareg.h: revision 1.6 via patch sys/dev/pci/twa.c: revision 1.13 via patch sys/dev/pci/twaio.h: revision 1.4 via patch sys/compat/freebsd/freebsd_sysctl.c: revision 1.5 via patch Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done. Fix a bug in FreeBSD sysctl emulation: use copyin for moving data to the kernel
1.9.2.2	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.9.2.1	30-Jul-2006	tron	file twa.c was added on branch netbsd-3 on 2006-07-31 12:35:23 +0000
1.12.6.2	10-Dec-2006	yamt	sync with head.
1.12.6.1	22-Oct-2006	yamt	sync with head
1.12.4.2	09-Sep-2006	rpaulo	sync with head
1.12.4.1	03-Sep-2006	rpaulo	file twa.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.12.2.1	18-Nov-2006	ad	Sync with head.
1.16.12.1	25-Aug-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #1349): sys/dev/pci/twa.c: revision 1.32 Disable completely bogus DIAGNOSTIC check.
1.16.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.16.2.1	25-Aug-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #1349): sys/dev/pci/twa.c: revision 1.32 Disable completely bogus DIAGNOSTIC check.
1.17.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.17.16.1	06-Nov-2007	matt	sync with HEAD
1.17.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.17.2.1	23-Oct-2007	ad	Sync with head.
1.18.16.4	28-Sep-2008	mjf	Sync with HEAD.
1.18.16.3	29-Jun-2008	mjf	Sync with HEAD.
1.18.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.18.16.1	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.19.4.4	19-Aug-2009	yamt	sync with head.
1.19.4.3	16-May-2009	yamt	sync with head
1.19.4.2	04-May-2009	yamt	sync with head.
1.19.4.1	16-May-2008	yamt	sync with head.
1.19.2.2	17-Jun-2008	yamt	sync with head.
1.19.2.1	18-May-2008	yamt	sync with head.
1.21.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.21.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.25.2.2	28-Jul-2008	simonb	Sync with head.
1.25.2.1	27-Jun-2008	simonb	Sync with head.
1.27.2.1	19-Oct-2008	haad	Sync with HEAD.
1.28.12.1	21-Apr-2010	matt	sync to netbsd-5
1.28.10.1	07-Aug-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #886): sys/dev/pci/twa.c: revision 1.32 Disable completely bogus DIAGNOSTIC check.
1.28.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.28.4.1	07-Aug-2009	snj	Pull up following revision(s) (requested by bouyer in ticket #886): sys/dev/pci/twa.c: revision 1.32 Disable completely bogus DIAGNOSTIC check.
1.33.4.2	31-May-2011	rmind	sync with head
1.33.4.1	05-Mar-2011	rmind	sync with head
1.36.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.38.6.1	18-Feb-2012	mrg	merge to -current.
1.38.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.38.2.3	16-Jan-2013	yamt	sync with (a bit old) head
1.38.2.2	30-Oct-2012	yamt	sync with head
1.38.2.1	17-Apr-2012	yamt	sync with head
1.40.2.1	22-Nov-2012	riz	Pull up following revision(s) (requested by chs in ticket #688): sys/dev/pci/twa.c: revision 1.43 fix autoconf output (don't print from the match routine).
1.42.2.3	03-Dec-2017	jdolecek	update from HEAD
1.42.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.42.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.43.2.1	18-May-2014	rmind	sync with head
1.49.2.1	10-Aug-2014	tls	Rebase.
1.50.2.1	30-Oct-2014	martin	Pull up following revision(s) (requested by maxv in ticket #164): sys/dev/ieee1394/fwcrom.c: revision 1.15 sys/dev/ieee1394/fwcrom.c: revision 1.16 sys/dev/marvell/if_gfe.c: revision 1.43 sys/dev/usb/stuirda.c: revision 1.16 sys/dev/pci/cxgb/cxgb_offload.c: revision 1.4 sys/dev/pci/twa.c: revision 1.51 sys/dev/pci/twa.c: revision 1.52 Various fixes in dev/: remove dead code and fix two inconsistencies.
1.52.2.2	05-Oct-2016	skrll	Sync with HEAD
1.52.2.1	09-Jul-2016	skrll	Sync with HEAD
1.53.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.55.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.55.16.1	10-Jun-2019	christos	Sync with HEAD
1.55.14.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.55.14.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.55.6.1	30-Apr-2017	pgoyette	More DEVSW_MODULE_INIT for modular drivers
1.58.10.3	22-Mar-2021	thorpej	Fix missing comma.
1.58.10.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.58.10.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.59.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.6	06-Sep-2015	dholland	More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.5	08-Sep-2008	gmcgarry	branches: 1.5.38; 1.5.58; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.4	23-Sep-2006	manu	branches: 1.4.48; 1.4.52; 1.4.54; 1.4.58; Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done.
1.3	11-Jul-2006	simonb	branches: 1.3.2; 1.3.6; 1.3.8; 1.3.10; Remove some incorrect licencing goop.
1.2	25-May-2006	wrstuden	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.10; Add $NetBSD: ids, which work better than $netbsd: ones.
1.1	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.2.10.3	11-Aug-2006	yamt	sync with head
1.2.10.2	26-Jun-2006	yamt	sync with head.
1.2.10.1	25-May-2006	yamt	file twaio.h was added on branch yamt-pdpolicy on 2006-06-26 12:51:23 +0000
1.2.8.3	30-Dec-2006	yamt	sync with head.
1.2.8.2	21-Jun-2006	yamt	sync with head.
1.2.8.1	25-May-2006	yamt	file twaio.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.2.6.2	19-Jun-2006	chap	Sync with head.
1.2.6.1	25-May-2006	chap	file twaio.h was added on branch chap-midi on 2006-06-19 04:01:37 +0000
1.2.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.2.2.2	01-Jun-2006	kardel	Sync with head.
1.2.2.1	25-May-2006	kardel	file twaio.h was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.3.10.1	22-Oct-2006	yamt	sync with head
1.3.8.2	09-Sep-2006	rpaulo	sync with head
1.3.8.1	11-Jul-2006	rpaulo	file twaio.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.3.6.1	18-Nov-2006	ad	Sync with head.
1.3.2.3	11-Nov-2006	bouyer	Pull up following revision(s) (requested by manu in ticket #1528): etc/MAKEDEV.tmpl: revision 1.66 via patch sys/conf/majors: revision 1.30 via patch sys/dev/pci/twareg.h: revision 1.6 via patch sys/dev/pci/twa.c: revision 1.13 via patch sys/dev/pci/twaio.h: revision 1.4 via patch sys/compat/freebsd/freebsd_sysctl.c: revision 1.5 via patch Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done. Fix a bug in FreeBSD sysctl emulation: use copyin for moving data to the kernel
1.3.2.2	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.3.2.1	11-Jul-2006	tron	file twaio.h was added on branch netbsd-3 on 2006-07-31 12:35:23 +0000
1.4.58.1	19-Oct-2008	haad	Sync with HEAD.
1.4.54.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.4.52.1	04-May-2009	yamt	sync with head.
1.4.48.1	28-Sep-2008	mjf	Sync with HEAD.
1.5.58.1	22-Sep-2015	skrll	Sync with HEAD
1.5.38.1	03-Dec-2017	jdolecek	update from HEAD
1.12	05-Feb-2024	andvar	s/bufffer/buffer/ in comments and log message.
1.11	22-Nov-2010	dholland	Extend the workaround for the 3ware 9650 full-queue bug to a full-fledged quirk, and use it for the 9690 as well where it's also needed. Patch by me, problem hunted down by Jiri Novotny and Zdenek Slavet and reported in PR 44002.
1.10	08-Sep-2008	gmcgarry	branches: 1.10.16; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.9	08-May-2008	joerg	branches: 1.9.4; Add support for 3ware 9650 and 9690 based on patches contributed by Wasabi Systems. Fix bus_dma usage to fix write operations in the light of bounce buffers. Use the 64bit bus_dma tag for 9550+ to avoid bouncing buffers.
1.8	07-May-2008	joerg	branches: 1.8.2; First part of twa(4) improvements: - more/better diagnostic messages - dynamically allocate units in preparation for 9650+ support - better distribute queue slots over disks according to size - drain large request queue on 9550 - keep track of resets, will be needed for 9650. This changes have been contributed by Wasabi Systems.
1.7	19-Oct-2007	ad	branches: 1.7.16; 1.7.18; 1.7.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.6	23-Sep-2006	manu	branches: 1.6.8; 1.6.22; 1.6.24; 1.6.28; Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done.
1.5	28-Jul-2006	wrstuden	branches: 1.5.2; 1.5.6; 1.5.8; 1.5.10; Updated twa driver contributed by Jordan Rhody of Wasabi systems. Correct AEN severity values.
1.4	11-Jul-2006	simonb	Use c99 uintN_t.
1.3	11-Jul-2006	simonb	Remove some incorrect licencing goop.
1.2	25-May-2006	wrstuden	branches: 1.2.2; 1.2.4; 1.2.6; 1.2.8; 1.2.10; Add $NetBSD: ids, which work better than $netbsd: ones.
1.1	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.2.10.3	11-Aug-2006	yamt	sync with head
1.2.10.2	26-Jun-2006	yamt	sync with head.
1.2.10.1	25-May-2006	yamt	file twareg.h was added on branch yamt-pdpolicy on 2006-06-26 12:51:23 +0000
1.2.8.4	27-Oct-2007	yamt	sync with head.
1.2.8.3	30-Dec-2006	yamt	sync with head.
1.2.8.2	21-Jun-2006	yamt	sync with head.
1.2.8.1	25-May-2006	yamt	file twareg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.2.6.2	19-Jun-2006	chap	Sync with head.
1.2.6.1	25-May-2006	chap	file twareg.h was added on branch chap-midi on 2006-06-19 04:01:37 +0000
1.2.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.2.2.2	01-Jun-2006	kardel	Sync with head.
1.2.2.1	25-May-2006	kardel	file twareg.h was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.5.10.1	22-Oct-2006	yamt	sync with head
1.5.8.2	09-Sep-2006	rpaulo	sync with head
1.5.8.1	28-Jul-2006	rpaulo	file twareg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.5.6.1	18-Nov-2006	ad	Sync with head.
1.5.2.3	11-Nov-2006	bouyer	Pull up following revision(s) (requested by manu in ticket #1528): etc/MAKEDEV.tmpl: revision 1.66 via patch sys/conf/majors: revision 1.30 via patch sys/dev/pci/twareg.h: revision 1.6 via patch sys/dev/pci/twa.c: revision 1.13 via patch sys/dev/pci/twaio.h: revision 1.4 via patch sys/compat/freebsd/freebsd_sysctl.c: revision 1.5 via patch Tweak the twa(4) driver so that 3ware command line utility works with it. This includes: - fixing various structure definitions so that the ioctl parameter match - adding a hw.twa*.driver_version sysctl - do not refuse multiple device openings, as the management tool will do it. I'm not sure we are safe. FreeBSD allows multiple openings, and use the open flag only when an attempt to detach the device is done. Fix a bug in FreeBSD sysctl emulation: use copyin for moving data to the kernel
1.5.2.2	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.5.2.1	28-Jul-2006	tron	file twareg.h was added on branch netbsd-3 on 2006-07-31 12:35:23 +0000
1.6.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.6.24.1	06-Nov-2007	matt	sync with HEAD
1.6.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.6.8.1	23-Oct-2007	ad	Sync with head.
1.7.20.2	04-May-2009	yamt	sync with head.
1.7.20.1	16-May-2008	yamt	sync with head.
1.7.18.1	18-May-2008	yamt	sync with head.
1.7.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.7.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.8.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.8.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.9.4.1	19-Oct-2008	haad	Sync with HEAD.
1.10.16.1	05-Mar-2011	rmind	sync with head
1.12	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.11	27-Jul-2012	jakllsch	branches: 1.11.2; Convert twa(4) to device_t (CFATTACH_DECL_NEW).
1.10	22-Nov-2010	dholland	branches: 1.10.8; Extend the workaround for the 3ware 9650 full-queue bug to a full-fledged quirk, and use it for the 9690 as well where it's also needed. Patch by me, problem hunted down by Jiri Novotny and Zdenek Slavet and reported in PR 44002.
1.9	06-May-2009	cegger	branches: 1.9.4; struct device * -> device_t, no functional changes intended.
1.8	07-May-2008	joerg	branches: 1.8.14; First part of twa(4) improvements: - more/better diagnostic messages - dynamically allocate units in preparation for 9650+ support - better distribute queue slots over disks according to size - drain large request queue on 9550 - keep track of resets, will be needed for 9650. This changes have been contributed by Wasabi Systems.
1.7	04-Mar-2007	christos	branches: 1.7.36; 1.7.38; 1.7.40; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.6	17-Aug-2006	christos	branches: 1.6.4; 1.6.10; Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.5	11-Jul-2006	simonb	branches: 1.5.2; Use c99 uintN_t.
1.4	11-Jul-2006	simonb	Remove some incorrect licencing goop.
1.3	25-May-2006	wrstuden	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.8; 1.3.10; Add $NetBSD: ids, which work better than $netbsd: ones.
1.2	25-May-2006	wrstuden	Style changes from simonb at wasabisystems dot com: Use config_stdsubmatch() instead of our own submatch function. Use "__inline" in header files and "inline" in .c files. Make some local function static.
1.1	24-May-2006	wrstuden	3ware 9000 driver, contributed by Wasabi Systems and written by Jordan Rhody (based on the FreeBSD driver). Contributed in NetBSD PR 33364.
1.3.10.4	03-Sep-2006	yamt	sync with head.
1.3.10.3	11-Aug-2006	yamt	sync with head
1.3.10.2	26-Jun-2006	yamt	sync with head.
1.3.10.1	25-May-2006	yamt	file twavar.h was added on branch yamt-pdpolicy on 2006-06-26 12:51:23 +0000
1.3.8.4	03-Sep-2007	yamt	sync with head.
1.3.8.3	30-Dec-2006	yamt	sync with head.
1.3.8.2	21-Jun-2006	yamt	sync with head.
1.3.8.1	25-May-2006	yamt	file twavar.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.3.6.2	19-Jun-2006	chap	Sync with head.
1.3.6.1	25-May-2006	chap	file twavar.h was added on branch chap-midi on 2006-06-19 04:01:37 +0000
1.3.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.3.2.2	01-Jun-2006	kardel	Sync with head.
1.3.2.1	25-May-2006	kardel	file twavar.h was added on branch simonb-timecounters on 2006-06-01 22:36:49 +0000
1.5.2.2	31-Jul-2006	tron	Apply patch (requested by bouyer in ticket #1441): Pullup the twa(4) driver for the 3ware Apache RAID controllers. Matching revisions in current: share/man/man4/twa.4 1.1 sys/dev/pci/ld_twa.c 1.2 sys/dev/pci/twa.c 1.9 sys/dev/pci/twaio.h 1.3 sys/dev/pci/twareg.h 1.5 sys/dev/pci/twavar.h 1.5
1.5.2.1	11-Jul-2006	tron	file twavar.h was added on branch netbsd-3 on 2006-07-31 12:35:23 +0000
1.6.10.1	12-Mar-2007	rmind	Sync with HEAD.
1.6.4.2	09-Sep-2006	rpaulo	sync with head
1.6.4.1	17-Aug-2006	rpaulo	file twavar.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.7.40.2	16-May-2009	yamt	sync with head
1.7.40.1	16-May-2008	yamt	sync with head.
1.7.38.1	18-May-2008	yamt	sync with head.
1.7.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.8.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.9.4.1	05-Mar-2011	rmind	sync with head
1.10.8.1	30-Oct-2012	yamt	sync with head
1.11.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.111	02-Feb-2024	andvar	s/mangement/management/ in comments.
1.110	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.109	24-Apr-2021	thorpej	branches: 1.109.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.108	10-Nov-2019	chs	branches: 1.108.10; in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.107	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.106	27-Sep-2016	pgoyette	branches: 1.106.6; 1.106.14; 1.106.16; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.105	14-Jul-2016	msaitoh	branches: 1.105.2; KNF. No functional change.
1.104	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.103	25-Jul-2014	dholland	branches: 1.103.4; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.102	29-Mar-2014	christos	branches: 1.102.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.101	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.100	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.99	15-Sep-2013	martin	Remove unused variable
1.98	02-Dec-2012	chs	branches: 1.98.2; fix device/softc split: initialize sc->sc_dev.
1.97	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.96	27-Jan-2012	para	branches: 1.96.6; extending vmem(9) to be able to allocated resources for it's own needs. simplifying uvm_map handling (no special kernel entries anymore no relocking) make malloc(9) a thin wrapper around kmem(9) (with private interface for interrupt safety reasons) releng@ acknowledged
1.95	30-Jun-2011	wiz	branches: 1.95.2; 1.95.6; dependant -> dependent
1.94	24-May-2011	joerg	Explicitly get rid of constness.
1.93	14-Nov-2010	uebayasi	branches: 1.93.2; Avoid pointless use of a sysctl definition, CTL_NAMES. Clean up whitespace.
1.92	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.91	27-Jul-2010	jakllsch	printf() size_t with 'z' format modifier in debug code.
1.90	12-May-2009	cegger	branches: 1.90.2; 1.90.4; use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.89	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.88	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.87	08-Jun-2008	tsutsui	branches: 1.87.12; Replace device_lookup() with device_lookup_private() on getting softc for future device_t/softc spilt.
1.86	28-Apr-2008	martin	branches: 1.86.2; Remove clause 3 and 4 from TNF licenses
1.85	10-Apr-2008	cegger	branches: 1.85.2; 1.85.4; use aprint_*_dev and device_xname
1.84	19-Oct-2007	ad	branches: 1.84.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.83	04-Mar-2007	christos	branches: 1.83.2; 1.83.14; 1.83.16; 1.83.20; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.82	02-Dec-2006	elad	branches: 1.82.2; Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.81	16-Nov-2006	christos	branches: 1.81.2; __unused removal on arguments; approved by core.
1.80	08-Nov-2006	elad	Replace securelevel checks with kauth(9) calls.
1.79	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.78	03-Sep-2006	christos	branches: 1.78.2; 1.78.4; - add missing initializer - add missing function decl
1.77	27-Aug-2006	xtraeme	Initialize a variable to 0 to appease gcc4 with -O3.
1.76	23-Aug-2006	christos	Acting as Elad's keyboard: Place the securelevel checks in their logical locations. This will be clearer in the future when code changes to use kauth(9) calls. input and okay ad@
1.75	11-Apr-2006	rpaulo	Coverity ID 2859: Don't leak ccb, move the allocation a few lines lower next to where it'll be used for the first time.
1.74	23-Feb-2006	simonb	branches: 1.74.2; 1.74.4; 1.74.6; Fix memory leak in the TWEIO_GET_PARAM ioctl. Tested by gendalia@.
1.73	29-Jan-2006	dsl	branches: 1.73.2; 1.73.4; Make almost everything #include <sys/bswap.h> instead of <machine/bswap.h> The bswap.h and endian.h files are all rather incestuous, but I want to get the constant folding stuff into one place - sys/bswap.h
1.72	24-Dec-2005	perry	branches: 1.72.2; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.71	11-Dec-2005	christos	merge ktrace-lwp.
1.70	26-Aug-2005	drochner	kill some more simple submatch() functions, use config_stdsubmatch()
1.69	25-Aug-2005	drochner	replace the "locdesc_t" structure carrying the number of locators explicitely by a plain integer array the length in now known to all relevant parties, so this avoids duplication of information, and we can allocate that thing in drivers without hacks
1.68	28-Jun-2005	thorpej	branches: 1.68.2; Use ANSI function decls and static.
1.67	20-Jun-2005	atatat	Change the rest of the sysctl subsystem to use const consistently. The __UNCONST macro is now used only where necessary and the RW macros are gone. Most of the changes here are consumers of the sysctl_createv(9) interface that now takes a pair of const pointers which used not to be.
1.66	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.65	01-Apr-2005	yamt	merge yamt-km branch. - don't use managed mappings/backing objects for wired memory allocations. save some resources like pv_entry. also fix (most of) PR/27030. - simplify kernel memory management API. - simplify pmap bootstrap of some ports. - some related cleanups.
1.64	16-Mar-2005	erh	branches: 1.64.2; 1.64.4; Fix PR#24302. Clear PCI abort and parity errors to give the card a chance at working. Patch provided by Jean-Luc Wasmer, based on the FreeBSD driver.
1.63	27-Feb-2005	perry	nuke trailing whitespace
1.62	20-Feb-2005	heas	Add sysctl node hw.tweN.driver_version whose value is FreeBSD's driver version number for now. This is used by 3ware's FreeBSD version of their CLI to find controllers and check the driver version.
1.61	15-Feb-2005	lukem	Add some more stripe sizes observed in the field. Add "JBOD" to twe_table_unittype[].
1.60	23-Sep-2004	heas	branches: 1.60.4; 1.60.6; Take the modulus of the AEN queue tail number rather than the mask of. Now the tail pointer advances properly.
1.59	13-Sep-2004	drochner	a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.58	04-Jun-2004	thorpej	twe_ccb_map(): Remove the DEBUG panic case of "unknown" S/G list offset. In all likelihood, this is a command passed from userspace management tools where no S/G list is needed because no data is being passed.
1.57	03-Jun-2004	heas	Double the time we wait for the attention bit following a soft reset (and make the timeout message just a bit clearer), as the latest firmware appears to be a bit slower. OK from Jason Thorpe.
1.56	27-May-2004	thorpej	Rename twe_tweio_command_handler() to twe_ccb_wait_handler() and expose it.
1.55	15-Apr-2004	thorpej	Assign a syslog priority to the AEN codes and use that priority when logging the error message. Potentially allows critical messages to be noticed much more quickly.
1.54	04-Dec-2003	thorpej	branches: 1.54.2; When performing command pass-through for management tools, don't poll for CCB completion. Instead, sleep, and use a CCB callback to wake the sleeper.
1.53	10-Nov-2003	thorpej	Correct the size passed to a malloc() call when performing command passthrough for userland management tools.
1.52	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.51	25-Sep-2003	thorpej	Fix up the TWEIO_COMMAND code to actually work. The 3ware management API library can now communicate with the card.
1.50	25-Sep-2003	thorpej	- Fix a bug I introduced in a previous commit (oops). - Add the twe_drivecommand ioctl structure.
1.49	23-Sep-2003	thorpej	- Make CCB allocation slightly more efficient by changing how the CCB is returned to the caller. - Make code paths that can use twe_ccb_alloc_wait() use it, and assert that a CCB is always returned from that function. - Assert that a CCB is always returned when allocating the reserved CCB for an AEN fetch.
1.48	23-Sep-2003	thorpej	Add a comment clarifying why we clear TWEF_AEN in twe_reset() now.
1.47	23-Sep-2003	thorpej	Separate the AEN fetching path into its own special path that uses the reserved CCB. This means that all remaining callers of twe_param_get*() are called from a valid thread context, and thus have no need to use a reserved CCB. This will allow for further cleanup in a future commit.
1.46	22-Sep-2003	thorpej	Add support for dynamically attaching and detaching RAID array units.
1.45	22-Sep-2003	thorpej	Improve AEN handling: - Make AENs use the generic code table stuff. - Add a few more AEN codes (from FreeBSD). - Correct the context of a few AEN codes (some were listed as "unit context" when they're really "port context"). - Add a queue of AENs that management tools in userspace can poll (from FreeBSD).
1.44	21-Sep-2003	thorpej	Fetch info about the drives during attach, and inject that info into the msgbuf using aprint_verbose().
1.43	21-Sep-2003	thorpej	Expose the "param get" functions.
1.42	21-Sep-2003	thorpej	- Record more information about the array unit, including array type and stripe depth. - Report array type and stripe depth when attaching the logical drive.
1.41	21-Sep-2003	thorpej	Add some helper routines to fetch 1, 2, and 4 byte parameter values. Modeled after code in FreeBSD.
1.40	21-Sep-2003	thorpej	Add some additional code -> description tables, and add a generic routine to translate code -> description. AENs will be converted to this mechanism in a future commit. Partially from FreeBSD.
1.39	03-Aug-2003	jdolecek	tweioctl(): store the result of twe_param_get() directly to the right variable rather than using (void *) temporary should fix strict aliasing warning of gcc 3.3
1.38	25-Jul-2003	jdolecek	print some additional information about the controller on boot (from FreeBSD) use aprint_* as appropriate change some debug #ifdef DIAGNOSTIC messages into #ifdef DEBUG make aenfmt[] more const change twe_param_get() signature to take (struct twe_param **) last arg, and remove now unnecessary casts
1.37	29-Jun-2003	fvdl	branches: 1.37.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.36	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.35	01-Jan-2003	thorpej	Use aprint_normal() in cfprint routines.
1.34	16-Dec-2002	fvdl	Remove comparison of u_int8_t value to (512 - a few bytes), since it's always false. As found by gcc-current.
1.33	13-Dec-2002	christos	Add /dev/twex support. To be used with FreeBSD's 3dmd. Initial port from Masaru OKI <oki@netbsd.org>, with KNF, memory leak fixes, denial of service problems, and null pointer fixes from me.
1.32	25-Nov-2002	thorpej	Avoid strict-alias warnings.
1.31	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.30	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.29	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.28	27-Sep-2002	thorpej	Introduce a new routine, config_match(), which invokes the cfattach->ca_match function in behalf of the caller. Use it rather than invoking cfattach->ca_match directly.
1.27	06-Aug-2002	kim	Missing comma in aenfmt array initialization.
1.26	10-Jul-2002	christos	Add missing messages from FreeBSD's driver. Re-organize a bit how the message is formatted and printed.
1.25	01-Jun-2002	lukem	SIMPLEQ rototill: - implement SIMPLEQ_REMOVE(head, elm, type, field). whilst it's O(n), this mirrors the functionality of SLIST_REMOVE() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE() - remove the unnecessary elm arg from SIMPLEQ_REMOVE_HEAD(). this mirrors the functionality of SLIST_REMOVE_HEAD() (the other singly-linked list type) and FreeBSD's STAILQ_REMOVE_HEAD() - remove notes about SIMPLEQ not supporting arbitrary element removal - use SIMPLEQ_FOREACH() instead of home-grown for loops - use SIMPLEQ_EMPTY() appropriately - use SIMPLEQ_*() instead of accessing sqh_first,sqh_last,sqe_next directly - reorder manual page; be consistent about how the types are listed - other minor cleanups
1.24	24-May-2002	christos	NBPG is not constant on the sparc, so don't use CPP tricks.
1.23	24-May-2002	christos	Deal with the NBPG problem in a better way.
1.22	18-May-2002	ad	branches: 1.22.2; - Use write barriers. - NBPG -> PAGE_SIZE.
1.21	13-Nov-2001	lukem	branches: 1.21.8; add RCSID
1.20	10-Oct-2001	ad	branches: 1.20.4; Back out previous - it causes panics.
1.19	20-Sep-2001	ad	Use bus_dma to get temporary I/O buffers.
1.18	16-Sep-2001	wiz	Spell 'occurred' with two 'r's.
1.17	15-Sep-2001	chs	a whole bunch of changes to improve performance and robustness under load: - remove special treatment of pager_map mappings in pmaps. this is required now, since I've removed the globals that expose the address range. pager_map now uses pmap_kenter_pa() instead of pmap_enter(), so there's no longer any need to special-case it. - eliminate struct uvm_vnode by moving its fields into struct vnode. - rewrite the pageout path. the pager is now responsible for handling the high-level requests instead of only getting control after a bunch of work has already been done on its behalf. this will allow us to UBCify LFS, which needs tighter control over its pages than other filesystems do. writing a page to disk no longer requires making it read-only, which allows us to write wired pages without causing all kinds of havoc. - use a new PG_PAGEOUT flag to indicate that a page should be freed on behalf of the pagedaemon when it's unlocked. this flag is very similar to PG_RELEASED, but unlike PG_RELEASED, PG_PAGEOUT can be cleared if the pageout fails due to eg. an indirect-block buffer being locked. this allows us to remove the "version" field from struct vm_page, and together with shrinking "loan_count" from 32 bits to 16, struct vm_page is now 4 bytes smaller. - no longer use PG_RELEASED for swap-backed pages. if the page is busy because it's being paged out, we can't release the swap slot to be reallocated until that write is complete, but unlike with vnodes we don't keep a count of in-progress writes so there's no good way to know when the write is done. instead, when we need to free a busy swap-backed page, just sleep until we can get it busy ourselves. - implement a fast-path for extending writes which allows us to avoid zeroing new pages. this substantially reduces cpu usage. - encapsulate the data used by the genfs code in a struct genfs_node, which must be the first element of the filesystem-specific vnode data for filesystems which use genfs_{get,put}pages(). - eliminate many of the UVM pagerops, since they aren't needed anymore now that the pager "put" operation is a higher-level operation. - enhance the genfs code to allow NFS to use the genfs_{get,put}pages instead of a modified copy. - clean up struct vnode by removing all the fields that used to be used by the vfs_cluster.c code (which we don't use anymore with UBC). - remove kmem_object and mb_object since they were useless. instead of allocating pages to these objects, we now just allocate pages with no object. such pages are mapped in the kernel until they are freed, so we can use the mapping to find the page to free it. this allows us to remove splvm() protection in several places. The sum of all these changes improves write throughput on my decstation 5000/200 to within 1% of the rate of NetBSD 1.5 and reduces the elapsed time for "make release" of a NetBSD 1.5 source tree on my 128MB pc to 10% less than a 1.5 kernel took.
1.16	19-Jul-2001	thorpej	branches: 1.16.2; Use BUS_DMA_READ and BUS_DMA_WRITE in some obvious places.
1.15	31-May-2001	ad	branches: 1.15.2; Reduce the delay when polling so that dumps go faster.
1.14	30-Mar-2001	ad	AEN 0x000c appears to mean `init started'. From lukem.
1.13	07-Mar-2001	thorpej	Add BUS_DMA_STREAMING to data transfer maps.
1.12	04-Mar-2001	ad	branches: 1.12.2; Don't choke when we get lots of AENs all at once.
1.11	25-Feb-2001	ad	Wait 10x longer for status bits.
1.10	11-Feb-2001	ad	Match an upcoming product.
1.9	23-Jan-2001	ad	Oops, be sure to clear the ATTN interrupt on error.
1.8	22-Jan-2001	ad	opt_twe.h isn't needed any more.
1.7	22-Jan-2001	ad	- Complain if interrupt-driven retrieval of AENs fails somehow. - Clamp the number of outstanding commands to 129. - Other tweaks.
1.6	14-Jan-2001	thorpej	splimp() -> splvm()
1.5	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.4	14-Nov-2000	thorpej	branches: 1.4.2; NBPG -> PAGE_SIZE
1.3	08-Nov-2000	ad	- Pull in some updates from FreeBSD - mainly clarifications. Thanks to Mike Smith <msmith@freebsd.org> for pointing these out. - Ensure that we get AENs: reserve one CCB/command block for parameter retrieval, and chain AEN retrievals that are the result of an interrupt.
1.2	20-Oct-2000	ad	- tsleep()/wakeup_one() will hurt when under load, so use a queue instead. - Don't copy to or from the aligned buffer unless there's a need to.
1.1	19-Oct-2000	ad	Driver for the 3ware Escalade family of RAID controllers.
1.4.2.8	21-Apr-2001	bouyer	Sync with HEAD
1.4.2.7	12-Mar-2001	bouyer	Sync with HEAD.
1.4.2.6	11-Feb-2001	bouyer	Sync with HEAD.
1.4.2.5	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.4.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.4.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.4.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	14-Nov-2000	bouyer	file twe.c was added on branch thorpej_scsipi on 2000-11-20 11:42:38 +0000
1.12.2.14	03-Jan-2003	thorpej	Sync with HEAD.
1.12.2.13	19-Dec-2002	thorpej	Sync with HEAD.
1.12.2.12	11-Dec-2002	thorpej	Sync with HEAD.
1.12.2.11	18-Oct-2002	nathanw	Catch up to -current.
1.12.2.10	13-Aug-2002	nathanw	Catch up to -current.
1.12.2.9	01-Aug-2002	nathanw	Catch up to -current.
1.12.2.8	20-Jun-2002	nathanw	Catch up to -current.
1.12.2.7	09-Jan-2002	nathanw	Fix merge botch.
1.12.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.12.2.5	22-Oct-2001	nathanw	Catch up to -current.
1.12.2.4	21-Sep-2001	nathanw	Catch up to -current.
1.12.2.3	24-Aug-2001	nathanw	Catch up with -current.
1.12.2.2	21-Jun-2001	nathanw	Catch up to -current.
1.12.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.15.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.15.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.15.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.15.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.15.2.1	03-Aug-2001	lukem	update to -current
1.16.2.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.16.2.1	01-Oct-2001	fvdl	Catch up with -current.
1.20.4.3	24-Nov-2001	he	Pull up revision 1.20 (via patch, requested by lukem): Improve stability of this driver by un-applying some earlier changes.
1.20.4.2	25-Oct-2001	he	Pull up revisions 1.1-1.19 (new, via patch, requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.20.4.1	10-Oct-2001	he	file twe.c was added on branch netbsd-1-5 on 2001-10-25 18:03:02 +0000
1.21.8.4	29-Aug-2002	gehenna	catch up with -current.
1.21.8.3	15-Jul-2002	gehenna	catch up with -current.
1.21.8.2	20-Jun-2002	gehenna	catch up with -current.
1.21.8.1	30-May-2002	gehenna	Catch up with -current.
1.22.2.1	26-May-2002	perry	partially resolve 1.6 pullup ticket #29 from grant pulls up revisions 1.23, 1.24 Original commit messages: > Deal with the NBPG problem in a better way. > NBPG is not constant on the sparc, so don't use CPP tricks. Note that grant erroneously requested a pullup of only 1.24, which would not apply cleanly. Normally I would have rejected the entire request, but this is needed to make the sparc & sparc64 ports compile.
1.37.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.37.2.8	01-Apr-2005	skrll	Sync with HEAD.
1.37.2.7	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.37.2.6	15-Feb-2005	skrll	Sync with HEAD.
1.37.2.5	24-Sep-2004	skrll	Sync with HEAD.
1.37.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.37.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.37.2.2	03-Aug-2004	skrll	Sync with HEAD
1.37.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.54.2.2	05-Jun-2004	jmc	Pullup rev 1.58 (requested by thorpej in ticket #441) twe_ccb_map(): Remove the DEBUG panic case of "unknown" S/G list offset.
1.54.2.1	30-May-2004	tron	Pull up revision 1.56 (requested by thorpej in ticket #410): Rename twe_tweio_command_handler() to twe_ccb_wait_handler() and expose it.
1.60.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.60.6.1	25-Jan-2005	yamt	convert to new apis.
1.60.4.1	29-Apr-2005	kent	sync with -current
1.64.4.1	26-Feb-2006	riz	Pull up following revision(s) (requested by simonb in ticket #1182): sys/dev/pci/twe.c: revision 1.74 Fix memory leak in the TWEIO_GET_PARAM ioctl. Tested by gendalia@. --
1.64.2.1	26-Feb-2006	riz	Pull up following revision(s) (requested by simonb in ticket #1182): sys/dev/pci/twe.c: revision 1.74 Fix memory leak in the TWEIO_GET_PARAM ioctl. Tested by gendalia@. --
1.68.2.4	27-Oct-2007	yamt	sync with head.
1.68.2.3	03-Sep-2007	yamt	sync with head.
1.68.2.2	30-Dec-2006	yamt	sync with head.
1.68.2.1	21-Jun-2006	yamt	sync with head.
1.72.2.2	01-Mar-2006	yamt	sync with head.
1.72.2.1	01-Feb-2006	yamt	sync with head.
1.73.4.1	22-Apr-2006	simonb	Sync with head.
1.73.2.1	09-Sep-2006	rpaulo	sync with head
1.74.6.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.74.4.1	19-Apr-2006	elad	sync with head.
1.74.2.2	03-Sep-2006	yamt	sync with head.
1.74.2.1	24-May-2006	yamt	sync with head.
1.78.4.2	10-Dec-2006	yamt	sync with head.
1.78.4.1	22-Oct-2006	yamt	sync with head
1.78.2.2	12-Jan-2007	ad	Sync with head.
1.78.2.1	18-Nov-2006	ad	Sync with head.
1.81.2.1	04-Dec-2006	tron	Pull up following revision(s) (requested by elad in ticket #247): sys/dev/ic/dpt.c: revision 1.55 sys/dev/pci/amr.c: revision 1.43 sys/secmodel/bsd44/secmodel_bsd44_securelevel.c: revision 1.19 sys/dev/pci/mly.c: revision 1.33 share/man/man9/kauth.9: revision 1.37 sys/dev/ic/mlx.c: revision 1.49 sys/dev/ic/icp_ioctl.c: revision 1.14 sys/dev/i2o/iop.c: revision 1.62 sys/dev/pci/twe.c: revision 1.82 sys/sys/kauth.h: revision 1.25 sys/dev/i2o/dpti.c: revision 1.31 sys/kern/kern_auth.c: revision 1.33 sys/dev/tc/stic.c: revision 1.37 Change kauth(9) KPI for kauth_authorize_device_passthru() to add another argument, u_long, serving as a bit-mask of generic requests for the passthru request. Discussed on tech-security@ and tech-kern@. Okay tls@.
1.82.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.83.20.1	25-Oct-2007	bouyer	Sync with HEAD.
1.83.16.1	06-Nov-2007	matt	sync with HEAD
1.83.14.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.83.2.1	23-Oct-2007	ad	Sync with head.
1.84.16.3	29-Jun-2008	mjf	Sync with HEAD.
1.84.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.84.16.1	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.85.4.4	11-Aug-2010	yamt	sync with head.
1.85.4.3	16-May-2009	yamt	sync with head
1.85.4.2	04-May-2009	yamt	sync with head.
1.85.4.1	16-May-2008	yamt	sync with head.
1.85.2.2	17-Jun-2008	yamt	sync with head.
1.85.2.1	18-May-2008	yamt	sync with head.
1.86.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.87.12.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.90.4.2	31-May-2011	rmind	sync with head
1.90.4.1	05-Mar-2011	rmind	sync with head
1.90.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.93.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.95.6.1	18-Feb-2012	mrg	merge to -current.
1.95.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.95.2.3	16-Jan-2013	yamt	sync with (a bit old) head
1.95.2.2	30-Oct-2012	yamt	sync with head
1.95.2.1	17-Apr-2012	yamt	sync with head
1.96.6.4	03-Dec-2017	jdolecek	update from HEAD
1.96.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.96.6.2	25-Feb-2013	tls	resync with head
1.96.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.98.2.1	18-May-2014	rmind	sync with head
1.102.2.1	10-Aug-2014	tls	Rebase.
1.103.4.2	05-Oct-2016	skrll	Sync with HEAD
1.103.4.1	09-Jul-2016	skrll	Sync with HEAD
1.105.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.106.16.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.106.16.1	10-Jun-2019	christos	Sync with HEAD
1.106.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.106.6.1	30-Apr-2017	pgoyette	More DEVSW_MODULE_INIT for modular drivers
1.108.10.3	28-Mar-2021	thorpej	Correct the names of the arguments passed to the rescan function.
1.108.10.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.108.10.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.109.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	11-Dec-2005	christos	merge ktrace-lwp.
1.4	25-Sep-2003	thorpej	- Protect against multiple inclusion. - Pull in <dev/pci/twereg.h> ourselves, as well.
1.3	25-Sep-2003	thorpej	- Fix a bug I introduced in a previous commit (oops). - Add the twe_drivecommand ioctl structure.
1.2	22-Sep-2003	thorpej	Add TWEIO_{ADD,DEL}_UNIT ioctls (used by 3ware management tools). From FreeBSD.
1.1	13-Dec-2002	christos	branches: 1.1.6; 1.1.8; Add /dev/twex support. To be used with FreeBSD's 3dmd. Initial port from Masaru OKI <oki@netbsd.org>, with KNF, memory leak fixes, denial of service problems, and null pointer fixes from me.
1.1.8.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.8.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.8.1	03-Aug-2004	skrll	Sync with HEAD
1.1.6.2	19-Dec-2002	thorpej	Sync with HEAD.
1.1.6.1	13-Dec-2002	thorpej	file tweio.h was added on branch nathanw_sa on 2002-12-19 00:48:15 +0000
1.16	08-Nov-2018	msaitoh	"s/ are are / are /" in comment. No functional change.
1.15	08-Sep-2008	gmcgarry	branches: 1.15.80; 1.15.82; Replace most gcc-specific __attribute__ uses with BSD-style sys/cdef.h preprocessor macros.
1.14	28-Apr-2008	martin	branches: 1.14.2; 1.14.6; Remove clause 3 and 4 from TNF licenses
1.13	21-May-2006	christos	branches: 1.13.56; 1.13.58; 1.13.60; Remove packed attributes gcc-4 dislikes.
1.12	11-Dec-2005	christos	branches: 1.12.4; 1.12.6; 1.12.8; 1.12.12; 1.12.14; merge ktrace-lwp.
1.11	15-Feb-2005	lukem	branches: 1.11.6; Add some more stripe sizes observed in the field. Add "JBOD" to twe_table_unittype[].
1.10	05-Oct-2004	heas	branches: 1.10.4; 1.10.6; Add TWE_PARAM_PROC_PERCENT; completion percentage per-sub-unit (2 bytes/ea) for initialize/verify/rebuild processes.
1.9	05-Oct-2004	heas	Add TWE_OP_PROBEUNIT (0x7), which probes a drive (not associated with a unit from the controller's POV) for unit information. It's arguments mirror that of TWE_OP_PROBE.
1.8	25-Sep-2004	heas	Add TWE_OP_PROBE, which probes (and attaches) a port/drive. It's argument is bytes 9 & 10 of the twe_cmd where the bit set corresponds to the port to probe + 1 (ie: port 0 == 0x1).
1.7	23-Sep-2004	heas	Typo in define; PhysSectorNym -> PhysSectorNum
1.6	22-Sep-2003	thorpej	Add a few additional control bits and opcodes. From FreeBSD.
1.5	22-Jan-2001	ad	branches: 1.5.10; 1.5.26; - Complain if interrupt-driven retrieval of AENs fails somehow. - Clamp the number of outstanding commands to 129. - Other tweaks.
1.4	14-Nov-2000	thorpej	branches: 1.4.2; NBPG -> PAGE_SIZE
1.3	08-Nov-2000	ad	- Pull in some updates from FreeBSD - mainly clarifications. Thanks to Mike Smith <msmith@freebsd.org> for pointing these out. - Ensure that we get AENs: reserve one CCB/command block for parameter retrieval, and chain AEN retrievals that are the result of an interrupt.
1.2	26-Oct-2000	ad	Work around a fatal firmware bug on some boards. From msmith@freebsd.org.
1.1	19-Oct-2000	ad	Driver for the 3ware Escalade family of RAID controllers.
1.4.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.4.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.4.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	14-Nov-2000	bouyer	file twereg.h was added on branch thorpej_scsipi on 2000-11-20 11:42:38 +0000
1.5.26.6	15-Feb-2005	skrll	Sync with HEAD.
1.5.26.5	19-Oct-2004	skrll	Sync with HEAD
1.5.26.4	24-Sep-2004	skrll	Sync with HEAD.
1.5.26.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.5.26.2	18-Sep-2004	skrll	Sync with HEAD.
1.5.26.1	03-Aug-2004	skrll	Sync with HEAD
1.5.10.2	25-Oct-2001	he	Pull up revisions 1.1-1.5 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.5.10.1	22-Jan-2001	he	file twereg.h was added on branch netbsd-1-5 on 2001-10-25 17:53:44 +0000
1.10.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.10.4.1	29-Apr-2005	kent	sync with -current
1.11.6.1	21-Jun-2006	yamt	sync with head.
1.12.14.1	19-Jun-2006	chap	Sync with head.
1.12.12.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.12.8.1	24-May-2006	yamt	sync with head.
1.12.6.1	01-Jun-2006	kardel	Sync with head.
1.12.4.1	09-Sep-2006	rpaulo	sync with head
1.13.60.2	04-May-2009	yamt	sync with head.
1.13.60.1	16-May-2008	yamt	sync with head.
1.13.58.1	18-May-2008	yamt	sync with head.
1.13.56.2	28-Sep-2008	mjf	Sync with HEAD.
1.13.56.1	02-Jun-2008	mjf	Sync with HEAD.
1.14.6.1	19-Oct-2008	haad	Sync with HEAD.
1.14.2.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.15.82.1	10-Jun-2019	christos	Sync with HEAD
1.15.80.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.30	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.29	07-May-2009	cegger	branches: 1.29.12; 1.29.22; struct device * -> device_t, no functional changes intended.
1.28	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.27	28-Apr-2008	martin	branches: 1.27.14; Remove clause 3 and 4 from TNF licenses
1.26	04-Mar-2007	christos	branches: 1.26.36; 1.26.38; 1.26.40; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.25	16-Feb-2006	perry	branches: 1.25.20; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.24	24-Dec-2005	perry	branches: 1.24.2; 1.24.4; 1.24.6; __inline__ -> inline
1.23	11-Dec-2005	christos	merge ktrace-lwp.
1.22	13-Sep-2004	drochner	branches: 1.22.12; a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.21	27-May-2004	thorpej	Rename twe_tweio_command_handler() to twe_ccb_wait_handler() and expose it.
1.20	23-Sep-2003	thorpej	branches: 1.20.2; - Make CCB allocation slightly more efficient by changing how the CCB is returned to the caller. - Make code paths that can use twe_ccb_alloc_wait() use it, and assert that a CCB is always returned from that function. - Assert that a CCB is always returned when allocating the reserved CCB for an AEN fetch.
1.19	23-Sep-2003	thorpej	Separate the AEN fetching path into its own special path that uses the reserved CCB. This means that all remaining callers of twe_param_get*() are called from a valid thread context, and thus have no need to use a reserved CCB. This will allow for further cleanup in a future commit.
1.18	22-Sep-2003	thorpej	Add support for dynamically attaching and detaching RAID array units.
1.17	22-Sep-2003	thorpej	Improve AEN handling: - Make AENs use the generic code table stuff. - Add a few more AEN codes (from FreeBSD). - Correct the context of a few AEN codes (some were listed as "unit context" when they're really "port context"). - Add a queue of AENs that management tools in userspace can poll (from FreeBSD).
1.16	21-Sep-2003	thorpej	Expose the "param get" functions.
1.15	21-Sep-2003	thorpej	- Record more information about the array unit, including array type and stripe depth. - Report array type and stripe depth when attaching the logical drive.
1.14	21-Sep-2003	thorpej	Add some additional code -> description tables, and add a generic routine to translate code -> description. AENs will be converted to this mechanism in a future commit. Partially from FreeBSD.
1.13	13-Dec-2002	christos	branches: 1.13.6; Add /dev/twex support. To be used with FreeBSD's 3dmd. Initial port from Masaru OKI <oki@netbsd.org>, with KNF, memory leak fixes, denial of service problems, and null pointer fixes from me.
1.12	24-May-2002	christos	NBPG is not constant on the sparc, so don't use CPP tricks.
1.11	24-May-2002	christos	Deal with the NBPG problem in a better way.
1.10	18-May-2002	ad	branches: 1.10.2; - Use write barriers. - NBPG -> PAGE_SIZE.
1.9	10-Oct-2001	ad	branches: 1.9.4; 1.9.12; Back out previous - it causes panics.
1.8	20-Sep-2001	ad	Use bus_dma to get temporary I/O buffers.
1.7	04-Mar-2001	ad	branches: 1.7.2; 1.7.4; 1.7.6; Don't choke when we get lots of AENs all at once.
1.6	23-Jan-2001	ad	Fix up license.
1.5	22-Jan-2001	ad	- Complain if interrupt-driven retrieval of AENs fails somehow. - Clamp the number of outstanding commands to 129. - Other tweaks.
1.4	14-Nov-2000	thorpej	branches: 1.4.2; NBPG -> PAGE_SIZE
1.3	08-Nov-2000	ad	- Pull in some updates from FreeBSD - mainly clarifications. Thanks to Mike Smith <msmith@freebsd.org> for pointing these out. - Ensure that we get AENs: reserve one CCB/command block for parameter retrieval, and chain AEN retrievals that are the result of an interrupt.
1.2	20-Oct-2000	ad	- tsleep()/wakeup_one() will hurt when under load, so use a queue instead. - Don't copy to or from the aligned buffer unless there's a need to.
1.1	19-Oct-2000	ad	Driver for the 3ware Escalade family of RAID controllers.
1.4.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.4.2.4	11-Feb-2001	bouyer	Sync with HEAD.
1.4.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.4.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.2.1	14-Nov-2000	bouyer	file twevar.h was added on branch thorpej_scsipi on 2000-11-20 11:42:38 +0000
1.7.6.2	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.7.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.7.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.7.2.4	19-Dec-2002	thorpej	Sync with HEAD.
1.7.2.3	20-Jun-2002	nathanw	Catch up to -current.
1.7.2.2	22-Oct-2001	nathanw	Catch up to -current.
1.7.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.9.12.1	30-May-2002	gehenna	Catch up with -current.
1.9.4.3	24-Nov-2001	he	Pull up revision 1.9 (requested by lukem): Improve stability of this driver by un-applying some earlier changes.
1.9.4.2	25-Oct-2001	he	Pull up revisions 1.1-1.8 (requested by ad): Add Mylex DACC960, CAC-EISA, and I2O block/SCSI drivers.
1.9.4.1	10-Oct-2001	he	file twevar.h was added on branch netbsd-1-5 on 2001-10-25 17:53:55 +0000
1.10.2.1	26-May-2002	perry	partially resolve 1.6 pullup ticket #29 from grant pulls up revisions 1.11, 1.12 Original commit messages: > Deal with the NBPG problem in a better way. > NBPG is not constant on the sparc, so don't use CPP tricks. Note that grant erroneously requested a pullup of only 1.12, which would not apply cleanly. Normally I would have rejected the entire request, but this is needed to make the sparc & sparc64 ports compile.
1.13.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.13.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.13.6.1	03-Aug-2004	skrll	Sync with HEAD
1.20.2.1	30-May-2004	tron	Pull up revision 1.21 (requested by thorpej in ticket #410): Rename twe_tweio_command_handler() to twe_ccb_wait_handler() and expose it.
1.22.12.2	03-Sep-2007	yamt	sync with head.
1.22.12.1	21-Jun-2006	yamt	sync with head.
1.24.6.1	22-Apr-2006	simonb	Sync with head.
1.24.4.1	09-Sep-2006	rpaulo	sync with head
1.24.2.1	18-Feb-2006	yamt	sync with head.
1.25.20.1	12-Mar-2007	rmind	Sync with HEAD.
1.26.40.2	16-May-2009	yamt	sync with head
1.26.40.1	16-May-2008	yamt	sync with head.
1.26.38.1	18-May-2008	yamt	sync with head.
1.26.36.1	02-Jun-2008	mjf	Sync with HEAD.
1.27.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.29.22.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.29.12.1	30-Oct-2012	yamt	sync with head
1.65	23-Feb-2024	andvar	s/opetions/options/ in hw.ubsec node description.
1.64	04-Aug-2023	riastradh	entropy(9): Simplify stages. Split interrupt vs non-interrupt paths. - Nix the entropy stage (cold, warm, hot). Just use the usual kernel `cold' (cold: single-core, single-thread; interrupts may happen), and don't make any three-way distinction about whether interrupts or threads or other CPUs can be running. Instead, while cold, use splhigh/splx or forbid paths to come from interrupt context, and while warm, use mutex or the per-CPU hard and soft interrupt paths for low latency. This comes at a small cost to some interrupt latency, since we may stir the pool in interrupt context -- but only for a very short window early at boot between configure and configure2, so it's hard to imagine it matters much. - Allow rnd_add_uint32 to run in hard interrupt context or with spin locks held, but defer processing to softint and drop samples on the floor if buffer is full. This is mainly used for cheaply tossing samples from drivers for non-HWRNG devices into the entropy pool, so it is often used from interrupt context and/or under spin locks. - New rnd_add_data_intr provides the interrupt-like data entry path for arbitrary buffers and driver-specified entropy estimates: defer processing to softint and drop samples on the floor if buffer is full. - Document that rnd_add_data is forbidden under spin locks outside interrupt context (will crash in LOCKDEBUG), and inadvisable in interrupt context (but technically permitted just in case there are compatibility issues for now); later we can forbid it altogether in interrupt context or under spin locks. - Audit all uses of rnd_add_data to use rnd_add_data_intr where it might be used in interrupt context or under a spin lock. This fixes a regression from last year when the global entropy lock was changed from IPL_VM (spin) to IPL_SOFTSERIAL (adaptive). Thought I'd caught all the problems from that, but another one bit three different people this week, presumably because of recent changes that led to more non-HWRNG drivers entering the entropy consolidation path from rnd_add_uint32. In my attempt to preserve the rnd(9) API for the (now long-since abandoned) prospect of pullup to netbsd-9 in my rewrite of the entropy subsystem in 2020, I didn't introduce a separate entry point for entering entropy from interrupt context or equivalent, i.e., spin locks held, and instead made rnd_add_data rely on cpu_intr_p() to decide whether to process the whole sample under a lock or only take as much as there's buffer space for before scheduling a softint. In retrospect, that was a mistake (though perhaps not as much of a mistake as other entropy API decisions...), a mistake which is finally getting rectified now by rnd_add_data_intr. XXX pullup-10
1.63	06-May-2023	andvar	s/Regiser/Register/ and s/regester/register/ in comments.
1.62	27-Aug-2022	skrll	branches: 1.62.4; Correct the abbreviation of approximately to 'approx.'
1.61	27-Aug-2022	skrll	Trailing whitespace
1.60	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.59	22-May-2022	riastradh	opencrypto: Make freesession callback return void. No functional change intended: all drivers already return zero unconditionally.
1.58	22-May-2022	riastradh	ubsec(4): Prune dead branches. Assert session id validity.
1.57	22-May-2022	riastradh	ubsec(4): Tidy up error branches of ubsec_process. Make sure to return zero, not error, when we've already done crypto_done and set crp_etype.
1.56	22-May-2022	riastradh	ubsec(4): Assert crp_sid is valid. If opencrypto passes a bad sid in, that's a bug in opencrypto that needs to be fixed, not a user-triggered invalid input that we need to fail gracefully on.
1.55	22-May-2022	riastradh	ubsec(4): ubsec_kprocess always returns zero. Prune dead branches.
1.54	22-May-2022	riastradh	ubsec(4): Fix error branch: call crypto_kdone, don't return error.
1.53	18-May-2022	riastradh	ubsec(4): Nix dead code. No functional change intended.
1.52	14-Jun-2020	riastradh	ubsec(4): Don't use prev msg's last block as IV for next msg in CBC. This violates the security contract of the CBC construction, which requires that the IV be unpredictable in advance; an adaptive adversary can exploit this to verify plaintext guesses. XXX Compile-tested only.
1.51	25-May-2020	thorpej	unifdef everything-but-NetBSD. Gets rid of a naked cfattach decl.
1.50	30-Apr-2020	riastradh	rnd_attach_source calls the callback itself now. No need for every driver to explicitly call it to prime the pool. Eliminate now-unused <sys/rndpool.h>.
1.49	30-Apr-2020	riastradh	Don't attach rndsource until it's actually ready to run.
1.48	16-Mar-2020	pgoyette	Use the module subsystem's ability to process SYSCTL_SETUP() entries to automate installation of sysctl nodes. Note that there are still a number of device and pseudo-device modules that create entries tied to individual device units, rather than to the module itself. These are not changed.
1.47	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.46	22-Dec-2018	maxv	Replace: M_COPY_PKTHDR -> m_copy_pkthdr. No functional change, since the former is a macro to the latter.
1.45	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.44	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.43	07-Jul-2016	msaitoh	branches: 1.43.16; 1.43.18; KNF. Remove extra spaces. No functional change.
1.42	13-Apr-2015	riastradh	Add header guards and necessary includes.
1.41	10-Aug-2014	tls	branches: 1.41.4; Merge tls-earlyentropy branch into HEAD.
1.40	19-Apr-2014	bad	Add support for accelerated AES_CBC in ubsec(4) for BCM5823 and newer. Update man-page and bump date. Adjust OpenBSD RCS IDs to reflect roughly the version we are in sync with.
1.39	18-Apr-2014	bad	Rewrite the dmamap handling to allocate and cache the dmamaps beforehand. Calling bus_dmamap_create/destroy is no longer possible in interrupt context. Move the dmamaps to the end of struct ubsec_q so the rest of the struct can be cleared with one call to memset(). As a bonus we get a 25% increase in throughput encrypting 8K blocks.
1.38	29-Mar-2014	christos	branches: 1.38.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.37	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.36	03-Jan-2014	pgoyette	If you're going to build this driver as a module, make sure that it depends on opencrypto module.
1.35	26-Dec-2013	bad	Also disable MCR4INT in ubsec_detach() when appropriate.
1.34	17-Nov-2013	bad	Add support for BCM5825, and BCM5860, 5861, 5862 from OpenBSD rev 1.143: Add support for the BCM5825 and the next-generation BCM5860, 5861, 5862 Broadcom CryptoNetX IPSec/SSL Security Processors. The 5825 is a faster version of the already supported 5823, and the even faster 586x series is a bit different and needed some more changes. AES support hasn't been pulled in yet.
1.33	17-Nov-2013	bad	Create the sysctl variables on module initialization. Create them under hw.ubsec as is hip these days.
1.32	17-Nov-2013	bad	Make ubsec(4) loadable as kmod.
1.31	17-Nov-2013	bad	Fix locking botch. Callers of ubsec_rng_locked() lock and release sc_mtx already.
1.30	17-Nov-2013	bad	Use callout_setfunc()/callout_schedule() instead of callout_reset(), it is cheaper.
1.29	13-Jun-2013	tls	branches: 1.29.2; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.28	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.27	30-Jan-2012	drochner	branches: 1.27.6; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.26	19-Nov-2011	tls	branches: 1.26.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.25	13-Nov-2010	uebayasi	branches: 1.25.8; Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.24	26-Nov-2009	njoly	branches: 1.24.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.23	12-May-2009	cegger	use device_private(). "looks good" ad@ XXX for the device_t/softc split, please check the driver that no cases have been missed.
1.22	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.21	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.20	18-Apr-2009	tsutsui	Remove extra whitespace added by a stupid tool. XXX: more in src/sys/arch
1.19	18-Mar-2009	cegger	bcopy -> memcpy
1.18	18-Mar-2009	cegger	bzero -> memset
1.17	19-Dec-2008	cegger	branches: 1.17.2; use M_ZERO on malloc() and remove subsequent bzero().
1.16	10-Apr-2008	cegger	branches: 1.16.4; 1.16.12; use aprint_*_dev and device_xname
1.15	02-Feb-2008	tls	branches: 1.15.6; From Darran Hunt at Coyote Point: don't truncate HMAC to 96 bits unless actually asked to. Fixed in FreeBSD a while ago, discussed on tech-kern and tech-crypto.
1.14	11-Dec-2007	lukem	use __KERNEL_RCSID()
1.13	09-Jul-2007	ad	branches: 1.13.8; 1.13.14; 1.13.16; 1.13.18; 1.13.20; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.12	04-Mar-2007	christos	branches: 1.12.2; 1.12.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.11	16-Nov-2006	christos	branches: 1.11.4; __unused removal on arguments; approved by core.
1.10	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.9	28-Mar-2006	thorpej	branches: 1.9.8; 1.9.10; Use device_unit().
1.8	25-Nov-2005	thorpej	branches: 1.8.4; 1.8.6; 1.8.8; 1.8.10; 1.8.12; - De-couple the software crypto implementation from the rest of the framework. There is no need to waste the space if you are only using algoritms provided by hardware accelerators. To get the software implementations, add "pseudo-device swcr" to your kernel config. - Lazily initialize the opencrypto framework when crypto drivers (either hardware or swcr) register themselves with the framework.
1.7	28-Jun-2005	thorpej	branches: 1.7.2; 1.7.8; Use ANSI function decls and static.
1.6	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	28-Aug-2003	thorpej	branches: 1.4.4; 1.4.10; 1.4.12; * Make matching and chip info table-driven. * Print product name and revision at attach time. * Use aprint_*().
1.3	27-Aug-2003	thorpej	Fix some diagnotic printfs.
1.2	21-Aug-2003	jonathan	Remove #ifdef _OpenBSD__ code which sets the softc pointer `sc' by passing an index into ubsec_cd.cd_devs[]: that causes too much confusion with the checks that sc is non-null.
1.1	01-Aug-2003	jonathan	Preliminary port of merged OpenBSD/FreeBSD ubsec driver for Bluesteel Networks/Broadcom line of cryptographic accelerators.
1.4.12.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.10.1	29-Apr-2005	kent	sync with -current
1.4.4.7	11-Dec-2005	christos	Sync with head.
1.4.4.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.4.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.4.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.4.4.2	03-Aug-2004	skrll	Sync with HEAD
1.4.4.1	28-Aug-2003	skrll	file ubsec.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.7.8.1	29-Nov-2005	yamt	sync with head.
1.7.2.5	04-Feb-2008	yamt	sync with head.
1.7.2.4	21-Jan-2008	yamt	sync with head
1.7.2.3	03-Sep-2007	yamt	sync with head.
1.7.2.2	30-Dec-2006	yamt	sync with head.
1.7.2.1	21-Jun-2006	yamt	sync with head.
1.8.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.8.10.1	19-Apr-2006	elad	sync with head.
1.8.8.1	01-Apr-2006	yamt	sync with head.
1.8.6.1	22-Apr-2006	simonb	Sync with head.
1.8.4.1	09-Sep-2006	rpaulo	sync with head
1.9.10.2	10-Dec-2006	yamt	sync with head.
1.9.10.1	22-Oct-2006	yamt	sync with head
1.9.8.1	18-Nov-2006	ad	Sync with head.
1.11.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.12.4.1	11-Jul-2007	mjf	Sync with head.
1.12.2.1	01-Jul-2007	ad	Adapt to callout API change.
1.13.20.1	13-Dec-2007	bouyer	Sync with HEAD
1.13.18.1	11-Dec-2007	yamt	sync with head.
1.13.16.1	26-Dec-2007	ad	Sync with head.
1.13.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.13.8.2	23-Mar-2008	matt	sync with HEAD
1.13.8.1	09-Jan-2008	matt	sync with HEAD
1.15.6.2	17-Jan-2009	mjf	Sync with HEAD.
1.15.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.16.12.2	28-Apr-2009	skrll	Sync with HEAD.
1.16.12.1	19-Jan-2009	skrll	Sync with HEAD.
1.16.4.3	11-Mar-2010	yamt	sync with head
1.16.4.2	16-May-2009	yamt	sync with head
1.16.4.1	04-May-2009	yamt	sync with head.
1.17.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.24.4.1	05-Mar-2011	rmind	sync with head
1.25.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.25.8.2	30-Oct-2012	yamt	sync with head
1.25.8.1	17-Apr-2012	yamt	sync with head
1.26.2.1	18-Feb-2012	mrg	merge to -current.
1.27.6.4	03-Dec-2017	jdolecek	update from HEAD
1.27.6.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.27.6.2	23-Jun-2013	tls	resync from head
1.27.6.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.29.2.1	18-May-2014	rmind	sync with head
1.38.2.2	10-Aug-2014	tls	Rebase.
1.38.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.41.4.2	09-Jul-2016	skrll	Sync with HEAD
1.41.4.1	06-Jun-2015	skrll	Sync with HEAD
1.43.18.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.43.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.43.18.1	10-Jun-2019	christos	Sync with HEAD
1.43.16.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.43.16.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.62.4.1	11-Aug-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #319): sys/dev/pci/ubsec.c: revision 1.64 sys/dev/pci/hifn7751.c: revision 1.82 lib/libc/gen/getentropy.3: revision 1.5 lib/libc/gen/getentropy.3: revision 1.6 share/man/man4/rnd.4: revision 1.41 lib/libc/sys/getrandom.2: revision 1.2 lib/libc/sys/getrandom.2: revision 1.3 share/man/man5/rc.conf.5: revision 1.193 share/man/man7/entropy.7: revision 1.5 share/man/man7/entropy.7: revision 1.6 share/man/man7/entropy.7: revision 1.7 share/man/man7/entropy.7: revision 1.8 etc/security: revision 1.130 share/man/man7/entropy.7: revision 1.9 etc/security: revision 1.131 sys/crypto/cprng_fast/cprng_fast.c: revision 1.19 sys/sys/rndio.h: revision 1.3 tests/lib/libc/sys/t_getrandom.c: revision 1.5 etc/defaults/rc.conf: revision 1.164 etc/defaults/rc.conf: revision 1.165 sys/sys/rndsource.h: revision 1.10 sys/kern/kern_entropy.c: revision 1.62 sys/kern/kern_entropy.c: revision 1.63 sys/kern/kern_entropy.c: revision 1.64 sys/kern/subr_cprng.c: revision 1.44 sys/kern/kern_entropy.c: revision 1.65 sys/kern/kern_clock.c: revision 1.149 sys/dev/pci/viornd.c: revision 1.22 share/man/man9/rnd.9: revision 1.32 sys/kern/subr_prf.c: revision 1.202 sys/sys/rndsource.h: revision 1.8 sys/sys/rndsource.h: revision 1.9 share/man/man7/entropy.7: revision 1.10 1. Reinstate netbsd<=9 entropy estimator to unblock /dev/random, in parallel with assessment of only confident entropy sources (seed, HWRNG) for security warnings like sshd keys in motd and daily insecurity report. 2. Make multiuser boot wait for first /dev/random output soon after loading a seed and configuring rndctl, so that getentropy(3) meets its contract starting early at boot without introducing blocking paths that could cause hangs in init(8) or single-user mode. Operators can choose to disable this wait in rc.conf. 3. Fix some bugs left over from reducing the global entropy lock from a spin lock at IPL_VM to an adaptive lock at IPL_SOFTSERIAL. 4. Update man pages.
1.5	13-Apr-2015	riastradh	Add header guards and necessary includes.
1.4	19-Apr-2014	bad	branches: 1.4.4; Add support for accelerated AES_CBC in ubsec(4) for BCM5823 and newer. Update man-page and bump date. Adjust OpenBSD RCS IDs to reflect roughly the version we are in sync with.
1.3	17-Nov-2013	bad	branches: 1.3.2; Add support for BCM5825, and BCM5860, 5861, 5862 from OpenBSD rev 1.143: Add support for the BCM5825 and the next-generation BCM5860, 5861, 5862 Broadcom CryptoNetX IPSec/SSL Security Processors. The 5825 is a faster version of the already supported 5823, and the even faster 586x series is a bit different and needed some more changes. AES support hasn't been pulled in yet.
1.2	11-Dec-2005	christos	branches: 1.2.110; 1.2.120; 1.2.126; merge ktrace-lwp.
1.1	01-Aug-2003	jonathan	branches: 1.1.4; Preliminary port of merged OpenBSD/FreeBSD ubsec driver for Bluesteel Networks/Broadcom line of cryptographic accelerators.
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	01-Aug-2003	skrll	file ubsecreg.h was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.2.126.1	18-May-2014	rmind	sync with head
1.2.120.2	03-Dec-2017	jdolecek	update from HEAD
1.2.120.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.110.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.2.1	10-Aug-2014	tls	Rebase.
1.4.4.1	06-Jun-2015	skrll	Sync with HEAD
1.11	14-Jun-2020	riastradh	ubsec(4): Don't use prev msg's last block as IV for next msg in CBC. This violates the security contract of the CBC construction, which requires that the IV be unpredictable in advance; an adaptive adversary can exploit this to verify plaintext guesses. XXX Compile-tested only.
1.10	13-Apr-2015	riastradh	Add header guards and necessary includes.
1.9	19-Apr-2014	bad	branches: 1.9.4; Add support for accelerated AES_CBC in ubsec(4) for BCM5823 and newer. Update man-page and bump date. Adjust OpenBSD RCS IDs to reflect roughly the version we are in sync with.
1.8	18-Apr-2014	bad	Rewrite the dmamap handling to allocate and cache the dmamaps beforehand. Calling bus_dmamap_create/destroy is no longer possible in interrupt context. Move the dmamaps to the end of struct ubsec_q so the rest of the struct can be cleared with one call to memset(). As a bonus we get a 25% increase in throughput encrypting 8K blocks.
1.7	17-Nov-2013	bad	branches: 1.7.2; Add support for BCM5825, and BCM5860, 5861, 5862 from OpenBSD rev 1.143: Add support for the BCM5825 and the next-generation BCM5860, 5861, 5862 Broadcom CryptoNetX IPSec/SSL Security Processors. The 5825 is a faster version of the already supported 5823, and the even faster 586x series is a bit different and needed some more changes. AES support hasn't been pulled in yet.
1.6	17-Nov-2013	bad	Make ubsec(4) loadable as kmod.
1.5	13-Jun-2013	tls	branches: 1.5.2; Convert the entropy pool framework from pseudo-callout-driven to soft interrupt driven operation. Add a polling mode of operation -- now we can ask hardware random number generators to top us up just when we need it (bcm2835_rng and amdpm converted as examples). Fix a stall noticed with repeated reads from /dev/random while testing.
1.4	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.3	04-Mar-2007	christos	branches: 1.3.76; 1.3.86; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.2	11-Dec-2005	christos	branches: 1.2.26; merge ktrace-lwp.
1.1	01-Aug-2003	jonathan	branches: 1.1.4; 1.1.18; Preliminary port of merged OpenBSD/FreeBSD ubsec driver for Bluesteel Networks/Broadcom line of cryptographic accelerators.
1.1.18.1	03-Sep-2007	yamt	sync with head.
1.1.4.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.4.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.4.2	03-Aug-2004	skrll	Sync with HEAD
1.1.4.1	01-Aug-2003	skrll	file ubsecvar.h was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.2.26.1	12-Mar-2007	rmind	Sync with HEAD.
1.3.86.4	03-Dec-2017	jdolecek	update from HEAD
1.3.86.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.3.86.2	23-Jun-2013	tls	resync from head
1.3.86.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.3.76.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.76.1	30-Oct-2012	yamt	sync with head
1.5.2.1	18-May-2014	rmind	sync with head
1.7.2.1	10-Aug-2014	tls	Rebase.
1.9.4.1	06-Jun-2015	skrll	Sync with HEAD
1.67	29-Jun-2022	mlelstv	Unmap device registers only when successfully mapped.
1.66	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.65	24-Apr-2021	thorpej	branches: 1.65.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.64	04-Jul-2020	jdolecek	branches: 1.64.4; switch to pci_intr_alloc() so this uses MSI/MSI-X if device supports it
1.63	10-May-2018	msaitoh	KNF. No binary change.
1.62	09-Apr-2018	jakllsch	Stop potential misuse of vendor names and USB vendor IDs in root hub device and string descriptors. Firstly: Few vendors have identical PCI-SIG vendor IDs and USB-IF vendor IDs. As such, using the PCI vendor ID as a USB vendor ID may trample on whomever is allocated that USB vendor ID. Secondly: The vendor of the host controller hardware implementation has little to nothing to do with our usbroothub implementation. Thus we should not potentially associate any problems therewith to such third party. This change will result in root hubs being identified by USB Vendor ID 0x0000. Root hub vendor string will now be "NetBSD" (or, specifically: ostype). Product ID (0x0000) and product strings remain unchanged.
1.61	27-Apr-2017	msaitoh	branches: 1.61.8; Use {isa,pci}_intr_establish"_xname"() for intrctl(8).
1.60	23-Apr-2016	skrll	branches: 1.60.8; Merge nick-nhusb - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix
1.59	19-Aug-2015	skrll	Interrupt handlers aren't MP-safe yet so use IPL_USB which is IPL_VM which takes the kernel lock.
1.58	21-Sep-2014	christos	branches: 1.58.2; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.57	29-Mar-2014	christos	branches: 1.57.4; 1.57.8; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.56	17-Mar-2013	martin	branches: 1.56.6; Track what parts of initialization were successfull and only revert those during detach - prevents a panic on shutdown when, for example, interrupt mapping failed. Problem reported by Raymond Phillips on port-i386.
1.55	10-Jun-2012	mrg	branches: 1.55.2; merge the jmcneill-usbmp branch. many thanks to jared for the initial work, and every one else who has tested things for me. this is largely my fault at this point :-) the main changes are something like: - usbd_bus_methods{} gains a get_lock() to enable the host controller to provide a lock for the USB code. if the lock isn't provided, old-style protection is (partially) applied. - ehci/ohci/uhci have been converted to the new interfaces, including mutex/cv/etc conversion. - usbdivar.h contains a discussion about locking and what locks are held for which method calls. more to come for usbdi(9) here. - audio drivers (uaudio, umidi, auvitek) have been properly SMPified now that USB is ready. - scsi drivers have been modified to take the kernel lock explicitly before calling into scsi code. - usb pipes are associated with a lock, that is the same as the controller lock. (this could be split up further in the future.) - several usbfoo_locked() or usbfoo_unlocked() functions have been added to the usbdi(9) to enable functionality with or without the USB lock (per controller) already being held. the TODO.usbmp file has specific details on what is left to do, including what device-specific changes should be done now that the whole framework is ready.
1.54	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.53	04-Apr-2011	dyoung	branches: 1.53.4; 1.53.8; Deregister PMF handlers during detachment. Complete the host controller shutdown.
1.52	25-May-2010	cegger	branches: 1.52.2; build fix: revert rev. 1.51
1.51	24-May-2010	pgoyette	Update all callers of the pci_find{vendor,product} routines to now call these routines through their global pointers.
1.50	24-Feb-2010	dyoung	branches: 1.50.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.49	08-Jan-2010	dyoung	branches: 1.49.2; Expand PMF_FN_* macros.
1.48	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.47	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.46	17-Apr-2009	dyoung	Disable interrupts on the h/w before detaching the interrupt handler. Protects against the problem described in <http://mail-index.netbsd.org/port-i386/2009/04/10/msg001317.html>, Xintr_ioapic_levelN jumps to 0 during shutdown.
1.45	07-Apr-2009	dyoung	Detach ehci(4) and uhci(4) at shutdown.
1.44	28-Apr-2008	martin	branches: 1.44.8; 1.44.14; Remove clause 3 and 4 from TNF licenses
1.43	28-Mar-2008	drochner	branches: 1.43.2; 1.43.4; split device/softc for USB host controllers and the usb (control) device, this is hairy stuff, and I've only tested with uhci/ehci at pci, please test the rest and report problems
1.42	07-Mar-2008	dyoung	Use device_t and its accessors. In uhci_{pci,cardbus}_attach, use `self' instead of casting the softc to void *.
1.41	22-Feb-2008	dyoung	Add a method to detach child. Cosmetic: use device_t and accessors. Use aprint_*_dev(). Use PMF_FN_{ARGS,CALL,PROTO}.
1.40	09-Dec-2007	jmcneill	branches: 1.40.6; 1.40.10; Merge jmcneill-pm branch.
1.39	25-Nov-2007	dsainty	branches: 1.39.2; 1.39.4; Minor typos in comments, no functional change.
1.38	24-Nov-2007	gavan	Be paranoid about disabling interrupts and acknowledging any pending interrupts while attaching uhci. Fixes recent problems with uvm_fault during uhci attach, which appear to be caused by a pending interrupt left over from the device while it was operating in legacy support mode under the control of the BIOS. Patch by joerg@ because his was better than mine.
1.37	19-Oct-2007	ad	branches: 1.37.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.36	01-Aug-2007	christos	branches: 1.36.2; 1.36.4; 1.36.8; 1.36.10; Establish powerhooks in the proper order. From Anon Ymous
1.35	10-Dec-2006	uwe	branches: 1.35.6; 1.35.14; 1.35.16; Convert to use aprint_*
1.34	16-Nov-2006	christos	branches: 1.34.2; 1.34.4; __unused removal on arguments; approved by core.
1.33	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.32	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.31	10-Mar-2006	jmcneill	branches: 1.31.10; 1.31.12; On suspend, capture PCI configuration registers. On resume, restore them as well as explicitly reclaiming ownership of the UHCI/EHCI from the BIOS.
1.30	11-Dec-2005	christos	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.10; merge ktrace-lwp.
1.29	28-Jun-2005	thorpej	branches: 1.29.2; Use ANSI function decls and static.
1.28	27-Feb-2005	perry	nuke trailing whitespace
1.27	02-Aug-2004	mycroft	branches: 1.27.4; 1.27.6; For the PCIVERBOSE case, separate vendors and products into separate tables. Eliminating redundant pointers in the tables saves nearly 20K (20% of the table size). In the process, add a pci_findproduct() and make that and pci_findvendor() return a "const char *".
1.26	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.25	22-Apr-2004	itojun	sprintf -> snprintf
1.24	02-Oct-2002	thorpej	branches: 1.24.6; Add trailing ; to CFATTACH_DECL.
1.23	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.22	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.21	13-Nov-2001	lukem	add RCSID
1.20	06-Nov-2001	augustss	Add some data structures to keep enable finding companion controllers for EHCI.
1.19	25-Oct-2001	augustss	Force BIOS out of the game by setting the LEGSUP register to its default value.
1.18	28-Dec-2000	sommerfeld	branches: 1.18.2; 1.18.4; 1.18.8; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.17	06-Sep-2000	augustss	ANSIfy.
1.16	27-Apr-2000	augustss	Change my email address.
1.15	25-Apr-2000	augustss	Move the size of the mapped bus_space region into the bus independent softc.
1.14	25-Jan-2000	augustss	Make sure PIRQD enable bit is set. From FreeBSD.
1.13	20-Nov-1999	augustss	Propagate the USB revision number to the usb driver.
1.12	12-Oct-1999	augustss	branches: 1.12.2; Deallocate resources (bus space and interrupts) when detaching from the bus.
1.11	15-Sep-1999	augustss	branches: 1.11.2; Add preliminary (untested) code for detaching the USB host controller (needed for CardBus based controllers).
1.10	14-Sep-1999	augustss	Don't do IO before we have set the tag and handle.
1.9	14-Sep-1999	augustss	And commit two files I forgot in the rearrangement for testing for interrupt context.
1.8	21-Aug-1999	augustss	Disable interrupts early.
1.7	20-May-1999	augustss	Keep track of vendor id and make sure the vendor string is NUL terminated.
1.6	13-May-1999	thorpej	Rework the way ukbd attaches itself as the console (again). We now allow the code to pick the first USB keyboard instance as the console, ignoring which USB controller it's on. Should eventually allow detaching of the console keyboard. From Jason Thorpe <thorpej@nas.nasa.gov>
1.5	06-May-1999	thorpej	Add a way for machine-dependent code to tell a USB controller that it has the console input device. The USB keyboard driver uses this to attach the first USB keyboard instance as the console keyboard. Unfortunately, this must still be deferred to autoconfiguration time, but there's not much we can do about that right now.
1.4	25-Nov-1998	augustss	branches: 1.4.4; Make the copyright header conform to the NetBSD template.
1.3	25-Jul-1998	augustss	Use PCI class/subclass/interface to match USB host controllers instead of vendor/product id. This makes the probe find more controllers without recompilation.
1.2	25-Jul-1998	augustss	I forgot to commit these when usb_mem was added.
1.1	12-Jul-1998	augustss	Add USB support. Supported so far: * UHCI and OHCI host controllers on PCI * Hubs * HID devices withe special drivers for mouse and keyboard * Printers
1.4.4.2	28-Jun-1999	perry	pullup 1.5->1.6 (wrstuden)
1.4.4.1	06-May-1999	perry	branches: 1.4.4.1.2; 1.4.4.1.4; pullup 1.4->1.5 (thorpej)
1.4.4.1.4.1	30-Nov-1999	itojun	bring in latest KAME (as of 19991130, KAME/NetBSD141) into kame branch just for reference purposes. This commit includes 1.4 -> 1.4.1 sync for kame branch. The branch does not compile at all (due to the lack of ALTQ and some other source code). Please do not try to modify the branch, this is just for referenre purposes. synchronization to latest KAME will take place on HEAD branch soon.
1.4.4.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.11.2.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.12.2.2	05-Jan-2001	bouyer	Sync with HEAD
1.12.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.18.8.1	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.18.4.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.18.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.18.2.2	18-Oct-2002	nathanw	Catch up to -current.
1.18.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.24.6.5	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.24.6.4	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.24.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.24.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.24.6.1	03-Aug-2004	skrll	Sync with HEAD
1.27.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.27.4.1	29-Apr-2005	kent	sync with -current
1.29.2.8	17-Mar-2008	yamt	sync with head.
1.29.2.7	27-Feb-2008	yamt	sync with head.
1.29.2.6	21-Jan-2008	yamt	sync with head
1.29.2.5	07-Dec-2007	yamt	sync with head
1.29.2.4	27-Oct-2007	yamt	sync with head.
1.29.2.3	03-Sep-2007	yamt	sync with head.
1.29.2.2	30-Dec-2006	yamt	sync with head.
1.29.2.1	21-Jun-2006	yamt	sync with head.
1.30.10.1	19-Apr-2006	elad	sync with head.
1.30.8.1	13-Mar-2006	yamt	sync with head.
1.30.6.1	22-Apr-2006	simonb	Sync with head.
1.30.4.1	09-Sep-2006	rpaulo	sync with head
1.31.12.2	10-Dec-2006	yamt	sync with head.
1.31.12.1	22-Oct-2006	yamt	sync with head
1.31.10.2	12-Jan-2007	ad	Sync with head.
1.31.10.1	18-Nov-2006	ad	Sync with head.
1.34.4.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.34.2.1	01-Feb-2008	riz	Pull up following revision(s) (requested by gavan in ticket #1000): sys/dev/pci/uhci_pci.c: revision 1.38 Be paranoid about disabling interrupts and acknowledging any pending interrupts while attaching uhci. Fixes recent problems with uvm_fault during uhci attach, which appear to be caused by a pending interrupt left over from the device while it was operating in legacy support mode under the control of the BIOS. Patch by joerg@ because his was better than mine.
1.35.16.1	15-Aug-2007	skrll	Sync with HEAD.
1.35.14.1	22-May-2007	itohy	Overhaul of USB stack, mostly DMA related This applies to NetBSD 4.99.13 (March 1, 2007) usbdi(9) interface is based on FreeBSD version, excluding - removal of portability code Patch most NetBSD changes, excluding - DMA memory "reserve", since we don't need contiguous buffers any longer - volatiles in DMA structure, since it should not be needed with proper bus_dmamap_sync(9)s DMA/non-DMA memory management overhaul - Move all DMA related code to usb_mem.[ch] (add usb_alloc_buffer_dma(), usb_free_buffer_dma(), etc.). XXX Should usb_mem.[ch] be renamed as usb_mem_dma.[ch] ? - Add corresponding non-DMA code to usb_mem_nodma.[ch] . Currently just use malloc(9). - Above files are conditionally used by config framework (added attributes to conf/files and dev/usb/files.usb). - Add diagnostic panics when resource allocation is requested on interrupt context. - Change memory allocations (that require context) from NOWAIT to WAITOK. Allocate DMA/non-DMA buffer per host interface, not globally. advantage: Buffers can be freed on detaching host interface. Activity of a host interface does not affect others. disadvantages: It possibly consumes more memory. API changes - usbd_alloc_xfer() is changed: old: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev); new: usbd_xfer_handle usbd_alloc_xfer(usbd_device_handle dev, usbd_pipe_handle pipe); - pipe argument of usbd_setup_xfer() are now unused XXX the pipe argument should be removed? - add mapping APIs - async request will be processed as a task (kernel thread context), and delayed to some extent - usbdivar.h: struct usbd_xfer: renamed a member "allocbuf" to "hcbuffer" (mapped/allocated/refered buffer for HCI driver) - usb_port.h: change usb_proc_ptr from struct ptoc to struct lwp * - usb_port.h: add usb_sigproc_ptr for psignal(9) (struct proc ) - usb.h: add UE_MAXPKTSZ(ep) and UE_MAXPKTSZ_MASK macros for USB 2.0 changes to USB device drivers - atu, aue, axe, cdce, cue, kue, rum, udav, upl, ural, url, uaudio, ubt, ucom, ugen, uhidev, uirda, ulpt, umidi, urio, uscanner, ustir, utoppy: catch up API change of usbd_alloc_xfer() - umass, usscanner: * catch up API change of usbd_alloc_xfer() * eliminate memory copy for large transfer ohci - free resources on detach - add lots of bus_dmamap_sync() operations - simplify the code of loading std chain - rewrite code of looking up TD/ITD from DMA addr by using allocation chunk - add workaround for CMD Tech 670 and 673 chipsets - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf slhci - allocate xfer and slhci_xfer at once, and simplify relevant code - add slhci_detach() - remove second arg of slhci_attach() since it is the same as the first arg. - add support for "mapping" (no, it doesn't map since it doesn't do DMA) buffer and mbuf - add pcmcia frontend - NOT TESTED, missing hardware ehci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf - done only simple test uhci - add lots of bus_dmamap_sync() operations, possibly too many - make sure resources are not allocated in interrupt context - add support for mapping buffer and mbuf To do - review, test, debug - rewrite network drivers to utilize usbd_map_buffer_mbuf() - rewrite uaudio(4) to eliminate memcpy - "pipe" argument of usbd_setup_*xfer() should eventually be removed
1.35.6.2	23-Oct-2007	ad	Sync with head.
1.35.6.1	20-Aug-2007	ad	Sync with HEAD.
1.36.10.2	01-Aug-2007	christos	Establish powerhooks in the proper order. From Anon Ymous
1.36.10.1	01-Aug-2007	christos	file uhci_pci.c was added on branch matt-mips64 on 2007-08-01 10:39:44 +0000
1.36.8.1	25-Oct-2007	bouyer	Sync with HEAD.
1.36.4.3	23-Mar-2008	matt	sync with HEAD
1.36.4.2	09-Jan-2008	matt	sync with HEAD
1.36.4.1	06-Nov-2007	matt	sync with HEAD
1.36.2.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.36.2.5	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.36.2.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.36.2.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.36.2.2	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.36.2.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.37.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.37.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.39.4.1	11-Dec-2007	yamt	sync with head.
1.39.2.1	26-Dec-2007	ad	Sync with head.
1.40.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.40.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.40.6.1	24-Mar-2008	keiichi	sync with head.
1.43.4.4	11-Mar-2010	yamt	sync with head
1.43.4.3	16-May-2009	yamt	sync with head
1.43.4.2	04-May-2009	yamt	sync with head.
1.43.4.1	16-May-2008	yamt	sync with head.
1.43.2.1	18-May-2008	yamt	sync with head.
1.44.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.44.8.1	28-Apr-2009	skrll	Sync with HEAD.
1.49.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.50.2.1	21-Apr-2011	rmind	sync with head
1.52.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.53.8.2	18-Feb-2012	mrg	merge to -current.
1.53.8.1	09-Dec-2011	mrg	- make pipe->close method take the thread lock - convert usb_taskq to use mutex/cv - convert needs_explore usage into a cv on the thread lock - remove KERNEL_*LOCK from uaudio and umidi, since we're supposedly MPSAFE here now - use IPL_SCHED instead of IPL_USB (aka biglocked) interrupts - drop the audio thread lock when calling into usb when it may sleep, avoiding a deadlock between audiowrite and audioioctl. this fixes mixerctl -a vs. playing hanging the system XXX probably need to check this in a bunch more places.
1.53.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.53.4.2	30-Oct-2012	yamt	sync with head
1.53.4.1	17-Apr-2012	yamt	sync with head
1.55.2.3	03-Dec-2017	jdolecek	update from HEAD
1.55.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.55.2.1	23-Jun-2013	tls	resync from head
1.56.6.1	18-May-2014	rmind	sync with head
1.57.8.2	14-Sep-2016	skrll	Use IPL_USB for interrupt handler
1.57.8.1	06-Sep-2016	skrll	First pass at netbsd-7 updated with USB code from HEAD
1.57.4.1	05-Apr-2017	snj	Pull up following revision(s) (requested by skrll in ticket #1395): share/man/man4/axe.4: netbsd-7-nhusb share/man/man4/axen.4: netbsd-7-nhusb share/man/man4/cdce.4: netbsd-7-nhusb share/man/man4/uaudio.4: netbsd-7-nhusb share/man/man4/ucom.4: netbsd-7-nhusb share/man/man4/uep.4: netbsd-7-nhusb share/man/man4/urtw.4: netbsd-7-nhusb share/man/man4/usb.4: netbsd-7-nhusb share/man/man4/uyap.4: netbsd-7-nhusb share/man/man4/xhci.4: netbsd-7-nhusb share/man/man9/usbdi.9: netbsd-7-nhusb sys/arch/amd64/conf/ALL: netbsd-7-nhusb sys/arch/amd64/conf/GENERIC: netbsd-7-nhusb sys/arch/amiga/dev/slhci_zbus.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_otg.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_usb.c: netbsd-7-nhusb sys/arch/arm/amlogic/amlogic_dwctwo.c: netbsd-7-nhusb sys/arch/arm/at91/at91ohci.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm2835_dwctwo.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm53xx_usb.c: netbsd-7-nhusb sys/arch/arm/ep93xx/epohci.c: netbsd-7-nhusb sys/arch/arm/gemini/obio_ehci.c: netbsd-7-nhusb sys/arch/arm/imx/files.imx23: netbsd-7-nhusb sys/arch/arm/imx/imxusb.c: netbsd-7-nhusb sys/arch/arm/imx/imxusbreg.h: netbsd-7-nhusb sys/arch/arm/omap/obio_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/omap3_ehci.c: netbsd-7-nhusb sys/arch/arm/omap/omapl1x_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/tiotg.c: netbsd-7-nhusb sys/arch/arm/s3c2xx0/ohci_s3c24x0.c: netbsd-7-nhusb sys/arch/arm/samsung/exynos_usb.c: netbsd-7-nhusb sys/arch/arm/xscale/pxa2x0_ohci.c: netbsd-7-nhusb sys/arch/arm/zynq/zynq_usb.c: netbsd-7-nhusb sys/arch/hpcarm/dev/nbp_slhci.c: netbsd-7-nhusb sys/arch/hpcmips/dev/plumohci.c: netbsd-7-nhusb sys/arch/i386/conf/ALL: netbsd-7-nhusb sys/arch/i386/conf/GENERIC: netbsd-7-nhusb sys/arch/i386/pci/gcscehci.c: netbsd-7-nhusb sys/arch/luna68k/conf/GENERIC: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahci.c: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahcivar.h: netbsd-7-nhusb sys/arch/mips/alchemy/dev/ohci_aubus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ehci_arbus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ohci_arbus.c: netbsd-7-nhusb sys/arch/mips/conf/files.adm5120: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ehci.c: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ohci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ehci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ohci.c: netbsd-7-nhusb sys/arch/playstation2/dev/ohci_sbus.c: netbsd-7-nhusb sys/arch/powerpc/booke/dev/pq3ehci.c: netbsd-7-nhusb sys/arch/powerpc/ibm4xx/dev/dwctwo_plb.c: netbsd-7-nhusb sys/arch/x68k/dev/slhci_intio.c: netbsd-7-nhusb sys/conf/files: netbsd-7-nhusb sys/dev/cardbus/ehci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/ohci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/uhci_cardbus.c: netbsd-7-nhusb sys/dev/ic/sl811hs.c: netbsd-7-nhusb sys/dev/ic/sl811hsvar.h: netbsd-7-nhusb sys/dev/isa/slhci_isa.c: netbsd-7-nhusb sys/dev/marvell/ehci_mv.c: netbsd-7-nhusb sys/dev/pci/ehci_pci.c: netbsd-7-nhusb sys/dev/pci/ohci_pci.c: netbsd-7-nhusb sys/dev/pci/uhci_pci.c: netbsd-7-nhusb sys/dev/pci/xhci_pci.c: netbsd-7-nhusb sys/dev/pcmcia/slhci_pcmcia.c: netbsd-7-nhusb sys/dev/usb/Makefile.usbdevs: netbsd-7-nhusb sys/dev/usb/TODO: netbsd-7-nhusb sys/dev/usb/TODO.usbmp: netbsd-7-nhusb sys/dev/usb/aubtfwl.c: netbsd-7-nhusb sys/dev/usb/auvitek.c: netbsd-7-nhusb sys/dev/usb/auvitek_audio.c: netbsd-7-nhusb sys/dev/usb/auvitek_dtv.c: netbsd-7-nhusb sys/dev/usb/auvitek_i2c.c: netbsd-7-nhusb sys/dev/usb/auvitek_video.c: netbsd-7-nhusb sys/dev/usb/auvitekvar.h: netbsd-7-nhusb sys/dev/usb/ehci.c: netbsd-7-nhusb sys/dev/usb/ehcireg.h: netbsd-7-nhusb sys/dev/usb/ehcivar.h: netbsd-7-nhusb sys/dev/usb/emdtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_dtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_ir.c: netbsd-7-nhusb sys/dev/usb/emdtvvar.h: netbsd-7-nhusb sys/dev/usb/ezload.c: netbsd-7-nhusb sys/dev/usb/ezload.h: netbsd-7-nhusb sys/dev/usb/files.usb: netbsd-7-nhusb sys/dev/usb/hid.c: netbsd-7-nhusb sys/dev/usb/hid.h: netbsd-7-nhusb sys/dev/usb/if_athn_usb.c: netbsd-7-nhusb sys/dev/usb/if_athn_usb.h: netbsd-7-nhusb sys/dev/usb/if_atu.c: netbsd-7-nhusb sys/dev/usb/if_atureg.h: netbsd-7-nhusb sys/dev/usb/if_aue.c: netbsd-7-nhusb sys/dev/usb/if_auereg.h: netbsd-7-nhusb sys/dev/usb/if_axe.c: netbsd-7-nhusb sys/dev/usb/if_axen.c: netbsd-7-nhusb sys/dev/usb/if_axenreg.h: netbsd-7-nhusb sys/dev/usb/if_axereg.h: netbsd-7-nhusb sys/dev/usb/if_cdce.c: netbsd-7-nhusb sys/dev/usb/if_cdcereg.h: netbsd-7-nhusb sys/dev/usb/if_cue.c: netbsd-7-nhusb sys/dev/usb/if_cuereg.h: netbsd-7-nhusb sys/dev/usb/if_kue.c: netbsd-7-nhusb sys/dev/usb/if_kuereg.h: netbsd-7-nhusb sys/dev/usb/if_otus.c: netbsd-7-nhusb sys/dev/usb/if_otusvar.h: netbsd-7-nhusb sys/dev/usb/if_rum.c: netbsd-7-nhusb sys/dev/usb/if_rumreg.h: netbsd-7-nhusb sys/dev/usb/if_rumvar.h: netbsd-7-nhusb sys/dev/usb/if_run.c: netbsd-7-nhusb sys/dev/usb/if_runvar.h: netbsd-7-nhusb sys/dev/usb/if_smsc.c: netbsd-7-nhusb sys/dev/usb/if_smscreg.h: netbsd-7-nhusb sys/dev/usb/if_smscvar.h: netbsd-7-nhusb sys/dev/usb/if_udav.c: netbsd-7-nhusb sys/dev/usb/if_udavreg.h: netbsd-7-nhusb sys/dev/usb/if_upgt.c: netbsd-7-nhusb sys/dev/usb/if_upgtvar.h: netbsd-7-nhusb sys/dev/usb/if_upl.c: netbsd-7-nhusb sys/dev/usb/if_ural.c: netbsd-7-nhusb sys/dev/usb/if_uralreg.h: netbsd-7-nhusb sys/dev/usb/if_uralvar.h: netbsd-7-nhusb sys/dev/usb/if_url.c: netbsd-7-nhusb sys/dev/usb/if_urlreg.h: netbsd-7-nhusb sys/dev/usb/if_urndis.c: netbsd-7-nhusb sys/dev/usb/if_urndisreg.h: netbsd-7-nhusb sys/dev/usb/if_urtw.c: netbsd-7-nhusb sys/dev/usb/if_urtwn.c: netbsd-7-nhusb sys/dev/usb/if_urtwn_data.h: netbsd-7-nhusb sys/dev/usb/if_urtwnreg.h: netbsd-7-nhusb sys/dev/usb/if_urtwnvar.h: netbsd-7-nhusb sys/dev/usb/if_urtwreg.h: netbsd-7-nhusb sys/dev/usb/if_zyd.c: netbsd-7-nhusb sys/dev/usb/if_zydreg.h: netbsd-7-nhusb sys/dev/usb/irmce.c: netbsd-7-nhusb sys/dev/usb/moscom.c: netbsd-7-nhusb sys/dev/usb/motg.c: netbsd-7-nhusb sys/dev/usb/motgvar.h: netbsd-7-nhusb sys/dev/usb/ohci.c: netbsd-7-nhusb sys/dev/usb/ohcireg.h: netbsd-7-nhusb sys/dev/usb/ohcivar.h: netbsd-7-nhusb sys/dev/usb/pseye.c: netbsd-7-nhusb sys/dev/usb/slurm.c: netbsd-7-nhusb sys/dev/usb/stuirda.c: netbsd-7-nhusb sys/dev/usb/u3g.c: netbsd-7-nhusb sys/dev/usb/uark.c: netbsd-7-nhusb sys/dev/usb/uatp.c: netbsd-7-nhusb sys/dev/usb/uaudio.c: netbsd-7-nhusb sys/dev/usb/uberry.c: netbsd-7-nhusb sys/dev/usb/ubsa.c: netbsd-7-nhusb sys/dev/usb/ubsa_common.c: netbsd-7-nhusb sys/dev/usb/ubsavar.h: netbsd-7-nhusb sys/dev/usb/ubt.c: netbsd-7-nhusb sys/dev/usb/uchcom.c: netbsd-7-nhusb sys/dev/usb/ucom.c: netbsd-7-nhusb sys/dev/usb/ucomvar.h: netbsd-7-nhusb sys/dev/usb/ucycom.c: netbsd-7-nhusb sys/dev/usb/udl.c: netbsd-7-nhusb sys/dev/usb/udl.h: netbsd-7-nhusb sys/dev/usb/udsbr.c: netbsd-7-nhusb sys/dev/usb/udsir.c: netbsd-7-nhusb sys/dev/usb/uep.c: netbsd-7-nhusb sys/dev/usb/uftdi.c: netbsd-7-nhusb sys/dev/usb/uftdireg.h: netbsd-7-nhusb sys/dev/usb/ugen.c: netbsd-7-nhusb sys/dev/usb/ugensa.c: netbsd-7-nhusb sys/dev/usb/uhci.c: netbsd-7-nhusb sys/dev/usb/uhcireg.h: netbsd-7-nhusb sys/dev/usb/uhcivar.h: netbsd-7-nhusb sys/dev/usb/uhid.c: netbsd-7-nhusb sys/dev/usb/uhidev.c: netbsd-7-nhusb sys/dev/usb/uhidev.h: netbsd-7-nhusb sys/dev/usb/uhmodem.c: netbsd-7-nhusb sys/dev/usb/uhso.c: netbsd-7-nhusb sys/dev/usb/uhub.c: netbsd-7-nhusb sys/dev/usb/uipad.c: netbsd-7-nhusb sys/dev/usb/uipaq.c: netbsd-7-nhusb sys/dev/usb/uirda.c: netbsd-7-nhusb sys/dev/usb/uirdavar.h: netbsd-7-nhusb sys/dev/usb/ukbd.c: netbsd-7-nhusb sys/dev/usb/ukbdmap.c: netbsd-7-nhusb sys/dev/usb/ukyopon.c: netbsd-7-nhusb sys/dev/usb/ukyopon.h: netbsd-7-nhusb sys/dev/usb/ulpt.c: netbsd-7-nhusb sys/dev/usb/umass.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.h: netbsd-7-nhusb sys/dev/usb/umass_quirks.c: netbsd-7-nhusb sys/dev/usb/umass_quirks.h: netbsd-7-nhusb sys/dev/usb/umass_scsipi.c: netbsd-7-nhusb sys/dev/usb/umass_scsipi.h: netbsd-7-nhusb sys/dev/usb/umassvar.h: netbsd-7-nhusb sys/dev/usb/umcs.c: netbsd-7-nhusb sys/dev/usb/umct.c: netbsd-7-nhusb sys/dev/usb/umidi.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.h: netbsd-7-nhusb sys/dev/usb/umodem.c: netbsd-7-nhusb sys/dev/usb/umodem_common.c: netbsd-7-nhusb sys/dev/usb/umodemvar.h: netbsd-7-nhusb sys/dev/usb/ums.c: netbsd-7-nhusb sys/dev/usb/uplcom.c: netbsd-7-nhusb sys/dev/usb/urio.c: netbsd-7-nhusb sys/dev/usb/urio.h: netbsd-7-nhusb sys/dev/usb/usb.c: netbsd-7-nhusb sys/dev/usb/usb.h: netbsd-7-nhusb sys/dev/usb/usb_mem.c: netbsd-7-nhusb sys/dev/usb/usb_mem.h: netbsd-7-nhusb sys/dev/usb/usb_quirks.c: netbsd-7-nhusb sys/dev/usb/usb_quirks.h: netbsd-7-nhusb sys/dev/usb/usb_subr.c: netbsd-7-nhusb sys/dev/usb/usbdevices.config: netbsd-7-nhusb sys/dev/usb/usbdevs: netbsd-7-nhusb sys/dev/usb/usbdevs.h: netbsd-7-nhusb sys/dev/usb/usbdevs_data.h: netbsd-7-nhusb sys/dev/usb/usbdi.c: netbsd-7-nhusb sys/dev/usb/usbdi.h: netbsd-7-nhusb sys/dev/usb/usbdi_util.c: netbsd-7-nhusb sys/dev/usb/usbdi_util.h: netbsd-7-nhusb sys/dev/usb/usbdivar.h: netbsd-7-nhusb sys/dev/usb/usbhid.h: netbsd-7-nhusb sys/dev/usb/usbhist.h: netbsd-7-nhusb sys/dev/usb/usbroothub.c: netbsd-7-nhusb sys/dev/usb/usbroothub.h: netbsd-7-nhusb sys/dev/usb/usbroothub_subr.c: delete sys/dev/usb/usbroothub_subr.h: delete sys/dev/usb/uscanner.c: netbsd-7-nhusb sys/dev/usb/uslsa.c: netbsd-7-nhusb sys/dev/usb/usscanner.c: netbsd-7-nhusb sys/dev/usb/ustir.c: netbsd-7-nhusb sys/dev/usb/uthum.c: netbsd-7-nhusb sys/dev/usb/utoppy.c: netbsd-7-nhusb sys/dev/usb/uts.c: netbsd-7-nhusb sys/dev/usb/uvideo.c: netbsd-7-nhusb sys/dev/usb/uvisor.c: netbsd-7-nhusb sys/dev/usb/uvscom.c: netbsd-7-nhusb sys/dev/usb/uyap.c: netbsd-7-nhusb sys/dev/usb/uyap_firmware.h: netbsd-7-nhusb sys/dev/usb/uyurex.c: netbsd-7-nhusb sys/dev/usb/x1input_rdesc.h: netbsd-7-nhusb sys/dev/usb/xhci.c: netbsd-7-nhusb sys/dev/usb/xhcireg.h: netbsd-7-nhusb sys/dev/usb/xhcivar.h: netbsd-7-nhusb sys/dev/usb/xinput_rdesc.h: netbsd-7-nhusb sys/external/bsd/common/conf/files.linux: netbsd-7-nhusb sys/external/bsd/common/include/linux/err.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/kernel.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/workqueue.h: netbsd-7-nhusb sys/external/bsd/common/linux/linux_work.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/atombios_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/drm/files.drmkms: netbsd-7-nhusb sys/external/bsd/drm2/i915drm/files.i915drmkms: netbsd-7-nhusb sys/external/bsd/drm2/include/linux/err.h: delete sys/external/bsd/drm2/include/linux/workqueue.h: delete sys/external/bsd/drm2/linux/files.drmkms_linux: netbsd-7-nhusb sys/external/bsd/drm2/linux/linux_work.c: delete sys/external/bsd/dwc2/dwc2.c: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2var.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwctwo2netbsd: netbsd-7-nhusb sys/external/bsd/dwc2/conf/files.dwc2: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_coreintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdddma.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdqueue.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hw.h: netbsd-7-nhusb sys/modules/drmkms_linux/Makefile: netbsd-7-nhusb sys/modules/i915drmkms/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libugenhc/ugenhc.c: netbsd-7-nhusb sys/rump/dev/lib/libusb/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libusb/USB.ioconf: netbsd-7-nhusb sys/rump/dev/lib/libusb/usb_at_ugenhc.c: delete sys/rump/dev/lib/libusb/opt/opt_usb.h: delete sys/rump/dev/lib/libusb/opt/opt_usbverbose.h: delete sys/sys/mbuf.h: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.8: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.c: netbsd-7-nhusb Merge netbsd-7-nhusb: - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix - Change the SOFTINT level from NET to SERIAL for the USB softint handler. This gives the callback a chance of running when another softint handler at SOFTINT_NET has blocked holding a lock, e.g. softnet_lock and most of the network stack. - kern/49065 - ifconfig tun0 ... sequence locks up system / lockup: softnet_lock held across usb xfr - kern/50491 - unkillable wait in usbd_transfer while using usmsc0 on raspberry pi 2 - kern/51395 - USB Ethernet makes xhci hang - Various improvements to slhci(4) - Various improvements to dwc2(4)
1.58.2.5	28-Aug-2017	skrll	Sync with HEAD
1.58.2.4	22-Sep-2015	skrll	Sync with HEAD
1.58.2.3	05-Dec-2014	skrll	Use int for return type for [eou]chi_init and motg_init.
1.58.2.2	05-Dec-2014	skrll	KNF. Remove ( ) from return statements.
1.58.2.1	03-Dec-2014	skrll	The grand renaming of structure members. No functional change.
1.60.8.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.61.8.2	21-May-2018	pgoyette	Sync with HEAD
1.61.8.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.64.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.65.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	13-Aug-2006	jmcneill	branches: 1.1.2; 1.1.6; 1.1.12; Acceleration definitions (missed in previous commit).
1.1.12.2	30-Dec-2006	yamt	sync with head.
1.1.12.1	13-Aug-2006	yamt	file unichromeaccel.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.1.6.2	09-Sep-2006	rpaulo	sync with head
1.1.6.1	13-Aug-2006	rpaulo	file unichromeaccel.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.2.2	03-Sep-2006	yamt	sync with head.
1.1.2.1	13-Aug-2006	yamt	file unichromeaccel.h was added on branch yamt-pdpolicy on 2006-09-03 15:24:23 +0000
1.2	13-Aug-2006	jmcneill	branches: 1.2.4; 1.2.10; Expose RGB palette tables.
1.1	02-Aug-2006	jmcneill	branches: 1.1.4; Add work-in-progress framebuffer console driver for the VIA UniChrome integrated graphics family.
1.1.4.3	03-Sep-2006	yamt	sync with head.
1.1.4.2	11-Aug-2006	yamt	sync with head
1.1.4.1	02-Aug-2006	yamt	file unichromeconfig.h was added on branch yamt-pdpolicy on 2006-08-11 15:44:26 +0000
1.2.10.2	30-Dec-2006	yamt	sync with head.
1.2.10.1	13-Aug-2006	yamt	file unichromeconfig.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.2.4.2	09-Sep-2006	rpaulo	sync with head
1.2.4.1	13-Aug-2006	rpaulo	file unichromeconfig.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.22	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.21	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.20	24-Apr-2021	thorpej	branches: 1.20.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.19	25-Oct-2015	khorben	branches: 1.19.34; Remove references to SPLASHSCREEN_PROGRESS To my knowledge this feature is no longer supported at the moment. The manual page for wsdisplay(4) should probably also be updated to reflect this situation.
1.18	22-Jan-2011	cegger	branches: 1.18.14; 1.18.32; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.17	16-Dec-2010	cegger	branches: 1.17.2; 1.17.4; ioctl: KNF switch-case, remove useless brackets
1.16	06-May-2009	cegger	branches: 1.16.4; struct device * -> device_t, no functional changes intended.
1.15	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.14	05-May-2008	jmcneill	branches: 1.14.14; Change autoconf message 'direct rendering on dev unsupported' to the more accurate 'drm at dev not configured'.
1.13	05-May-2008	jmcneill	Use 2-clause license.
1.12	27-Mar-2008	jmcneill	branches: 1.12.2; 1.12.4; Split device_t and softc, aprintify
1.11	12-Mar-2008	phx	Make console drivers support PCI_IOC_CFGREAD/WRITE ioctls. This is useful for XFree86, which can alternatively find the graphics card through /dev/ttyE0. Approved by macallan.
1.10	01-Dec-2007	ad	branches: 1.10.10; 1.10.14; Back out previous.
1.9	01-Dec-2007	ad	Noisy printfs.
1.8	19-Oct-2007	ad	branches: 1.8.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.7	23-Mar-2007	jmcneill	branches: 1.7.8; 1.7.10; 1.7.14; Make unichromefb(4) peacefully co-exist with viadrm(4). Basically, when we switch from WSDISPLAYIO_MODE_EMUL to anything else, we release any bus_space resources. When we return to WSDISPLAYIO_MODE_EMUL, reclaim these resources and reinitialize the display. Idea from macallan@
1.6	04-Mar-2007	christos	branches: 1.6.2; 1.6.4; 1.6.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.5	12-Jan-2007	ober	branches: 1.5.2; Update to compile properly. OK jmcneill
1.4	13-Aug-2006	jmcneill	branches: 1.4.2; 1.4.4; 1.4.8; 1.4.10; 1.4.12; Reclaim resources from an initial vga console on attach.
1.3	13-Aug-2006	jmcneill	Add accelerated putchar and cursor ops. Also fix a bug in copyrows that caused scrolling down to always fail.
1.2	13-Aug-2006	jmcneill	Add acceleration support for the VIA UniChrome driver. While here, fix a bug that caused the screen to be filled with annoying vertical lines.
1.1	02-Aug-2006	jmcneill	branches: 1.1.2; 1.1.4; Add work-in-progress framebuffer console driver for the VIA UniChrome integrated graphics family.
1.1.4.3	03-Sep-2006	yamt	sync with head.
1.1.4.2	11-Aug-2006	yamt	sync with head
1.1.4.1	02-Aug-2006	yamt	file unichromefb.c was added on branch yamt-pdpolicy on 2006-08-11 15:44:26 +0000
1.1.2.1	15-Aug-2006	ghen	Pull up following revision(s) (requested by jmcneill in ticket #19): sys/dev/pci/unichromefb.c: revision 1.2 Add acceleration support for the VIA UniChrome driver. While here, fix a bug that caused the screen to be filled with annoying vertical lines.
1.4.12.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.4.10.6	17-Mar-2008	yamt	sync with head.
1.4.10.5	27-Oct-2007	yamt	sync with head.
1.4.10.4	03-Sep-2007	yamt	sync with head.
1.4.10.3	26-Feb-2007	yamt	sync with head.
1.4.10.2	30-Dec-2006	yamt	sync with head.
1.4.10.1	13-Aug-2006	yamt	file unichromefb.c was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.4.8.1	21-Jun-2007	liamjfoy	Pull up following revision(s) (requested by sborrill in ticket #737): sys/dev/pci/unichromefb.c: revision 1.5 Update to compile properly. OK jmcneill
1.4.4.2	09-Sep-2006	rpaulo	sync with head
1.4.4.1	13-Aug-2006	rpaulo	file unichromefb.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.4.2.1	01-Feb-2007	ad	Sync with head.
1.5.2.2	24-Mar-2007	yamt	sync with head.
1.5.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.6.6.1	29-Mar-2007	reinoud	Pullup to -current
1.6.4.1	11-Jul-2007	mjf	Sync with head.
1.6.2.2	23-Oct-2007	ad	Sync with head.
1.6.2.1	10-Apr-2007	ad	Sync with head.
1.7.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.7.10.2	23-Mar-2008	matt	sync with HEAD
1.7.10.1	06-Nov-2007	matt	sync with HEAD
1.7.8.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.8.2.1	08-Dec-2007	mjf	Sync with HEAD.
1.10.14.2	02-Jun-2008	mjf	Sync with HEAD.
1.10.14.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.10.1	24-Mar-2008	keiichi	sync with head.
1.12.4.2	16-May-2009	yamt	sync with head
1.12.4.1	16-May-2008	yamt	sync with head.
1.12.2.1	18-May-2008	yamt	sync with head.
1.14.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.16.4.1	05-Mar-2011	rmind	sync with head
1.17.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.17.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.18.32.1	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.18.14.1	03-Dec-2017	jdolecek	update from HEAD
1.19.34.3	24-Apr-2021	thorpej	Make sure to explcitly specify the "wsemuldisplaydev" interface attribute when attaching the wscons display instance, because these devices can also attach children using the "drm" interface attribute.
1.19.34.2	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.19.34.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.20.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	02-Aug-2006	jmcneill	branches: 1.1.4; 1.1.8; 1.1.14; Add work-in-progress framebuffer console driver for the VIA UniChrome integrated graphics family.
1.1.14.2	30-Dec-2006	yamt	sync with head.
1.1.14.1	02-Aug-2006	yamt	file unichromehw.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:48 +0000
1.1.8.2	09-Sep-2006	rpaulo	sync with head
1.1.8.1	02-Aug-2006	rpaulo	file unichromehw.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.4.2	11-Aug-2006	yamt	sync with head
1.1.4.1	02-Aug-2006	yamt	file unichromehw.h was added on branch yamt-pdpolicy on 2006-08-11 15:44:26 +0000
1.2	02-Feb-2024	andvar	fix typos, mainly s/unsupport/unsupported/ in log messages.
1.1	02-Aug-2006	jmcneill	branches: 1.1.4; 1.1.8; 1.1.14; Add work-in-progress framebuffer console driver for the VIA UniChrome integrated graphics family.
1.1.14.2	30-Dec-2006	yamt	sync with head.
1.1.14.1	02-Aug-2006	yamt	file unichromemode.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:49 +0000
1.1.8.2	09-Sep-2006	rpaulo	sync with head
1.1.8.1	02-Aug-2006	rpaulo	file unichromemode.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.4.2	11-Aug-2006	yamt	sync with head
1.1.4.1	02-Aug-2006	yamt	file unichromemode.h was added on branch yamt-pdpolicy on 2006-08-11 15:44:26 +0000
1.1	02-Aug-2006	jmcneill	branches: 1.1.4; 1.1.8; 1.1.14; Add work-in-progress framebuffer console driver for the VIA UniChrome integrated graphics family.
1.1.14.2	30-Dec-2006	yamt	sync with head.
1.1.14.1	02-Aug-2006	yamt	file unichromereg.h was added on branch yamt-lazymbuf on 2006-12-30 20:48:49 +0000
1.1.8.2	09-Sep-2006	rpaulo	sync with head
1.1.8.1	02-Aug-2006	rpaulo	file unichromereg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.4.2	11-Aug-2006	yamt	sync with head
1.1.4.1	02-Aug-2006	yamt	file unichromereg.h was added on branch yamt-pdpolicy on 2006-08-11 15:44:26 +0000
1.13	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.12	29-Mar-2014	christos	branches: 1.12.28; 1.12.30; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.11	26-Nov-2009	njoly	branches: 1.11.12; 1.11.22; 1.11.26; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.10	16-Mar-2009	dsl	ANSIfy functions with function-pointer arguments
1.9	14-Mar-2009	dsl	ANSIfy another 1261 function definitions. The only ones left in sys are beyond by sed script! (or in sys/dist or sys/external) Mostly they have function pointer parameters.
1.8	19-Oct-2007	ad	branches: 1.8.20; 1.8.28; 1.8.34; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.7	11-Dec-2005	christos	branches: 1.7.30; 1.7.44; 1.7.46; 1.7.50; merge ktrace-lwp.
1.6	04-Feb-2005	perry	branches: 1.6.6; de-__P
1.5	31-Jan-2003	thorpej	branches: 1.5.2; 1.5.10; 1.5.12; Use aprint_*().
1.4	13-Nov-2001	lukem	add RCSID
1.3	28-Dec-2000	sommerfeld	branches: 1.3.2; 1.3.4; Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.2	12-Mar-2000	drochner	branches: 1.2.6; add support for DMA mapping, bus probes ala "badaddr" and VME interrupts, improve diagnostic output
1.1	25-Feb-2000	drochner	add (minimal, PIO only) support for the popular "Tundra Universe" PCI-VME interface chip and a frontend for the "Bit3 Mod. 2706" PCI-VME adapter
1.2.6.3	05-Jan-2001	bouyer	Sync with HEAD
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	12-Mar-2000	bouyer	file universe_pci.c was added on branch thorpej_scsipi on 2000-11-20 11:42:38 +0000
1.3.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.5.12.1	12-Feb-2005	yamt	sync with head.
1.5.10.1	29-Apr-2005	kent	sync with -current
1.5.2.1	04-Feb-2005	skrll	Sync with HEAD.
1.6.6.1	27-Oct-2007	yamt	sync with head.
1.7.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.7.46.1	06-Nov-2007	matt	sync with HEAD
1.7.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.7.30.1	23-Oct-2007	ad	Sync with head.
1.8.34.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.8.28.1	28-Apr-2009	skrll	Sync with HEAD.
1.8.20.2	11-Mar-2010	yamt	sync with head
1.8.20.1	04-May-2009	yamt	sync with head.
1.11.26.1	18-May-2014	rmind	sync with head
1.11.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.11.12.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.12.30.1	10-Jun-2019	christos	Sync with HEAD
1.12.28.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.5	30-Jun-2011	wiz	dependant -> dependent
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	04-Feb-2005	perry	de-__P
1.2	12-Mar-2000	drochner	branches: 1.2.6; 1.2.30; 1.2.38; 1.2.40; add support for DMA mapping, bus probes ala "badaddr" and VME interrupts, improve diagnostic output
1.1	25-Feb-2000	drochner	add (minimal, PIO only) support for the popular "Tundra Universe" PCI-VME interface chip and a frontend for the "Bit3 Mod. 2706" PCI-VME adapter
1.2.40.1	12-Feb-2005	yamt	sync with head.
1.2.38.1	29-Apr-2005	kent	sync with -current
1.2.30.1	04-Feb-2005	skrll	Sync with HEAD.
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	12-Mar-2000	bouyer	file universe_pci_var.h was added on branch thorpej_scsipi on 2000-11-20 11:42:39 +0000
1.9	31-Mar-2025	riastradh	ehci(4): Make usb__rem work as noop on zero-initialized input. This way, if some hci_attach function fails before usb__add, the caller need not maintain boolean state to decide whether to call usb__rem on detach -- it'll just work without extra effort. Cleanup in preparation for: PR port-amd64/59180: System reboots instead of shutting down
1.8	31-Mar-2025	riastradh	ehci(4): Sprinkle KERNEL_LOCKED_P and KNF on usb__add/rem. Mark with XXXSMP comments to be fixed later without the kernel lock. No functional change intended: callers generally do this from autoconf _attach/detach routines, which run kernel-locked anyway (for the moment). Cleanup in preparation for: PR port-amd64/59180: System reboots instead of shutting down
1.7	28-Apr-2008	martin	branches: 1.7.126; Remove clause 3 and 4 from TNF licenses
1.6	28-Mar-2008	drochner	branches: 1.6.2; 1.6.4; split device/softc for USB host controllers and the usb (control) device, this is hairy stuff, and I've only tested with uhci/ehci at pci, please test the rest and report problems
1.5	11-Dec-2005	christos	branches: 1.5.70; merge ktrace-lwp.
1.4	21-Dec-2004	fvdl	Implement an USB memory reserve. Allocate some memory per host controller in the wanted bus_dma space. If an allocation fails during operation (i.e. when hotplugging an umass device), fall back to using the reserve. The amount can be configured as USB_MEM_RESERVE. The default value is 256k. Ideally, there would be a way to steal pages in the desired area from UVM, but that's far more complicated, and this is not intrusive, plus it works.
1.3	14-Feb-2002	augustss	branches: 1.3.16; 1.3.22; Use the correct years for copyright.
1.2	12-Jan-2002	enami	Make ehci_pci.c compiles again.
1.1	09-Jan-2002	augustss	branches: 1.1.2; 1.1.4; Split out some code so you can have cardbus ehci without pci.
1.1.4.3	17-Apr-2002	nathanw	Catch up to -current.
1.1.4.2	11-Jan-2002	nathanw	More catchup.
1.1.4.1	09-Jan-2002	nathanw	file usb_pci.c was added on branch nathanw_sa on 2002-01-11 23:39:28 +0000
1.1.2.4	16-Mar-2002	jdolecek	Catch up with -current.
1.1.2.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.2.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.2.1	09-Jan-2002	thorpej	file usb_pci.c was added on branch kqueue on 2002-01-10 19:57:08 +0000
1.3.22.1	13-May-2005	riz	Pull up revision 1.4 (requested by bad in ticket #1479): Implement an USB memory reserve. Allocate some memory per host controller in the wanted bus_dma space. If an allocation fails during operation (i.e. when hotplugging an umass device), fall back to using the reserve. The amount can be configured as USB_MEM_RESERVE. The default value is 256k. Ideally, there would be a way to steal pages in the desired area from UVM, but that's far more complicated, and this is not intrusive, plus it works.
1.3.16.1	17-Jan-2005	skrll	Sync with HEAD.
1.5.70.2	02-Jun-2008	mjf	Sync with HEAD.
1.5.70.1	03-Apr-2008	mjf	Sync with HEAD.
1.6.4.1	16-May-2008	yamt	sync with head.
1.6.2.1	18-May-2008	yamt	sync with head.
1.7.126.1	02-Aug-2025	perseant	Sync with HEAD
1.7	09-Oct-2024	andvar	s/compainion/companion/ in comments.
1.6	11-Dec-2010	matt	branches: 1.6.98; Make ehci deal with pci devices that have multiple companion functions and ehci functions. We can't assume that there will only be one ehci per device. The existing code could not deal with: Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 0 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 1 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 2 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 3 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 4 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 5 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x10) at pci1 dev 0 function 6 not configured Netmos MCS9990 Quad USB 2.0 Port (USB serial bus, interface 0x20) at pci1 dev 0 function 7 not configured
1.5	28-Apr-2008	martin	branches: 1.5.18; 1.5.22; 1.5.26; Remove clause 3 and 4 from TNF licenses
1.4	28-Mar-2008	drochner	branches: 1.4.2; 1.4.4; split device/softc for USB host controllers and the usb (control) device, this is hairy stuff, and I've only tested with uhci/ehci at pci, please test the rest and report problems
1.3	14-Feb-2002	augustss	branches: 1.3.114; Use the correct years for copyright.
1.2	12-Jan-2002	enami	Make ehci_pci.c compiles again.
1.1	06-Nov-2001	augustss	branches: 1.1.2; 1.1.4; 1.1.6; Add some data structures to keep enable finding companion controllers for EHCI.
1.1.6.4	16-Mar-2002	jdolecek	Catch up with -current.
1.1.6.3	11-Feb-2002	jdolecek	Sync w/ -current.
1.1.6.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.6.1	06-Nov-2001	thorpej	file usb_pci.h was added on branch kqueue on 2002-01-10 19:57:08 +0000
1.1.4.3	28-Feb-2002	nathanw	Catch up to -current.
1.1.4.2	14-Nov-2001	nathanw	Catch up to -current.
1.1.4.1	06-Nov-2001	nathanw	file usb_pci.h was added on branch nathanw_sa on 2001-11-14 19:15:32 +0000
1.1.2.2	12-Nov-2001	thorpej	Sync the thorpej-mips-cache branch with -current.
1.1.2.1	06-Nov-2001	thorpej	file usb_pci.h was added on branch thorpej-mips-cache on 2001-11-12 21:18:23 +0000
1.3.114.2	02-Jun-2008	mjf	Sync with HEAD.
1.3.114.1	03-Apr-2008	mjf	Sync with HEAD.
1.4.4.1	16-May-2008	yamt	sync with head.
1.4.2.1	18-May-2008	yamt	sync with head.
1.5.26.1	07-Jan-2011	matt	Deal with multifunction pci devices better in choosing companion devices.
1.5.22.1	05-Mar-2011	rmind	sync with head
1.5.18.1	04-Jan-2012	matt	Take changes from matt-nb5-pq3 which fixes how multiple ehci functions on a pci device claim companion ohci controllers.
1.6.98.1	02-Aug-2025	perseant	Sync with HEAD
1.59	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.58	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.57	24-Apr-2021	thorpej	branches: 1.57.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.56	17-Jun-2020	jdolecek	branches: 1.56.4; explicitly disable INTx interrupts to avoid interrupt storm triggered by unhandled adapter interrupts fixes PR kern/46596 by Andreas Gustafsson, fix adopted from FreeBSD
1.55	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.54	30-Jan-2012	drochner	branches: 1.54.6; 1.54.24; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.53	22-Jan-2011	cegger	branches: 1.53.4; 1.53.8; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.52	16-Dec-2010	cegger	branches: 1.52.2; 1.52.4; ioctl: KNF switch-case, remove useless brackets
1.51	19-Nov-2010	macallan	add PCI_MAGIC_IO_RANGE so non-x86 machines can map the IO range belonging to the console vga thanks to nick for reminding me to commit this and for confirming that it works as intended
1.50	09-Nov-2010	shattered	PR/10000: disable fast scrolling on Cyrix GXLV (CS5530 chipset) OK by wiz@
1.49	24-Feb-2010	dyoung	branches: 1.49.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.48	08-Jan-2010	dyoung	branches: 1.48.2; Expand PMF_FN_* macros.
1.47	18-Aug-2009	jmcneill	Adapt for NACPI -> NACPICA change.
1.46	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.45	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.44	03-Aug-2008	joerg	branches: 1.44.8; Allow using VGA_POST without ACPI again.
1.43	31-Jul-2008	joerg	machdep.acpi_vbios_reset = 2 --> vga_pci_resume will use x86emu to do a POST when options VGA_POST is present.
1.42	18-May-2008	jmcneill	branches: 1.42.4; Add 'rescan' function for vga.
1.41	05-May-2008	jmcneill	branches: 1.41.2; Change autoconf message 'direct rendering on dev unsupported' to the more accurate 'drm at dev not configured'.
1.40	14-Mar-2008	cube	branches: 1.40.2; 1.40.4; Split device_t and softc for all attachments of vga(4).
1.39	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.38	25-Dec-2007	joerg	branches: 1.38.2; 1.38.6; Add initial version of calling VGA POST from vga_resume. This is the equivalent to "vbetool post" using x86emu in the kernel.
1.37	09-Dec-2007	jmcneill	branches: 1.37.2; Merge jmcneill-pm branch.
1.36	01-Dec-2007	ad	branches: 1.36.2; 1.36.4; Back out previous.
1.35	01-Dec-2007	ad	Noisy printfs.
1.34	01-Dec-2007	jmcneill	aprintify
1.33	26-Aug-2007	martin	branches: 1.33.2; 1.33.8; Allow 64bit memory BARs if we are compiling on 64bit archs.
1.32	20-Mar-2007	drochner	branches: 1.32.4; 1.32.8; Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11. Minor modifications by me: -use an mi device major number -(coarsly) divided into pci card specific and less specific parts, moved the latter to dev/drm -renamed autoconf attributes to reflect this Todo: -adapt all card frontends but i915 to drm include file location -review the mtrr change -make the change to agp_i810.c coexist with the fix for buggy VESA BIOSes which is commented out temporarily -RCS IDs etc style stuff -LKM support (rescan support for vga) -test
1.31	04-Mar-2007	christos	branches: 1.31.2; 1.31.4; 1.31.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.30	16-Nov-2006	christos	branches: 1.30.2; 1.30.4; 1.30.8; __unused removal on arguments; approved by core.
1.29	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.28	11-Dec-2005	christos	branches: 1.28.20; 1.28.22; merge ktrace-lwp.
1.27	28-Jun-2005	thorpej	branches: 1.27.2; Use ANSI function decls and static.
1.26	27-Feb-2005	perry	nuke trailing whitespace
1.25	11-May-2004	wiz	branches: 1.25.4; 1.25.6; Fix typo in comment; from Sergey Svishchev.
1.24	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.23	29-Jun-2003	fvdl	branches: 1.23.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.22	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.21	26-Feb-2003	salo	Regen.
1.20	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.19	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.18	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.17	08-Jul-2002	drochner	there is at least one VGA chip (Silicon Motion Lynx EM+) where fast scrolling by moving the display offset pointer doesn't work
1.16	02-Jul-2002	drochner	consider all ATI cards broken wrt font handling, as suggested by Bang Jun-Young <junyoung@mogua.com>, but do it in way which allows single cards to be declared O.K. in case we find one
1.15	28-Jun-2002	drochner	clean up font handling: -treat the builtin font like any other font at runtime -for that, copy it to malloc()'d memory during attach() -in early console initialization, if we have to consider a broken card (VGA_CONSOLE_ATI_BROKEN_FONTSEL), copy the builtin font to another location in font ram; the attach() code will do the rest put the "quirk" code into effect again
1.14	27-Jun-2002	junyoung	Whitespace KNF.
1.13	27-Jun-2002	junyoung	- Work around a hardware bug that loaded fonts don't work, which is found on many (all?) of PCI-based ATI graphics cards. It is fully optional and can be enabled by adding `options VGA_CONSOLE_ATI_BROKEN_FONTSEL' to config file. - Temporarily remove `quirk' mechanism. Similar code already exists in pci_quirks.c.
1.12	25-Jun-2002	drochner	allow to overwrite the builtin VGA font if necessary, make the number of available font slots variable, set up a "quirk" mechanism to tell the generic vga code about crippled VGA adapters which ignore the "fontsel" TS register, initiate the quirk table with an ATI chip which happened to be on a board I tested with. Afaik quite a number of ATI chips suffers from the "loaded fonts don't work" problem - these should be added. Bad side effect of my change: The builtin font will be kicked out always if a VGA_CONSOLE_SCREENTYPE is specified which needs a loaded font. In early console initialization, we don't know much about the graphics card, so we have to assume the worst (ie ATI:-).
1.11	30-May-2002	drochner	implement a check whether a BAR is present at all at a given configuration space address and use it where the mappings of the VGA card are registered before descenting too deep into "memory" type specific code (pci_mem_find() gets noisy if it doesn't like the register)
1.10	17-Mar-2002	atatat	branches: 1.10.4; Convert ioctl code to use EPASSTHROUGH instead of -1 or ENOTTY for indicating an unhandled "command". ERESTART is -1, which can lead to confusion. ERESTART has been moved to -3 and EPASSTHROUGH has been placed at -4. No ioctl code should now return -1 anywhere. The ioctl() system call is now properly restartable.
1.9	13-Dec-2001	junyoung	ANSIfication & KNF fixes.
1.8	13-Nov-2001	lukem	add RCSID
1.7	14-Sep-2001	thorpej	Gather information about the mappable regions of the card at attach time. Allow access to these regions in vga_pci_mmap(). Note: we also have to allow access to the legacy ISA "hole" in vga_pci_mmap(), since the X server needs to be able to read the legacy BIOS on many cards. XXX Need to address the PCI Expansion ROM. Based on changes from Simon Burge <simonb@wasabisystems.com>.
1.6	14-Sep-2001	thorpej	Add ioctl and mmap entry points for PCI VGA. ioctl provides passthrough for PCI config space read/write. mmap currently returns error (to be addressed in a future revision).
1.5	14-Sep-2001	thorpej	Allow bus front-end to provide both ioctl and mmap entry points, and also the wsdisplay type. Based on changes from Simon Burge <simonb@wasabisystems.com>.
1.4	14-Aug-2000	thorpej	branches: 1.4.2; 1.4.4; 1.4.6; Just add the mmap argument to vga_common_attach(), don't bother with vga_extended_attach().
1.3	08-Jun-1998	thorpej	branches: 1.3.14; Nuke __BROKEN_INDIRECT_CONFIG.
1.2	28-May-1998	drochner	Put definitions and subroutines needed for all PC display adapters (MGA/GCA and compatibles) into global headers / source files. Let the VGA driver use them.
1.1	22-Mar-1998	drochner	PCI frontend for sys/dev/ic/vga.c
1.3.14.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.4.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.4.4.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.4.4.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.4.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.4.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.4.2.7	18-Oct-2002	nathanw	Catch up to -current.
1.4.2.6	01-Aug-2002	nathanw	Catch up to -current.
1.4.2.5	20-Jun-2002	nathanw	Catch up to -current.
1.4.2.4	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.4.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.4.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.4.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.10.4.2	15-Jul-2002	gehenna	catch up with -current.
1.10.4.1	20-Jun-2002	gehenna	catch up with -current.
1.23.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.23.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.23.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.23.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.23.2.2	03-Aug-2004	skrll	Sync with HEAD
1.23.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.25.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.25.4.1	29-Apr-2005	kent	sync with -current
1.27.2.6	17-Mar-2008	yamt	sync with head.
1.27.2.5	21-Jan-2008	yamt	sync with head
1.27.2.4	07-Dec-2007	yamt	sync with head
1.27.2.3	03-Sep-2007	yamt	sync with head.
1.27.2.2	30-Dec-2006	yamt	sync with head.
1.27.2.1	21-Jun-2006	yamt	sync with head.
1.28.22.2	10-Dec-2006	yamt	sync with head.
1.28.22.1	22-Oct-2006	yamt	sync with head
1.28.20.1	18-Nov-2006	ad	Sync with head.
1.30.8.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.30.4.2	24-Mar-2007	yamt	sync with head.
1.30.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.30.2.1	28-Aug-2007	liamjfoy	Pull up following revision(s) (requested by martin in ticket #848): sys/dev/pci/vga_pci.c: revision 1.33 Allow 64bit memory BARs if we are compiling on 64bit archs.
1.31.6.1	29-Mar-2007	reinoud	Pullup to -current
1.31.4.1	11-Jul-2007	mjf	Sync with head.
1.31.2.2	09-Oct-2007	ad	Sync with head.
1.31.2.1	10-Apr-2007	ad	Sync with head.
1.32.8.6	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.32.8.5	01-Dec-2007	jmcneill	Sync with HEAD.
1.32.8.4	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.32.8.3	01-Oct-2007	joerg	Extend device API by device_power_private and device_power_set_private. The latter is a temporary mean until the pnp_register API itself is overhault. This functions allow a generic power handler to store its state independent of the driver. Use this and revamp the PCI power handling. Pretty much all PCI devices had power handlers that did the same thing, generalize this in pci_generic_power_register/deregister and the handler. This interface offers callbacks for the drivers to save and restore state on transistions. After a long discussion with jmcneill@ it was considered to be powerful enough until evidence is shown that devices can handle D1/D2 with less code and higher speed than without the full save/restore. The generic code is carefully written to handle device without PCI-PM support and ensure that the correct registers are written to when D3 loses all state. Reimplement the generic PCI network device handling on top of PCI generic power handling. Introduce pci_disable_retry as used and implemented locally at least by ath(4) and iwi(4). Use it in this drivers to restore behaviour from before the introduction of generic PCI network handling. Convert all PCI drivers that were using pnp_register to the new framework. The only exception is vga(4) as it is commonly used as console device. Add a note therein that this should be fixed later.
1.32.8.2	03-Sep-2007	jmcneill	Sync with HEAD.
1.32.8.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.32.4.1	03-Sep-2007	skrll	Sync with HEAD.
1.33.8.3	18-Feb-2008	mjf	Sync with HEAD.
1.33.8.2	27-Dec-2007	mjf	Sync with HEAD.
1.33.8.1	08-Dec-2007	mjf	Sync with HEAD.
1.33.2.2	23-Mar-2008	matt	sync with HEAD
1.33.2.1	09-Jan-2008	matt	sync with HEAD
1.36.4.1	11-Dec-2007	yamt	sync with head.
1.36.2.1	26-Dec-2007	ad	Sync with head.
1.37.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.38.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.38.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.38.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.38.2.1	24-Mar-2008	keiichi	sync with head.
1.40.4.5	11-Mar-2010	yamt	sync with head
1.40.4.4	19-Aug-2009	yamt	sync with head.
1.40.4.3	16-May-2009	yamt	sync with head
1.40.4.2	04-May-2009	yamt	sync with head.
1.40.4.1	16-May-2008	yamt	sync with head.
1.40.2.2	04-Jun-2008	yamt	sync with head
1.40.2.1	18-May-2008	yamt	sync with head.
1.41.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.41.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.42.4.1	19-Oct-2008	haad	Sync with HEAD.
1.44.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.48.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.49.2.1	05-Mar-2011	rmind	sync with head
1.52.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.52.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.53.8.1	18-Feb-2012	mrg	merge to -current.
1.53.4.1	17-Apr-2012	yamt	sync with head
1.54.24.1	09-Jul-2016	skrll	Sync with HEAD
1.54.6.1	03-Dec-2017	jdolecek	update from HEAD
1.56.4.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.57.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	20-Mar-2007	drochner	Import DRM drivers, brought into shape by Yorick Hardy, posted to tech-x11. Minor modifications by me: -use an mi device major number -(coarsly) divided into pci card specific and less specific parts, moved the latter to dev/drm -renamed autoconf attributes to reflect this Todo: -adapt all card frontends but i915 to drm include file location -review the mtrr change -make the change to agp_i810.c coexist with the fix for buggy VESA BIOSes which is commented out temporarily -RCS IDs etc style stuff -LKM support (rescan support for vga) -test
1.4	11-Dec-2005	christos	branches: 1.4.26; 1.4.30; 1.4.32; 1.4.34; merge ktrace-lwp.
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	04-Feb-2005	perry	de-__P
1.1	22-Mar-1998	drochner	branches: 1.1.48; 1.1.56; 1.1.58; PCI frontend for sys/dev/ic/vga.c
1.1.58.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.1.58.1	12-Feb-2005	yamt	sync with head.
1.1.56.1	29-Apr-2005	kent	sync with -current
1.1.48.2	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.48.1	04-Feb-2005	skrll	Sync with HEAD.
1.3.4.1	03-Sep-2007	yamt	sync with head.
1.4.34.1	29-Mar-2007	reinoud	Pullup to -current
1.4.32.1	11-Jul-2007	mjf	Sync with head.
1.4.30.1	10-Apr-2007	ad	Sync with head.
1.4.26.1	24-Mar-2007	yamt	sync with head.
1.34	04-Mar-2018	jdolecek	according to VT82C686A chip specs, the VIA Technologies device 0x3057 is more a Power Management controller, rename the pcidevs entry and device macro PR kern/31963 by Nicolas Joly
1.33	11-Aug-2014	ozaki-r	Add missing curly braces It unbreaks the i386 build of ALL.
1.32	10-Aug-2014	tls	Merge tls-earlyentropy branch into HEAD.
1.31	20-Jun-2011	pgoyette	branches: 1.31.12; 1.31.26; Initialize sensors states before registering.
1.30	10-Dec-2010	phx	branches: 1.30.6; Initialize sc_iot early enough to avoid a crash when the hardware monitor function was found disabled.
1.29	30-Apr-2008	ad	branches: 1.29.22; Make various bits of debug code compile again.
1.28	10-Apr-2008	cegger	branches: 1.28.2; 1.28.4; use aprint_*_dev and device_xname
1.27	05-Mar-2008	xtraeme	Split device_t/softc and other related cosmetic changes.
1.26	19-Nov-2007	njoly	branches: 1.26.10; 1.26.14; Do sensors inits before trying to attach them on the sysmon_envsys device.
1.25	16-Nov-2007	xtraeme	Extend the envsys2 API (one more time, sorry) as defined in: http://mail-index.netbsd.org/tech-kern/2007/11/09/0001.html sysmon_envsys_create() and sysmon_envsys_destroy() were added to create/destroy sysmon_envsys objects (and its TAILQ/LIST for sensors/events). sysmon_envsys_sensor_attach() and sysmon_envsys_sensor_detach() were added to attach/detach sensors to a specified sysmon_envsys device. The events framework is now per device and configurable via the ENVSYS_SETDICTIONARY ioctl or /etc/envsys.conf and envstat(8). Update all users and documentation to reflect these changes.
1.24	19-Oct-2007	ad	branches: 1.24.2; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.23	26-Aug-2007	xtraeme	branches: 1.23.2; 1.23.6; There's no need to add timecounter code inside of __HAVE_TIMECOUNTER ifdefs.
1.22	01-Jul-2007	xtraeme	branches: 1.22.2; 1.22.6; In the gtredata function, use the edata pointer rather than using directly the edata array index in the softc.
1.21	01-Jul-2007	xtraeme	Imported envsys 2, a brief description of the new features: (Part 2: drivers) * Support for detachable sensors. * Cleaned up the API for simplicity and efficiency. * Ability to send capacity/critical/warning events to powerd(8). * Adapted all the code to the new locking order. * Compatibility with the old envsys API: the ENVSYS_GTREINFO and ENVSYS_GTREDATA ioctl(2)s are supported. * Added support for a 'dictionary based communication channel' between sysmon_power(9) and powerd(8), that means there is no 32 bytes event size restriction anymore. * Binary compatibility with old envstat(8) and powerd(8) via COMPAT_40. * All drivers with the n^2 gtredata bug were fixed, PR kern/36226. Tested by: blymn: smsc(4). bouyer: ipmi(4), mfi(4). kefren: ug(4). njoly: viaenv(4), adt7463.c. riz: owtemp(4). xtraeme: acpiacad(4), acpibat(4), acpitz(4), aiboost(4), it(4), lm(4).
1.20	18-Feb-2007	xtraeme	branches: 1.20.4; 1.20.6; There's no need to use a simple_lock here, ok ad@.
1.19	16-Feb-2007	njoly	branches: 1.19.2; Fix incorrect voltage sensors range. It was using 7 numbers, where the driver support only 5 voltage sensors. ok by xtraeme
1.18	20-Jan-2007	xtraeme	Updated viaenv(4) driver: * Support for the VIA VT8231 Hardware monitor. * Power Management Timer available for timecounters in both VT86C686A and VT8231 (code simplified thanks to dev/ic/acpipmtimer). * Remove viapm(4) code and manpage (which was a link to viaenv.4 anyway). From OpenBSD, tested by some users.
1.17	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.16	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.15	29-Jul-2006	riz	branches: 1.15.4; 1.15.6; Correct an error introduced in rev 1.3; if there are no sensors of a given type, this is indicated by "low" > "high" in the envsys_range struct - the comments (and revision 1.2 of this file) indicate there are no sensors of type ENVSYS_SVOLTS_AC, so fix viaenv_ranges[] to agree with this.
1.14	07-Jun-2006	kardel	merge FreeBSD timecounters from branch simonb-timecounters - struct timeval time is gone time.tv_sec -> time_second - struct timeval mono_time is gone mono_time.tv_sec -> time_uptime - access to time via {get,}{micro,nano,bin}time() get* versions are fast but less precise - support NTP nanokernel implementation (NTP API 4) - further reading: Timecounter Paper: http://phk.freebsd.dk/pubs/timecounter.pdf NTP Nanokernel: http://www.eecis.udel.edu/~mills/ntp/html/kern.html
1.13	11-Dec-2005	christos	branches: 1.13.4; 1.13.6; 1.13.8; 1.13.14; merge ktrace-lwp.
1.12	23-Oct-2005	jmcneill	If the temperature sensors are disabled, enable them rather than giving up.
1.11	28-Jun-2005	thorpej	branches: 1.11.2; 1.11.4; Use ANSI function decls and static.
1.10	27-Feb-2005	perry	nuke trailing whitespace
1.9	02-Oct-2002	thorpej	branches: 1.9.6; 1.9.14; 1.9.16; Add trailing ; to CFATTACH_DECL.
1.8	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.7	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.6	25-Mar-2002	thorpej	There is no need to use a kthread to read the VIA environment sensors.
1.5	13-Nov-2001	lukem	add RCSID
1.4	27-Aug-2001	ad	Use PWAIT for the lock, not 0.
1.3	24-Jun-2000	thorpej	branches: 1.3.2; 1.3.4; 1.3.6; Add a clearing-house pseudo-device for system monitoring devices such as the LM78 and VT82C686A (and eventually ACPI). Multiple sensor devices can be hooked registered with `sysmon', and eventually sysmon will also handle hardware (and software) watchdog timers. Convert the `lm' and `viaenv' drivers to the new interface.
1.2	12-May-2000	thorpej	branches: 1.2.4; Apply indent(1).
1.1	08-May-2000	joda	VIA VT82C686A hardware monitor
1.2.4.2	30-Sep-2001	he	Pull up revision 1.4 (requested by ad): Use PWAIT for the lock, not 0.
1.2.4.1	30-Jul-2000	bouyer	Pull up (requested by thorpej): lm_pnpbios.c 1.2->1.3 files 1.374->1.375 nslm7x.c 1.3->1.4 nslm7xvar.h 1.2->1.3 lm_isa.c 1.2->1.3 files.pci 1.96->1.97 viaenv.c 1.2->1.3 sysmon.c 1.1 sysmonvar.h 1.1 Add a clearing-house pseudo-device for system monitoring devices such as the LM78 and VT82C686A (and eventually ACPI). Multiple sensor devices can be hooked registered with `sysmon', and eventually sysmon will also handle hardware (and software) watchdog timers. Convert the `lm' and `viaenv' drivers to the new interface.
1.3.6.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.3.6.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.3.6.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.3.6.1	13-Sep-2001	thorpej	Update the kqueue branch to HEAD.
1.3.4.4	18-Oct-2002	nathanw	Catch up to -current.
1.3.4.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.3.4.2	14-Nov-2001	nathanw	Catch up to -current.
1.3.4.1	21-Sep-2001	nathanw	Catch up to -current.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	24-Jun-2000	bouyer	file viaenv.c was added on branch thorpej_scsipi on 2000-11-20 11:42:39 +0000
1.9.16.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.9.14.1	29-Apr-2005	kent	sync with -current
1.9.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.6.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.11.4.1	26-Oct-2005	yamt	sync with head
1.11.2.7	17-Mar-2008	yamt	sync with head.
1.11.2.6	07-Dec-2007	yamt	sync with head
1.11.2.5	27-Oct-2007	yamt	sync with head.
1.11.2.4	03-Sep-2007	yamt	sync with head.
1.11.2.3	26-Feb-2007	yamt	sync with head.
1.11.2.2	30-Dec-2006	yamt	sync with head.
1.11.2.1	21-Jun-2006	yamt	sync with head.
1.13.14.1	19-Jun-2006	chap	Sync with head.
1.13.8.2	11-Aug-2006	yamt	sync with head
1.13.8.1	26-Jun-2006	yamt	sync with head.
1.13.6.1	04-Feb-2006	simonb	Adapt for timecounters: mostly use get*time() and use "time_second" instead of "time.tv_sec".
1.13.4.1	09-Sep-2006	rpaulo	sync with head
1.15.6.2	10-Dec-2006	yamt	sync with head.
1.15.6.1	22-Oct-2006	yamt	sync with head
1.15.4.2	01-Feb-2007	ad	Sync with head.
1.15.4.1	18-Nov-2006	ad	Sync with head.
1.19.2.1	27-Feb-2007	yamt	- sync with head. - move sched_changepri back to kern_synch.c as it doesn't know PPQ anymore.
1.20.6.1	11-Jul-2007	mjf	Sync with head.
1.20.4.3	23-Oct-2007	ad	Sync with head.
1.20.4.2	09-Oct-2007	ad	Sync with head.
1.20.4.1	15-Jul-2007	ad	Sync with head.
1.22.6.3	21-Nov-2007	joerg	Sync with HEAD.
1.22.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.22.6.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.22.2.1	03-Sep-2007	skrll	Sync with HEAD.
1.23.6.3	21-Nov-2007	bouyer	Sync with HEAD
1.23.6.2	18-Nov-2007	bouyer	Sync with HEAD
1.23.6.1	25-Oct-2007	bouyer	Sync with HEAD.
1.23.2.3	23-Mar-2008	matt	sync with HEAD
1.23.2.2	09-Jan-2008	matt	sync with HEAD
1.23.2.1	06-Nov-2007	matt	sync with HEAD
1.24.2.2	08-Dec-2007	mjf	Sync with HEAD.
1.24.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.26.14.2	02-Jun-2008	mjf	Sync with HEAD.
1.26.14.1	03-Apr-2008	mjf	Sync with HEAD.
1.26.10.1	24-Mar-2008	keiichi	sync with head.
1.28.4.1	16-May-2008	yamt	sync with head.
1.28.2.1	18-May-2008	yamt	sync with head.
1.29.22.1	05-Mar-2011	rmind	sync with head
1.30.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.31.26.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.31.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.105	17-Oct-2025	andvar	viaide(4): add UDMA support for VT8233C IDE controller. The VT8233C is another obscure VIA southbridge and appears to be a variant of the VT8233 rather than the VT8233A (which supports ATA133). The C variant was paired with a 3Com network controller instead of VIA’s own NIC. Detect the ISA–PCI bridge and enable UDMA100 capability for improved IDE performance. Update the viaide(4) man page and refine the list of supported VIA controllers.
1.104	04-Oct-2025	thorpej	Use device_getprop_bool() for "use-compat-native-irq".
1.103	01-Jun-2025	andvar	branches: 1.103.2; viaide(4): give a bit more context on why ISA bridge is used to identify VIA IDE controllers. NFCI.
1.102	01-Jun-2025	andvar	viaide(4): explicitly enable channels for VT6410 add-in cards. VT6410 and VT6415 have different reasons for "missing" enable bits. VT6410 uses the standard 0x40 offset for channel enable bits, but some add-in cards have both channels disabled by default. This requires explicitly writing the enable bits; otherwise, devices will not be detected. The same bug is present in Linux and OpenBSD, where the bits are also simulated. It was reported in Linux but closed as no-fix due to an unidentified root cause. VT6415 returns 0 at offsets 0x40-0x43 and likely uses a different location for its single channel enable bit, or it may not expose it at all. Until the correct offset is identified (if ever), the driver must simulate the enable bits.
1.101	31-May-2025	andvar	viaide(4): use via_chip_map() to attach VT8261 SATA controller in RAID mode. Similarly to VT8251 and CX700, VT8261 has two ports per channel, while via_sata_chip_map supports only one drive per channel. This caused disks on secondary ports (2, 4) to not be detected, which is addressed by this change.
1.100	09-May-2025	andvar	remove leading tab in the empty line. NFCI.
1.99	20-Apr-2025	andvar	viaide(4): use __func__ in via_setup_channel() ATADEBUG_PRINT message. Fixes function name in debug message, otherwise NFCI.
1.98	20-Apr-2025	andvar	viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them).
1.97	20-Apr-2025	andvar	viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad.
1.96	05-Apr-2025	andvar	viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID.
1.95	03-Apr-2025	andvar	viaide(4): Add support for VIA VT6415/VT6330 single-channel IDE controllers. Beyond adding the usual identification code, several adjustments were required: 1) Introduced 'single_channel` to set 'nchannels` to 1 for this controller. 2) Added 'APO_IDECONF_ALWAYS_EN` definition and the 'no_ideconf` flag to mimic enabled channel bits, as this controller's enable chip register is unset. Applied the same to VT6410, following OpenBSD and Linux, since some controllers are known to not set this register as well. 3) Combined all VIA controllers identified by PCI ID (not ISA bus) using switch fallthrough, as they currently share the same UDMA setting. Consequently, the VX900 name printing was moved to the device description and adjusted to better reflect the device. 4) Moved setting interface bits for RAID controllers under the RAID capability check, enabling the above fallthrough and paving the way for support of other RAID-capable controllers in upcoming changes. 5) The VT6330 is a combo FireWire/IDE controller. Added its FireWire controller PCI ID to pcidevs. Tested on ASRock P5B-DE. Reviewed by jak and bad. Addresses PR kern/45917
1.94	16-Mar-2025	andvar	viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. Pullups to netbsd-9 and netbsd-10 are needed.
1.93	11-Mar-2025	andvar	viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. Worth pulling up to netbsd-9, 10.
1.92	03-Mar-2025	andvar	viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Pull-ups to netbsd-9, 10 are needed. Reviewed and approved by jakllsch@.
1.91	23-Jan-2025	andvar	viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaur's demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board.
1.90	17-Oct-2024	andvar	Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855.
1.89	02-Jun-2019	jdolecek	branches: 1.89.2; 1.89.28; 1.89.34; updated patch from Andrius - update the chip map for CX700 to use via_sata_chip_map_new, this seems to eliminate the ghost non-existing viaide1 XXX guess should use via_sata_chip_map_new for PCI_PRODUCT_VIATECH_CX700M2_IDE XXX too, but this needs test on real hw remove again the entries for VX800 and VX855 from pciide_via_products[], as that part is untested with real hw update for PR kern/43309
1.88	01-Jun-2019	jdolecek	match VIA Technologies VX800, VX855, and VT8237S in RAID mode patch provided and testing done by Andrius Varanavicius via PR kern/43309
1.87	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.86	20-Oct-2017	jdolecek	branches: 1.86.2; 1.86.4; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.85	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.84	29-Mar-2014	christos	branches: 1.84.18; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.83	31-Jul-2012	bouyer	branches: 1.83.2; 1.83.4; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.82	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.81	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.80	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.79	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.78	18-Mar-2012	tsutsui	Add UDMA support for VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. This should be pulled up to netbsd-6, since ZOTAC ZBOX nano VD01 http://www.zotacusa.com/zbox-nano-vd01.html exhibited at NetBSD booth in Open Source Conference 2012 Tokyo Spring was favored by many people.
1.77	28-Dec-2011	phx	branches: 1.77.2; Add VT6410 support, heavily based on a patch by Takahiro Kambe: http://mail-index.netbsd.org/port-i386/2007/03/30/0000.html The VT6410 is special, because it may exist without a VIA PCI-ISA bridge.
1.76	10-Jul-2011	jakllsch	branches: 1.76.2; 1.76.6; Add complete support for all channels on VT6421A SATA/PATA controller to viaide(4).
1.75	17-May-2011	dyoung	PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED changed their functional role in NetBSD (drivers are no longer supposed to write these to pa_flags) without changing name. Correct that. Rename PCI_FLAGS_IO_ENABLED to PCI_FLAGS_IO_OKAY and PCI_FLAGS_MEM_ENABLED to PCI_FLAGS_MEM_OKAY, thus making their names consistent with the other PCI flags and poisoning 3rd-party driver sources that use the flags in the old bad way. This patch produces no binary changes in this set of PCI kernels when they are compiled w/o 'options DIAGNOSTIC' and w/ -V MKREPRO=yes: algor P4032 P5064 P6032 alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE evbarm-el GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 evbarm-el IXDP425 IXM1200 KUROBOX_PRO evbarm-el LUBBOCK MARVELL_NAS NAPPI NSLU2 SHEEVAPLUG SMDK2800 TEAMASA_NPWR evbarm-el TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 evbppc OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC iyonix GENERIC landisk GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sbmips-el GENERIC sgimips GENERIC32_IP2x GENERIC32_IP3x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC
1.74	10-May-2011	dyoung	Stop abuse of PCI_FLAGS_IO_ENABLED and PCI_FLAGS_MEM_ENABLED. The flags specifically tell a driver that bridges upstream forward PCI I/O (or memory) transactions to the device." Only bus drivers have any business modifying these.
1.73	10-Apr-2011	jakllsch	Fix regression introduced in 1.72. (Attach arguments need to be copied before using the copy.)
1.72	04-Apr-2011	dyoung	Neither pci_dma64_available(), pci_probe_device(), pci_mapreg_map(9), pci_find_rom(), pci_intr_map(9), pci_enumerate_bus(), nor the match predicate passed to pciide_compat_intr_establish() should ever modify their pci_attach_args argument, so make their pci_attach_args arguments const and deal with the fallout throughout the kernel. For the most part, these changes add a 'const' where there was no 'const' before, however, some drivers and MD code used to modify pci_attach_args. Now those drivers either copy their pci_attach_args and modify the copy, or refrain from modifying pci_attach_args: Xen: according to Manuel Bouyer, writing to pci_attach_args in pci_intr_map() was a leftover from Xen 2. Probably a bug. I stopped writing it. I have not tested this change. siside(4): sis_hostbr_match() needlessly wrote to pci_attach_args. Probably a bug. I use a temporary variable. I have not tested this change. slide(4): sl82c105_chip_map() overwrote the caller's pci_attach_args. Probably a bug. Use a local pci_attach_args. I have not tested this change. viaide(4): via_sata_chip_map() and via_sata_chip_map_new() overwrote the caller's pci_attach_args. Probably a bug. Make a local copy of the caller's pci_attach_args and modify the copy. I have not tested this change. While I'm here, make pci_mapreg_submap() static. With these changes in place, I have tested the compilation of these kernels: alpha GENERIC amd64 GENERIC XEN3_DOM0 arc GENERIC atari HADES MILAN-PCIIDE bebox GENERIC cats GENERIC cobalt GENERIC evbarm-eb NSLU2 evbarm-el ADI_BRH ARMADILLO9 CP3100 GEMINI GEMINI_MASTER GEMINI_SLAVE GUMSTIX HDL_G IMX31LITE INTEGRATOR IQ31244 IQ80310 IQ80321 IXDP425 IXM1200 KUROBOX_PRO LUBBOCK MARVELL_NAS NAPPI SHEEVAPLUG SMDK2800 TEAMASA_NPWR TEAMASA_NPWR_FC TS7200 TWINTAIL ZAO425 evbmips-el AP30 DBAU1500 DBAU1550 MALTA MERAKI MTX-1 OMSAL400 RB153 WGT624V3 evbmips64-el XLSATX evbppc EV64260 MPC8536DS MPC8548CDS OPENBLOCKS200 OPENBLOCKS266 OPENBLOCKS266_OPT P2020RDB PMPPC RB800 WALNUT hp700 GENERIC i386 ALL XEN3_DOM0 XEN3_DOMU ibmnws GENERIC macppc GENERIC mvmeppc GENERIC netwinder GENERIC ofppc GENERIC prep GENERIC sandpoint GENERIC sgimips GENERIC32_IP2x sparc GENERIC_SUN4U KRUPS sparc64 GENERIC As of Sun Apr 3 15:26:26 CDT 2011, I could not compile these kernels with or without my patches in place: ### evbmips-el GDIUM nbmake: nbmake: don't know how to make /home/dyoung/pristine-nbsd/src/sys/arch/mips/mips/softintr.c. Stop ### evbarm-el MPCSA_GENERIC src/sys/arch/evbarm/conf/MPCSA_GENERIC:318: ds1672rtc*: unknown device `ds1672rtc' ### ia64 GENERIC /tmp/genassym.28085/assym.c: In function 'f111': /tmp/genassym.28085/assym.c:67: error: invalid application of 'sizeof' to incomplete type 'struct pcb' /tmp/genassym.28085/assym.c:76: error: dereferencing pointer to incomplete type ### sgimips GENERIC32_IP3x crmfb.o: In function `crmfb_attach': crmfb.c:(.text+0x2304): undefined reference to `ddc_read_edid' crmfb.c:(.text+0x2304): relocation truncated to fit: R_MIPS_26 against `ddc_read_edid' crmfb.c:(.text+0x234c): undefined reference to `edid_parse' crmfb.c:(.text+0x234c): relocation truncated to fit: R_MIPS_26 against `edid_parse' crmfb.c:(.text+0x2354): undefined reference to `edid_print' crmfb.c:(.text+0x2354): relocation truncated to fit: R_MIPS_26 against `edid_print'
1.71	06-Nov-2010	jakllsch	branches: 1.71.2; Enable detachment. Tested on nForce 4 IDE (compatiblity mapped) and SATA (native mapped).
1.70	05-Nov-2010	jakllsch	Pave the way for detachment of pciide(4)-family ATA controllers.
1.69	22-Mar-2010	dyoung	Activate PCI memory-mapped I/O access in the PCI Control and Status Register if the PCI BAR is the memory-mapped type. The driver used to decide whether to activate memory-mapped access by testing the truth of pci_attach_args.pa_memt; that isn't MI, and it doesn't make sense.
1.68	24-Feb-2010	dyoung	branches: 1.68.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.67	08-Jan-2010	dyoung	branches: 1.67.2; Expand PMF_FN_* macros.
1.66	23-Dec-2009	mrg	enable mem space if it isn't already enabled. several bioses leave this not enabled, and viaide kind of depends upon it. fixes viaide(4) on several machines. idea from PR 38011, but reworked a little.
1.65	22-Dec-2009	tsutsui	Don't forget to initialize data32iot and data32ioh in via_sata_chip_map_new(). This makes 32bit UDMA xfer work properly on VIA VT6421.
1.64	21-Nov-2009	jakllsch	Fix matching of CX700. Should address PR/39063. Also, Change bus_space_subregion() size for each SATA PHY register from 1 to 4, as we do a bus_space_read_4() within these subregions. Use PCI_MAPREG_START where appropriate.
1.63	20-Nov-2009	jakllsch	Switch VT8237R SATA to via_sata_chip_map_7(). via_sata_chip_map_7() works fine and is consistent with the other VIA SATA controllers. Tested on VT8237R+. (Somewhat amazing that via_sata_chip_map_0() worked at all for the second port.)
1.62	18-Nov-2009	jakllsch	Match VT8251 IDE controller and enable UDMA Mode 6 thereon. Should fix PATA speed issue in PR/37517. Due to lack of hardware, only compile tested at this point.
1.61	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.60	26-Sep-2009	jmcneill	Add PCI_PRODUCT_VIATECH_VT8237A_SATA_2 to list of product IDs.
1.59	23-Aug-2009	jmcneill	s/Advanced Micro Devices/AMD/ for device descriptions.
1.58	21-Dec-2008	nonaka	Add VT8237S Integrated SATA Controller support.
1.57	06-Sep-2008	rmind	branches: 1.57.2; 1.57.4; 1.57.10; PR/39119: Jan Thorsson: Support for VIA CX700M2 IDE controller.
1.56	17-Jun-2008	phx	branches: 1.56.2; Renamed via_pegasos_mapregs_native to via_mapregs_compat_native, after discussion with simonb. There may be other architectures who want to use this function.
1.55	15-Jun-2008	phx	Fixed system lockup in native mode with compat irq 14/15, which can be selected with the use-compat-native-irq property. It was caused by handling irqs from both channels at once with the wrong interrupt handler routine (pciide_pci_intr). Now pciide_compat_intr is installed for each channel. Also renamed the function to via_pegasos_mapregs_native(), which seems more appropriate.
1.54	14-May-2008	dholland	branches: 1.54.2; fix typo in error message
1.53	18-Mar-2008	cube	branches: 1.53.2; 1.53.4; 1.53.6; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.52	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.51	18-Jan-2008	joerg	branches: 1.51.2; 1.51.6; Add power management support. Untested due to lack of feedback.
1.50	20-Dec-2007	phx	VT8231 IDE workaround for native mode with "compatibility" interrupts 14/15. This time done right by calling pciide_machdep_compat_intr_establish() to get a MD interrupt handler. Verified that it compiles for XEN again.
1.49	17-Dec-2007	phx	xen did no longer build. Temporarily reverted to 1.47. Will need some time to find a good solution.
1.48	16-Dec-2007	phx	At least under certain (mis)configurations (e.g. on the "Pegasos" board) the VT8231-IDE's native mode only works with irq 14/15, and cannot be programmed to use a single native PCI irq alone. So we install an interrupt handler for each channel, as in compatibility mode. This behaviour has to be activated by the "use-compat-native-irq" property, otherwise the driver works as before.
1.47	14-Nov-2007	xtraeme	branches: 1.47.2; 1.47.6; emo-style: forgot to update the switch for MCP7[13] IDE.
1.46	14-Nov-2007	xtraeme	Attach to the NVIDIA MCP7[37] IDE Controllers.
1.45	04-Oct-2007	xtraeme	branches: 1.45.2; 1.45.4; AMD CS5536 Companion Device is able to work in Ultra DMA Mode 5, patch from Andreas Wrede in PR kern/37059.
1.44	31-Aug-2007	xtraeme	branches: 1.44.2; Use a different case for the CX700 IDE Controller to print the attach message and when setting the UDMA mode.
1.43	31-Aug-2007	xtraeme	- Proper VIA CX700 IDE Controller support (missed in previous). - Attach to the NVIDIA MCP67 IDE and SATA controllers.
1.42	31-Aug-2007	xtraeme	Attach to the VIA CX700 IDE Controller.
1.41	27-Mar-2007	garbled	branches: 1.41.4; 1.41.8; 1.41.10; The VT8237A SATA Controller uses chip_map_7, not 0, tested and confirmed working on my machine. Also when a via controller is set to RAID mode, it sets the pci_subclass to raid. Notice this and set ATAC_CAP_RAID.
1.40	10-Feb-2007	mlelstv	branches: 1.40.2; 1.40.6; 1.40.8; 1.40.10; Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.39	09-Feb-2007	ad	Merge newlock2 to head.
1.38	21-Jan-2007	isaki	Add support for CS5536. Reviewed by tsutsui@ on port-i386.
1.37	16-Nov-2006	christos	branches: 1.37.2; __unused removal on arguments; approved by core.
1.36	12-Nov-2006	bouyer	Properly detect if BAR5 is IO or MEM type; avoid a message about not being able to map the BAR on boot (note that it's just cosmetic, the registers were mapped anyway). Tested by paul at whooppee.com; thanks !
1.35	27-Oct-2006	bouyer	Add SATA native registers support for VIA SATA controllers, so that drives are probed using the SATA way (from FreeBSD). While here add the VT8237A SATA Controller to the tables, should fix PR kern/34927. Thanks to the C�sar Catri�n Carre�o and paul (at) whooppee.com for tests.
1.34	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.33	03-Sep-2006	xtraeme	branches: 1.33.2; 1.33.4; Add support for nvidia MCP61/65 IDE/SATA Controllers.
1.32	01-Aug-2006	xtraeme	branches: 1.32.2; Match nForce3 250 SATA2.
1.31	17-Jun-2006	xtraeme	s/SATA/Serial ATA/
1.30	17-Jun-2006	xtraeme	Enable support for NVIDIA MCP04 and MCP55 IDE/SATA Controllers, tested by rillig with MCP55.
1.29	30-Apr-2006	xtraeme	branches: 1.29.2; 1.29.4; PR port-amd64/33393: viaide driver patch for VT8237R SATA.
1.28	31-Jan-2006	xtraeme	branches: 1.28.2; 1.28.4; 1.28.6; 1.28.8; 1.28.10; The NVidia nForce430 IDE Controller supports Ultra-DMA Mode 6, enable it. Patch from MASUDA Hideo via tech-kern.
1.27	02-Jan-2006	manu	Add support for nForce430 ATA133 and SATA controllers. My disks now run at a decent speed.
1.26	11-Dec-2005	christos	branches: 1.26.2; merge ktrace-lwp.
1.25	24-May-2005	lukem	branches: 1.25.2; add __KERNEL_RCSID
1.24	27-Feb-2005	perry	branches: 1.24.2; nuke trailing whitespace
1.23	02-Feb-2005	abs	Reminder not to trust cheap no-name packaging. VIA product 0x3249 is a VT6421 not VT642. Also add to list in viaide manpage and note that drives need to be configured into RAID/JOBS sets via its BIOS. http://www.via.com.tw/en/products/peripherals/serial-ata_raid/vt6421/
1.22	01-Feb-2005	abs	Add 'VT642 SATA RAID Controller'. It looks like a viaide, it smells like a viaide... but you can only address the drives once you have put two or more of them into a RAID or JBOD.
1.21	27-Jan-2005	kent	support for nForce 4 PATA and SATA
1.20	13-Jan-2005	jdolecek	branches: 1.20.2; 1.20.4; match NVIDIA nForce2 Ultra 400 IDE and Serial ATA controller; only the IDE controller actually tested
1.19	06-Nov-2004	xtraeme	Add support for: NVIDIA nForce3 250 IDE Controller NVIDIA nForce3 250 Serial ATA Controller From Kouichirou Hiratsuka in PR 27843, ok'ed christos@
1.18	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.17	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.16	19-Aug-2004	thorpej	- Add and use a CHAN_TO_WDC() macro to get the wdc_softc from an ata_channel. - Add and use a CHAN_TO_WDC_REGS() macro to get the wdc_regs from an ata_channel. - Add and use a CHAN_TO_PCIIDE() macro to get the pciide_softc from an ata_channel. - Add and use a CHAN_TO_PCHAN() macro to get the pciide_channel from an ata_channel. (This one just hides a cast, and is really just for consistency with the others.)
1.15	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.14	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.13	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.12	22-Jul-2004	drochner	branches: 1.12.2; Replace the hooribly overloaded use of pci_enumerate_bus() - to find out the chipset revision - by pci_find_device(). While the latter isn't better technically (works around the otherwise hierarchical device tree), using it doesn't require PCI configuration functions to support stuff which philosophically doesn't belong there. So we get the hands free to restructure things for better loadable driver support. (Actually, since this is about chipset internal IDE interfaces where the PCI device/function numbers are well known, this all is unnecessary.)
1.11	10-Mar-2004	bouyer	branches: 1.11.4; Enable ATA-133 for AMD8111. From Nicolas Joly in kern/24729. pciide_apollo_reg.h already contains the Ultra133 timming values for amd.
1.10	20-Feb-2004	fvdl	Actually match the nforce3 as well..
1.9	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.8	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.7	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.6	24-Oct-2003	mycroft	Remove IDE_PCI_CLASS_OVERRIDE -- it wasn't actually used anywhere. In lieu of it, add a class/subclass check to drivers that do not (appear) to have a unique ID for the IDE controller. This includes aceride, cypide and optiide.
1.5	23-Oct-2003	fvdl	* match the nForce3 controller * search the whole current PCI bus for the VIA bridge, can't rely on it to be function 0 of the same device
1.4	18-Oct-2003	enami	Factor out common code.
1.3	18-Oct-2003	enami	Whitespace nits.
1.2	11-Oct-2003	thorpej	Cosmetic changes: - Use static to unexport symbols that don't need to be exported. - ANSI'ify. - Some whitespace nits.
1.1	08-Oct-2003	bouyer	Split pciide in per-chip family driver, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0007.html We now have: acardide* at pci? dev ? function ? # Acard IDE controllers aceride* at pci? dev ? function ? # Acer Lab IDE controllers cmdide* at pci? dev ? function ? # CMD tech IDE controllers cypide* at pci? dev ? function ? # Cypress IDE controllers hptide* at pci? dev ? function ? # Triones/HighPoint IDE controllers optiide* at pci? dev ? function ? # Opti IDE controllers piixide* at pci? dev ? function ? # Intel IDE controllers pdcide* at pci? dev ? function ? # Promise IDE controllers siside* at pci? dev ? function ? # SiS IDE controllers slide* at pci? dev ? function ? # Symphony Labs IDE controllers viaide* at pci? dev ? function ? # VIA/AMD/Nvidia IDE controllers pciide* at pci? dev ? function ? flags 0x0000 # GENERIC pciide driver serverworks driver not commited yet; there are still copyright issues about it.
1.11.4.2	05-May-2005	riz	Pull up revision 1.22-1.23 (requested by abs in ticket #1153): Add 'VT642 SATA RAID Controller'. It looks like a viaide, it smells like a viaide... but you can only address the drives once you have put two or more of them into a RAID or JBOD.
1.11.4.1	07-Apr-2005	tron	Pull up revision 1.19 via patch (requested by hira in ticket #1103): Add support for: NVIDIA nForce3 250 IDE Controller NVIDIA nForce3 250 Serial ATA Controller From Kouichirou Hiratsuka in PR 27843, ok'ed christos@
1.12.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.12.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.12.2.8	04-Feb-2005	skrll	Sync with HEAD.
1.12.2.7	17-Jan-2005	skrll	Sync with HEAD.
1.12.2.6	14-Nov-2004	skrll	Sync with HEAD.
1.12.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.12.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.12.2.2	03-Aug-2004	skrll	Sync with HEAD
1.12.2.1	22-Jul-2004	skrll	file viaide.c was added on branch ktrace-lwp on 2004-08-03 10:49:12 +0000
1.20.4.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.20.4.1	12-Feb-2005	yamt	sync with head.
1.20.2.1	29-Apr-2005	kent	sync with -current
1.24.2.4	03-Aug-2006	tron	Pull up following revision(s) (requested by xtraeme in ticket #1446): sys/dev/pci/viaide.c: revision 1.32 Match nForce3 250 SATA2.
1.24.2.3	30-Apr-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #1295): sys/dev/pci/pcidevs: revision 1.785 sys/dev/pci/viaide.c: revision 1.29 PR port-amd64/33393: viaide driver patch for VT8237R SATA.
1.24.2.2	01-Feb-2006	tron	Pull up following revision(s) (requested by xtraeme in ticket #1149): sys/dev/pci/viaide.c: revision 1.28 The NVidia nForce430 IDE Controller supports Ultra-DMA Mode 6, enable it. Patch from MASUDA Hideo via tech-kern.
1.24.2.1	05-Jan-2006	riz	Pull up following revision(s) (requested by manu in ticket #1086): sys/dev/pci/pcidevs: revision 1.755 sys/dev/pci/viaide.c: revision 1.27 Add support for nForce430 ATA133 and SATA controllers. My disks now run at a decent speed.
1.25.2.9	24-Mar-2008	yamt	sync with head.
1.25.2.8	17-Mar-2008	yamt	sync with head.
1.25.2.7	21-Jan-2008	yamt	sync with head
1.25.2.6	15-Nov-2007	yamt	sync with head.
1.25.2.5	27-Oct-2007	yamt	sync with head.
1.25.2.4	03-Sep-2007	yamt	sync with head.
1.25.2.3	26-Feb-2007	yamt	sync with head.
1.25.2.2	30-Dec-2006	yamt	sync with head.
1.25.2.1	21-Jun-2006	yamt	sync with head.
1.26.2.2	01-Feb-2006	yamt	sync with head.
1.26.2.1	15-Jan-2006	yamt	sync with head.
1.28.10.1	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.28.8.1	11-May-2006	elad	sync with head
1.28.6.4	03-Sep-2006	yamt	sync with head.
1.28.6.3	11-Aug-2006	yamt	sync with head
1.28.6.2	26-Jun-2006	yamt	sync with head.
1.28.6.1	24-May-2006	yamt	sync with head.
1.28.4.1	01-Jun-2006	kardel	Sync with head.
1.28.2.1	09-Sep-2006	rpaulo	sync with head
1.29.4.1	13-Jul-2006	gdamore	Merge from HEAD.
1.29.2.1	19-Jun-2006	chap	Sync with head.
1.32.2.1	07-Sep-2006	riz	Pull up following revision(s) (requested by xtraeme in ticket #113): sys/dev/pci/viaide.c: revision 1.33 share/man/man4/viaide.4: revision 1.7 sys/dev/pci/pcidevs: revision 1.825 sys/dev/pci/pcidevs: revision 1.826 - Add all pciids for nvidia MCP61/65 controllers, from obsd. - Improve description in all them. Erm forgot to add MCP61/65 SATA Controllers and sort them. Add support for nvidia MCP61/65 IDE/SATA Controllers. Mention MCP61/65... and bump date.
1.33.4.2	10-Dec-2006	yamt	sync with head.
1.33.4.1	22-Oct-2006	yamt	sync with head
1.33.2.3	06-Feb-2007	ad	Quiten noisy boot messages.
1.33.2.2	01-Feb-2007	ad	Sync with head.
1.33.2.1	18-Nov-2006	ad	Sync with head.
1.37.2.4	11-Oct-2007	pavel	Pull up following revision(s) (requested by xtraeme in ticket #922): sys/dev/pci/viaide.c: revision 1.45 AMD CS5536 Companion Device is able to work in Ultra DMA Mode 5, patch from Andreas Wrede in PR kern/37059.
1.37.2.3	10-Jun-2007	bouyer	Pull up following revision(s) (requested by hira in ticket #713): sys/dev/pci/pcidevs: revision 1.863 sys/dev/pci/viaide.c: revision 1.38 share/man/man4/viaide.4: revision 1.8 Add PCI Ids for AMD Geode LX and CS5536. Add support for CS5536 IDE.
1.37.2.2	31-Mar-2007	bouyer	branches: 1.37.2.2.2; Pull up following revision(s) (requested by garbled in ticket #541): sys/dev/pci/viaide.c: revision 1.41 via patch The VT8237A SATA Controller uses chip_map_7, not 0, tested and confirmed working on my machine. Also when a via controller is set to RAID mode, it sets the pci_subclass to raid. Notice this and set ATAC_CAP_RAID.
1.37.2.1	17-Feb-2007	tron	Pull up following revision(s) (requested by mlelstv in ticket #449): sys/dev/pci/pcidevs: revision 1.867 sys/dev/pci/viaide.c: revision 1.40 Add support for VT8237A/VT82C586A PCI-ISA-Bridge Fixes PR kern/35314
1.37.2.2.2.2	29-Oct-2007	wrstuden	Catch up with 4.0 RC3
1.37.2.2.2.1	03-Sep-2007	wrstuden	Sync w/ NetBSD-4-RC_1
1.40.10.1	29-Mar-2007	reinoud	Pullup to -current
1.40.8.1	11-Jul-2007	mjf	Sync with head.
1.40.6.2	09-Oct-2007	ad	Sync with head.
1.40.6.1	10-Apr-2007	ad	Sync with head.
1.40.2.1	15-Apr-2007	yamt	sync with head.
1.41.10.3	23-Mar-2008	matt	sync with HEAD
1.41.10.2	09-Jan-2008	matt	sync with HEAD
1.41.10.1	06-Nov-2007	matt	sync with HEAD
1.41.8.3	14-Nov-2007	joerg	Sync with HEAD.
1.41.8.2	04-Oct-2007	joerg	Sync with HEAD.
1.41.8.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.41.4.1	03-Sep-2007	skrll	Sync with HEAD.
1.44.2.1	06-Oct-2007	yamt	sync with head.
1.45.4.3	18-Feb-2008	mjf	Sync with HEAD.
1.45.4.2	27-Dec-2007	mjf	Sync with HEAD.
1.45.4.1	19-Nov-2007	mjf	Sync with HEAD.
1.45.2.1	18-Nov-2007	bouyer	Sync with HEAD
1.47.6.2	19-Jan-2008	bouyer	Sync with HEAD
1.47.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.47.2.1	26-Dec-2007	ad	Sync with head.
1.51.6.5	17-Jan-2009	mjf	Sync with HEAD.
1.51.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.51.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.51.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.51.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.51.2.1	24-Mar-2008	keiichi	sync with head.
1.53.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.53.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.53.4.5	11-Aug-2010	yamt	sync with head.
1.53.4.4	11-Mar-2010	yamt	sync with head
1.53.4.3	16-Sep-2009	yamt	sync with head
1.53.4.2	04-May-2009	yamt	sync with head.
1.53.4.1	16-May-2008	yamt	sync with head.
1.53.2.2	17-Jun-2008	yamt	sync with head.
1.53.2.1	18-May-2008	yamt	sync with head.
1.54.2.1	18-Jun-2008	simonb	Sync with head.
1.56.2.1	19-Oct-2008	haad	Sync with HEAD.
1.57.10.1	21-Apr-2010	matt	sync to netbsd-5
1.57.4.4	20-Jan-2013	bouyer	Apply patch, requested by msaitoh in ticket #1842: sys/dev/pci/pcidevs 1.969 via patch sys/dev/pci/viaide.c 1.58 Add VT8237S Integrated SATA Controller support. PR#47452.
1.57.4.3	09-Jan-2010	snj	Pull up following revision(s) (requested by mrg in ticket #1227): sys/dev/pci/viaide.c: revision 1.66 enable mem space if it isn't already enabled. several bioses leave this not enabled, and viaide kind of depends upon it. fixes viaide(4) on several machines. idea from PR 38011, but reworked a little.
1.57.4.2	09-Jan-2010	snj	Pull up following revision(s) (requested by tsutsui in ticket #1210): sys/dev/pci/viaide.c: revision 1.65 Don't forget to initialize data32iot and data32ioh in via_sata_chip_map_new(). This makes 32bit UDMA xfer work properly on VIA VT6421.
1.57.4.1	26-Sep-2009	snj	Pull up following revision(s) (requested by jmcneill in ticket #1021): sys/dev/pci/viaide.c: revision 1.60 Add PCI_PRODUCT_VIATECH_VT8237A_SATA_2 to list of product IDs.
1.57.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.67.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.67.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.68.2.4	31-May-2011	rmind	sync with head
1.68.2.3	21-Apr-2011	rmind	sync with head
1.68.2.2	05-Mar-2011	rmind	sync with head
1.68.2.1	30-May-2010	rmind	sync with head
1.71.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.76.6.2	05-Apr-2012	mrg	sync to latest -current.
1.76.6.1	18-Feb-2012	mrg	merge to -current.
1.76.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.76.2.2	30-Oct-2012	yamt	sync with head
1.76.2.1	17-Apr-2012	yamt	sync with head
1.77.2.1	03-Apr-2012	riz	Pull up following revision(s) (requested by tsutsui in ticket #151): sys/dev/pci/viaide.c: revision 1.78 sys/dev/pci/pcidevs: revision 1.1106 Add VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. Regen from pcidevs rev 1.1106: Add VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. Add UDMA support for VIA VX900 media system processor IDE controller. PR/46166 from Hiroshi Tokuda. This should be pulled up to netbsd-6, since ZOTAC ZBOX nano VD01 http://www.zotacusa.com/zbox-nano-vd01.html exhibited at NetBSD booth in Open Source Conference 2012 Tokyo Spring was favored by many people.
1.83.4.1	18-May-2014	rmind	sync with head
1.83.2.3	03-Dec-2017	jdolecek	update from HEAD
1.83.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.83.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.84.18.2	27-Sep-2017	jdolecek	change wdc_init_shadow_regs() to accept only struct wdc_regs, it doesn't touch anything else factor out the probe-only struct ata_channel initialization to wdcprobe(), to reduce duplication of logic in individual drivers, and to actually work now that more init is needed beyond the memset()
1.84.18.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.86.4.1	10-Jun-2019	christos	Sync with HEAD
1.86.2.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.89.34.1	02-Aug-2025	perseant	Sync with HEAD
1.89.28.3	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1184): share/man/man4/viaide.4: revision 1.13 sys/dev/pci/viaide.c: revision 1.105 viaide(4): add UDMA support for VT8233C IDE controller. The VT8233C is another obscure VIA southbridge and appears to be a variant of the VT8233 rather than the VT8233A (which supports ATA133). The C variant was paired with a 3Com network controller instead of VIA's own NIC. Detect the ISA-PCI bridge and enable UDMA100 capability for improved IDE performance. Update the viaide(4) man page and refine the list of supported VIA controllers.
1.89.28.2	07-Jun-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1128): sys/dev/pci/viaide.c: revision 1.101 sys/dev/pci/viaide.c: revision 1.102 viaide(4): use via_chip_map() to attach VT8261 SATA controller in RAID mode. Similarly to VT8251 and CX700, VT8261 has two ports per channel, while via_sata_chip_map supports only one drive per channel. This caused disks on secondary ports (2, 4) to not be detected, which is addressed by this change. viaide(4): explicitly enable channels for VT6410 add-in cards. VT6410 and VT6415 have different reasons for "missing" enable bits. VT6410 uses the standard 0x40 offset for channel enable bits, but some add-in cards have both channels disabled by default. This requires explicitly writing the enable bits; otherwise, devices will not be detected. The same bug is present in Linux and OpenBSD, where the bits are also simulated. It was reported in Linux but closed as no-fix due to an unidentified root cause. VT6415 returns 0 at offsets 0x40-0x43 and likely uses a different location for its single channel enable bit, or it may not expose it at all. Until the correct offset is identified (if ever), the driver must simulate the enable bits.
1.89.28.1	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1101): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.89.2.3	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1978): share/man/man4/viaide.4: revision 1.13 sys/dev/pci/viaide.c: revision 1.105 viaide(4): add UDMA support for VT8233C IDE controller. The VT8233C is another obscure VIA southbridge and appears to be a variant of the VT8233 rather than the VT8233A (which supports ATA133). The C variant was paired with a 3Com network controller instead of VIA's own NIC. Detect the ISA-PCI bridge and enable UDMA100 capability for improved IDE performance. Update the viaide(4) man page and refine the list of supported VIA controllers.
1.89.2.2	07-Jun-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1959): sys/dev/pci/viaide.c: revision 1.101 sys/dev/pci/viaide.c: revision 1.102 viaide(4): use via_chip_map() to attach VT8261 SATA controller in RAID mode. Similarly to VT8251 and CX700, VT8261 has two ports per channel, while via_sata_chip_map supports only one drive per channel. This caused disks on secondary ports (2, 4) to not be detected, which is addressed by this change. viaide(4): explicitly enable channels for VT6410 add-in cards. VT6410 and VT6415 have different reasons for "missing" enable bits. VT6410 uses the standard 0x40 offset for channel enable bits, but some add-in cards have both channels disabled by default. This requires explicitly writing the enable bits; otherwise, devices will not be detected. The same bug is present in Linux and OpenBSD, where the bits are also simulated. It was reported in Linux but closed as no-fix due to an unidentified root cause. VT6415 returns 0 at offsets 0x40-0x43 and likely uses a different location for its single channel enable bit, or it may not expose it at all. Until the correct offset is identified (if ever), the driver must simulate the enable bits.
1.89.2.1	09-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1944): sys/dev/pci/viaide.c: revision 1.90 sys/dev/pci/ahcisata_pci.c: revision 1.72 sys/dev/pci/pciide_apollo_reg.h: revision 1.21 sys/dev/pci/viaide.c: revision 1.91 sys/dev/pci/viaide.c: revision 1.92 sys/dev/pci/viaide.c: revision 1.93 sys/dev/pci/viaide.c: revision 1.94 sys/dev/pci/viaide.c: revision 1.96 sys/dev/pci/viaide.c: revision 1.97 sys/dev/pci/viaide.c: revision 1.98 share/man/man4/viaide.4: revision 1.11 share/man/man4/viaide.4: revision 1.12 Use the chipset name to describe the IDE controller in order to avoid confusion. VT800->VX800 and VT855->VX855. Add VIA VT8261 southbridge SATA controller IDs and PCIB. viaide(4): use via_chip_map() instead of via_sata_chip_map_new() to attach CX700/VX800 IDE/SATA RAID controllers and set interface flags the same way as the VT6410_RAID case for native interrupts required by RAID mode. Device descriptions were moved to pciide_via_products and updated to cascade VT6410_RAID and CX700_IDE options. via_sata_chip_map_new previously worked more by "accident," as it did not return early on failure to map SATA registers and defaulted to the generic drive probe function. However, it failed to attach PATA drives (endless timeouts) and incorrectly detected "three" channels as VT6421, despite the controller having only two. Fixes PR kern/59010. Reviewed and approved by jakllsch@. Update CX700_IDE and CX700M2_IDE descriptions to better reflect their actual purpose. These are not separate IDE controllers for CX700 and CX700M2 but rather the IDE and RAID modes of the same controller in these chipsets (and few more). viaide(4): check and add ATA RAID capability in via_sata_chip_map_new() used by VT6241(A) RAID controller. This is required for ataraid(4) to attach on this controller if RAID was configured using its firwmare. The conditions may be redundant for this controller, but it is uncertain what variations exist in the wild. viaide(4): add IDE controller identification by the VT8237S ISA bridge. The controller's PATA interface uses PCI ID 0x0571, which is shared among many VIA southbridges. The ISA bridge is used to identify capabilities, including for the VT8237S. Unknown PATA controllers disable UDMA by default. This change ensures PATA drives attach at full speed. viaide(4): Add support for VIA VT8261 IDE/SATA integrated controller. The VT8261 southbridge, paired with the VN1000 chipset, is obscure and hard to find. Besides some evaluation boards surfacing after Centaurs demise, at least one rare production motherboard is known to exist. Despite its rarity, the VT8261 is very similar to the VT8251. It uses separate PCI IDs for PATA and SATA (IDE/RAID modes) with 4 SATA ports in total, 2 ports sharing one channel, and maps registers through BAR5. This similarity made adding basic support relatively simple (sharing same issues too). Tested briefly on VIA VT8591B eval board. viaide(4): Add RAID mode support for VIA VX900/VX11 SATA controllers. The BIOS option to enable RAID is uncommon on systems using these chipsets; however, at least two motherboards mention it in their manuals. The RAID firmware does not provide management features, but it can recognize and boot from preconfigured VIA RAID arrays. Tested on ECS VX900-I. Also add the VX11 chipset to the controller name, as it shares the same PCI ID. viaide(4): Improve VT8251 integrated SATA controller support in IDE/RAID modes. Use via_chip_map() to attach this controller in IDE and RAID modes, similarly to VX800. This allows drives to be identified and attached successfully on all four ports (two channels with master/slave ports). Switch channel setup to sata_setup_channel(), replacing via_setup_channel(). This avoids writes to the 0x50 register, which is only relevant for IDE controllers. Writes to SATA controller registers caused drives on the IDE controller (0x0571) to fail to attach. Apply this fix also to VX900 and VT8261. Rename VT8237R_SATA to VT8251_SATA (0x3349 is not used for VT8237R, AFAIK). This PCI ID is used in RAID mode on newer (CE) chipset revision and in all modes (IDE/RAID/AHCI) on older revision. Add the 0x5287 PCI ID, used in IDE mode on newer (CE) chipset revision. Rename VT8251_SATA to VT8251_AHCI (0x6287 is used only in AHCI mode). Add VT8251_AHCI to the ahcisata(4) quirk list, same as VT8251_SATA. For more details, see the tech-kern thread: https://mail-index.netbsd.org/tech-kern/2025/04/13/msg030365.html Should fix and close PR kern/37517. Tested on Asus A8V-VM (CD) and MSI MS-7318 (CE) motherboards. Reviewed by bad. viaide(4): check if chip enable register returns 0 before emulating enable bits. At least some VT6410 controllers have register exposed. In fact, some motherboards allow to control IDE channels (enable/disable them). viaide(4): update the list of supported VIA controllers. viaide(4): remove notes section about VT6421, it is not currently required to setup RAID/JBOD sets to access drives.
1.103.2.1	26-Oct-2025	martin	Pull up following revision(s) (requested by andvar in ticket #71): share/man/man4/viaide.4: revision 1.13 sys/dev/pci/viaide.c: revision 1.105 viaide(4): add UDMA support for VT8233C IDE controller. The VT8233C is another obscure VIA southbridge and appears to be a variant of the VT8233 rather than the VT8233A (which supports ATA133). The C variant was paired with a 3Com network controller instead of VIA's own NIC. Detect the ISA-PCI bridge and enable UDMA100 capability for improved IDE performance. Update the viaide(4) man page and refine the list of supported VIA controllers.
1.17	20-Jan-2007	xtraeme	Updated viaenv(4) driver: * Support for the VIA VT8231 Hardware monitor. * Power Management Timer available for timecounters in both VT86C686A and VT8231 (code simplified thanks to dev/ic/acpipmtimer). * Remove viapm(4) code and manpage (which was a link to viaenv.4 anyway). From OpenBSD, tested by some users.
1.16	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.15	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.14	11-Dec-2005	christos	branches: 1.14.20; 1.14.22; merge ktrace-lwp.
1.13	14-Sep-2004	drochner	branches: 1.13.12; fix botch in config_found conversion, found by Stefan Krueger
1.12	13-Sep-2004	drochner	a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.11	25-Jul-2004	christos	jason says that it is unclear if the new device is compatible with the old one so disable it for now.
1.10	25-Jul-2004	christos	Match viatech's power controller (Sergio Jimenez)
1.9	01-Jan-2003	thorpej	branches: 1.9.2; Use aprint_normal() in cfprint routines.
1.8	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.7	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.6	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.5	27-Sep-2002	thorpej	Introduce a new routine, config_match(), which invokes the cfattach->ca_match function in behalf of the caller. Use it rather than invoking cfattach->ca_match directly.
1.4	25-Mar-2002	thorpej	There is no need to use a kthread to read the VIA environment sensors.
1.3	13-Nov-2001	lukem	add RCSID
1.2	12-May-2000	thorpej	branches: 1.2.6; 1.2.8; 1.2.10; Apply indent(1).
1.1	08-May-2000	joda	VIA VT82C686A hardware monitor
1.2.10.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.2.10.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.2.10.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.2.8.4	03-Jan-2003	thorpej	Sync with HEAD.
1.2.8.3	18-Oct-2002	nathanw	Catch up to -current.
1.2.8.2	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.2.8.1	14-Nov-2001	nathanw	Catch up to -current.
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	12-May-2000	bouyer	file viapm.c was added on branch thorpej_scsipi on 2000-11-20 11:42:39 +0000
1.9.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.9.2.1	03-Aug-2004	skrll	Sync with HEAD
1.13.12.2	26-Feb-2007	yamt	sync with head.
1.13.12.1	30-Dec-2006	yamt	sync with head.
1.14.22.2	10-Dec-2006	yamt	sync with head.
1.14.22.1	22-Oct-2006	yamt	sync with head
1.14.20.2	01-Feb-2007	ad	Sync with head.
1.14.20.1	18-Nov-2006	ad	Sync with head.
1.5	20-Jan-2007	xtraeme	Updated viaenv(4) driver: * Support for the VIA VT8231 Hardware monitor. * Power Management Timer available for timecounters in both VT86C686A and VT8231 (code simplified thanks to dev/ic/acpipmtimer). * Remove viapm(4) code and manpage (which was a link to viaenv.4 anyway). From OpenBSD, tested by some users.
1.4	11-Dec-2005	christos	branches: 1.4.20; merge ktrace-lwp.
1.3	27-Feb-2005	perry	branches: 1.3.4; nuke trailing whitespace
1.2	12-May-2000	thorpej	branches: 1.2.6; 1.2.30; 1.2.38; 1.2.40; Apply indent(1).
1.1	08-May-2000	joda	VIA VT82C686A hardware monitor
1.2.40.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.2.38.1	29-Apr-2005	kent	sync with -current
1.2.30.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.2.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.6.1	12-May-2000	bouyer	file viapmvar.h was added on branch thorpej_scsipi on 2000-11-20 11:42:39 +0000
1.3.4.1	26-Feb-2007	yamt	sync with head.
1.4.20.1	01-Feb-2007	ad	Sync with head.
1.12	22-Apr-2025	ozaki-r	vio9p: export tag via sysctl With this change, we can see tags of virtio-9p devices via sysctl: # sysctl hw.vio9p hw.vio9p.vio9p0.tag = test hw.vio9p.vio9p1.tag = test2 PR kern/59165
1.11	23-Mar-2023	yamaguchi	branches: 1.11.6; Added functions to set interrupt handler and index into virtqueue
1.10	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.9	20-Apr-2022	uwe	branches: 1.9.4; vio9p: avoid null pointer dereference on attach. Pass "vqs" argument to virtio_child_attach_start(). The virtio API is not quite clear to me here - there's also a separate setter virtio_child_attach_set_vqs that performs some additional actions. Queue setup for Virtio 1.0 accesses virtio_softc::sc_vqs and we didn't set it. QEMU exposes only the old virtio interface, so it doesn't trigger the problem.
1.8	13-Apr-2022	uwe	vio9p: Bump max tag len to match QEMU. QEMU defines it to 32 but includes the final zero byte into the limit. Our code counts the final zero byte separately, so adjust this define appropriately.
1.7	13-Apr-2022	uwe	vio9p: Properly negotiate MOUNT_TAG feature. I _think_ that this is what is intended here. The code used to actually test for the presence of any random device independent feature bits (the test would happen to succeed b/c of INDIRECT_DESC).
1.6	13-Apr-2022	uwe	vio9p: Add definition for the MOUNT_TAG feature bit. From QEMU sources.
1.5	31-Mar-2022	pgoyette	For device modules that provide both auto-config and /dev/xxx interfaces, make sure that initialization and destruction follow the proper sequence. This is triggered by the recent changes to the devsw stuff; per riastradh@ the required call sequence is: devsw_attach() config_init_component() or config_cf_attach() ... config_fini_component() or config_cf_detach() devsw_detach() While here, add a few missing calls to some of the detach routines. Testing of these changes has been limited to: 1. compile without build break 2. no related test failures from atf 3. modload/modunload work as well as before. No functional device testing done, since I don't have any of these devices. Let me know of any damage I might cause here! XXX Some of the modules affected by this commit are already XXX broken; see kern/56772. This commit does not break any additional modules (as far as I know).
1.4	26-Sep-2021	thorpej	Change the kqueue filterops::f_isfd field to filterops::f_flags, and define a flag FILTEROP_ISFD that has the meaning of the prior f_isfd. Field and flag name aligned with OpenBSD. This does not constitute a functional or ABI change, as the field location and size, and the value placed in that field, are the same as the previous code, but we're bumping __NetBSD_Version__ so 3rd-party module source code can adapt, as needed. NetBSD 9.99.89
1.3	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.2	18-Dec-2020	thorpej	Use sel{record,remove}_knote().
1.1	28-Oct-2019	ozaki-r	branches: 1.1.8; 1.1.10; Implement a front-end driver of virtio-9p called vio9p In conjunction with mount_9p, it enables a NetBSD system running as a VM guest to mount an exported filesystem by the host via virtio-9p. It exports a 9p end-point of virtio-9p via a character device file for mount_9p. Reviewed by yamaguchi@
1.1.10.2	03-Apr-2021	thorpej	Sync with HEAD.
1.1.10.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	28-Oct-2019	martin	file vio9p.c was added on branch phil-wifi on 2020-04-13 08:04:45 +0000
1.9.4.2	15-May-2025	martin	Pull up following revision(s) (requested by ozaki-r in ticket #1115): sys/dev/pci/vio9p.c: revision 1.12 vio9p: export tag via sysctl With this change, we can see tags of virtio-9p devices via sysctl: # sysctl hw.vio9p hw.vio9p.vio9p0.tag = test hw.vio9p.vio9p1.tag = test2 PR kern/59165
1.9.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.11.6.1	02-Aug-2025	perseant	Sync with HEAD
1.2	04-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "is_console".
1.1	26-Jul-2025	martin	branches: 1.1.4; PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.1.4.2	02-Aug-2025	perseant	Sync with HEAD
1.1.4.1	26-Jul-2025	perseant	file viogpu.c was added on branch perseant-exfatfs on 2025-08-02 05:57:00 +0000
1.1	26-Jul-2025	martin	branches: 1.1.4; PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.1.4.2	02-Aug-2025	perseant	Sync with HEAD
1.1.4.1	26-Jul-2025	perseant	file viogpu.h was added on branch perseant-exfatfs on 2025-08-02 05:57:00 +0000
1.17	25-Mar-2023	mlelstv	Mark as MPSAFE.
1.16	25-Mar-2023	mlelstv	Fix 32bit overflow when calculating balloon size. Don't make uvm_pglistalloc wait, it will wait forever. Instead rely on the retries by this driver. This also allows to cancel the request. Increase inflate speed by factor 10 (same as deflate).
1.15	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.14	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.13	13-Apr-2022	uwe	branches: 1.13.4; virtio: use the new syntax for snprintb(3) format strings. The old syntax is limited to 32 bits only (and has 1-based bit numbers which is rather incovenient too).
1.12	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.11	13-Jan-2021	reinoud	Fix for virtios viomb memory balloon driver. The inflate_done() and deflate_done() issued a wrong extent to bus_dmamap_sync() giving rise to panics on aarch64s mmio backend.
1.10	10-Dec-2019	ad	branches: 1.10.8; p->phys_addr -> VM_PAGE_TO_PHYS(p)
1.9	10-Jun-2018	jakllsch	branches: 1.9.2; remove irrelevant pci(9) #includes from virtio child drivers
1.8	25-Mar-2017	jdolecek	branches: 1.8.12; reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.7	27-Sep-2016	pgoyette	branches: 1.7.2; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.6	07-Jul-2016	msaitoh	branches: 1.6.2; KNF. Remove extra spaces. No functional change.
1.5	27-Oct-2015	christos	Print the negotiated feature bits
1.4	04-May-2015	ozaki-r	Add NULL check for TAILQ_FIRST
1.3	22-Jul-2014	ozaki-r	branches: 1.3.4; Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio It is set by a child driver, e.g., if_vioif. If set, virtio sets PCI_INTR_MPSAFE for pci_intr_establish.
1.2	17-Oct-2013	christos	branches: 1.2.2; - remove unused variables - move debugging code inside debugging sections
1.1	30-Oct-2011	hannken	branches: 1.1.2; 1.1.8; 1.1.14; 1.1.18; Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.1.18.1	18-May-2014	rmind	sync with head
1.1.14.2	03-Dec-2017	jdolecek	update from HEAD
1.1.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.8.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.1.8.1	30-Oct-2011	riz	file viomb.c was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.2.1	10-Aug-2014	tls	Rebase.
1.3.4.5	28-Aug-2017	skrll	Sync with HEAD
1.3.4.4	05-Oct-2016	skrll	Sync with HEAD
1.3.4.3	09-Jul-2016	skrll	Sync with HEAD
1.3.4.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.3.4.1	06-Jun-2015	skrll	Sync with HEAD
1.6.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.6.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.7.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.8.12.1	25-Jun-2018	pgoyette	Sync with HEAD
1.9.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.10.8.1	03-Apr-2021	thorpej	Sync with HEAD.
1.13.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.23	04-Nov-2025	nia	KNF
1.22	04-Aug-2023	riastradh	entropy(9): Simplify stages. Split interrupt vs non-interrupt paths. - Nix the entropy stage (cold, warm, hot). Just use the usual kernel `cold' (cold: single-core, single-thread; interrupts may happen), and don't make any three-way distinction about whether interrupts or threads or other CPUs can be running. Instead, while cold, use splhigh/splx or forbid paths to come from interrupt context, and while warm, use mutex or the per-CPU hard and soft interrupt paths for low latency. This comes at a small cost to some interrupt latency, since we may stir the pool in interrupt context -- but only for a very short window early at boot between configure and configure2, so it's hard to imagine it matters much. - Allow rnd_add_uint32 to run in hard interrupt context or with spin locks held, but defer processing to softint and drop samples on the floor if buffer is full. This is mainly used for cheaply tossing samples from drivers for non-HWRNG devices into the entropy pool, so it is often used from interrupt context and/or under spin locks. - New rnd_add_data_intr provides the interrupt-like data entry path for arbitrary buffers and driver-specified entropy estimates: defer processing to softint and drop samples on the floor if buffer is full. - Document that rnd_add_data is forbidden under spin locks outside interrupt context (will crash in LOCKDEBUG), and inadvisable in interrupt context (but technically permitted just in case there are compatibility issues for now); later we can forbid it altogether in interrupt context or under spin locks. - Audit all uses of rnd_add_data to use rnd_add_data_intr where it might be used in interrupt context or under a spin lock. This fixes a regression from last year when the global entropy lock was changed from IPL_VM (spin) to IPL_SOFTSERIAL (adaptive). Thought I'd caught all the problems from that, but another one bit three different people this week, presumably because of recent changes that led to more non-HWRNG drivers entering the entropy consolidation path from rnd_add_uint32. In my attempt to preserve the rnd(9) API for the (now long-since abandoned) prospect of pullup to netbsd-9 in my rewrite of the entropy subsystem in 2020, I didn't introduce a separate entry point for entering entropy from interrupt context or equivalent, i.e., spin locks held, and instead made rnd_add_data rely on cpu_intr_p() to decide whether to process the whole sample under a lock or only take as much as there's buffer space for before scheduling a softint. In retrospect, that was a mistake (though perhaps not as much of a mistake as other entropy API decisions...), a mistake which is finally getting rectified now by rnd_add_data_intr. XXX pullup-10
1.21	25-Mar-2023	mlelstv	Mark as MPSAFE.
1.20	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.19	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.18	14-Apr-2022	riastradh	branches: 1.18.4; Revert "viornd(4): Process host entropy in softint context." Apparently this has the effect of sometimes making the network hang on Google Compute Engine as used by syzbot, which has held up all the syzkaller testing for weeks now. Let's revert this for now, and separately try to figure out what's wrong with it.
1.17	23-Mar-2022	riastradh	viornd(4): Process host entropy in softint context. Samples added to the entropy pool in hard interrupt context are only buffered, never processed directly, and if they fill the buffer, the sample is dropped -- this serves to encourage taking timing samples in hard interrupt context because it's cheap, and we have no idea how many samples we really need for full entropy so it's safer to err on the side of `as many as we can get'. But for viornd(4), we assume the host has full entropy so we only need a single 32-byte sample, and we want to avoid dropping it so we get full entropy ASAP. Entering the sample in a soft interrupt rather than hard interrupt achieves this.
1.16	19-Mar-2022	riastradh	viornd(4): Revert IPL change for lock. This lock is taken in hard interrupt context, so it needs to remain at IPL_VM.
1.15	19-Mar-2022	riastradh	rnd(9): Adjust IPL of locks used by rndsource callbacks. These no longer ever run from hard interrupt context or with a spin lock held, so there is no longer any need to have them at IPL_VM to block hard interrupts. Instead, lower them to IPL_SOFTSERIAL.
1.14	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.13	30-Apr-2020	riastradh	branches: 1.13.2; rnd_attach_source calls the callback itself now. No need for every driver to explicitly call it to prime the pool. Eliminate now-unused <sys/rndpool.h>.
1.12	10-Jun-2018	jakllsch	remove irrelevant pci(9) #includes from virtio child drivers
1.11	25-Mar-2017	jdolecek	branches: 1.11.12; 1.11.14; reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.10	30-Nov-2016	christos	branches: 1.10.2; Don't call virtio_enqueue_abort when virtio_enqueue_reserve fails. Pointed out by uwe@
1.9	27-Oct-2015	christos	branches: 1.9.2; Print the negotiated feature bits
1.8	05-May-2015	ozaki-r	Use NULL for initialization of sc_config_change
1.7	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.6	08-Apr-2015	riastradh	Don't forget to bus_dmamap_unload.
1.5	06-Nov-2014	pooka	branches: 1.5.2; put the "asked for" print also behind VIORNG_DEBUG
1.4	06-Nov-2014	pooka	Abort only actually initialized slots.
1.3	06-Nov-2014	pooka	disable debug print by default
1.2	06-Nov-2014	pooka	Print attachment \n's so that we don't get junky lines (this driver now matches ld)
1.1	26-Oct-2014	tls	branches: 1.1.2; Add viornd(4), a driver for the VirtIO entropy source available on QEMU, KVM, and Google Compute Engine. From OpenBSD.
1.1.2.3	06-Nov-2015	riz	Pull up following revision(s) (requested by riastradh in ticket #995): sys/dev/pci/viornd.c: revision 1.3 disable debug print by default
1.1.2.2	02-Nov-2014	martin	Pull up following revision(s) (requested by tls in ticket #174): doc/CHANGES: revision 1.2005 sys/dev/pci/viornd.c: revision 1.1 sys/dev/pci/files.pci: revision 1.372 sys/arch/amd64/conf/GENERIC: revision 1.401 share/man/man4/viornd.4: revision 1.1 Add viornd(4), a driver for the VirtIO entropy source available on QEMU, KVM, and Google Compute Engine. From OpenBSD.
1.1.2.1	26-Oct-2014	martin	file viornd.c was added on branch netbsd-7 on 2014-11-02 10:09:44 +0000
1.5.2.4	28-Aug-2017	skrll	Sync with HEAD
1.5.2.3	05-Dec-2016	skrll	Sync with HEAD
1.5.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.5.2.1	06-Jun-2015	skrll	Sync with HEAD
1.9.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.9.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.10.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.11.14.1	25-Jun-2018	pgoyette	Sync with HEAD
1.11.12.2	03-Dec-2017	jdolecek	update from HEAD
1.11.12.1	25-Mar-2017	jdolecek	file viornd.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.13.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.18.4.2	11-Aug-2023	martin	Pull up following revision(s) (requested by riastradh in ticket #319): sys/dev/pci/ubsec.c: revision 1.64 sys/dev/pci/hifn7751.c: revision 1.82 lib/libc/gen/getentropy.3: revision 1.5 lib/libc/gen/getentropy.3: revision 1.6 share/man/man4/rnd.4: revision 1.41 lib/libc/sys/getrandom.2: revision 1.2 lib/libc/sys/getrandom.2: revision 1.3 share/man/man5/rc.conf.5: revision 1.193 share/man/man7/entropy.7: revision 1.5 share/man/man7/entropy.7: revision 1.6 share/man/man7/entropy.7: revision 1.7 share/man/man7/entropy.7: revision 1.8 etc/security: revision 1.130 share/man/man7/entropy.7: revision 1.9 etc/security: revision 1.131 sys/crypto/cprng_fast/cprng_fast.c: revision 1.19 sys/sys/rndio.h: revision 1.3 tests/lib/libc/sys/t_getrandom.c: revision 1.5 etc/defaults/rc.conf: revision 1.164 etc/defaults/rc.conf: revision 1.165 sys/sys/rndsource.h: revision 1.10 sys/kern/kern_entropy.c: revision 1.62 sys/kern/kern_entropy.c: revision 1.63 sys/kern/kern_entropy.c: revision 1.64 sys/kern/subr_cprng.c: revision 1.44 sys/kern/kern_entropy.c: revision 1.65 sys/kern/kern_clock.c: revision 1.149 sys/dev/pci/viornd.c: revision 1.22 share/man/man9/rnd.9: revision 1.32 sys/kern/subr_prf.c: revision 1.202 sys/sys/rndsource.h: revision 1.8 sys/sys/rndsource.h: revision 1.9 share/man/man7/entropy.7: revision 1.10 1. Reinstate netbsd<=9 entropy estimator to unblock /dev/random, in parallel with assessment of only confident entropy sources (seed, HWRNG) for security warnings like sshd keys in motd and daily insecurity report. 2. Make multiuser boot wait for first /dev/random output soon after loading a seed and configuring rndctl, so that getentropy(3) meets its contract starting early at boot without introducing blocking paths that could cause hangs in init(8) or single-user mode. Operators can choose to disable this wait in rc.conf. 3. Fix some bugs left over from reducing the global entropy lock from a spin lock at IPL_VM to an adaptive lock at IPL_SOFTSERIAL. 4. Update man pages.
1.18.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.36	25-Mar-2023	mlelstv	Mark as MPSAFE.
1.35	25-Mar-2023	mlelstv	A single SCSI request may require multiple slots in the virtio queue but the queue isn't sized for the theoretical maximum. So just use XS_BUSY to pace the scsipi layer.
1.34	25-Mar-2023	mlelstv	restore fetch of qsize.
1.33	24-Mar-2023	yamaguchi	Added check of pointer for allocated memory before release of resource
1.32	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.31	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.30	11-Oct-2022	andvar	branches: 1.30.2; fix typos in log messages s/bus_dmamem_create/bus_dmamap_create/ and s/bus_dmamem_load/bus_dmamap_load/. Inspired by recent similar fixes in OpenBSD.
1.29	27-Jan-2022	jakllsch	We no need for PQUIRK_FORCELUNS now that we try REPORT LUNS.
1.28	12-Nov-2021	skrll	Trailing whitespace
1.27	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.26	24-Apr-2021	thorpej	branches: 1.26.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.25	20-Jan-2021	reinoud	branches: 1.25.2; Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.24	19-Sep-2020	kim	branches: 1.24.2; Update the Proxmox comment with link to their bugzilla
1.23	18-Sep-2020	jakllsch	Probe all LUNs on vioscsi(4) until scsi(4) can do REPORT LUNS.
1.22	12-Jul-2020	kim	Fix off-by-one SCSI target reporting Use max_target as the controller ID instead of zero, so that the device located at SCSI ID zero (e.g. a disk) is not obscured through not being probed by scsi_probe_bus() (which skips the controller ID). Copy the target requested onto the wire without decrementing it by one. ok christos@
1.21	13-Apr-2019	maxv	branches: 1.21.4; Fix use-after-free. If we're not polling, virtio_enqueue_commit() will send the transaction, and it means 'xs' can be immediately freed. So, save the value of xs_control beforehand. Detected by KASAN, ok jdolecek@. Fixes PR/54008 Reported-by: syzbot+6513c4afe66237d7207f@syzkaller.appspotmail.com
1.20	10-Jun-2018	jakllsch	branches: 1.20.2; remove irrelevant pci(9) #includes from virtio child drivers
1.19	15-May-2017	jdolecek	branches: 1.19.2; 1.19.8; 1.19.10; free the virtio slot _before_ calling scsipi_done(), so that driver is able to accept another request immediatelly and doesn't trigger spurious XS_RESOURCE_SHORTAGE part of fix for PR kern/52199
1.18	13-May-2017	jdolecek	refactor error handling in vioscsi_scsipi_request() to avoid the goto maze, and add missing virtio_enqueue_abort() call when bus_dmamap_load() fails; abort is only done implicitely when virtio_enqueue_reserve() fails, must do it explicitely if there is other reason also now always print error when bus_dmamap_load() or virtio_enqueue_reserve() fail - the former shouldn't fail due to BUS_DMA_ALLOCNOW, and the latter shouldn't ever fail now after the maxnsegs fix for virtio_alloc_vq(), so error there means driver bug
1.17	13-May-2017	jdolecek	fix number of maximum segments in the queue allocation so that MAXPHYS-sized tranfer fits inside the indirect descriptor; avoids the !indirect path in virtio_enqueue_reserve() for these transfers, which runs out of the vq_entries and triggers XS_RESOURCE_SHORTAGE for bigger number of outstanding tranfers fixes PR kern/52199 by MB
1.16	25-Mar-2017	christos	branches: 1.16.4; fix printf format.
1.15	25-Mar-2017	jdolecek	add support for loading vioscsi driver dynamically
1.14	25-Mar-2017	jdolecek	cap also adapt_nchannels to 1, we only attach one
1.13	25-Mar-2017	jdolecek	cap adapt_openings to minimum of cmd_per_lun and qsize, so that we don't overpromise to the upper layer cap ntargets and nluns to 16 and 1024 as avoid spending excessive time probing for nonexistent devices; e.g. qemu defaults to 255/16383 causing the probe to take about a minute
1.12	25-Mar-2017	jdolecek	reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.11	13-Mar-2017	jdolecek	several small tweaks: - use MSIX if available - fix off-by-one for nluns/ntargets (code inspection, not actual effect) - when virtio_enqueue_reserve() fails, do not execute vioscsi_req_put() and hence virtio_dequeue_commit() as there is nothing to commit, just unload the xs and return; also report XS_RESOURCE_SHORTAGE, it's semi-normal situation - set status/resid in vioscsi_req_done() before the switch, so that the override for VIRTIO_SCSI_S_BAD_TARGET actually has effect - g/c vioscsi_req_put(), call the appropriate cleanup routines directly - stop initializing id/task_attr in vioscsi_req_get(), it's overwritten in vioscsi_scsipi_request() anyway
1.10	13-Mar-2017	jdolecek	replace req queue offset with define, only set & start the req interrupt for that queue; add some masks for the lun for visibility; use slot as tag number no actual functional change, just small cleanup
1.9	07-Mar-2017	jdolecek	allocate bus dma maps during attachment, rather than creating and destroying them for each request; besides being faster, bus_dmamap_destroy() is not safe to be called from interrupt context adresses PR kern/52034 by Benny Siegert
1.8	04-Oct-2016	jdolecek	branches: 1.8.2; 1.8.4; 1.8.6; react on ADAPTER_REQ_SET_XFER_MODE so that we set tagged queuing pass tag type and set id
1.7	04-Oct-2016	jdolecek	vioscsi_req_get()/virtio_enqueue_prep() failing is actually perfectly normal - observed failures included 10, 27, 61 in-flight commands, so probably depends on particular command mix; return with XS_RESOURCE_SHORTAGE rather then panic do vioscsi_req_put() when initial bus_dmamap_load() fails, as suggested by the XXX; the vq_done hook is called by virtio, but in that case we never get to commit the request to it
1.6	01-Nov-2015	pooka	branches: 1.6.2; 1.6.4; Use SCSIPI_CHAN_NOSETTLE.
1.5	31-Oct-2015	christos	fix the status return.
1.4	30-Oct-2015	christos	Fake mode sense data for illegal targets. qemu reports 256 targets...
1.3	30-Oct-2015	christos	kill some more debugging.
1.2	30-Oct-2015	christos	this is working now, remove debugging.
1.1	29-Oct-2015	christos	Add vioscsi, compile tested only (toxic)
1.6.4.3	26-Apr-2017	pgoyette	Sync with HEAD
1.6.4.2	20-Mar-2017	pgoyette	Sync with HEAD
1.6.4.1	04-Nov-2016	pgoyette	Sync with HEAD
1.6.2.4	28-Aug-2017	skrll	Sync with HEAD
1.6.2.3	05-Dec-2016	skrll	Sync with HEAD
1.6.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.6.2.1	01-Nov-2015	skrll	file vioscsi.c was added on branch nick-nhusb on 2015-12-27 12:09:57 +0000
1.8.6.2	18-Jan-2017	skrll	Sync with netbsd-5
1.8.6.1	04-Oct-2016	skrll	file vioscsi.c was added on branch netbsd-7-nhusb on 2017-01-18 08:46:42 +0000
1.8.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.8.2.2	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1338): sys/arch/amd64/conf/GENERIC: revision 1.418 sys/dev/pci/files.pci: revision 1.379 sys/dev/pci/vioscsi.c: revisions 1.1-1.8 sys/dev/pci/vioscsireg.h: revision 1.1 sys/dev/pci/virtioreg.h: revisions 1.5, 1.6 add the common flag bits -- Add vioscsi, compile tested only (toxic) -- this is working now, remove debugging. -- add vioscsi -- kill some more debugging. -- Fake mode sense data for illegal targets. qemu reports 256 targets... -- fix the status return. -- Use SCSIPI_CHAN_NOSETTLE. -- vioscsi_req_get()/virtio_enqueue_prep() failing is actually perfectly normal - observed failures included 10, 27, 61 in-flight commands, so probably depends on particular command mix; return with XS_RESOURCE_SHORTAGE rather then panic do vioscsi_req_put() when initial bus_dmamap_load() fails, as suggested by the XXX; the vq_done hook is called by virtio, but in that case we never get to commit the request to it -- react on ADAPTER_REQ_SET_XFER_MODE so that we set tagged queuing pass tag type and set id
1.8.2.1	04-Oct-2016	snj	file vioscsi.c was added on branch netbsd-7 on 2016-12-23 06:33:56 +0000
1.16.4.1	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.19.10.1	25-Jun-2018	pgoyette	Sync with HEAD
1.19.8.2	03-Dec-2017	jdolecek	update from HEAD
1.19.8.1	15-May-2017	jdolecek	file vioscsi.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.19.2.2	13-Jul-2020	martin	Pull up following revision(s) (requested by kim in ticket #1572): sys/dev/pci/vioscsi.c: revision 1.22 Fix off-by-one SCSI target reporting Use max_target as the controller ID instead of zero, so that the device located at SCSI ID zero (e.g. a disk) is not obscured through not being probed by scsi_probe_bus() (which skips the controller ID). Copy the target requested onto the wire without decrementing it by one. ok christos@
1.19.2.1	15-Jul-2019	martin	Pull up following revision(s) (requested by maxv in ticket #1289): sys/dev/pci/vioscsi.c: revision 1.21 Fix use-after-free. If we're not polling, virtio_enqueue_commit() will send the transaction, and it means 'xs' can be immediately freed. So, save the value of xs_control beforehand. Detected by KASAN, ok jdolecek@. Fixes PR/54008
1.20.2.1	10-Jun-2019	christos	Sync with HEAD
1.21.4.1	13-Jul-2020	martin	Pull up following revision(s) (requested by kim in ticket #1001): sys/dev/pci/vioscsi.c: revision 1.22 Fix off-by-one SCSI target reporting Use max_target as the controller ID instead of zero, so that the device located at SCSI ID zero (e.g. a disk) is not obscured through not being probed by scsi_probe_bus() (which skips the controller ID). Copy the target requested onto the wire without decrementing it by one. ok christos@
1.24.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.25.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.26.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.30.2.2	02-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #914): sys/dev/pci/virtio_pci.c: revision 1.50 sys/dev/pci/virtio_pci.c: revision 1.51 sys/dev/pci/virtio_pci.c: revision 1.52 sys/dev/pci/virtio_pci.c: revision 1.53 sys/dev/pci/virtio_pci.c: revision 1.54 sys/dev/pci/vioscsi.c: revision 1.35 sys/dev/pci/ld_virtio.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.34 sys/dev/virtio/viocon.c: revision 1.9 sys/dev/pci/ld_virtio.c: revision 1.35 sys/dev/pci/virtio.c: revision 1.80 sys/dev/pci/virtio.c: revision 1.81 sys/dev/pci/virtio.c: revision 1.82 sys/dev/pci/virtio_pci.c: revision 1.45 sys/dev/pci/virtio_pci.c: revision 1.46 sys/dev/pci/virtio_pci.c: revision 1.47 sys/dev/pci/virtio_pci.c: revision 1.48 sys/dev/pci/virtio_pci.c: revision 1.49 sys/dev/pci/virtio.c: revision 1.79 A single SCSI request may require multiple slots in the virtio queue but the queue isn't sized for the theoretical maximum. So just use XS_BUSY to pace the scsipi layer. viocon(4): Show error in assertion failure. virtio_read_device_config_le_[24]: Fix for {aarch64,arm}eb Stop byte-swapping for big-endian aarch64 and arm to fix corrupted read for, e.g., sc_taglen for vio9p(4). As described as comments in virtio_pci.c, big-endian aarch64 and armv7 (armeb) are somewhat special. At the moment, all supported virtual/real machines are configured as little-endian, and only CPU cores are switched to big-endian mode during early boot stage. Most peripheral buses for armeb are LSB as a result, and their default bus_space(9) functions swap byte-order. Therefore, PIOed data from memory-mapped devices, as well as pci(4) ones, are actually LSB, but observed as MSB for armeb. Therefore, we should not swap byte-order further in virtio_read_device_config_le_[24] on armeb. See virtio_pci.c and sys/dev/fdt/virtio_mmio_fdt.c also. fix spelling mistakes, mainly in comments and log messages. Split KASSERT(A && B) into KASSERT(A); KASSERT(B). Fix typo in error message. Modify a confused expression in ld_virtio_attach(). VIRTIO_BLK_MIN_SEGMENTS should be the total number of non-data segments, so I rename it to VIRTIO_BLK_CTRL_SEGMENTS. PR kern/57981. ld@virtio(4): Fix maximum size parameters. - SEG_MAX is the maximum number of segments. - SIZE_MAX is the maximum number of bytes in a single segment. The maximum transfer size is, therefore, SEG_MAX * SIZE_MAX. => Don't add two extra segments in the dmamap vr_payload for the header and status -- we already have a separate dmamap vr_cmdsts for that. => Don't recalculate payload dmamap parameters based on division by NBPG, just use the ones specified by the host. => Allow SIZE_MAX below MAXPHYS as long as SIZE_MAX*SEG_MAX >= MAXPHYS. Even though ldattach clamps ld->sc_maxxfer to MAXPHYS, make sure to clamp it in ld_virtio_attach before ld_virtio_alloc_reqs since it determines the dmamap sizes and bounce buffer allocation and there's no sense in allocating those larger than ld will use anyway. PR kern/58338 virtio_pci.c: Nix trailing whitespace. virtio@pci: Unmap the BARs we actually mapped on >=1.0 detach. PR kern/58335 virtio@pci: Nix useless sc_mapped_iosize. This appears to have been a mistaken attempt to pass the subregion length through to bus_space_unmap, which doesn't make any sense -- bus_space_subregion does not allocate anything and the resulting handle must not be passed to bus_space_unmap. Once we pass the size that was actually returned by pci_mapreg_map to bus_space_unmap, which matches the size passed internally by pci_mapreg_map to bus_space_map, sc_mapped_iosize is no longer used and can be flushed. Prompted by PR kern/58335. virtio@pci: Don't detach on shutdown. The virtio bus does not keep state in memory that needs to flushed to persistent storage. Typically DVF_DETACH_SHUTDOWN is only needed by, e.g., disk drivers to flush disk caches (although really, vfs sync should do that anyway, so maybe we should just nix the option altogether). Prompted by PR kern/58335. virtio@pci: Zero sc_bars_iosize[i] on unmap. No functional change intended, but this makes the code less brittle. Prompted by PR kern/58335. virtio_pci.c: Sort includes. No functional change intended. virtio_pci.c: Omit needless blank lines and commented code. No functional change intended. virtio_pci.c: Sprinkle KNF. - No parameter names in function prototypes. - Fix indentation. - Break overlong lines. - Blank line between declarations and rest of block. virtio_pci.c: Sprinkle const. No functional change intended. virtio_pci.c: Use container_of, not cast. virtio(4): Assert child attach finished before we allow queueing. Should help catch mistakes like: PR kern/58555: Kernel panic during boot when using viocon with PCIe
1.30.2.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.2	26-Aug-2021	andvar	Add missing RCSID in header files and one __KERNEL_RCSID().
1.1	29-Oct-2015	christos	branches: 1.1.2; 1.1.6; 1.1.10; 1.1.22; Add vioscsi, compile tested only (toxic)
1.1.22.2	03-Dec-2017	jdolecek	update from HEAD
1.1.22.1	29-Oct-2015	jdolecek	file vioscsireg.h was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.1.10.2	18-Jan-2017	skrll	Sync with netbsd-5
1.1.10.1	29-Oct-2015	skrll	file vioscsireg.h was added on branch netbsd-7-nhusb on 2017-01-18 08:46:42 +0000
1.1.6.2	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1338): sys/arch/amd64/conf/GENERIC: revision 1.418 sys/dev/pci/files.pci: revision 1.379 sys/dev/pci/vioscsi.c: revisions 1.1-1.8 sys/dev/pci/vioscsireg.h: revision 1.1 sys/dev/pci/virtioreg.h: revisions 1.5, 1.6 add the common flag bits -- Add vioscsi, compile tested only (toxic) -- this is working now, remove debugging. -- add vioscsi -- kill some more debugging. -- Fake mode sense data for illegal targets. qemu reports 256 targets... -- fix the status return. -- Use SCSIPI_CHAN_NOSETTLE. -- vioscsi_req_get()/virtio_enqueue_prep() failing is actually perfectly normal - observed failures included 10, 27, 61 in-flight commands, so probably depends on particular command mix; return with XS_RESOURCE_SHORTAGE rather then panic do vioscsi_req_put() when initial bus_dmamap_load() fails, as suggested by the XXX; the vq_done hook is called by virtio, but in that case we never get to commit the request to it -- react on ADAPTER_REQ_SET_XFER_MODE so that we set tagged queuing pass tag type and set id
1.1.6.1	29-Oct-2015	snj	file vioscsireg.h was added on branch netbsd-7 on 2016-12-23 06:33:56 +0000
1.1.2.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.1.2.1	29-Oct-2015	skrll	file vioscsireg.h was added on branch nick-nhusb on 2015-12-27 12:09:57 +0000
1.84	06-Sep-2025	riastradh	virtio(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs
1.83	26-Jul-2025	martin	branches: 1.83.2; PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.82	05-Aug-2024	riastradh	virtio(4): Assert child attach finished before we allow queueing. Should help catch mistakes like: PR kern/58555: Kernel panic during boot when using viocon with PCIe
1.81	10-Feb-2024	isaki	branches: 1.81.2; Split KASSERT(A && B) into KASSERT(A); KASSERT(B).
1.80	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.79	07-Nov-2023	rin	virtio_read_device_config_le_[24]: Fix for {aarch64,arm}eb Stop byte-swapping for big-endian aarch64 and arm to fix corrupted read for, e.g., sc_taglen for vio9p(4). As described as comments in virtio_pci.c, big-endian aarch64 and armv7 (armeb) are somewhat special. At the moment, all supported virtual/real machines are configured as little-endian, and only CPU cores are switched to big-endian mode during early boot stage. Most peripheral buses for armeb are LSB as a result, and their default bus_space(9) functions swap byte-order. Therefore, PIOed data from memory-mapped devices, as well as pci(4) ones, are actually LSB, but observed as MSB for armeb. Therefore, we should not swap byte-order further in virtio_read_device_config_le_[24] on armeb. Thanks ozaki-r@ for comments. XXX Centralize? See virtio_pci.c and sys/dev/fdt/virtio_mmio_fdt.c also.
1.78	21-Apr-2023	yamaguchi	virtio(4): change members of struct vring_desc_extra before free a slot This prevents the following race condition. 1. Thread-A: calls virtio_dequeue_commit() and puts a slot into free descriptor chain in vq_free_slot() 2. Thread-B: calls virtio_enqueue_prep() and get the slot stored by Thread-A 3. Thread-B: calls virtio_enqueue_reserve() and changes desc_base and desc_free_idx for the slot 4. Thread-A: changes the same members updated by Thread-B reported by hannken, thanks.
1.77	19-Apr-2023	yamaguchi	virtio(4): change variable name, nfc
1.76	19-Apr-2023	yamaguchi	Fix not to allocate unnecessary descriptor fixes PR/57358
1.75	19-Apr-2023	yamaguchi	Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357
1.74	31-Mar-2023	yamaguchi	Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed().
1.73	31-Mar-2023	yamaguchi	Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler.
1.72	29-Mar-2023	riastradh	virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 Reported-by: syzbot+7fb1047f5dfa33b26331@syzkaller.appspotmail.com
1.71	27-Mar-2023	nakayama	Use PRIuBUSSIZE to print bus_size_t variables.
1.70	27-Mar-2023	riastradh	virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended.
1.69	25-Mar-2023	yamaguchi	Setup virtqueues after registering them to virtio_softc
1.68	24-Mar-2023	yamaguchi	Added check of pointer for allocated memory before release of resource
1.67	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.66	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.65	03-Jan-2023	jakllsch	Subtly adjust criteria for notification of Virtio devices in VIRTIO_F_RING_EVENT_IDX mode. This fixes stalls in the vioif(4) transmit path that would happen sooner or later, depending on interface transmit utilization, and were particularly noticable with a NetBSD guest transmitting to its x86 KVM host at multi-gigabit speeds.
1.64	30-Dec-2022	jakllsch	Fix dmamap_syncs more; the number of ring elements is vq->vq_num, not sc->sc_nvqs.
1.63	31-Oct-2022	simonb	branches: 1.63.2; Use PRIx64 in a debug print macro to print a uint64_t.
1.62	18-Oct-2022	skrll	KNF
1.61	17-Oct-2022	skrll	Fix previous by using roundup (and not howmany).
1.60	15-Oct-2022	riastradh	virtio(4): Use howmany from sys/param.h instead of open-coding it. howmany will divide by VIRTIO_PAGE_SIZE instead of doing & ~(VIRTIO_PAGE_SIZE - 1), but it's a constant 4096 so this should make no difference in the compiled output except possibly at -O0. No functional change intended.
1.59	15-Oct-2022	riastradh	virtio(4): Sprinkle KNF. No functional change intended.
1.58	14-Aug-2022	riastradh	virtio(4): Print numeric device type, even if unrecognized.
1.57	12-Aug-2022	riastradh	virtio(4): Membar and bus_dmamap_sync audit. - Don't use membar_* for DMA. - Sync only the header and payload of the rings separately as needed. If we bounce, this avoids large memcpy when we only care about the header. - Sync uring with PREREAD before triggering anything that will return data in it. => Move uring PREREAD in virtio_enqueue_commit to _before_ updating vq->vq_avail->idx, which is the pointat which the DMA read is triggered in the `device' (host). => Omit needless membar_consumer in virtio_enqueue_commit -- may not work with DMA memory, and even if it does, redundant with bus_dmamap_sync uring PREREAD here. => XXX Does the device/host ever return unsolicited entries in the queue, or only solicited ones? If only solicited ones, the PREREAD in virtio_init_vq is redundant. - Sync uring with POSTREAD before we read from it. This way the DMA read into our buffer has finished before we read from the buffer. => Add missing uring POSTREAD in virtio_vq_is_enqueued, between read of vq->vq_used_idx and return to caller, so that the caller can safely use virtio_dequeue. => Add missing uring POSTREADs in virtio_start_vq_intr: . between entry from caller and the read of vq->vq_used_idx . between the read of vq->vq_used_idx and return to caller, so that the caller can safely use virtio_dequeue, just like virtio_vq_is_enqueued => Move uring POSTREADs in virtio_enqueue_commit to _before_ reading vq->vq_used->flags or *vq->vq_avail_event, not after. - After we write to aring, sync it with PREWRITE. This way we finish writing to our buffer before the DMA write from it. => Omit needless PREWRITE in virtio_init_vq -- we do the appropriate PREWRITE in virtio_enqueue_commit now. => Convert membar_producer to bus_dmamap_sync PREWRITE in virtio_enqueue_commit. => Omit incorrect aring POSTWRITE in virtio_enqueue_commit -- no need because the DMA write may not have completed yet at this point, and we already do a POSTWRITE in virtio_vq_is_enqueued. => Omit needless membar_producer in virtio_postpone_intr -- may not work with DMA memory, and even if it does, redundant with bus_dmamap_sync PREWRITE here. - After xfers to aring have completed, sync it with POSTWRITE. => Add missing aring POSTWRITE in virtio_free_vq, in case there are paths from virtio_enqueue_commit to here that don't go through virtio_is_enqueued. (If there are no such paths, then maybe we should KASSERT(vq->vq_queued == 0) in virtio_free_vq.)
1.56	09-Aug-2022	riastradh	virtio(4): Move comment for virtio_vq_intr. No functional change intended.
1.55	18-Jun-2022	andvar	fix typos in word "functions" in comments, mainly s/fuctions/functions/.
1.54	24-Apr-2022	uwe	virtio: cosmetics - use (*pfn)(...) syntax. Some people prefer to call function pointers without an explicit dereference (that is purely cosmetic in this position), especially to create faux c++ s->pfn(...) with function pointers in struct members. Some prefer explicit dereference (that requires parens around it). (pfn)(...) without dereference but with parens looks odd to both, so make it conform to one of the established alternatives. Same object code is generated.
1.53	28-Oct-2021	yamaguchi	virtio: stop reinit for safety when a device resetting is failed
1.52	21-Oct-2021	yamaguchi	virtio: fix missing release of interrupt resources
1.51	21-Oct-2021	yamaguchi	virtio: reconfigure MSI-X vector on re-initialization This may fix a problem that some interrupts, for example link-state change of vioif(4) are not notified.
1.50	21-Oct-2021	yamaguchi	divide setup routine of virtio interrupts into establishment and device configuration
1.49	07-Feb-2021	skrll	In virtio_alloc_vq change BUS_DMA_NOWAIT to BUS_DMA_WAITOK. It's always called from attach routines.
1.48	07-Feb-2021	skrll	Trailing whitespace
1.47	05-Feb-2021	reinoud	Better reading of 4.1.3.1, it seems that using 32 bit reads/writes is mandatory for non-legacy devices.
1.46	05-Feb-2021	reinoud	Second round of cleaning up endian code. No more tailored code to maintain.
1.45	28-Jan-2021	reinoud	Rewrite and streamline virtio device config read/write and explicitly cater for the Aarch64-eb bus problem with Qemu. This removes lots of bus_space `magic' and cleans up the code.
1.44	20-Jan-2021	reinoud	Remove the virtio child driver finalisation check KASSERT that, while it should never trigger, a possible bug in a child driver shouldn't have to panic the kernel. Instead report the internal error.
1.43	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.42	17-Sep-2020	jakllsch	branches: 1.42.2; Fix word-o of function name in comment
1.41	25-May-2020	yamaguchi	Set handlers implemented in child device of virtio(4) to virtqueue instead of the commonized function
1.40	25-May-2020	yamaguchi	Register an interrupt handler for virtqueue to that for pci to reduce function call
1.39	25-May-2020	yamaguchi	Remove VIRTIO_F_PCI_INTR_SOFTINT support for multiqueue The functionality is no longer used by obsolating VIOIF_SOFTINT_INTR.
1.38	01-Oct-2019	chs	in many device attach paths, allocate memory with KM_SLEEP instead of KM_NOSLEEP and remove code to handle failures that can no longer happen.
1.37	14-Jan-2019	yamaguchi	branches: 1.37.4; Add multiqueue support, virtio(4)
1.36	30-Sep-2018	jmcneill	Suppress "not configured" message when no device is present
1.35	15-Jun-2018	jakllsch	branches: 1.35.2; Add cfprint_t virtiobusprint() and use it for virtio_mmio.
1.34	15-Jun-2018	jakllsch	Fix queue address setup bug introduced in 1.31. This use-before-initialization problem resulted in non-zero queues never being initialized with their virtqueue pointer, and queue zero's pointer to be clobbered with the virtqueue pointer for the higest-indexed ring.
1.33	06-Jun-2018	jakllsch	remove now-irrelevant <dev/pci> includes
1.32	06-Jun-2018	jakllsch	fix up virtio module for current structure; add virtio_pci module
1.31	02-Jun-2018	jakllsch	Begin to detangle virtio from its PCI attachment
1.30	14-Feb-2018	uwe	branches: 1.30.2; Cosmetic - consistently use #define<SPACE>
1.29	02-Aug-2017	cherry	Reorg the virtio(4) config(9) code to be explicitly pci specific. We'll later use this for a non-pci virtio(4) usecase. ok martin
1.28	01-Jun-2017	chs	branches: 1.28.2; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.27	28-Mar-2017	ozaki-r	Handle config change interrupts to inhibit sending packets while link down PR kern/52103 by s-yamaguchi@IIJ
1.26	26-Mar-2017	jdolecek	fix pasto for nbo_bus_space_write_4() - value needs to be uint32_t
1.25	26-Mar-2017	martin	Duh, remove accidently left over debug printfs
1.24	26-Mar-2017	martin	Backout previous, no need for these accessor in MD code
1.23	26-Mar-2017	martin	Provide an accessor for the pci attach args of a virtio device (if available - future non-pci attachments may return NULL). Sparc64 MD code needs this to match the OF node of the virtio device for boot device recognition.
1.22	25-Mar-2017	jdolecek	reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.21	25-Mar-2017	jdolecek	use ipl level specified by child when attaching interrupts, instead of always using IPL_NET
1.20	25-Mar-2017	martin	PR kern/52107: for now (while we still use the legacy virtio spec) at least hack around the endianess issues with that.
1.19	29-Nov-2016	uwe	branches: 1.19.2; Fix the "typical usage" example for enqueuing a request - if bus_dmamap_load(9) fails, we shouldn't call bus_dmamap_unload(9).
1.18	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.17	14-Aug-2016	tron	Prevent a panic during system shutdown when vioif(4) is used
1.16	11-Jul-2016	knakahara	branches: 1.16.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.15	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.14	10-Jan-2016	christos	PR/50636: Ryo ONODERA: Fix incorrect kmem_free()
1.13	30-Oct-2015	christos	more device names
1.12	27-Oct-2015	christos	use dma64 if available so we don't have do bounce around too much.
1.11	26-Oct-2015	ozaki-r	Support MSI-X in virtio Currently only vioif(4) uses the feature. knakahara@ helped to migrate to pci_intr_alloc(9). Thanks!
1.10	15-Oct-2015	ozaki-r	Name the interrupt handler for intrctl
1.9	05-May-2015	ozaki-r	Remove a garbage blank line
1.8	19-Dec-2014	ozaki-r	Implement softint-based interrupt handling in if_vioif Softint-based interrupt handling is considered as a future direction of the (network) device driver architecture in NetBSD. pq3etsec of ppc is already implemented based on the architecture (unlike pq3etsec, this change doesn't include softint-based if_start). In this architecture, a hardware interrupt handler just schedules a softint and the softint performs actual interrupt processing. It reduces processing in hardware interrupt context and allows Layer 2 network stack (e.g., bridge, vlan and even bpf) run in softint context, which makes it easy to implement fine-grain locking in the layer. This is an experimental implementation of the architecture in if_viof. virtio introduces a new flag VIRTIO_F_PCI_INTR_SOFTINT. If a driver of virtio sets it to sc_flags of virtio_softc, virtio calls softint_schedule in virtio_intr instead of directly calling the interrupt handler of the driver. When VIOIF_SOFTINT_INTR is on, vioif doesn't use the existing softint (vioif_rx_softint) that is called from vioif_rx_vq_done. Because vioif_rx_softint already runs in softint context and another softint isn't needed. This change actually improves performance in some cases. The feature is disabled by default and enabled when SOFTINT_INTR is set somewhere (normally in a kernel configuration).
1.7	06-Oct-2014	mlelstv	branches: 1.7.2; use correct macro to handle subsys ids. Functional there is no change, as both macros select the same bits.
1.6	22-Jul-2014	ozaki-r	branches: 1.6.2; 1.6.6; Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio It is set by a child driver, e.g., if_vioif. If set, virtio sets PCI_INTR_MPSAFE for pci_intr_establish.
1.5	29-Mar-2014	christos	branches: 1.5.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.4	10-Aug-2013	tsutsui	Make sure to check if the driver has a valid intr handler in virtio_detach(). Fixes a panic during shutdown on KVM under ubuntu 13.04 with virtio, as reported in PR kern/48105 by Richard Hansen. Should be pulled up to netbsd-6 branches.
1.3	02-Nov-2011	njoly	branches: 1.3.6; 1.3.8; 1.3.12; 1.3.14; 1.3.16; 1.3.22; Small typo in symbolic name.
1.2	02-Nov-2011	jakllsch	branches: 1.2.2; Use symbolic PCI_PRODUCTs.
1.1	30-Oct-2011	hannken	Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.2.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.2.1	10-Nov-2011	yamt	sync with head
1.3.22.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #931): sys/dev/pci/virtio.c: revision 1.4 Make sure to check if the driver has a valid intr handler in virtio_detach(). Fixes a panic during shutdown on KVM under ubuntu 13.04 with virtio, as reported in PR kern/48105 by Richard Hansen. Should be pulled up to netbsd-6 branches.
1.3.16.2	18-May-2014	rmind	sync with head
1.3.16.1	28-Aug-2013	rmind	sync with head
1.3.14.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #931): sys/dev/pci/virtio.c: revision 1.4 Make sure to check if the driver has a valid intr handler in virtio_detach(). Fixes a panic during shutdown on KVM under ubuntu 13.04 with virtio, as reported in PR kern/48105 by Richard Hansen. Should be pulled up to netbsd-6 branches.
1.3.12.2	03-Dec-2017	jdolecek	update from HEAD
1.3.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.3.8.1	07-Sep-2013	bouyer	Pull up following revision(s) (requested by tsutsui in ticket #931): sys/dev/pci/virtio.c: revision 1.4 Make sure to check if the driver has a valid intr handler in virtio_detach(). Fixes a panic during shutdown on KVM under ubuntu 13.04 with virtio, as reported in PR kern/48105 by Richard Hansen. Should be pulled up to netbsd-6 branches.
1.3.6.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.3.6.1	02-Nov-2011	riz	file virtio.c was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.5.2.1	10-Aug-2014	tls	Rebase.
1.6.6.1	18-Jan-2017	skrll	Sync with netbsd-5
1.6.2.1	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1337): sys/dev/pci/if_vioif.c: revision 1.21 sys/dev/pci/virtio.c: revision 1.14 PR/50636: Ryo ONODERA: Fix incorrect kmem_free() -- PR/50636: Ryo ONODERA: Reduce memory use
1.7.2.8	28-Aug-2017	skrll	Sync with HEAD
1.7.2.7	05-Dec-2016	skrll	Sync with HEAD
1.7.2.6	05-Oct-2016	skrll	Sync with HEAD
1.7.2.5	09-Jul-2016	skrll	Sync with HEAD
1.7.2.4	19-Mar-2016	skrll	Sync with HEAD
1.7.2.3	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.7.2.2	06-Jun-2015	skrll	Sync with HEAD
1.7.2.1	06-Apr-2015	skrll	Sync with HEAD
1.16.2.3	26-Apr-2017	pgoyette	Sync with HEAD
1.16.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.16.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.19.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.28.2.3	23-Sep-2020	martin	Pull up following revision(s) (requested by jakllsch in ticket #1609): sys/dev/pci/virtio_pci.c: revision 1.13 (patch) (applied to sys/dev/pci/virtio.c) Ensure MSI-X is disabled if allocation of MSI-X interrupts fail. The virtio device config space moves out from under us when MSI-X remains enabled, and/or INTx interrupts are masked if we don't ensure this. This un-breaks virtio devices that run out of MSI-X interrupts. Particularly a problem on uniproc x86, where there are only 8 or 9 vectors available, allowing for only about 4 virtio devices to use MSI-X.
1.28.2.2	20-Sep-2020	martin	Pull up following revision(s) (requested by jakllsch in ticket #1608): sys/dev/pci/virtio_pci.c: revision 1.12 Ensure interrupt handles buffer is zeroed on allocation Prevents crashes trying to deallocate interrupts at shutdown. Found by kim and mlelstv, confirmed by me
1.28.2.1	19-Feb-2018	snj	Pull up following revision(s) (requested by uwe in ticket #553): sys/dev/pci/virtio_pci.c: 1.2 via patch to sys/dev/pci/virtio.c Explicitly enable PCI_COMMAND_MASTER_ENABLE and PCI_COMMAND_IO_ENABLE. Recent versions of VirtualBox do not enable bus-mastering by default - this will be fixed, but it doesn't hurt to do it ourselves too.
1.30.2.3	18-Jan-2019	pgoyette	Synch with HEAD
1.30.2.2	20-Oct-2018	pgoyette	Sync with head
1.30.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.35.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.35.2.1	10-Jun-2019	christos	Sync with HEAD
1.37.4.1	19-Oct-2025	martin	Pull up following revision(s) (requested by riastradh in ticket #1977): sys/dev/pci/virtio.c: revision 1.84 (patch) virtio(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs
1.42.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.63.2.7	19-Oct-2025	martin	Pull up following revision(s) (requested by riastradh in ticket #1178): sys/dev/pci/virtio.c: revision 1.84 virtio(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs
1.63.2.6	02-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #914): sys/dev/pci/virtio_pci.c: revision 1.50 sys/dev/pci/virtio_pci.c: revision 1.51 sys/dev/pci/virtio_pci.c: revision 1.52 sys/dev/pci/virtio_pci.c: revision 1.53 sys/dev/pci/virtio_pci.c: revision 1.54 sys/dev/pci/vioscsi.c: revision 1.35 sys/dev/pci/ld_virtio.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.34 sys/dev/virtio/viocon.c: revision 1.9 sys/dev/pci/ld_virtio.c: revision 1.35 sys/dev/pci/virtio.c: revision 1.80 sys/dev/pci/virtio.c: revision 1.81 sys/dev/pci/virtio.c: revision 1.82 sys/dev/pci/virtio_pci.c: revision 1.45 sys/dev/pci/virtio_pci.c: revision 1.46 sys/dev/pci/virtio_pci.c: revision 1.47 sys/dev/pci/virtio_pci.c: revision 1.48 sys/dev/pci/virtio_pci.c: revision 1.49 sys/dev/pci/virtio.c: revision 1.79 A single SCSI request may require multiple slots in the virtio queue but the queue isn't sized for the theoretical maximum. So just use XS_BUSY to pace the scsipi layer. viocon(4): Show error in assertion failure. virtio_read_device_config_le_[24]: Fix for {aarch64,arm}eb Stop byte-swapping for big-endian aarch64 and arm to fix corrupted read for, e.g., sc_taglen for vio9p(4). As described as comments in virtio_pci.c, big-endian aarch64 and armv7 (armeb) are somewhat special. At the moment, all supported virtual/real machines are configured as little-endian, and only CPU cores are switched to big-endian mode during early boot stage. Most peripheral buses for armeb are LSB as a result, and their default bus_space(9) functions swap byte-order. Therefore, PIOed data from memory-mapped devices, as well as pci(4) ones, are actually LSB, but observed as MSB for armeb. Therefore, we should not swap byte-order further in virtio_read_device_config_le_[24] on armeb. See virtio_pci.c and sys/dev/fdt/virtio_mmio_fdt.c also. fix spelling mistakes, mainly in comments and log messages. Split KASSERT(A && B) into KASSERT(A); KASSERT(B). Fix typo in error message. Modify a confused expression in ld_virtio_attach(). VIRTIO_BLK_MIN_SEGMENTS should be the total number of non-data segments, so I rename it to VIRTIO_BLK_CTRL_SEGMENTS. PR kern/57981. ld@virtio(4): Fix maximum size parameters. - SEG_MAX is the maximum number of segments. - SIZE_MAX is the maximum number of bytes in a single segment. The maximum transfer size is, therefore, SEG_MAX * SIZE_MAX. => Don't add two extra segments in the dmamap vr_payload for the header and status -- we already have a separate dmamap vr_cmdsts for that. => Don't recalculate payload dmamap parameters based on division by NBPG, just use the ones specified by the host. => Allow SIZE_MAX below MAXPHYS as long as SIZE_MAX*SEG_MAX >= MAXPHYS. Even though ldattach clamps ld->sc_maxxfer to MAXPHYS, make sure to clamp it in ld_virtio_attach before ld_virtio_alloc_reqs since it determines the dmamap sizes and bounce buffer allocation and there's no sense in allocating those larger than ld will use anyway. PR kern/58338 virtio_pci.c: Nix trailing whitespace. virtio@pci: Unmap the BARs we actually mapped on >=1.0 detach. PR kern/58335 virtio@pci: Nix useless sc_mapped_iosize. This appears to have been a mistaken attempt to pass the subregion length through to bus_space_unmap, which doesn't make any sense -- bus_space_subregion does not allocate anything and the resulting handle must not be passed to bus_space_unmap. Once we pass the size that was actually returned by pci_mapreg_map to bus_space_unmap, which matches the size passed internally by pci_mapreg_map to bus_space_map, sc_mapped_iosize is no longer used and can be flushed. Prompted by PR kern/58335. virtio@pci: Don't detach on shutdown. The virtio bus does not keep state in memory that needs to flushed to persistent storage. Typically DVF_DETACH_SHUTDOWN is only needed by, e.g., disk drivers to flush disk caches (although really, vfs sync should do that anyway, so maybe we should just nix the option altogether). Prompted by PR kern/58335. virtio@pci: Zero sc_bars_iosize[i] on unmap. No functional change intended, but this makes the code less brittle. Prompted by PR kern/58335. virtio_pci.c: Sort includes. No functional change intended. virtio_pci.c: Omit needless blank lines and commented code. No functional change intended. virtio_pci.c: Sprinkle KNF. - No parameter names in function prototypes. - Fix indentation. - Break overlong lines. - Blank line between declarations and rest of block. virtio_pci.c: Sprinkle const. No functional change intended. virtio_pci.c: Use container_of, not cast. virtio(4): Assert child attach finished before we allow queueing. Should help catch mistakes like: PR kern/58555: Kernel panic during boot when using viocon with PCIe
1.63.2.5	03-Jun-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #186): sys/dev/pci/virtio_pci.c: revision 1.41 sys/dev/pci/virtio_pci.c: revision 1.42 sys/dev/virtio/virtio_mmio.c: revision 1.10 sys/dev/pci/virtiovar.h: revision 1.29 sys/dev/pci/virtio.c: revision 1.75 sys/dev/pci/virtio.c: revision 1.76 sys/dev/pci/virtio.c: revision 1.77 sys/dev/pci/virtio.c: revision 1.78 virtio@pci: Fix assertion on detach. If the child never attached in the first place, it's OK for it to not have detached. XXX This should not be a set of flags; this should be a state enumeration, because some flags make no sense, like FINISHED\|FAILED. XXX This should not be asserted separately in each bus; there should be a single place in virtio.c to assert this, uniformly in all buses. PR kern/57357 Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357 Fix not to allocate unnecessary descriptor fixes PR/57358 virtio(4): change variable name, nfc virtio(4): change members of struct vring_desc_extra before free a slot This prevents the following race condition. 1. Thread-A: calls virtio_dequeue_commit() and puts a slot into free descriptor chain in vq_free_slot() 2. Thread-B: calls virtio_enqueue_prep() and get the slot stored by Thread-A 3. Thread-B: calls virtio_enqueue_reserve() and changes desc_base and desc_free_idx for the slot 4. Thread-A: changes the same members updated by Thread-B reported by hannken, thanks.
1.63.2.4	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.63.2.3	02-Apr-2023	martin	Backout attempts to build fix for ticket #128 for this file, needs other changes.
1.63.2.2	01-Apr-2023	martin	Additionally pull up the following for ticket #128 to fix the build: sys/dev/pci/if_vioif.c: revision 1.107 sys/dev/pci/virtio.c: revision 1.71 Use PRIuBUSSIZE to print bus_size_t variables.
1.63.2.1	13-Jan-2023	martin	Pull up following revision(s) (requested by jakllsch in ticket #48): sys/dev/pci/virtio.c: revision 1.64 sys/dev/pci/virtio.c: revision 1.65 Fix dmamap_syncs more; the number of ring elements is vq->vq_num, not sc->sc_nvqs. Subtly adjust criteria for notification of Virtio devices in VIRTIO_F_RING_EVENT_IDX mode. This fixes stalls in the vioif(4) transmit path that would happen sooner or later, depending on interface transmit utilization, and were particularly noticable with a NetBSD guest transmitting to its x86 KVM host at multi-gigabit speeds.
1.81.2.1	02-Aug-2025	perseant	Sync with HEAD
1.83.2.1	19-Oct-2025	martin	Pull up following revision(s) (requested by riastradh in ticket #62): sys/dev/pci/virtio.c: revision 1.84 virtio(4): Use paravirt_membar_sync(9) where needed. PR kern/59618: occasional virtio block device lock ups/hangs
1.55	25-Sep-2024	christos	Wrong byte order of config space write in big-endian (gorg) In virtio_pci.c, the function virtio_pci_bus_space_write_8 writes the first 4 bytes and second 4 bytes of an 8-byte value in the same or swapped order within the device's config space depending on the endianness of the system it is compiled for. However, the BUS_ADDR_LO32 and BUS_ADDR_HI32 always produce the least-significant and most-significant 4 bytes of an 8-byte value, respectively, regardless of the endianness of the system. Since virtio_pci_bus_space_write_8 is always used only for writing to little-endian parts of the config space, the least significant 4 bytes must always be written to the first 4 bytes in the config space value, and the most significant 4 bytes must always be written to the last 4 bytes of the config space value. Therefore, swapping the byte order produces incorrect behaviour on big-endian systems. The incorrect behaviour seems to have been introduced in revision 1.19 of virtio_pci.c, which this change would mostly reversed
1.54	25-Jun-2024	riastradh	branches: 1.54.2; virtio_pci.c: Use container_of, not cast.
1.53	25-Jun-2024	riastradh	virtio_pci.c: Sprinkle const. No functional change intended.
1.52	25-Jun-2024	riastradh	virtio_pci.c: Sprinkle KNF. - No parameter names in function prototypes. - Fix indentation. - Break overlong lines. - Blank line between declarations and rest of block.
1.51	25-Jun-2024	riastradh	virtio_pci.c: Omit needless blank lines and commented code. No functional change intended.
1.50	25-Jun-2024	riastradh	virtio_pci.c: Sort includes. No functional change intended.
1.49	25-Jun-2024	riastradh	virtio@pci: Zero sc_bars_iosize[i] on unmap. No functional change intended, but this makes the code less brittle. Prompted by PR kern/58335.
1.48	25-Jun-2024	riastradh	virtio@pci: Don't detach on shutdown. The virtio bus does not keep state in memory that needs to flushed to persistent storage. Typically DVF_DETACH_SHUTDOWN is only needed by, e.g., disk drivers to flush disk caches (although really, vfs sync should do that anyway, so maybe we should just nix the option altogether). Prompted by PR kern/58335.
1.47	25-Jun-2024	riastradh	virtio@pci: Nix useless sc_mapped_iosize. This appears to have been a mistaken attempt to pass the subregion length through to bus_space_unmap, which doesn't make any sense -- bus_space_subregion does not allocate anything and the resulting handle must not be passed to bus_space_unmap. Once we pass the size that was actually returned by pci_mapreg_map to bus_space_unmap, which matches the size passed internally by pci_mapreg_map to bus_space_map, sc_mapped_iosize is no longer used and can be flushed. Prompted by PR kern/58335.
1.46	25-Jun-2024	riastradh	virtio@pci: Unmap the BARs we actually mapped on >=1.0 detach. PR kern/58335
1.45	25-Jun-2024	riastradh	virtio_pci.c: Nix trailing whitespace.
1.44	19-Nov-2023	thorpej	On alpha and sparc64, use VirtIO 1.0 and VIRTIO_F_ACCESS_PLATFORM so that DMA to PCI-attached VirtIO devices works properly. This is needed for Qemu to select the appropriate address space for PCI DMA.
1.43	07-Jul-2023	rin	virtio(4): Fix byte order of DMA data for armeb in the same manner as aarch64eb. Fix comments to explain what is going on for {arm,aarch64}eb on QEMU. This is not due to QEMU bugs; it correctly configures everything for little-endian environment, and we forcibly change byte order only for CPU cores during kernel bootstrap.
1.42	19-Apr-2023	yamaguchi	Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357
1.41	16-Apr-2023	riastradh	virtio@pci: Fix assertion on detach. If the child never attached in the first place, it's OK for it to not have detached. XXX This should not be a set of flags; this should be a state enumeration, because some flags make no sense, like FINISHED\|FAILED. XXX This should not be asserted separately in each bus; there should be a single place in virtio.c to assert this, uniformly in all buses. PR kern/57357 XXX pullup-10
1.40	31-Mar-2023	yamaguchi	Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed().
1.39	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.38	30-May-2022	riastradh	branches: 1.38.4; virtio at pci: Provide attribution in debug message. Also make it only happen in the error case so success doesn't clutter up the console output.
1.37	13-Apr-2022	uwe	virtio_pci: Don't use a cast for upcasting. Same object code is generated.
1.36	17-Mar-2022	uwe	virtio_pci_bus_space_write_8: don't use bus_space_write_8. The standard says: 4.1.3.1 Driver Requirements: PCI Device Layout For device configuration access, the driver MUST use ... 32-bit wide and aligned accesses for ... 64-bit wide fields. For 64-bit fields, the driver MAY access each of the high and low 32-bit parts of the field independently. NB: "MAY" in this text refers to "independently" (i.e. the order of accesses) not "32-bit" (which is restricted by the earlier "MUST"). Note also that virtio_{read,write}_device_config_8 in virtio.c already uses two 32-bit accesses.
1.35	17-Mar-2022	uwe	virtio_pci.c: use PCI_BAR macro instead of open-coding it.
1.34	14-Mar-2022	uwe	virtio_pci_match: add TODO about PCI Revision ID. The standard says: Transitional devices MUST have a PCI Revision ID of 0. Non-transitional devices SHOULD have a PCI Revision ID of 1 or higher. Drivers MUST match any PCI Revision ID value. so we must not check the revision id for non-transitional devices. The code in attach relies on the revision being specifically(NB!) 1 so this calls for a revision, but I can't test this properly at the moment, so just leave a reminder. Comment change only.
1.33	28-Oct-2021	yamaguchi	virtio: stop reinit for safety when a device resetting is failed
1.32	21-Oct-2021	yamaguchi	virtio: reconfigure MSI-X vector on re-initialization This may fix a problem that some interrupts, for example link-state change of vioif(4) are not notified.
1.31	21-Oct-2021	yamaguchi	divide setup routine of virtio interrupts into establishment and device configuration
1.30	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.29	24-Apr-2021	thorpej	branches: 1.29.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.28	05-Feb-2021	reinoud	branches: 1.28.2; Second round of cleaning up endian code. No more tailored code to maintain.
1.27	28-Jan-2021	reinoud	Rewrite and streamline virtio device config read/write and explicitly cater for the Aarch64-eb bus problem with Qemu. This removes lots of bus_space `magic' and cleans up the code.
1.26	26-Jan-2021	reinoud	Fix indexing bug in clean up code on error in virtio PCI v1.0 attach code. Thanks go to Rin Okuyama for spotting it.
1.25	24-Jan-2021	reinoud	Remove incorrect comment. It would violate the specs. VirtIO PCI v1.0 attachments can only happen on revision 1 devices as they have a radical different register layout. Transitional devices have to use revision 0 and have to use the VirtIO PCI v0.9 attachment method but can allow for features to be negotiated normally found in later versions.
1.24	24-Jan-2021	thorpej	Redefining bus_space functions in drivers is a bad idea, and we just should't be in the habit of doing so. Besides, the previous "solutions" still did not compile correctly, and this does, so let's be done with this nonsense, shall we?
1.23	24-Jan-2021	reinoud	On error unmap the pci_mapreg_map()d regions using bus_space_unmap() as suggested by jak@
1.22	24-Jan-2021	reinoud	Move definition of NMAPREG to the start
1.21	24-Jan-2021	reinoud	Prevent potential buffer over runs in number of BARS
1.20	24-Jan-2021	christos	Add a comment explaining why bus_space_write_8 is limited to virtio.
1.19	23-Jan-2021	christos	Provide a generic bus_space_write_8 function that is bi-endian.
1.18	21-Jan-2021	reinoud	Remove dependency on bus_space_write_8() for i386 and instead implement it as two bus_space_write_4()'s as allowed in the spec.
1.17	21-Jan-2021	martin	Use PRIxBUSSIZE to print a bus_size_t value.
1.16	20-Jan-2021	reinoud	Remove the virtio child driver finalisation check KASSERT that, while it should never trigger, a possible bug in a child driver shouldn't have to panic the kernel. Instead report the internal error.
1.15	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.14	23-Sep-2020	jakllsch	branches: 1.14.2; whitespace fix
1.13	17-Sep-2020	jakllsch	Ensure MSI-X is disabled if allocation of MSI-X interrupts fail. The virtio device config space moves out from under us when MSI-X remains enabled, and/or INTx interrupts are masked if we don't ensure this. This un-breaks virtio devices that run out of MSI-X interrupts. Particularly a problem on uniproc x86, where there are only 8 or 9 vectors available, allowing for only about 4 virtio devices to use MSI-X.
1.12	17-Sep-2020	jakllsch	Ensure interrupt handles buffer is zeroed on allocation Prevents crashes trying to deallocate interrupts at shutdown. Found by kim and mlelstv, confirmed by me
1.11	27-May-2020	yamaguchi	Fix to allocate resources actuall used around interrupts
1.10	25-May-2020	yamaguchi	Set handlers implemented in child device of virtio(4) to virtqueue instead of the commonized function
1.9	25-May-2020	yamaguchi	Register an interrupt handler for virtqueue to that for pci to reduce function call
1.8	25-May-2020	yamaguchi	Remove VIRTIO_F_PCI_INTR_SOFTINT support for multiqueue The functionality is no longer used by obsolating VIOIF_SOFTINT_INTR.
1.7	27-Jan-2019	pgoyette	branches: 1.7.4; Merge the [pgoyette-compat] branch
1.6	14-Jan-2019	yamaguchi	Add multiqueue support, virtio(4)
1.5	06-Jun-2018	jakllsch	branches: 1.5.2; fix up virtio module for current structure; add virtio_pci module
1.4	02-Jun-2018	jakllsch	Begin to detangle virtio from its PCI attachment
1.3	02-Jun-2018	jakllsch	fix CVS keyword in __KERNEL_RCSID
1.2	15-Feb-2018	uwe	branches: 1.2.2; Explicitly enable PCI_COMMAND_MASTER_ENABLE and PCI_COMMAND_IO_ENABLE. Recent versions of VirtualBox do not enable bus-mastering by default - this will be fixed, but it doesn't hurt to do it ourselves too.
1.1	02-Aug-2017	cherry	branches: 1.1.2; 1.1.4; Reorg the virtio(4) config(9) code to be explicitly pci specific. We'll later use this for a non-pci virtio(4) usecase. ok martin
1.1.4.2	03-Dec-2017	jdolecek	update from HEAD
1.1.4.1	02-Aug-2017	jdolecek	file virtio_pci.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.1.2.2	28-Aug-2017	skrll	Sync with HEAD
1.1.2.1	02-Aug-2017	skrll	file virtio_pci.c was added on branch nick-nhusb on 2017-08-28 17:52:25 +0000
1.2.2.2	18-Jan-2019	pgoyette	Synch with HEAD
1.2.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.5.2.1	10-Jun-2019	christos	Sync with HEAD
1.7.4.2	20-Sep-2020	martin	Pull up following revision(s) (requested by jakllsch in ticket #1090): sys/dev/pci/virtio_pci.c: revision 1.13 Ensure MSI-X is disabled if allocation of MSI-X interrupts fail. The virtio device config space moves out from under us when MSI-X remains enabled, and/or INTx interrupts are masked if we don't ensure this. This un-breaks virtio devices that run out of MSI-X interrupts. Particularly a problem on uniproc x86, where there are only 8 or 9 vectors available, allowing for only about 4 virtio devices to use MSI-X.
1.7.4.1	20-Sep-2020	martin	Pull up following revision(s) (requested by jakllsch in ticket #1089): sys/dev/pci/virtio_pci.c: revision 1.12 Ensure interrupt handles buffer is zeroed on allocation Prevents crashes trying to deallocate interrupts at shutdown. Found by kim and mlelstv, confirmed by me
1.14.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.28.2.1	27-Mar-2021	thorpej	- config_found_ia() -> config_found(). - virtio only carries a single interface attribute, so no need to specify it explicitly.
1.29.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.38.4.6	31-Dec-2024	snj	Pull up following revision(s) (requested by martin in ticket #1032): sys/dev/pci/virtioreg.h: 1.13 sys/dev/pci/virtio_pci.c: 1.44 Define the VIRTIO_F_ACCESS_PLATFORM, VIRTIO_F_RING_PACKED, VIRTIO_F_ORDER_PLATFORM, and VIRTIO_F_SR_IOV feature bits. -- On alpha and sparc64, use VirtIO 1.0 and VIRTIO_F_ACCESS_PLATFORM so that DMA to PCI-attached VirtIO devices works properly. This is needed for Qemu to select the appropriate address space for PCI DMA.
1.38.4.5	28-Nov-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #1018): sys/dev/pci/virtio_pci.c: revision 1.55 Wrong byte order of config space write in big-endian (gorg) In virtio_pci.c, the function virtio_pci_bus_space_write_8 writes the first 4 bytes and second 4 bytes of an 8-byte value in the same or swapped order within the device's config space depending on the endianness of the system it is compiled for. However, the BUS_ADDR_LO32 and BUS_ADDR_HI32 always produce the least-significant and most-significant 4 bytes of an 8-byte value, respectively, regardless of the endianness of the system. Since virtio_pci_bus_space_write_8 is always used only for writing to little-endian parts of the config space, the least significant 4 bytes must always be written to the first 4 bytes in the config space value, and the most significant 4 bytes must always be written to the last 4 bytes of the config space value. Therefore, swapping the byte order produces incorrect behaviour on big-endian systems. The incorrect behaviour seems to have been introduced in revision 1.19 of virtio_pci.c, which this change would mostly reversed
1.38.4.4	02-Oct-2024	martin	Pull up following revision(s) (requested by rin in ticket #914): sys/dev/pci/virtio_pci.c: revision 1.50 sys/dev/pci/virtio_pci.c: revision 1.51 sys/dev/pci/virtio_pci.c: revision 1.52 sys/dev/pci/virtio_pci.c: revision 1.53 sys/dev/pci/virtio_pci.c: revision 1.54 sys/dev/pci/vioscsi.c: revision 1.35 sys/dev/pci/ld_virtio.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.34 sys/dev/virtio/viocon.c: revision 1.9 sys/dev/pci/ld_virtio.c: revision 1.35 sys/dev/pci/virtio.c: revision 1.80 sys/dev/pci/virtio.c: revision 1.81 sys/dev/pci/virtio.c: revision 1.82 sys/dev/pci/virtio_pci.c: revision 1.45 sys/dev/pci/virtio_pci.c: revision 1.46 sys/dev/pci/virtio_pci.c: revision 1.47 sys/dev/pci/virtio_pci.c: revision 1.48 sys/dev/pci/virtio_pci.c: revision 1.49 sys/dev/pci/virtio.c: revision 1.79 A single SCSI request may require multiple slots in the virtio queue but the queue isn't sized for the theoretical maximum. So just use XS_BUSY to pace the scsipi layer. viocon(4): Show error in assertion failure. virtio_read_device_config_le_[24]: Fix for {aarch64,arm}eb Stop byte-swapping for big-endian aarch64 and arm to fix corrupted read for, e.g., sc_taglen for vio9p(4). As described as comments in virtio_pci.c, big-endian aarch64 and armv7 (armeb) are somewhat special. At the moment, all supported virtual/real machines are configured as little-endian, and only CPU cores are switched to big-endian mode during early boot stage. Most peripheral buses for armeb are LSB as a result, and their default bus_space(9) functions swap byte-order. Therefore, PIOed data from memory-mapped devices, as well as pci(4) ones, are actually LSB, but observed as MSB for armeb. Therefore, we should not swap byte-order further in virtio_read_device_config_le_[24] on armeb. See virtio_pci.c and sys/dev/fdt/virtio_mmio_fdt.c also. fix spelling mistakes, mainly in comments and log messages. Split KASSERT(A && B) into KASSERT(A); KASSERT(B). Fix typo in error message. Modify a confused expression in ld_virtio_attach(). VIRTIO_BLK_MIN_SEGMENTS should be the total number of non-data segments, so I rename it to VIRTIO_BLK_CTRL_SEGMENTS. PR kern/57981. ld@virtio(4): Fix maximum size parameters. - SEG_MAX is the maximum number of segments. - SIZE_MAX is the maximum number of bytes in a single segment. The maximum transfer size is, therefore, SEG_MAX * SIZE_MAX. => Don't add two extra segments in the dmamap vr_payload for the header and status -- we already have a separate dmamap vr_cmdsts for that. => Don't recalculate payload dmamap parameters based on division by NBPG, just use the ones specified by the host. => Allow SIZE_MAX below MAXPHYS as long as SIZE_MAX*SEG_MAX >= MAXPHYS. Even though ldattach clamps ld->sc_maxxfer to MAXPHYS, make sure to clamp it in ld_virtio_attach before ld_virtio_alloc_reqs since it determines the dmamap sizes and bounce buffer allocation and there's no sense in allocating those larger than ld will use anyway. PR kern/58338 virtio_pci.c: Nix trailing whitespace. virtio@pci: Unmap the BARs we actually mapped on >=1.0 detach. PR kern/58335 virtio@pci: Nix useless sc_mapped_iosize. This appears to have been a mistaken attempt to pass the subregion length through to bus_space_unmap, which doesn't make any sense -- bus_space_subregion does not allocate anything and the resulting handle must not be passed to bus_space_unmap. Once we pass the size that was actually returned by pci_mapreg_map to bus_space_unmap, which matches the size passed internally by pci_mapreg_map to bus_space_map, sc_mapped_iosize is no longer used and can be flushed. Prompted by PR kern/58335. virtio@pci: Don't detach on shutdown. The virtio bus does not keep state in memory that needs to flushed to persistent storage. Typically DVF_DETACH_SHUTDOWN is only needed by, e.g., disk drivers to flush disk caches (although really, vfs sync should do that anyway, so maybe we should just nix the option altogether). Prompted by PR kern/58335. virtio@pci: Zero sc_bars_iosize[i] on unmap. No functional change intended, but this makes the code less brittle. Prompted by PR kern/58335. virtio_pci.c: Sort includes. No functional change intended. virtio_pci.c: Omit needless blank lines and commented code. No functional change intended. virtio_pci.c: Sprinkle KNF. - No parameter names in function prototypes. - Fix indentation. - Break overlong lines. - Blank line between declarations and rest of block. virtio_pci.c: Sprinkle const. No functional change intended. virtio_pci.c: Use container_of, not cast. virtio(4): Assert child attach finished before we allow queueing. Should help catch mistakes like: PR kern/58555: Kernel panic during boot when using viocon with PCIe
1.38.4.3	09-Jul-2023	martin	Pull up following revision(s) (requested by rin in ticket #229): sys/dev/pci/virtio_pci.c: revision 1.43 sys/dev/virtio/virtio_mmio.c: revision 1.11 virtio(4): Fix byte order of DMA data for armeb in the same manner as aarch64eb. Fix comments to explain what is going on for {arm,aarch64}eb on QEMU. This is not due to QEMU bugs; it correctly configures everything for little-endian environment, and we forcibly change byte order only for CPU cores during kernel bootstrap.
1.38.4.2	03-Jun-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #186): sys/dev/pci/virtio_pci.c: revision 1.41 sys/dev/pci/virtio_pci.c: revision 1.42 sys/dev/virtio/virtio_mmio.c: revision 1.10 sys/dev/pci/virtiovar.h: revision 1.29 sys/dev/pci/virtio.c: revision 1.75 sys/dev/pci/virtio.c: revision 1.76 sys/dev/pci/virtio.c: revision 1.77 sys/dev/pci/virtio.c: revision 1.78 virtio@pci: Fix assertion on detach. If the child never attached in the first place, it's OK for it to not have detached. XXX This should not be a set of flags; this should be a state enumeration, because some flags make no sense, like FINISHED\|FAILED. XXX This should not be asserted separately in each bus; there should be a single place in virtio.c to assert this, uniformly in all buses. PR kern/57357 Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357 Fix not to allocate unnecessary descriptor fixes PR/57358 virtio(4): change variable name, nfc virtio(4): change members of struct vring_desc_extra before free a slot This prevents the following race condition. 1. Thread-A: calls virtio_dequeue_commit() and puts a slot into free descriptor chain in vq_free_slot() 2. Thread-B: calls virtio_enqueue_prep() and get the slot stored by Thread-A 3. Thread-B: calls virtio_enqueue_reserve() and changes desc_base and desc_free_idx for the slot 4. Thread-A: changes the same members updated by Thread-B reported by hannken, thanks.
1.38.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.54.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1	20-Jan-2021	reinoud	branches: 1.1.4; Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.1.4.2	03-Apr-2021	thorpej	Sync with HEAD.
1.1.4.1	20-Jan-2021	thorpej	file virtio_pcireg.h was added on branch thorpej-futex on 2021-04-03 22:28:49 +0000
1.14	26-Jul-2025	martin	PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.13	19-Nov-2023	thorpej	branches: 1.13.2; Define the VIRTIO_F_ACCESS_PLATFORM, VIRTIO_F_RING_PACKED, VIRTIO_F_ORDER_PLATFORM, and VIRTIO_F_SR_IOV feature bits.
1.12	29-Mar-2023	riastradh	virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304.
1.11	15-Oct-2022	riastradh	branches: 1.11.2; virtio(4): Use __BIT, not inline shift. Reduces need for cast.
1.10	13-Apr-2022	uwe	virtio: Use UINT64_C() instead of explicit ULL suffix.
1.9	13-Apr-2022	uwe	virtio: G/c accidentally committed comment in previous. The format spec for the V1 bit was to be committed as a separate followup commit, but I forgot to take it out. Sorry...
1.8	13-Apr-2022	uwe	virtio: use the new syntax for snprintb(3) format strings. The old syntax is limited to 32 bits only (and has 1-based bit numbers which is rather incovenient too).
1.7	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.6	29-Oct-2015	christos	branches: 1.6.32; Add vioscsi, compile tested only (toxic)
1.5	27-Oct-2015	christos	add the common flag bits
1.4	26-Oct-2015	ozaki-r	Support MSI-X in virtio Currently only vioif(4) uses the feature. knakahara@ helped to migrate to pci_intr_alloc(9). Thanks!
1.3	04-May-2015	ozaki-r	Fix configuration offset when MSI is enabled 24 is correct.
1.2	06-Dec-2013	yamt	branches: 1.2.4; 1.2.6; 1.2.10; virtio: move a comment to a more appropriate place no functional changes.
1.1	30-Oct-2011	hannken	branches: 1.1.2; 1.1.8; 1.1.14; 1.1.18; Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.1.18.1	18-May-2014	rmind	sync with head
1.1.14.2	03-Dec-2017	jdolecek	update from HEAD
1.1.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.8.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.1.8.1	30-Oct-2011	riz	file virtioreg.h was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.10.1	18-Jan-2017	skrll	Sync with netbsd-5
1.2.6.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.2.6.1	06-Jun-2015	skrll	Sync with HEAD
1.2.4.1	23-Dec-2016	snj	Pull up following revision(s) (requested by bsiegert in ticket #1338): sys/arch/amd64/conf/GENERIC: revision 1.418 sys/dev/pci/files.pci: revision 1.379 sys/dev/pci/vioscsi.c: revisions 1.1-1.8 sys/dev/pci/vioscsireg.h: revision 1.1 sys/dev/pci/virtioreg.h: revisions 1.5, 1.6 add the common flag bits -- Add vioscsi, compile tested only (toxic) -- this is working now, remove debugging. -- add vioscsi -- kill some more debugging. -- Fake mode sense data for illegal targets. qemu reports 256 targets... -- fix the status return. -- Use SCSIPI_CHAN_NOSETTLE. -- vioscsi_req_get()/virtio_enqueue_prep() failing is actually perfectly normal - observed failures included 10, 27, 61 in-flight commands, so probably depends on particular command mix; return with XS_RESOURCE_SHORTAGE rather then panic do vioscsi_req_put() when initial bus_dmamap_load() fails, as suggested by the XXX; the vq_done hook is called by virtio, but in that case we never get to commit the request to it -- react on ADAPTER_REQ_SET_XFER_MODE so that we set tagged queuing pass tag type and set id
1.6.32.1	03-Apr-2021	thorpej	Sync with HEAD.
1.11.2.2	31-Dec-2024	snj	Pull up following revision(s) (requested by martin in ticket #1032): sys/dev/pci/virtioreg.h: 1.13 sys/dev/pci/virtio_pci.c: 1.44 Define the VIRTIO_F_ACCESS_PLATFORM, VIRTIO_F_RING_PACKED, VIRTIO_F_ORDER_PLATFORM, and VIRTIO_F_SR_IOV feature bits. -- On alpha and sparc64, use VirtIO 1.0 and VIRTIO_F_ACCESS_PLATFORM so that DMA to PCI-attached VirtIO devices works properly. This is needed for Qemu to select the appropriate address space for PCI DMA.
1.11.2.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.13.2.1	02-Aug-2025	perseant	Sync with HEAD
1.30	26-Jul-2025	martin	PR 58681: add a port of the OpenBSD viogpu(4) driver written by Joshua Stein <jcs@openbsd.org>. Port by George Matsumura with help from Jiaxun Yang.
1.29	19-Apr-2023	yamaguchi	branches: 1.29.6; Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357
1.28	31-Mar-2023	yamaguchi	Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed().
1.27	31-Mar-2023	yamaguchi	Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler.
1.26	23-Mar-2023	yamaguchi	Added functions to set interrupt handler and index into virtqueue
1.25	23-Mar-2023	yamaguchi	Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function.
1.24	24-Mar-2022	andvar	branches: 1.24.4; s/interupt/interrupt/ in comments. also add missing RCS ID in bcm53xx_reg.h and fix one more typo in there.
1.23	28-Oct-2021	yamaguchi	virtio: stop reinit for safety when a device resetting is failed
1.22	21-Oct-2021	yamaguchi	virtio: reconfigure MSI-X vector on re-initialization This may fix a problem that some interrupts, for example link-state change of vioif(4) are not notified.
1.21	21-Oct-2021	yamaguchi	divide setup routine of virtio interrupts into establishment and device configuration
1.20	05-Feb-2021	reinoud	Second round of cleaning up endian code. No more tailored code to maintain.
1.19	28-Jan-2021	reinoud	Rewrite and streamline virtio device config read/write and explicitly cater for the Aarch64-eb bus problem with Qemu. This removes lots of bus_space `magic' and cleans up the code.
1.18	20-Jan-2021	reinoud	Remove the virtio child driver finalisation check KASSERT that, while it should never trigger, a possible bug in a child driver shouldn't have to panic the kernel. Instead report the internal error.
1.17	20-Jan-2021	reinoud	Add VirtIO PCI v1.0 attachments and fix the drivers affected. The vioif, ld, scsi, viornd and viomb devices were adjusted when needed and tested both in legacy 0.9 and v1.0 attachments trough PCI on amd64, sparc64, aarch64 and aarch64-eb. ACPI/FDT attachments also tested on aarch64/aarch64-eb. Known issues * viomb on aarch64 works only with ACPI/FDT attachment but not with PCI attachment. PCI and ACPI/FDT attachment works on aarch64-eb. * virtio on sparc64 attaches but is it not functioning though not a regression.
1.16	25-May-2020	yamaguchi	branches: 1.16.2; Set handlers implemented in child device of virtio(4) to virtqueue instead of the commonized function
1.15	25-May-2020	yamaguchi	Register an interrupt handler for virtqueue to that for pci to reduce function call
1.14	25-May-2020	yamaguchi	Remove VIRTIO_F_PCI_INTR_SOFTINT support for multiqueue The functionality is no longer used by obsolating VIOIF_SOFTINT_INTR.
1.13	14-Jan-2019	yamaguchi	Add multiqueue support, virtio(4)
1.12	15-Jun-2018	jakllsch	branches: 1.12.2; Add cfprint_t virtiobusprint() and use it for virtio_mmio.
1.11	02-Jun-2018	jakllsch	Begin to detangle virtio from its PCI attachment
1.10	02-Aug-2017	cherry	branches: 1.10.2; Reorg the virtio(4) config(9) code to be explicitly pci specific. We'll later use this for a non-pci virtio(4) usecase. ok martin
1.9	26-Mar-2017	martin	Backout previous, no need for these accessor in MD code
1.8	26-Mar-2017	martin	Provide an accessor for the pci attach args of a virtio device (if available - future non-pci attachments may return NULL). Sparc64 MD code needs this to match the OF node of the virtio device for boot device recognition.
1.7	25-Mar-2017	jdolecek	reorganize the attachment process for virtio child devices, so that more common code is shared among the drivers, and it's possible for the drivers to be correctly dynamically loaded; forbid direct access to struct virtio_softc from the child driver code
1.6	27-Sep-2016	pgoyette	branches: 1.6.2; Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.5	26-Oct-2015	ozaki-r	branches: 1.5.2; Support MSI-X in virtio Currently only vioif(4) uses the feature. knakahara@ helped to migrate to pci_intr_alloc(9). Thanks!
1.4	19-Dec-2014	ozaki-r	Implement softint-based interrupt handling in if_vioif Softint-based interrupt handling is considered as a future direction of the (network) device driver architecture in NetBSD. pq3etsec of ppc is already implemented based on the architecture (unlike pq3etsec, this change doesn't include softint-based if_start). In this architecture, a hardware interrupt handler just schedules a softint and the softint performs actual interrupt processing. It reduces processing in hardware interrupt context and allows Layer 2 network stack (e.g., bridge, vlan and even bpf) run in softint context, which makes it easy to implement fine-grain locking in the layer. This is an experimental implementation of the architecture in if_viof. virtio introduces a new flag VIRTIO_F_PCI_INTR_SOFTINT. If a driver of virtio sets it to sc_flags of virtio_softc, virtio calls softint_schedule in virtio_intr instead of directly calling the interrupt handler of the driver. When VIOIF_SOFTINT_INTR is on, vioif doesn't use the existing softint (vioif_rx_softint) that is called from vioif_rx_vq_done. Because vioif_rx_softint already runs in softint context and another softint isn't needed. This change actually improves performance in some cases. The feature is disabled by default and enabled when SOFTINT_INTR is set somewhere (normally in a kernel configuration).
1.3	22-Jul-2014	ozaki-r	branches: 1.3.4; Introduce VIRTIO_F_PCI_INTR_MPSAFE for virtio It is set by a child driver, e.g., if_vioif. If set, virtio sets PCI_INTR_MPSAFE for pci_intr_establish.
1.2	06-Dec-2013	yamt	branches: 1.2.2; virtio: move a comment to a more appropriate place no functional changes.
1.1	30-Oct-2011	hannken	branches: 1.1.2; 1.1.8; 1.1.14; 1.1.18; Import of the virtio driver written by MINOURA Makoto <minoura@netbsd.org> with minor changes to make it compile an run on -current. This driver speeds up disk and network access in virtual environments like KVM. Enabled on i386 and amd64. Tested with a CentOS 5.7 x86_64 host. See http://ozlabs.org/~rusty/virtio-spec/virtio.pdf for the specification.
1.1.18.1	18-May-2014	rmind	sync with head
1.1.14.2	03-Dec-2017	jdolecek	update from HEAD
1.1.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.8.2	25-Jan-2012	riz	Pull up following revision(s) (requested by hannken in ticket #1715): - Be robust against an invalid timer period value. sys/dev/ic/hpetreg.h Rev. 1.4 sys/dev/ic/hpet.c Rev. 1.8 - Fix wrong definition of LAPIC_LEVEL_ASSERT / _MASK sys/arch/x86/include/i82489reg.h Rev. 1.11 - Add virtio driver - speed up disk and network access in virtual environments sys/arch/i386/conf/GENERIC Rev. 1.1055 sys/arch/i386/conf/ALL Rev. 1.325 sys/arch/amd64/conf/GENERIC Rev. 1.338 sys/dev/pci/files.pci Rev. 1.350 sys/dev/pci/if_vioif.c Rev. 0-1.2 sys/dev/pci/ld_virtio.c Rev. 0-1.4 sys/dev/pci/viomb.c Rev. 0-1.1 sys/dev/pci/virtio.c Rev. 0-1.3 sys/dev/pci/virtioreg.h Rev. 0-1.1 sys/dev/pci/virtiovar.h Rev. 0-1.1 distrib/sets/lists/man/mi Rev. 1.1352 and 1.1358 share/man/man4/Makefile Rev. 1.573 and 1.575 share/man/man4/ld.4 Rev. 1.19 share/man/man4/virtio.4 Rev. 0-1.4 share/man/man4/vioif.4 Rev. 0-1.2 share/man/man4/viomb.4 Rev. 0-1.2 Allow NetBSD to run unmodified under Linux/kvm.
1.1.8.1	30-Oct-2011	riz	file virtiovar.h was added on branch netbsd-5 on 2012-01-25 21:18:15 +0000
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.2.1	10-Aug-2014	tls	Rebase.
1.3.4.4	28-Aug-2017	skrll	Sync with HEAD
1.3.4.3	05-Oct-2016	skrll	Sync with HEAD
1.3.4.2	27-Dec-2015	skrll	Sync with HEAD (as of 26th Dec)
1.3.4.1	06-Apr-2015	skrll	Sync with HEAD
1.5.2.2	26-Apr-2017	pgoyette	Sync with HEAD
1.5.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.6.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.10.2.2	18-Jan-2019	pgoyette	Synch with HEAD
1.10.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.12.2.1	10-Jun-2019	christos	Sync with HEAD
1.16.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.24.4.2	03-Jun-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #186): sys/dev/pci/virtio_pci.c: revision 1.41 sys/dev/pci/virtio_pci.c: revision 1.42 sys/dev/virtio/virtio_mmio.c: revision 1.10 sys/dev/pci/virtiovar.h: revision 1.29 sys/dev/pci/virtio.c: revision 1.75 sys/dev/pci/virtio.c: revision 1.76 sys/dev/pci/virtio.c: revision 1.77 sys/dev/pci/virtio.c: revision 1.78 virtio@pci: Fix assertion on detach. If the child never attached in the first place, it's OK for it to not have detached. XXX This should not be a set of flags; this should be a state enumeration, because some flags make no sense, like FINISHED\|FAILED. XXX This should not be asserted separately in each bus; there should be a single place in virtio.c to assert this, uniformly in all buses. PR kern/57357 Use enumeration for state of a child driver instead of flags and check its detaching by using sc->sc_child in virtio_softc pointed out by riastradh, thanks. fixes PR/57357 Fix not to allocate unnecessary descriptor fixes PR/57358 virtio(4): change variable name, nfc virtio(4): change members of struct vring_desc_extra before free a slot This prevents the following race condition. 1. Thread-A: calls virtio_dequeue_commit() and puts a slot into free descriptor chain in vq_free_slot() 2. Thread-B: calls virtio_enqueue_prep() and get the slot stored by Thread-A 3. Thread-B: calls virtio_enqueue_reserve() and changes desc_base and desc_free_idx for the slot 4. Thread-A: changes the same members updated by Thread-B reported by hannken, thanks.
1.24.4.1	13-May-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #139): sys/dev/pci/vioscsi.c: revision 1.31 sys/dev/pci/vio9p.c: revision 1.10 sys/dev/pci/vioscsi.c: revision 1.32 sys/dev/pci/vio9p.c: revision 1.11 sys/dev/pci/vioscsi.c: revision 1.33 sys/dev/pci/ld_virtio.c: revision 1.31 sys/dev/virtio/viocon.c: revision 1.6 sys/dev/pci/vioscsi.c: revision 1.34 sys/dev/pci/ld_virtio.c: revision 1.32 sys/dev/virtio/viocon.c: revision 1.7 sys/dev/virtio/viocon.c: revision 1.8 sys/dev/pci/vioscsi.c: revision 1.36 sys/dev/pci/virtioreg.h: revision 1.12 sys/dev/pci/viornd.c: revision 1.19 sys/dev/pci/virtio.c: revision 1.66 sys/dev/pci/virtio.c: revision 1.67 sys/dev/pci/virtio.c: revision 1.68 sys/dev/pci/if_vioif.c: revision 1.103 sys/dev/pci/virtio.c: revision 1.69 sys/dev/pci/if_vioif.c: revision 1.104 sys/dev/pci/virtio_pci.c: revision 1.40 sys/dev/virtio/virtio_mmio.c: revision 1.8 sys/dev/virtio/virtio_mmio.c: revision 1.9 sys/dev/pci/viomb.c: revision 1.14 sys/dev/pci/viomb.c: revision 1.15 sys/dev/pci/viomb.c: revision 1.17 sys/dev/pci/viornd.c: revision 1.20 sys/dev/pci/viornd.c: revision 1.21 sys/dev/pci/virtiovar.h: revision 1.25 sys/dev/pci/virtiovar.h: revision 1.26 sys/dev/pci/virtiovar.h: revision 1.27 sys/dev/pci/virtiovar.h: revision 1.28 sys/dev/pci/virtio.c: revision 1.70 sys/dev/pci/virtio.c: revision 1.71 sys/dev/pci/virtio.c: revision 1.72 sys/dev/pci/virtio.c: revision 1.73 sys/dev/pci/virtio.c: revision 1.74 sys/dev/pci/virtio_pci.c: revision 1.39 Set virtqueues in virtio_child_attach_finish The number of virtqueue maybe change in a part of VirtIO devices (e.g. vioif(4)). And it is fixed after negotiation of features. So the configuration is moved into the function. viocon(4): fix not to allocate unused virtqueue viocon(4) allocates 4 virtqueues but it only uses 2 (0 and 1) queues. Added functions to set interrupt handler and index into virtqueue Added check of pointer for allocated memory before release of resource Setup virtqueues after registering them to virtio_softc restore fetch of qsize. Mark as MPSAFE. virtio(4): Avoid name collision with global intrhand on sparc64. Pacifies -Werror=shadow. No functional change intended. Use PRIuBUSSIZE to print bus_size_t variables. virtio(4): Fix sizing of virtqueue allocation. vq->vq_avail[0].ring is a zero-length array, and thus sizeof is zero; likewise vq->vq_used[0].ring. Use vq->vq_avail[0].ring[0] and vq->vq_used[0].ring[0] to fix this and restore the previous allocation sizing logic. XXX We shouldn't use zero-length arrays here -- they are asking for trouble like this, and C99 has a standard way to express what we're actually trying to get at it, flexible array members. PR kern/57304 virtio(4): Use flexible array members, not zero-length arrays. This enables the compiler to detect sizeof mistakes like PR kern/57304. Use descriptor chain for free slots instead of vq_entry list Descriptors can be chained by themself. And descriptors added to avail ring or used ring are already chained. But it was not used for unused descriptors and another linked list structure named vq_entry was used. The chain is also used for unused descriptors to make virtio(4) simpler. Added flags to store status of attaching a virtio device This prevents a panic on reboot after a virtio device had called virtio_child_attach_failed(). Fix wrong variable names This fixes build errors in virtio_mmio.c
1.29.6.1	02-Aug-2025	perseant	Sync with HEAD
1.56	25-Sep-2022	thorpej	Remove unnecessary include of <sys/malloc.h>.
1.55	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.54	24-Apr-2021	thorpej	branches: 1.54.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.53	22-Dec-2019	thorpej	branches: 1.53.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.52	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.51	11-Jul-2016	msaitoh	branches: 1.51.16; 1.51.18; KNF. No functional change.
1.50	16-Sep-2015	macallan	prop_dictionary_get_bool() doesn't return FALSE if the property doesn't exist, so initialize things first. No more console stealing on sparc64.
1.49	24-Jul-2014	macallan	branches: 1.49.4; fix a bunch of bugs & tpyos: - keep track of registers being mapped or not to avoid double (un)mapping when switching between WSDISPLAYIO_MODE_DUMBFB and _MAPPED - distinguish between _DUMBFB and _MAPPED - actually map the framebuffer at offset 0 in _DUMBFB - set BUS_SPACE_MAP_PREFETCHABLE when mmap()ing video memory now Xorg + xf86-video-wsfb work on top of voodoofb
1.48	29-Mar-2014	christos	branches: 1.48.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.47	19-Nov-2013	macallan	(hopefully) shut up warnings about set-but-unused variables.
1.46	09-Oct-2013	macallan	support WSDISPLAYIO_GET_FBINFO
1.45	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.44	02-Jul-2013	joerg	Fix memset arguments.
1.43	09-Nov-2012	rkujawa	branches: 1.43.2; Differentiate between Voodoo 3 and Voodoo Banshee (MAX_CLOCK). Ok'd by macallan@.
1.42	09-Nov-2012	rkujawa	The voodoofb driver also works with Voodoo Banshee. Tested using card from martin@. Should be modified to take different MAX_CLOCK value into account.
1.41	23-Apr-2012	macallan	branches: 1.41.2; initialize a variable before using it found by uebayasi@ now why did gcc warn him on x86 but not me on macppc?
1.40	15-Mar-2012	macallan	we need to zap the glyph cache when re-initializing after X
1.39	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.38	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.37	25-Jan-2012	macallan	add a simple glyph cacheing scheme Glyphs which use the default attribute will be saved into offscreen video memory the first time they're used and subsequent uses will just copy them from there instead of rendering them every time. This should give a nice speedup especially on slow CPUs.
1.36	25-Jan-2012	macallan	sprinkle sc_ to make it consistent
1.35	22-Jan-2012	macallan	clear the screen when returning from X
1.34	19-Jan-2012	macallan	support anti-aliased fonts in 8 bit using host blits
1.33	18-Jan-2012	macallan	do some mode validation - for now reject everything above 1680x1200 until I find out why these don't work ( 1680x1200 works but not 1920x1080 ). Also use sync polarity info from videomode instead of guessing. Finally, if there is no useble preferred mode, be a little smarter trying to find the next best mode.
1.32	17-Jan-2012	macallan	we don't access video memory directly, no need to map it
1.31	17-Jan-2012	macallan	support DDC2, pick an appropriate video mode if a data are found
1.30	17-Jan-2012	macallan	get rid of HAVE_OPENFIRMWARE
1.29	11-Jan-2012	macallan	use rasops_init(0, 0)
1.28	08-Oct-2011	kiyohara	branches: 1.28.2; 1.28.6; Change get parameters(width,height...) from prop_dictionary, like genfb. tested on ofppc only.
1.27	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.26	25-Dec-2010	macallan	branches: 1.26.2; 1.26.4; remove some ancient, unused debug code
1.25	16-Dec-2010	cegger	ioctl: KNF switch-case, remove useless brackets
1.24	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.23	12-Oct-2010	macallan	make higher resolutions than 1024x768 work
1.22	04-May-2010	macallan	use alternate font if available ( for autogenerated box drawing characters )
1.21	26-Nov-2009	njoly	branches: 1.21.2; 1.21.4; Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.20	20-Aug-2009	macallan	use vcons_replay_msgbuf() when appropriate
1.19	06-May-2009	elad	Replace curlwp->l_cred with kauth_cred_get(). Mailing list reference: http://mail-index.netbsd.org/tech-kern/2009/05/05/msg005038.html
1.18	09-Jul-2008	joerg	branches: 1.18.8; - device/softc split
1.17	08-May-2008	macallan	branches: 1.17.2; 1.17.4; nuke clause 3
1.16	05-May-2008	jmcneill	branches: 1.16.2; Change autoconf message 'direct rendering on dev unsupported' to the more accurate 'drm at dev not configured'.
1.15	10-Apr-2008	cegger	branches: 1.15.2; 1.15.4; use aprint_*_dev and device_xname
1.14	24-Mar-2008	elad	Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. This and similar requests will be replaced Soon, but these two in particular should be pulled up to netbsd-4, and to allow doing that smoothly, we're first introducing the "issuser" version. Discussed with and okay macallan@.
1.13	27-Feb-2008	macallan	get rid of the last #ifdef macppc and use PCI_MAGIC_IO_RANGE to select at which offset we mmap PCI IO space
1.12	01-Dec-2007	ad	branches: 1.12.10; 1.12.14; Back out previous.
1.11	01-Dec-2007	ad	Noisy printfs.
1.10	12-Jun-2007	xtraeme	branches: 1.10.6; 1.10.8; 1.10.14; Fix a typo in voodoofb_drm_unmap(), requested by garbled. cvs: ----------------------------------------------------------------------
1.9	27-Mar-2007	jmcneill	Make this compatible with tdfxdrm.
1.8	04-Mar-2007	christos	branches: 1.8.2; 1.8.4; 1.8.6; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.7	20-Jan-2007	he	branches: 1.7.2; Unify the declaration of OF_interpret() between sparc{,64} and macppc by adding the "nargs" argument to the macppc version, and fix the macppc ports uses of OF_interpret() accordingly. Also move the declaration of OF_interpt() from macppc's autoconf.h to ofw/openfirm.h. This fixes the build of the macppc port. Approved by macallan@.
1.6	13-Jan-2007	cube	Complete initializer, and make it compile on (lesser?) archs that don't have OpenFirmware.
1.5	04-May-2006	simonb	branches: 1.5.4; 1.5.6; 1.5.10; 1.5.14; 1.5.16; 1.5.20; Add missing "return 0;" from WSDISPLAYIO_SMODE case in voodoofb_ioctl().
1.4	15-Apr-2006	jmmv	branches: 1.4.2; 1.4.4; Remove the getwschar and putwschar accessops from wsdisplay drivers as requested by uwe@. These were wrong because they were receiving an emulcookie yet they were accessops (thus having to receive an accesscookie). Instead, just handle the WSDISPLAYIO_{GET,PUT}WSCHAR ioctls from the driver's ioctl accessop. As this reduces the amount of code needed to handle these operations to two small functions in each driver, remove the WSDISPLAY_CHARFUNCS kernel option. Reviewed by, at least, uwe@ and macallan@. No objections in tech-kern@.
1.3	12-Apr-2006	macallan	whitespace nits ( no, perry didn't steal my account )
1.2	12-Apr-2006	jmmv	Add an extra cookie to the ioctl and mmap wsdisplay accessops that points to the screen on which they are being called. The driver cannot guess this by itself but it is needed to implement, at least, the getwschar and putwschar functions in the correct place. There are no functional changes yet. Tested on i386 (vga, vga_raster, machfb, vesafb), macppc and sparc64. Suggested and reviewed by macallan@.
1.1	11-Apr-2006	macallan	A driver for 3Dfx Voodoo3 graphics boards, may or may not work with Voodoo4, Voodoo5 or Banshee boards. So far it supports: - full acceleration in 8bit graphics mode - video mode switching - virtual consoles via vcons So far it hasn't been tested on anything else than macppc and even there it needs a hack to overload ofb. TODO: - test on i386 - don't hardcode video mode
1.4.4.3	01-Jun-2006	kardel	Sync with head.
1.4.4.2	22-Apr-2006	simonb	Sync with head.
1.4.4.1	15-Apr-2006	simonb	file voodoofb.c was added on branch simonb-timecounters on 2006-04-22 11:39:16 +0000
1.4.2.3	11-May-2006	elad	sync with head
1.4.2.2	19-Apr-2006	elad	sync with head.
1.4.2.1	15-Apr-2006	elad	file voodoofb.c was added on branch elad-kernelauth on 2006-04-19 03:25:38 +0000
1.5.20.2	24-Mar-2008	bouyer	Pull up following revision(s) (requested by elad in ticket #1107): sys/dev/pci/machfb.c: revision 1.51 sys/dev/pci/voodoofb.c: revision 1.14 Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. Discussed with and okay macallan@.
1.5.20.1	10-Mar-2007	bouyer	branches: 1.5.20.1.2; 1.5.20.1.6; Apply patch (requested by macallan in ticket #503): sys/dev/pci/voodoofb.c patch Add missing include; should fix macppc build.
1.5.20.1.6.1	24-Mar-2008	bouyer	Pull up following revision(s) (requested by elad in ticket #1107): sys/dev/pci/machfb.c: revision 1.51 sys/dev/pci/voodoofb.c: revision 1.14 Introduce two missing KAUTH_GENERIC_ISSUSER check in the voodoo and machfb mmap() code. Discussed with and okay macallan@.
1.5.20.1.2.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.5.16.2	09-Sep-2006	rpaulo	sync with head
1.5.16.1	04-May-2006	rpaulo	file voodoofb.c was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.5.14.1	01-Feb-2007	ad	Sync with head.
1.5.10.5	17-Mar-2008	yamt	sync with head.
1.5.10.4	03-Sep-2007	yamt	sync with head.
1.5.10.3	26-Feb-2007	yamt	sync with head.
1.5.10.2	21-Jun-2006	yamt	sync with head.
1.5.10.1	04-May-2006	yamt	file voodoofb.c was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.5.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.5.6.1	04-May-2006	tron	file voodoofb.c was added on branch peter-altq on 2006-05-24 15:50:28 +0000
1.5.4.2	24-May-2006	yamt	sync with head.
1.5.4.1	04-May-2006	yamt	file voodoofb.c was added on branch yamt-pdpolicy on 2006-05-24 10:58:01 +0000
1.7.2.2	15-Apr-2007	yamt	sync with head.
1.7.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.8.6.1	29-Mar-2007	reinoud	Pullup to -current
1.8.4.1	11-Jul-2007	mjf	Sync with head.
1.8.2.2	15-Jul-2007	ad	Sync with head.
1.8.2.1	10-Apr-2007	ad	Sync with head.
1.10.14.1	08-Dec-2007	mjf	Sync with HEAD.
1.10.8.1	23-Mar-2008	matt	sync with HEAD
1.10.6.1	03-Dec-2007	joerg	Sync with HEAD.
1.12.14.3	28-Sep-2008	mjf	Sync with HEAD.
1.12.14.2	02-Jun-2008	mjf	Sync with HEAD.
1.12.14.1	03-Apr-2008	mjf	Sync with HEAD.
1.12.10.1	24-Mar-2008	keiichi	sync with head.
1.15.4.6	11-Aug-2010	yamt	sync with head.
1.15.4.5	11-Mar-2010	yamt	sync with head
1.15.4.4	16-Sep-2009	yamt	sync with head
1.15.4.3	16-May-2009	yamt	sync with head
1.15.4.2	04-May-2009	yamt	sync with head.
1.15.4.1	16-May-2008	yamt	sync with head.
1.15.2.1	18-May-2008	yamt	sync with head.
1.16.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.16.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.17.4.1	19-Oct-2008	haad	Sync with HEAD.
1.17.2.1	18-Jul-2008	simonb	Sync with head.
1.18.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.21.4.2	05-Mar-2011	rmind	sync with head
1.21.4.1	30-May-2010	rmind	sync with head
1.21.2.2	22-Oct-2010	uebayasi	Sync with HEAD (-D20101022).
1.21.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.26.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.26.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.28.6.3	29-Apr-2012	mrg	sync to latest -current.
1.28.6.2	05-Apr-2012	mrg	sync to latest -current.
1.28.6.1	18-Feb-2012	mrg	merge to -current.
1.28.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.28.2.3	16-Jan-2013	yamt	sync with (a bit old) head
1.28.2.2	23-May-2012	yamt	sync with head.
1.28.2.1	17-Apr-2012	yamt	sync with head
1.41.2.3	03-Dec-2017	jdolecek	update from HEAD
1.41.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.41.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.43.2.2	18-May-2014	rmind	sync with head
1.43.2.1	28-Aug-2013	rmind	sync with head
1.48.2.1	10-Aug-2014	tls	Rebase.
1.49.4.2	05-Oct-2016	skrll	Sync with HEAD
1.49.4.1	22-Sep-2015	skrll	Sync with HEAD
1.51.18.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.51.18.1	10-Jun-2019	christos	Sync with HEAD
1.51.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.53.10.2	31-Mar-2021	thorpej	There are two interface attriubtues on this device, so we need to be explcit about it in BOTH config_found() calls.
1.53.10.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.54.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.4	21-Jan-2012	jakllsch	Silence "large integer implicitly truncated to unsigned type" warning on LP64.
1.3	19-Jan-2012	macallan	support anti-aliased fonts in 8 bit using host blits
1.2	18-Jan-2012	macallan	do some mode validation - for now reject everything above 1680x1200 until I find out why these don't work ( 1680x1200 works but not 1920x1080 ). Also use sync polarity info from videomode instead of guessing. Finally, if there is no useble preferred mode, be a little smarter trying to find the next best mode.
1.1	11-Apr-2006	macallan	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; 1.1.14; 1.1.20; 1.1.110; 1.1.114; A driver for 3Dfx Voodoo3 graphics boards, may or may not work with Voodoo4, Voodoo5 or Banshee boards. So far it supports: - full acceleration in 8bit graphics mode - video mode switching - virtual consoles via vcons So far it hasn't been tested on anything else than macppc and even there it needs a hack to overload ofb. TODO: - test on i386 - don't hardcode video mode
1.1.114.1	18-Feb-2012	mrg	merge to -current.
1.1.110.1	17-Apr-2012	yamt	sync with head
1.1.20.2	09-Sep-2006	rpaulo	sync with head
1.1.20.1	11-Apr-2006	rpaulo	file voodoofbreg.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:52:19 +0000
1.1.14.2	21-Jun-2006	yamt	sync with head.
1.1.14.1	11-Apr-2006	yamt	file voodoofbreg.h was added on branch yamt-lazymbuf on 2006-06-21 15:05:07 +0000
1.1.10.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.1.10.1	11-Apr-2006	tron	file voodoofbreg.h was added on branch peter-altq on 2006-05-24 15:50:28 +0000
1.1.8.2	24-May-2006	yamt	sync with head.
1.1.8.1	11-Apr-2006	yamt	file voodoofbreg.h was added on branch yamt-pdpolicy on 2006-05-24 10:58:01 +0000
1.1.4.2	22-Apr-2006	simonb	Sync with head.
1.1.4.1	11-Apr-2006	simonb	file voodoofbreg.h was added on branch simonb-timecounters on 2006-04-22 11:39:16 +0000
1.1.2.2	19-Apr-2006	elad	sync with head.
1.1.2.1	11-Apr-2006	elad	file voodoofbreg.h was added on branch elad-kernelauth on 2006-04-19 03:25:38 +0000
1.20	15-Sep-2025	thorpej	Encapsulate what's needed to attach an I2C bus into a iicbus_attach() inline.
1.19	15-Jul-2025	macallan	no need to provide our own i2c bus lock and acquire/releade methods anymore
1.18	25-Sep-2022	thorpej	branches: 1.18.10; Remove unnecessary include of <sys/malloc.h>.
1.17	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.16	24-Apr-2021	thorpej	branches: 1.16.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.15	22-Dec-2019	thorpej	branches: 1.15.10; Cleanup i2c bus acquire / release, centralizing all of the logic into iic_acquire_bus() / iic_release_bus(). "acquire" and "release" hooks no longer need to be provided by back-end controller drivers (only if they need special handling, e.g. powering on the i2c controller). This results in the removal of a bunch of rendundant code from each back-end controller driver. Assert that we are not in hard interrupt context in iic_acquire_bus(), iic_exec(), and iic_release_bus().
1.14	21-Apr-2019	maya	wether -> whether
1.13	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.12	16-Nov-2016	macallan	branches: 1.12.14; 1.12.16; pass clockframe aropund in voyager_intr()
1.11	01-Jan-2016	macallan	branches: 1.11.2; zero out struct i2cbus_attach_args and i2c_controller before messing with them
1.10	29-Mar-2014	christos	branches: 1.10.6; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.9	30-Jan-2012	drochner	branches: 1.9.6; 1.9.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.8	13-Dec-2011	macallan	only try to attach child drivers if they're actually present
1.7	07-Dec-2011	macallan	consolidate bit twiddling functions into one generic case and a bunch of macros, use them
1.6	20-Oct-2011	macallan	branches: 1.6.2; 1.6.6; include opt_voyager.h
1.5	18-Oct-2011	macallan	add functions to control gpio usage and to program PWM timers also add basic interrupt support ( not used yet )
1.4	06-Sep-2011	macallan	add a function to mess with gpios which locks against the i2c stuff
1.3	01-Sep-2011	macallan	some i2c devices we're going to deal with need a little delay after each write. Also make sure the gpios we're using are outputs.
1.2	01-Sep-2011	macallan	only attach the i2c bus if the gpio pins we're using are actually configured as gpio while there remove some accidentially left in debug code
1.1	31-Aug-2011	macallan	split up voyagerfb into the framebuffer portion and a pseudo bus that we can attach all the other sub-devices to while there attach an i2c bus
1.6.6.1	18-Feb-2012	mrg	merge to -current.
1.6.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6.2.1	17-Apr-2012	yamt	sync with head
1.9.10.1	18-May-2014	rmind	sync with head
1.9.6.2	03-Dec-2017	jdolecek	update from HEAD
1.9.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.10.6.2	05-Dec-2016	skrll	Sync with HEAD
1.10.6.1	19-Mar-2016	skrll	Sync with HEAD
1.11.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.12.16.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.12.16.1	10-Jun-2019	christos	Sync with HEAD
1.12.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.15.10.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.16.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.18.10.1	02-Aug-2025	perseant	Sync with HEAD
1.10	31-Aug-2011	macallan	the framebuffer part lives on in voyager/voyagerfb.c
1.9	30-Aug-2011	macallan	contrary to the manual the right-to-left flag also causes the blitter to work bottom-to-top Now all copy*() methods should work properly in all directions.
1.8	30-Aug-2011	macallan	fix thinko in downward scrolling ( never had a chance to test it until now... )
1.7	22-Jan-2011	cegger	Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.6	23-Dec-2010	cegger	branches: 1.6.2; 1.6.4; read 32bits instead of 64bits and truncate the upper 32bits
1.5	16-Dec-2010	cegger	ioctl: KNF switch-case, remove useless brackets
1.4	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.3	04-May-2010	macallan	use alternate font if available ( for autogenerated box drawing characters )
1.2	20-Aug-2009	macallan	branches: 1.2.2; 1.2.4; use vcons_replay_msgbuf() when appropriate
1.1	12-Aug-2009	macallan	branches: 1.1.2; add voyagerfb, a driver for the Gdium Liberty 1000's onboard graphics chip
1.1.2.4	11-Aug-2010	yamt	sync with head.
1.1.2.3	16-Sep-2009	yamt	sync with head
1.1.2.2	19-Aug-2009	yamt	sync with head.
1.1.2.1	12-Aug-2009	yamt	file voyagerfb.c was added on branch yamt-nfs-mp on 2009-08-19 18:47:18 +0000
1.2.4.2	05-Mar-2011	rmind	sync with head
1.2.4.1	30-May-2010	rmind	sync with head
1.2.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.6.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.6.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.7	16-Apr-2020	rin	Revert previous for now: http://mail-index.netbsd.org/source-changes/2020/04/16/msg116278.html The reasoning turned out to be wrong; __KERNEL_RCSID() in header files does not overwrite RCSID in main source files. The real problem is that it inserts its RCSID into every object files. However, it can be still useful even if heavily duplicated.
1.6	16-Apr-2020	rin	Stop using __KERNEL_RCSID() in header files; it confuses ident(1) by overwriting RCSID in main source files. XXX The first argument of __KERNEL_RCSID() is neglected for ELF. If we wish to have RCSID of header files in kernel binary, we need something like __FBSDID() macro in FreeBSD.
1.5	19-Apr-2018	christos	branches: 1.5.14; s/static inline/static __inline/g for consistency.
1.4	07-Dec-2011	macallan	branches: 1.4.46; consolidate bit twiddling functions into one generic case and a bunch of macros, use them
1.3	18-Oct-2011	macallan	branches: 1.3.2; 1.3.6; add functions to control gpio usage and to program PWM timers also add basic interrupt support ( not used yet )
1.2	06-Sep-2011	macallan	add a function to mess with gpios which locks against the i2c stuff
1.1	31-Aug-2011	macallan	split up voyagerfb into the framebuffer portion and a pseudo bus that we can attach all the other sub-devices to while there attach an i2c bus
1.3.6.1	18-Feb-2012	mrg	merge to -current.
1.3.2.1	17-Apr-2012	yamt	sync with head
1.4.46.1	22-Apr-2018	pgoyette	Sync with HEAD
1.5.14.1	20-Apr-2020	bouyer	Sync with HEAD
1.22	04-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "is_console".
1.21	17-Jul-2022	riastradh	wsdisplay(4): Factor out WSDISPLAY_SCROLLSUPPORT logic. Should find a way to avoid the #ifdefs in the .h file, but this makes the code a good deal more legible and easier to maitain, at least. No functional change intended.
1.20	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.19	24-Apr-2021	thorpej	branches: 1.19.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.18	01-Jun-2017	chs	branches: 1.18.26; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.17	24-Mar-2017	macallan	support screen blanking
1.16	24-Mar-2017	macallan	missed one line of debug spam...
1.15	24-Mar-2017	macallan	- clean up some testing & debugging goop - 0xff is transparent, so restrict ourself to just 4 bit colour, all we need for wscons - support WSDISPLAYIO_GET_FBINFO etc. for wsfb, allow mmap()ing the fb etc.
1.14	14-Jul-2016	msaitoh	branches: 1.14.2; 1.14.4; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.13	20-Mar-2015	macallan	fix copyright & license, make it explicit that this driver is based on OpenBSD's ifb driver
1.12	09-Oct-2012	macallan	branches: 1.12.14; add acceleration support, for Sun XVR-1200 and compatible only so far adapted from OpenBSD with various improvements, no support for Expert3D has been added this is still suboptimal since we have to sync the engine all the time but it's an improvement over software-only operation
1.11	13-Mar-2012	elad	branches: 1.11.2; Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.10	30-Jan-2012	drochner	Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.9	11-Jan-2012	macallan	use rasops_init(0, 0)
1.8	22-Jan-2011	cegger	branches: 1.8.4; 1.8.8; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.7	16-Dec-2010	cegger	branches: 1.7.2; 1.7.4; ioctl: KNF switch-case, remove useless brackets
1.6	10-Mar-2010	macallan	branches: 1.6.2; 1.6.4; 1.6.6; - move register definition in their own header, reduce magic numbers - get geometry and framebuffer layout from the chip instead of hardcoding - get rid of some now superfluous leftovers - remove some debug code - clean things up a bit - dump registers with WCFB_DEBUG even if we're not the console
1.5	10-Mar-2010	macallan	forgot to include opt_wsdisplay_compat.h for WSDISPLAY_SCROLLSUPPORT
1.4	09-Mar-2010	macallan	now we also draw a cursor
1.3	09-Mar-2010	macallan	Use a shadow framebuffer instead of doing every operation twice ( once for each framebuffer ) - this gives quite a dramatic speedup and hides the funky effects previously seen. Almost there, now we need to actually draw a cursor.
1.2	25-Feb-2010	macallan	some cleanup
1.1	25-Feb-2010	macallan	a very preliminary driver for Wildcat 5110 / Sun XVR-500 graphics boards Since we don't have any documentationthis driver is unaccelerated and thanks to the insane state we get the hardware in it's also slow and scrolling looks weird. Some hardware info came from OpenBSD's ifb driver.
1.6.6.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.6.6.1	10-Mar-2010	uebayasi	file wcfb.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.6.4.1	05-Mar-2011	rmind	sync with head
1.6.2.2	11-Mar-2010	yamt	sync with head
1.6.2.1	10-Mar-2010	yamt	file wcfb.c was added on branch yamt-nfs-mp on 2010-03-11 15:03:59 +0000
1.7.4.1	08-Feb-2011	bouyer	Sync with HEAD
1.7.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.8.8.2	05-Apr-2012	mrg	sync to latest -current.
1.8.8.1	18-Feb-2012	mrg	merge to -current.
1.8.4.2	30-Oct-2012	yamt	sync with head
1.8.4.1	17-Apr-2012	yamt	sync with head
1.11.2.2	03-Dec-2017	jdolecek	update from HEAD
1.11.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.12.14.3	28-Aug-2017	skrll	Sync with HEAD
1.12.14.2	05-Oct-2016	skrll	Sync with HEAD
1.12.14.1	06-Apr-2015	skrll	Sync with HEAD
1.14.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.14.2.1	26-Apr-2017	pgoyette	Sync with HEAD
1.18.26.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.19.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.2	09-Oct-2012	macallan	add acceleration support, for Sun XVR-1200 and compatible only so far adapted from OpenBSD with various improvements, no support for Expert3D has been added this is still suboptimal since we have to sync the engine all the time but it's an improvement over software-only operation
1.1	10-Mar-2010	macallan	branches: 1.1.2; 1.1.6; 1.1.14; 1.1.24; - move register definition in their own header, reduce magic numbers - get geometry and framebuffer layout from the chip instead of hardcoding - get rid of some now superfluous leftovers - remove some debug code - clean things up a bit - dump registers with WCFB_DEBUG even if we're not the console
1.1.24.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.1.14.1	30-Oct-2012	yamt	sync with head
1.1.6.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.6.1	10-Mar-2010	uebayasi	file wcfbreg.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.1.2.2	11-Mar-2010	yamt	sync with head
1.1.2.1	10-Mar-2010	yamt	file wcfbreg.h was added on branch yamt-nfs-mp on 2010-03-11 15:03:59 +0000
1.3	10-Nov-1998	thorpej	File is obsolete.
1.2	08-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.1	04-Mar-1998	cgd	PCI IDE glue. Right now, just glues 'wdc's to PCI IDE controller channels. Eventually should do things like support PCI IDE DMA (it _DOES NOT_ do that now).
1.16	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.15	17-Oct-2013	christos	branches: 1.15.6; - remove unused variables - move debugging code inside debugging sections
1.14	18-Mar-2009	cegger	branches: 1.14.12; 1.14.22; 1.14.26; bzero -> memset
1.13	09-Jul-2008	joerg	branches: 1.13.2; 1.13.8; - device/softc split
1.12	28-Apr-2008	martin	branches: 1.12.2; 1.12.4; 1.12.6; Remove clause 3 and 4 from TNF licenses
1.11	10-Apr-2008	cegger	branches: 1.11.2; 1.11.4; use aprint_*_dev and device_xname
1.10	19-Oct-2007	ad	branches: 1.10.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.9	16-Nov-2006	christos	branches: 1.9.8; 1.9.22; 1.9.24; 1.9.28; __unused removal on arguments; approved by core.
1.8	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.7	11-Dec-2005	christos	branches: 1.7.20; 1.7.22; merge ktrace-lwp.
1.6	28-Jun-2005	thorpej	branches: 1.6.2; Use ANSI function decls and static.
1.5	27-Feb-2005	perry	nuke trailing whitespace
1.4	02-Oct-2002	thorpej	branches: 1.4.6; 1.4.14; 1.4.16; Add trailing ; to CFATTACH_DECL.
1.3	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.2	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.1	01-Jan-2002	hpeyerl	branches: 1.1.2; 1.1.4; Watchdog back-end driver for PC Weasel/PCI serial console and remote management board. Written by hpeyerl@realweasel.com and thorpej@wasabisystems.com. Requires Weasel firmware v2.04 and later.
1.1.4.3	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.1.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.1	01-Jan-2002	thorpej	file weasel_pci.c was added on branch kqueue on 2002-01-10 19:57:09 +0000
1.1.2.3	18-Oct-2002	nathanw	Catch up to -current.
1.1.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.1	01-Jan-2002	nathanw	file weasel_pci.c was added on branch nathanw_sa on 2002-01-08 00:31:19 +0000
1.4.16.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.14.1	29-Apr-2005	kent	sync with -current
1.4.6.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.6.1	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.6.2.2	27-Oct-2007	yamt	sync with head.
1.6.2.1	30-Dec-2006	yamt	sync with head.
1.7.22.2	10-Dec-2006	yamt	sync with head.
1.7.22.1	22-Oct-2006	yamt	sync with head
1.7.20.1	18-Nov-2006	ad	Sync with head.
1.9.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.9.24.1	06-Nov-2007	matt	sync with HEAD
1.9.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.9.8.1	23-Oct-2007	ad	Sync with head.
1.10.16.2	28-Sep-2008	mjf	Sync with HEAD.
1.10.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.11.4.2	04-May-2009	yamt	sync with head.
1.11.4.1	16-May-2008	yamt	sync with head.
1.11.2.1	18-May-2008	yamt	sync with head.
1.12.6.1	19-Oct-2008	haad	Sync with HEAD.
1.12.4.1	18-Jul-2008	simonb	Sync with head.
1.12.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.13.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.13.2.1	28-Apr-2009	skrll	Sync with HEAD.
1.14.26.1	18-May-2014	rmind	sync with head
1.14.22.2	03-Dec-2017	jdolecek	update from HEAD
1.14.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.14.12.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.15.6.1	09-Jul-2016	skrll	Sync with HEAD
1.2	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.1	01-Jan-2002	hpeyerl	branches: 1.1.2; 1.1.4; 1.1.118; 1.1.120; 1.1.122; Watchdog back-end driver for PC Weasel/PCI serial console and remote management board. Written by hpeyerl@realweasel.com and thorpej@wasabisystems.com. Requires Weasel firmware v2.04 and later.
1.1.122.1	16-May-2008	yamt	sync with head.
1.1.120.1	18-May-2008	yamt	sync with head.
1.1.118.1	02-Jun-2008	mjf	Sync with HEAD.
1.1.4.2	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.4.1	01-Jan-2002	thorpej	file weaselreg.h was added on branch kqueue on 2002-01-10 19:57:09 +0000
1.1.2.2	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.1	01-Jan-2002	nathanw	file weaselreg.h was added on branch nathanw_sa on 2002-01-08 00:31:19 +0000
1.1	22-Jan-2011	cegger	branches: 1.1.2; 1.1.4; 1.1.8; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.1.8.2	06-Jun-2011	jruoho	Sync with HEAD.
1.1.8.1	22-Jan-2011	jruoho	file wsdisplay_pci.c was added on branch jruoho-x86intr on 2011-06-06 09:08:28 +0000
1.1.4.2	05-Mar-2011	rmind	sync with head
1.1.4.1	22-Jan-2011	rmind	file wsdisplay_pci.c was added on branch rmind-uvmplock on 2011-03-05 20:53:59 +0000
1.1.2.2	08-Feb-2011	bouyer	Sync with HEAD
1.1.2.1	22-Jan-2011	bouyer	file wsdisplay_pci.c was added on branch bouyer-quota2 on 2011-02-08 16:19:53 +0000
1.1	22-Jan-2011	cegger	branches: 1.1.2; 1.1.4; 1.1.8; Implement new WSDISPLAYIO_GET_BUSID ioctl. It returns the bus id and allows userland (like Xorg) to create mapping of ttyE? and bus id. For now only PCI is implemented. First discussed with macallan@ then public on tech-kern@ and tech-x11@
1.1.8.2	06-Jun-2011	jruoho	Sync with HEAD.
1.1.8.1	22-Jan-2011	jruoho	file wsdisplay_pci.h was added on branch jruoho-x86intr on 2011-06-06 09:08:29 +0000
1.1.4.2	05-Mar-2011	rmind	sync with head
1.1.4.1	22-Jan-2011	rmind	file wsdisplay_pci.h was added on branch rmind-uvmplock on 2011-03-05 20:53:59 +0000
1.1.2.2	08-Feb-2011	bouyer	Sync with HEAD
1.1.2.1	22-Jan-2011	bouyer	file wsdisplay_pci.h was added on branch bouyer-quota2 on 2011-02-08 16:19:53 +0000
1.33	24-Jan-2023	mlelstv	Consistently use cached chipset tag value.
1.32	28-Oct-2022	riastradh	branches: 1.32.2; xhci(4): Mark PCI interrupt handler MP-safe. xhci_intr has its own intr lock to coordinate with the MP-safe softint it defers all its work to, other than reading and writing a few registers to get and acknowledge the interrupt status.
1.31	11-Oct-2022	msaitoh	There is an xHCI device which has USB 2 port only. Support it. - Example: xhci4 at pci17 dev 0 function 0: AMD product 15b8 (rev. 0x00) xhci4: 64-bit DMA allocated pic msix10 type edge pin 0 level 6 to cpu0 slot 32 idt entry 107 xhci4: interrupting at msix10 vec 0 xhci4: xHCI version 1.20 xhci4: hcs1=1000840 hcs2=140000f1 hcs3=7000a xhci4: hcc=0x110ffc5<XECP=0x110,MAXPSA=0xf,CFC,SEC,SPC,PAE,NSS,LTC,CSZ,AC64> xhci4: xECP 440 xhci4: hcc2=0x3f<CIC,LEC,CTC,FSC,CMC,U3C> xhci4: ECR: 0x00000401 xhci4: ECR: 0x02000402 xhci4: SP: 0x02000402 0x20425355 0x00180101 0x00000000 xhci4: hs ports 1 - 1 xhci4: ECR: 0x000f000a xhci4: PAGESIZE 0x00000001 xhci4: sc_pgsz 0x00001000 xhci4: sc_maxslots 0x00000040 xhci4: sc_maxports 1 xhci4: sc_maxspbuf 2 xhci4: eventst: 0x000000013ee60fc0 0xffffb08826f5afc0 1000 xhci4: dcbaa: 0x000000013ee63000 0xffffb08826f5b000 1000 xhci4: current IMOD 0 (snip) usb8 at xhci4: USB revision 3.1 usb9 at xhci4: USB revision 2.0 uhub8 at usb8: NetBSD (0x0000) xHCI root hub (0x0000), class 9/0, rev 3.00/1.00, addr 0 uhub8: 0 ports with 0 removable, self powered uhub8: no ports, hub ignored uhub8: WARNING: power management not supported autoconfiguration error: usb8: root device is not a hub usb8: WARNING: power management not supported uhub9 at usb9: NetBSD (0x0000) xHCI root hub (0x0000), class 9/0, rev 2.00/1.00, addr 0 uhub9: 1 port with 1 removable, self powered - To resolve this problem, keep number of ports of SS and HS and use it to attach child device(s). - Tested on ASUS TUF GAMING X670E-PLUS. - OK'd by skrll@.
1.30	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.29	24-Apr-2021	thorpej	branches: 1.29.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.28	07-Feb-2021	ryoon	branches: 1.28.2; Allow MSIX and MSI for interrupt type
1.27	07-Feb-2021	dbj	xhci_pci.c: reimplement XHCI_DISABLE_MSI and XHCI_DISABLE_MSIX
1.26	09-Dec-2020	jakllsch	xhci_pci: avoid potential double free of interrupt handles Found by Kouichi Hashikawa in PR 55855.
1.25	27-Oct-2020	skrll	branches: 1.25.2; Report HW DMA capability and if the DMA tag limits it.
1.24	02-Dec-2019	msaitoh	Use PCI_MSIX_"TBL"BIR_MASK instead of PCI_MSIX_"PBA"BIR_MASK for MSI-X table. This is not a real bug because both macros have the same value.
1.23	14-Nov-2019	msaitoh	Revert previous. Sorry.
1.22	14-Nov-2019	msaitoh	Add D-Link DGE-530T C1 and TP-Link TG-3468 v2.
1.21	23-Jan-2019	msaitoh	branches: 1.21.4; Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them expilicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.20	18-Jan-2019	skrll	Default to USBREV_3_0 if usbrev is unknown.
1.19	07-Dec-2018	msaitoh	xHCI uses BAR0 to map memory space. If a chip has MSI-X capability and it points to BAR0, reserve the space. Almost the same as nvme_pci.c. XXX Is it possible to make a nice API?
1.18	30-Nov-2018	jdolecek	simplify intr establish code - rely on pci_intr_alloc() to allow also MSI-X, and to return interrupt types which are possible for pci_intr_establish(); remove fallbacks to retry with MSI/MSI-X explicitly disabled discussed on tech-kern@ https://mail-index.netbsd.org/tech-kern/2018/11/27/msg024240.html
1.17	24-Nov-2018	skrll	Fixup the retry loop wrt with XHCI_DISABLE_MSI{,X}
1.16	15-Nov-2018	jdolecek	remove #ifdef DEBUG printf, it seems to have outlived it's usefulness
1.15	31-Oct-2018	jmcneill	Add MSI-X support.
1.14	18-Sep-2018	mrg	deal with partial attach failures in usb_attach vs usb_detach aka PR 53598. - make sure xhci's sc->sc_ios is NULL if failure happens. - rearrange usb_attach() / usb_doattach() to make it simpler to clean up. - move usb_async_intr softint into usb_once_init(). previously, each USB controller would start a new one, and leave the old one leaked. - handle controller interrupts without a bus attached
1.13	29-Jun-2018	msaitoh	Detect USB 3.1.
1.12	09-Apr-2018	jakllsch	branches: 1.12.2; Stop potential misuse of vendor names and USB vendor IDs in root hub device and string descriptors. Firstly: Few vendors have identical PCI-SIG vendor IDs and USB-IF vendor IDs. As such, using the PCI vendor ID as a USB vendor ID may trample on whomever is allocated that USB vendor ID. Secondly: The vendor of the host controller hardware implementation has little to nothing to do with our usbroothub implementation. Thus we should not potentially associate any problems therewith to such third party. This change will result in root hubs being identified by USB Vendor ID 0x0000. Root hub vendor string will now be "NetBSD" (or, specifically: ostype). Product ID (0x0000) and product strings remain unchanged.
1.11	28-Dec-2017	msaitoh	branches: 1.11.2; - Fix panic in xhci_pci_detach() if xhci_pci_attach() failed. - Fallback from MSI to INTx correctly.
1.10	25-Dec-2017	msaitoh	- Fix a panic while cleaning PCI interrupt. - Fallback to INTx if MSI allocation succeeded but the establish failed.
1.9	05-Sep-2017	skrll	Add an XHCI_DISABLE_MSI option
1.8	19-Jan-2017	skrll	branches: 1.8.6; Pull across xhci(4) improvemnts from nick-nhusb
1.7	13-Oct-2016	jdolecek	branches: 1.7.2; provide intr xname
1.6	03-May-2016	skrll	branches: 1.6.2; Kill XHCI_QUIRK_FORCE_INTR and surrounding code
1.5	23-Apr-2016	skrll	Merge nick-nhusb - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix
1.4	21-Sep-2014	christos	branches: 1.4.2; Merge the 3 copies of devlist2h.awk that deal with 16 bit key and value pairs to the compressed one that matt wrote.
1.3	29-Mar-2014	christos	branches: 1.3.4; 1.3.6; 1.3.8; 1.3.10; 1.3.14; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.2	10-Mar-2014	skrll	Don't abuse usbd_status in xhci_init
1.1	14-Sep-2013	jakllsch	branches: 1.1.2; Add work-in-progress xhci(4) driver code. Currently (mostly) supports interrupt-driven control, interrupt and bulk transfers at the three USB 2.0 speeds on root hub ports.
1.1.2.2	05-Nov-2013	matt	Pull down xhci support from HEAD
1.1.2.1	14-Sep-2013	matt	file xhci_pci.c was added on branch matt-nb5-mips64 on 2013-11-05 18:36:31 +0000
1.3.14.2	26-Jan-2017	skrll	Sync with HEAD/nhusb
1.3.14.1	06-Sep-2016	skrll	First pass at netbsd-7 updated with USB code from HEAD
1.3.10.3	03-Dec-2017	jdolecek	update from HEAD
1.3.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.3.10.1	29-Mar-2014	tls	file xhci_pci.c was added on branch tls-maxphys on 2014-08-20 00:03:48 +0000
1.3.8.1	05-Apr-2017	snj	Pull up following revision(s) (requested by skrll in ticket #1395): share/man/man4/axe.4: netbsd-7-nhusb share/man/man4/axen.4: netbsd-7-nhusb share/man/man4/cdce.4: netbsd-7-nhusb share/man/man4/uaudio.4: netbsd-7-nhusb share/man/man4/ucom.4: netbsd-7-nhusb share/man/man4/uep.4: netbsd-7-nhusb share/man/man4/urtw.4: netbsd-7-nhusb share/man/man4/usb.4: netbsd-7-nhusb share/man/man4/uyap.4: netbsd-7-nhusb share/man/man4/xhci.4: netbsd-7-nhusb share/man/man9/usbdi.9: netbsd-7-nhusb sys/arch/amd64/conf/ALL: netbsd-7-nhusb sys/arch/amd64/conf/GENERIC: netbsd-7-nhusb sys/arch/amiga/dev/slhci_zbus.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_otg.c: netbsd-7-nhusb sys/arch/arm/allwinner/awin_usb.c: netbsd-7-nhusb sys/arch/arm/amlogic/amlogic_dwctwo.c: netbsd-7-nhusb sys/arch/arm/at91/at91ohci.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm2835_dwctwo.c: netbsd-7-nhusb sys/arch/arm/broadcom/bcm53xx_usb.c: netbsd-7-nhusb sys/arch/arm/ep93xx/epohci.c: netbsd-7-nhusb sys/arch/arm/gemini/obio_ehci.c: netbsd-7-nhusb sys/arch/arm/imx/files.imx23: netbsd-7-nhusb sys/arch/arm/imx/imxusb.c: netbsd-7-nhusb sys/arch/arm/imx/imxusbreg.h: netbsd-7-nhusb sys/arch/arm/omap/obio_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/omap3_ehci.c: netbsd-7-nhusb sys/arch/arm/omap/omapl1x_ohci.c: netbsd-7-nhusb sys/arch/arm/omap/tiotg.c: netbsd-7-nhusb sys/arch/arm/s3c2xx0/ohci_s3c24x0.c: netbsd-7-nhusb sys/arch/arm/samsung/exynos_usb.c: netbsd-7-nhusb sys/arch/arm/xscale/pxa2x0_ohci.c: netbsd-7-nhusb sys/arch/arm/zynq/zynq_usb.c: netbsd-7-nhusb sys/arch/hpcarm/dev/nbp_slhci.c: netbsd-7-nhusb sys/arch/hpcmips/dev/plumohci.c: netbsd-7-nhusb sys/arch/i386/conf/ALL: netbsd-7-nhusb sys/arch/i386/conf/GENERIC: netbsd-7-nhusb sys/arch/i386/pci/gcscehci.c: netbsd-7-nhusb sys/arch/luna68k/conf/GENERIC: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahci.c: netbsd-7-nhusb sys/arch/mips/adm5120/dev/ahcivar.h: netbsd-7-nhusb sys/arch/mips/alchemy/dev/ohci_aubus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ehci_arbus.c: netbsd-7-nhusb sys/arch/mips/atheros/dev/ohci_arbus.c: netbsd-7-nhusb sys/arch/mips/conf/files.adm5120: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ehci.c: netbsd-7-nhusb sys/arch/mips/ralink/ralink_ohci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ehci.c: netbsd-7-nhusb sys/arch/mips/rmi/rmixl_ohci.c: netbsd-7-nhusb sys/arch/playstation2/dev/ohci_sbus.c: netbsd-7-nhusb sys/arch/powerpc/booke/dev/pq3ehci.c: netbsd-7-nhusb sys/arch/powerpc/ibm4xx/dev/dwctwo_plb.c: netbsd-7-nhusb sys/arch/x68k/dev/slhci_intio.c: netbsd-7-nhusb sys/conf/files: netbsd-7-nhusb sys/dev/cardbus/ehci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/ohci_cardbus.c: netbsd-7-nhusb sys/dev/cardbus/uhci_cardbus.c: netbsd-7-nhusb sys/dev/ic/sl811hs.c: netbsd-7-nhusb sys/dev/ic/sl811hsvar.h: netbsd-7-nhusb sys/dev/isa/slhci_isa.c: netbsd-7-nhusb sys/dev/marvell/ehci_mv.c: netbsd-7-nhusb sys/dev/pci/ehci_pci.c: netbsd-7-nhusb sys/dev/pci/ohci_pci.c: netbsd-7-nhusb sys/dev/pci/uhci_pci.c: netbsd-7-nhusb sys/dev/pci/xhci_pci.c: netbsd-7-nhusb sys/dev/pcmcia/slhci_pcmcia.c: netbsd-7-nhusb sys/dev/usb/Makefile.usbdevs: netbsd-7-nhusb sys/dev/usb/TODO: netbsd-7-nhusb sys/dev/usb/TODO.usbmp: netbsd-7-nhusb sys/dev/usb/aubtfwl.c: netbsd-7-nhusb sys/dev/usb/auvitek.c: netbsd-7-nhusb sys/dev/usb/auvitek_audio.c: netbsd-7-nhusb sys/dev/usb/auvitek_dtv.c: netbsd-7-nhusb sys/dev/usb/auvitek_i2c.c: netbsd-7-nhusb sys/dev/usb/auvitek_video.c: netbsd-7-nhusb sys/dev/usb/auvitekvar.h: netbsd-7-nhusb sys/dev/usb/ehci.c: netbsd-7-nhusb sys/dev/usb/ehcireg.h: netbsd-7-nhusb sys/dev/usb/ehcivar.h: netbsd-7-nhusb sys/dev/usb/emdtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_dtv.c: netbsd-7-nhusb sys/dev/usb/emdtv_ir.c: netbsd-7-nhusb sys/dev/usb/emdtvvar.h: netbsd-7-nhusb sys/dev/usb/ezload.c: netbsd-7-nhusb sys/dev/usb/ezload.h: netbsd-7-nhusb sys/dev/usb/files.usb: netbsd-7-nhusb sys/dev/usb/hid.c: netbsd-7-nhusb sys/dev/usb/hid.h: netbsd-7-nhusb sys/dev/usb/if_athn_usb.c: netbsd-7-nhusb sys/dev/usb/if_athn_usb.h: netbsd-7-nhusb sys/dev/usb/if_atu.c: netbsd-7-nhusb sys/dev/usb/if_atureg.h: netbsd-7-nhusb sys/dev/usb/if_aue.c: netbsd-7-nhusb sys/dev/usb/if_auereg.h: netbsd-7-nhusb sys/dev/usb/if_axe.c: netbsd-7-nhusb sys/dev/usb/if_axen.c: netbsd-7-nhusb sys/dev/usb/if_axenreg.h: netbsd-7-nhusb sys/dev/usb/if_axereg.h: netbsd-7-nhusb sys/dev/usb/if_cdce.c: netbsd-7-nhusb sys/dev/usb/if_cdcereg.h: netbsd-7-nhusb sys/dev/usb/if_cue.c: netbsd-7-nhusb sys/dev/usb/if_cuereg.h: netbsd-7-nhusb sys/dev/usb/if_kue.c: netbsd-7-nhusb sys/dev/usb/if_kuereg.h: netbsd-7-nhusb sys/dev/usb/if_otus.c: netbsd-7-nhusb sys/dev/usb/if_otusvar.h: netbsd-7-nhusb sys/dev/usb/if_rum.c: netbsd-7-nhusb sys/dev/usb/if_rumreg.h: netbsd-7-nhusb sys/dev/usb/if_rumvar.h: netbsd-7-nhusb sys/dev/usb/if_run.c: netbsd-7-nhusb sys/dev/usb/if_runvar.h: netbsd-7-nhusb sys/dev/usb/if_smsc.c: netbsd-7-nhusb sys/dev/usb/if_smscreg.h: netbsd-7-nhusb sys/dev/usb/if_smscvar.h: netbsd-7-nhusb sys/dev/usb/if_udav.c: netbsd-7-nhusb sys/dev/usb/if_udavreg.h: netbsd-7-nhusb sys/dev/usb/if_upgt.c: netbsd-7-nhusb sys/dev/usb/if_upgtvar.h: netbsd-7-nhusb sys/dev/usb/if_upl.c: netbsd-7-nhusb sys/dev/usb/if_ural.c: netbsd-7-nhusb sys/dev/usb/if_uralreg.h: netbsd-7-nhusb sys/dev/usb/if_uralvar.h: netbsd-7-nhusb sys/dev/usb/if_url.c: netbsd-7-nhusb sys/dev/usb/if_urlreg.h: netbsd-7-nhusb sys/dev/usb/if_urndis.c: netbsd-7-nhusb sys/dev/usb/if_urndisreg.h: netbsd-7-nhusb sys/dev/usb/if_urtw.c: netbsd-7-nhusb sys/dev/usb/if_urtwn.c: netbsd-7-nhusb sys/dev/usb/if_urtwn_data.h: netbsd-7-nhusb sys/dev/usb/if_urtwnreg.h: netbsd-7-nhusb sys/dev/usb/if_urtwnvar.h: netbsd-7-nhusb sys/dev/usb/if_urtwreg.h: netbsd-7-nhusb sys/dev/usb/if_zyd.c: netbsd-7-nhusb sys/dev/usb/if_zydreg.h: netbsd-7-nhusb sys/dev/usb/irmce.c: netbsd-7-nhusb sys/dev/usb/moscom.c: netbsd-7-nhusb sys/dev/usb/motg.c: netbsd-7-nhusb sys/dev/usb/motgvar.h: netbsd-7-nhusb sys/dev/usb/ohci.c: netbsd-7-nhusb sys/dev/usb/ohcireg.h: netbsd-7-nhusb sys/dev/usb/ohcivar.h: netbsd-7-nhusb sys/dev/usb/pseye.c: netbsd-7-nhusb sys/dev/usb/slurm.c: netbsd-7-nhusb sys/dev/usb/stuirda.c: netbsd-7-nhusb sys/dev/usb/u3g.c: netbsd-7-nhusb sys/dev/usb/uark.c: netbsd-7-nhusb sys/dev/usb/uatp.c: netbsd-7-nhusb sys/dev/usb/uaudio.c: netbsd-7-nhusb sys/dev/usb/uberry.c: netbsd-7-nhusb sys/dev/usb/ubsa.c: netbsd-7-nhusb sys/dev/usb/ubsa_common.c: netbsd-7-nhusb sys/dev/usb/ubsavar.h: netbsd-7-nhusb sys/dev/usb/ubt.c: netbsd-7-nhusb sys/dev/usb/uchcom.c: netbsd-7-nhusb sys/dev/usb/ucom.c: netbsd-7-nhusb sys/dev/usb/ucomvar.h: netbsd-7-nhusb sys/dev/usb/ucycom.c: netbsd-7-nhusb sys/dev/usb/udl.c: netbsd-7-nhusb sys/dev/usb/udl.h: netbsd-7-nhusb sys/dev/usb/udsbr.c: netbsd-7-nhusb sys/dev/usb/udsir.c: netbsd-7-nhusb sys/dev/usb/uep.c: netbsd-7-nhusb sys/dev/usb/uftdi.c: netbsd-7-nhusb sys/dev/usb/uftdireg.h: netbsd-7-nhusb sys/dev/usb/ugen.c: netbsd-7-nhusb sys/dev/usb/ugensa.c: netbsd-7-nhusb sys/dev/usb/uhci.c: netbsd-7-nhusb sys/dev/usb/uhcireg.h: netbsd-7-nhusb sys/dev/usb/uhcivar.h: netbsd-7-nhusb sys/dev/usb/uhid.c: netbsd-7-nhusb sys/dev/usb/uhidev.c: netbsd-7-nhusb sys/dev/usb/uhidev.h: netbsd-7-nhusb sys/dev/usb/uhmodem.c: netbsd-7-nhusb sys/dev/usb/uhso.c: netbsd-7-nhusb sys/dev/usb/uhub.c: netbsd-7-nhusb sys/dev/usb/uipad.c: netbsd-7-nhusb sys/dev/usb/uipaq.c: netbsd-7-nhusb sys/dev/usb/uirda.c: netbsd-7-nhusb sys/dev/usb/uirdavar.h: netbsd-7-nhusb sys/dev/usb/ukbd.c: netbsd-7-nhusb sys/dev/usb/ukbdmap.c: netbsd-7-nhusb sys/dev/usb/ukyopon.c: netbsd-7-nhusb sys/dev/usb/ukyopon.h: netbsd-7-nhusb sys/dev/usb/ulpt.c: netbsd-7-nhusb sys/dev/usb/umass.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.c: netbsd-7-nhusb sys/dev/usb/umass_isdata.h: netbsd-7-nhusb sys/dev/usb/umass_quirks.c: netbsd-7-nhusb sys/dev/usb/umass_quirks.h: netbsd-7-nhusb sys/dev/usb/umass_scsipi.c: netbsd-7-nhusb sys/dev/usb/umass_scsipi.h: netbsd-7-nhusb sys/dev/usb/umassvar.h: netbsd-7-nhusb sys/dev/usb/umcs.c: netbsd-7-nhusb sys/dev/usb/umct.c: netbsd-7-nhusb sys/dev/usb/umidi.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.c: netbsd-7-nhusb sys/dev/usb/umidi_quirks.h: netbsd-7-nhusb sys/dev/usb/umodem.c: netbsd-7-nhusb sys/dev/usb/umodem_common.c: netbsd-7-nhusb sys/dev/usb/umodemvar.h: netbsd-7-nhusb sys/dev/usb/ums.c: netbsd-7-nhusb sys/dev/usb/uplcom.c: netbsd-7-nhusb sys/dev/usb/urio.c: netbsd-7-nhusb sys/dev/usb/urio.h: netbsd-7-nhusb sys/dev/usb/usb.c: netbsd-7-nhusb sys/dev/usb/usb.h: netbsd-7-nhusb sys/dev/usb/usb_mem.c: netbsd-7-nhusb sys/dev/usb/usb_mem.h: netbsd-7-nhusb sys/dev/usb/usb_quirks.c: netbsd-7-nhusb sys/dev/usb/usb_quirks.h: netbsd-7-nhusb sys/dev/usb/usb_subr.c: netbsd-7-nhusb sys/dev/usb/usbdevices.config: netbsd-7-nhusb sys/dev/usb/usbdevs: netbsd-7-nhusb sys/dev/usb/usbdevs.h: netbsd-7-nhusb sys/dev/usb/usbdevs_data.h: netbsd-7-nhusb sys/dev/usb/usbdi.c: netbsd-7-nhusb sys/dev/usb/usbdi.h: netbsd-7-nhusb sys/dev/usb/usbdi_util.c: netbsd-7-nhusb sys/dev/usb/usbdi_util.h: netbsd-7-nhusb sys/dev/usb/usbdivar.h: netbsd-7-nhusb sys/dev/usb/usbhid.h: netbsd-7-nhusb sys/dev/usb/usbhist.h: netbsd-7-nhusb sys/dev/usb/usbroothub.c: netbsd-7-nhusb sys/dev/usb/usbroothub.h: netbsd-7-nhusb sys/dev/usb/usbroothub_subr.c: delete sys/dev/usb/usbroothub_subr.h: delete sys/dev/usb/uscanner.c: netbsd-7-nhusb sys/dev/usb/uslsa.c: netbsd-7-nhusb sys/dev/usb/usscanner.c: netbsd-7-nhusb sys/dev/usb/ustir.c: netbsd-7-nhusb sys/dev/usb/uthum.c: netbsd-7-nhusb sys/dev/usb/utoppy.c: netbsd-7-nhusb sys/dev/usb/uts.c: netbsd-7-nhusb sys/dev/usb/uvideo.c: netbsd-7-nhusb sys/dev/usb/uvisor.c: netbsd-7-nhusb sys/dev/usb/uvscom.c: netbsd-7-nhusb sys/dev/usb/uyap.c: netbsd-7-nhusb sys/dev/usb/uyap_firmware.h: netbsd-7-nhusb sys/dev/usb/uyurex.c: netbsd-7-nhusb sys/dev/usb/x1input_rdesc.h: netbsd-7-nhusb sys/dev/usb/xhci.c: netbsd-7-nhusb sys/dev/usb/xhcireg.h: netbsd-7-nhusb sys/dev/usb/xhcivar.h: netbsd-7-nhusb sys/dev/usb/xinput_rdesc.h: netbsd-7-nhusb sys/external/bsd/common/conf/files.linux: netbsd-7-nhusb sys/external/bsd/common/include/linux/err.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/kernel.h: netbsd-7-nhusb sys/external/bsd/common/include/linux/workqueue.h: netbsd-7-nhusb sys/external/bsd/common/linux/linux_work.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/atombios_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/dist/drm/radeon/radeon_legacy_encoders.c: netbsd-7-nhusb sys/external/bsd/drm2/drm/files.drmkms: netbsd-7-nhusb sys/external/bsd/drm2/i915drm/files.i915drmkms: netbsd-7-nhusb sys/external/bsd/drm2/include/linux/err.h: delete sys/external/bsd/drm2/include/linux/workqueue.h: delete sys/external/bsd/drm2/linux/files.drmkms_linux: netbsd-7-nhusb sys/external/bsd/drm2/linux/linux_work.c: delete sys/external/bsd/dwc2/dwc2.c: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwc2var.h: netbsd-7-nhusb sys/external/bsd/dwc2/dwctwo2netbsd: netbsd-7-nhusb sys/external/bsd/dwc2/conf/files.dwc2: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_core.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_coreintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcd.h: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdddma.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdintr.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hcdqueue.c: netbsd-7-nhusb sys/external/bsd/dwc2/dist/dwc2_hw.h: netbsd-7-nhusb sys/modules/drmkms_linux/Makefile: netbsd-7-nhusb sys/modules/i915drmkms/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libugenhc/ugenhc.c: netbsd-7-nhusb sys/rump/dev/lib/libusb/Makefile: netbsd-7-nhusb sys/rump/dev/lib/libusb/USB.ioconf: netbsd-7-nhusb sys/rump/dev/lib/libusb/usb_at_ugenhc.c: delete sys/rump/dev/lib/libusb/opt/opt_usb.h: delete sys/rump/dev/lib/libusb/opt/opt_usbverbose.h: delete sys/sys/mbuf.h: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.8: netbsd-7-nhusb usr.sbin/usbdevs/usbdevs.c: netbsd-7-nhusb Merge netbsd-7-nhusb: - API / infrastructure changes to support memory management changes. - Memory management improvements and bug fixes. - HCDs should now be MP safe - conversion to KERNHIST based debug - FS/LS isoc support on ehci(4). - conversion to kmem(9) - Some USB 3 support - mostly from Takahiro HAYASHI (t-hash). - interrupt transfers now get proper DMA operations - general bug fixes - kern/48308 - uhub status notification improvements - umass(4) probe fix (applied to HEAD already) - ohci(4) short transfer fix - Change the SOFTINT level from NET to SERIAL for the USB softint handler. This gives the callback a chance of running when another softint handler at SOFTINT_NET has blocked holding a lock, e.g. softnet_lock and most of the network stack. - kern/49065 - ifconfig tun0 ... sequence locks up system / lockup: softnet_lock held across usb xfr - kern/50491 - unkillable wait in usbd_transfer while using usmsc0 on raspberry pi 2 - kern/51395 - USB Ethernet makes xhci hang - Various improvements to slhci(4) - Various improvements to dwc2(4)
1.3.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.6.1	29-Mar-2014	yamt	file xhci_pci.c was added on branch yamt-pagecache on 2014-05-22 11:40:34 +0000
1.3.4.2	18-May-2014	rmind	sync with head
1.3.4.1	29-Mar-2014	rmind	file xhci_pci.c was added on branch rmind-smpnet on 2014-05-18 17:45:44 +0000
1.4.2.8	02-Jan-2017	skrll	Parse the extended capabilies to and log each controller port to SS/HS bus root hub ports. Create/attach the two buses and adapt the xhci_roothub_ctrl to deal with both buses and sets of roothub ports. XXX the roothub ub_devices entry needs work to interact with usbdevs(1) XXX correctly
1.4.2.7	05-Dec-2016	skrll	Sync with HEAD
1.4.2.6	29-May-2016	skrll	Sync with HEAD
1.4.2.5	05-Apr-2016	skrll	Switch PCI xhci(4) attachment from pci_intr_map() to pci_intr_alloc()/pci_intr_release(). This enables MSI where available.
1.4.2.4	31-Aug-2015	skrll	Various improvements from t-hash (very very slightly tweaked by me) + Use usbd_xfer_isread(). + Change mutex to be initialized at IPL_USB. + Add vendor init/portsc hook. + Modify xhci_trb_put() to take host byte order arguments and convert them to little endian byte order. + Return PCI vendor ID of xhci instead of NetBSD(0x0) as a root hub vendor ID like other HCs do. + Move sc_ih in struct xhci_softc to struct xhci_pci_softc. + Improve debug message.
1.4.2.3	27-May-2015	skrll	Add patch for intel chips quirk from ryoon@ in https://mail-index.netbsd.org/netbsd-bugs/2014/08/31/msg038109.html slightly modified to set quirks before calling xhci_init. Set special IMOD value and route ports to xhci for intel chips. From Takahiro HAYASHI.
1.4.2.2	06-Apr-2015	skrll	Remove useless break;
1.4.2.1	03-Dec-2014	skrll	The grand renaming of structure members. No functional change.
1.6.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.6.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.7.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.8.6.4	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1787: sys/dev/pci/xhci_pci.c 1.31 via patch sys/dev/usb/xhci.c 1.173-1.175 sys/dev/usb/xhcivar.h 1.22 Support xHCI device which has USB 2 port only.
1.8.6.3	16-Nov-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1443: sys/arch/arm/nvidia/tegra_xusb.c 1.13-1.14 via patch sys/dev/pci/xhci_pci.c 1.13 sys/dev/usb/usb.c 1.169 sys/dev/usb/usbdivar.h 1.116 sys/dev/usb/xhci.c 1.93, 1.95, 1.97 sys/dev/usb/xhcireg.h 1.11-1.12 Detect USB 3.1
1.8.6.2	27-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1172): sys/dev/pci/nvme_pci.c: revision 1.26 sys/dev/pci/xhci_pci.c: revision 1.21 sys/dev/pci/ixgbe/ixv.c: revision 1.108 sys/dev/pci/ixgbe/ixgbe.c: revision 1.171 sys/dev/pci/if_fxp_pci.c: revision 1.84 sys/dev/pci/if_fxp_pci.c: revision 1.85 sys/dev/pci/xhci_pci.c: revision 1.16 remove #ifdef DEBUG printf, it seems to have outlived it's usefulness - KNF. No functional change. - Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them explicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.8.6.1	27-Sep-2018	martin	Pull up following revision(s) (requested by mrg in ticket #1037): sys/dev/usb/uhub.c: revision 1.139 sys/external/bsd/dwc2/dwc2.c: revision 1.55 sys/ddb/db_output.c: revision 1.34 sys/ddb/db_command.c: revision 1.160 sys/dev/usb/ehci.c: revision 1.264 sys/dev/usb/xhci.c: revision 1.99 sys/dev/usb/ehci.c: revision 1.265 sys/kern/subr_userconf.c: revision 1.27 sys/dev/usb/ehcivar.h: revision 1.46 sys/dev/usb/ohci.c: revision 1.287 sys/dev/usb/uhci.c: revision 1.284 sys/dev/usb/usbdi.c: revision 1.178 sys/dev/usb/usb.c: revision 1.172 sys/dev/pci/xhci_pci.c: revision 1.14 sys/dev/usb/usb.c: revision 1.173 sys/dev/usb/usb.c: revision 1.174 share/man/man4/usb.4: revision 1.110 sys/ddb/db_command.c: revision 1.159 sys/dev/usb/usb_subr.c: revision 1.227 sys/dev/usb/uhcivar.h: revision 1.56 (all via patch) consolidate the handling of polling across HC drivers, and generic USB: - don't take mutexes if polling - normalise the code across all drivers - add some not yet code to block discovery to/from polling - minor CSE - adjust comment for usbd_set_polling() to reality now i properly understand what it is used for and why. this, with a hack to make RB_ASKNAME to wait 5 seconds allows boot -a work with USB keyboards. there are still multiple issues remaining: - discovery and polling need to be mutually exclusive - attachment of ukbd and wskbd is not handled by config_pending, and the 5 second delay isn't going to always be enough. call cnpollc(1) and cnpollc(0) around cngetc(). (christos has a good idea to add a function that does all 3, and we should switch all the callers in this sequence to use it (and fix the MD ones missing it still). not all can, as eg, line-grabbing functions can use cngetsn(), which only calls cnpollc() twice.) When this file is used when not building the kernel (eg: /usr/sbin/crash) make cnpollc() go away. reorder some struct members to remove holes. add config_pending usage to uhub and general USB device attachment. - call config_pending_incr() and config_pending_decr() around attaching devices against "usbdevif" attribute. uhub: - convert sc_explorepending and sc_running to bool. add new sc_first_explore. - call config_pending_incr() at the start of uhub_attach(). dropped in uhub_explore(), if this is the first explore. implement a gross hack to fix "boot -a" on systems with usb keyboards on systems with ehci handover to uhci (and maybe ohci), and fix a similar problem for "boot -s". there is effort to ensure that all devices attached via USB are probed before RB_ASKNAME or RB_SINGLE attempts to ask any questions on the console, and largely this works, often by chance, today, for USB disks and root. i've recently pushed this more into uhub and general USB device attachment as well, and kept a config_pending reference across the first explore of a bus. these fix many issues with directly attached hubs. however, on systems where devices connected to ehci ports are handed over to a companion uhci or ohci port, it may not be the first, or even second, bus explore that finds the device finally before attachment, and at this point all config_pending references are dropped. there is no direct communication between drivers, the potentials are looked up but their device_t is only used for generic things like the name, so informing the correct companion to expect a device and deal with the config_pending references is not possible without some fairly ugly layer violations or multi-level callbacks (eg, we have "ehci0", and usually an the relevant companion, eg, "uhci2", but it is the uhub that uhci2 has attached that will deal with the device attachment.) with the above fixes to generic USB code, the disown happens during the first explore. the hack works by, at this point, checking if (a) root is not mounted, (b) single user or ask name are set, and (c) if the hack as not been triggered already. if all 3 conditions are true, then a config_pending_incr() is called and a callback is triggered for (default) 5 seconds to call config_pending_decr(). ehci detach pauses waiting for this callback if scheduled. this allows enough time for the uhub and the ukbd/wskbd to attach before the RK_ASKROOT prompts appear. testing shows it takes between 1.5 and 2 seconds for the keyboard to appear after the disown occurs. Index: dev/usb/ehcivar.c - new sc_compcallout, sc_compcallout, sc_complock, and a state for th handover hack. Index: dev/usb/ehci.c ehci_init(): - use aprint_normal_dev() instead of manual device_xname(). - initialise sc_compcallout, sc_compcallout, sc_complock, and sc_comp_state. ehci_detach(): - if there are companion controllers, tear own the above, including waiting if there is a callback scheduled. ehci_disown_callback(): - new callout to call config_pending_decr() in the the future. schedule this ca ehci_disown_sched_callback(): - if booting to single user or asking names, call config_pending_incr() and schedule the callout above, default 5 second delay. ehci_disown(): - if disowning a port call ehci_disown_sched_callback(). deal with partial attach failures in usb_attach vs usb_detach aka PR 53598. - make sure xhci's sc->sc_ios is NULL if failure happens. - rearrange usb_attach() / usb_doattach() to make it simpler to clean up. - move usb_async_intr softint into usb_once_init(). previously, each USB controller would start a new one, and leave the old one leaked. - handle controller interrupts without a bus attached remove usb(4)'s "flags 1" code. it has been dead for a while, as it runs during the interrupts part of configuration now, and all the devices try attach as early as possible, including any root or boot required disk or keyboard device, which is what this flag was for.
1.11.2.8	26-Jan-2019	pgoyette	Sync with HEAD
1.11.2.7	18-Jan-2019	pgoyette	Synch with HEAD
1.11.2.6	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.11.2.5	26-Nov-2018	pgoyette	And one more missed conflict. (I think this is the last one.)
1.11.2.4	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.11.2.3	30-Sep-2018	pgoyette	Ssync with HEAD
1.11.2.2	28-Jul-2018	pgoyette	Sync with HEAD
1.11.2.1	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.12.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.12.2.1	10-Jun-2019	christos	Sync with HEAD
1.21.4.2	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1570: sys/dev/pci/xhci_pci.c 1.31 via patch sys/dev/usb/xhci.c 1.173-1.175 sys/dev/usb/xhcivar.h 1.22 Support xHCI device which has USB 2 port only.
1.21.4.1	17-May-2021	martin	Pull up following revision(s) (requested by jakllsch in ticket #1273): sys/dev/pci/xhci_pci.c: revision 1.26 xhci_pci: avoid potential double free of interrupt handles Found by Kouichi Hashikawa in PR 55855.
1.25.2.2	03-Apr-2021	thorpej	Sync with HEAD.
1.25.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.28.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.29.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.32.2.1	11-Mar-2024	martin	Pull up following revision(s) (requested by riastradh in ticket #632): sys/dev/pci/xhci_pci.c: revision 1.33 Consistently use cached chipset tag value.
1.17	27-Oct-2022	riastradh	branches: 1.17.6; wwanc(4): Convert to ttylock/ttyunlock.
1.16	12-Feb-2022	thorpej	Don't dig into the guts of selinfo.
1.15	12-Feb-2022	thorpej	Add inline functions to manipulate the klists that link up knotes via kn_selnext: - klist_init() - klist_fini() - klist_insert() - klist_remove() These provide some API insulation from the implementation details of these lists (but not completely; see vn_knote_attach() and vn_knote_detach()). Currently just a wrapper around SLIST(9). This will make it significantly easier to switch kn_selnext linkage to a different kind of list.
1.14	12-Feb-2022	riastradh	sys: Fix various abuse of struct device internals. Will help to make struct device opaque later.
1.13	18-Oct-2021	hannken	Use a local static variable to hold "pktq_rps_hash_default" like the other devices do. Kernel ALL/amd64 compiles again. OK: Kengo NAKAHARA <knakahara@netbsd.org>
1.12	11-Oct-2021	knakahara	Make pktq_rps_hash() pluggable for each interface type. Reviewed by gdt@n.o, thorpej@n.o, and riastradh@n.o, thanks.
1.11	11-Oct-2021	thorpej	Setting EV_EOF requires modifying kn->kn_flags. However, that relies on holding the kq_lock of that note's kq. Rather than exposing this directly, add new knote_set_eof() and knote_clear_eof() functions that handle the necessary locking and don't leak as many implementation details to modules. NetBSD 9.99.91
1.10	26-Sep-2021	thorpej	Change the kqueue filterops::f_isfd field to filterops::f_flags, and define a flag FILTEROP_ISFD that has the meaning of the prior f_isfd. Field and flag name aligned with OpenBSD. This does not constitute a functional or ABI change, as the field location and size, and the value placed in that field, are the same as the previous code, but we're bumping __NetBSD_Version__ so 3rd-party module source code can adapt, as needed. NetBSD 9.99.89
1.9	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.8	16-Jul-2021	andvar	branches: 1.8.2; Add missing RCSID, remove non-breaking space characters in copyright, remove vim modeline (from yambo in IRC). ok riastradh
1.7	24-Apr-2021	thorpej	branches: 1.7.2; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.6	29-Jul-2020	jdolecek	branches: 1.6.4; fix xmm7360_os_msleep() macro to actually work when not wrapped inside a block from genua
1.5	29-Jul-2020	jdolecek	sprinkle __diagused so this compiles without DIAGNOSTIC
1.4	29-Jul-2020	jdolecek	adjust OpenBSD mbuf handling so tcpdump(1) works there; from genua keep code as-is for NetBSD, DLT_RAW works there
1.3	27-Jul-2020	riastradh	Nix trailing whitespace. Helpful hint: (setq show-trailing-whitespace t)
1.2	26-Jul-2020	jdolecek	fix attribution - I was confused what was the correct final version
1.1	26-Jul-2020	jdolecek	Add driver for Intel XMM7360 LTE modem, based upon Linux driver available at https://github.com/xmm7360/xmm7360-pci This version works on Linux, OpenBSD, and NetBSD. OpenBSD port written for genua GmbH Modem requires python script from the master site to initialize the network, it will be added to pkgsrc shortly
1.6.4.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.7.2.1	01-Aug-2021	thorpej	Sync with HEAD.
1.8.2.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.17.6.1	15-Nov-2023	thorpej	Rename ifq_enqueue() -> if_enqueue(), ifq_enqueue2() -> if_enqueue2().
1.70	09-Feb-2024	andvar	fix spelling mistakes, mainly in comments and log messages.
1.69	08-Jan-2024	chs	mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.68	07-Aug-2021	thorpej	branches: 1.68.6; Merge thorpej-cfargs2.
1.67	24-Apr-2021	thorpej	branches: 1.67.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.66	06-Feb-2021	isaki	branches: 1.66.2; Remove an extra mutex_spin_exit() in yds_resume(). This one has been moved to before yds_init() in rev 1.59.
1.65	29-Feb-2020	isaki	branches: 1.65.6; Remove rounding by 4 bytes on round_blocksize(). For drivers which supports only 16bit * 2channels sampling, rounding by 4 bytes no longer meaningful.
1.64	08-Jun-2019	isaki	branches: 1.64.4; Clean get_props(). - Make get_props() return AUDIO_PROP_{PLAYBACK,CAPTURE} properly. This eliminates need for audio.c to take care of such (old) drivers which don't return both of PLAYBACK and CAPTURE. - All get_props() doesn't need to return AUDIO_PROP_MMAP. It is handled in the audio layer now.
1.63	25-May-2019	isaki	Fix yds_intr. It has not worked well since netbsd-8. - Read data offset before set ACTV2. - The interrupts occur every hardware block, not blocksize specified by round_blocksize. The hardware block may span (upper layer's) two blocks. So bus_dmamap_sync'ing each (upper layer's) block is not enough. XXX To fix this, Use only 48kHz and drop other frequencies. It can make blocksize fixed. Then the hardware block aligns to upper layer's block alignment. Analyzed by Y.Sugahara.
1.62	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.61	08-Apr-2019	isaki	branches: 1.61.2; Fix minor bugs. - sc_lock should use an adaptive lock. - yds_round_blocksize: fix typo. - yds_trigger_output: fix typo. SlotControlData is pbankoff. ptbloff points to ControlData. - yds_halt_input: fix wrong and extra sc_rec.intr.
1.60	09-Dec-2018	jdolecek	use pci_intr_establish_xname() everywhere
1.59	25-Jun-2017	christos	branches: 1.59.4; 1.59.6; PR/52331: ydc driver: sleep-under-spin-mutex bugs in yds_allocmem Don't hold the spin interrupt mutex while calling yds_init from resume. Instead use a flag to short-circuit the interrupt while disabled.
1.58	01-Jun-2017	chs	remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.57	14-Jul-2016	msaitoh	KNF. No functional change.
1.56	17-Aug-2014	tsutsui	branches: 1.56.2; Use C99 struct initializer. Also tweak some inconsistent TAB/space. No binary change on i386.
1.55	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.54	17-Oct-2013	christos	__USE a debugging var
1.53	30-Jan-2012	drochner	branches: 1.53.6; 1.53.10; Use pci_aprint_devinfo(9) instead of pci_devinfo+aprint_{normal,naive} where it looks straightforward, and pci_aprint_devinfo_fancy in a few others where drivers want to supply their own device names instead of the pcidevs generated one. More complicated cases, where names are composed at runtime, are left alone for now. It certainly makes sense to simplify the drivers here rather than inventing a catch-all API. This should serve as as example for new drivers, and also ensure consistent output in the AB_QUIET ("boot -q") case. Also, it avoids excessive stack usage where drivers attach child devices because the buffer for the device name is not kept on the local stack anymore.
1.52	24-Nov-2011	mrg	branches: 1.52.2; put back IPL_AUDIO and splaudio(), at the request of rmind. they are aliases and now we can easily revert audio to IPL_VM if necessary, without having to revert the whole branch.
1.51	23-Nov-2011	jmcneill	Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.50	26-Jul-2010	jym	branches: 1.50.8; 1.50.10; Add PAE to ALL kernel, so that most paddr_t format string errors get caught during compilation. While here, fix the compilation for ALL.
1.49	24-Feb-2010	dyoung	branches: 1.49.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.48	08-Jan-2010	dyoung	branches: 1.48.2; Expand PMF_FN_* macros.
1.47	26-Nov-2009	njoly	Cleanup interrupt establish error messages. Do not mix aprint_error/aprint_normal/printf calls for a single line.
1.46	06-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.45	17-Dec-2008	cegger	branches: 1.45.2; make this compile with AUDIO_DEBUG
1.44	16-Dec-2008	christos	replace bitmask_snprintf(9) with snprintb(3)
1.43	09-Jul-2008	joerg	branches: 1.43.2; 1.43.6; - device/softc split
1.42	10-Apr-2008	cegger	branches: 1.42.4; 1.42.6; 1.42.8; 1.42.10; use aprint_*_dev and device_xname
1.41	01-Apr-2008	xtraeme	Fix fallout from device_t/softc split to mpu(4): always pass the mpu_softc to mpu_intr(). Found by Takeshi Nakayama.
1.40	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.39	09-Dec-2007	jmcneill	branches: 1.39.6; 1.39.10; Merge jmcneill-pm branch.
1.38	19-Oct-2007	ad	branches: 1.38.4; 1.38.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.37	16-Nov-2006	christos	branches: 1.37.6; 1.37.8; 1.37.22; 1.37.24; 1.37.28; __unused removal on arguments; approved by core.
1.36	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.35	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.34	03-Sep-2006	christos	branches: 1.34.2; 1.34.4; add missing initializer.
1.33	17-Aug-2006	christos	Fix all the -DDEBUG code that it was rotting away and did not even compile. Mostly from Arnaud Lacombe, many thanks!
1.32	19-Jun-2006	jmcneill	Restore yds(4) state on resume.
1.31	07-Jun-2006	nakayama	branches: 1.31.2; Add void cast to functions whose return values are ignored.
1.30	11-Dec-2005	christos	branches: 1.30.4; 1.30.6; 1.30.8; 1.30.14; merge ktrace-lwp.
1.29	28-Jun-2005	thorpej	branches: 1.29.2; Use ANSI function decls and static.
1.28	30-May-2005	christos	- const poisoning - avoid variable shadowing.
1.27	15-Jan-2005	kent	ansify and KNF
1.26	10-Jan-2005	kent	merge kent-audio1 branch, which introduces audio filter pipeline to the MI audio framework Summary of changes: * struct audio_params - remove sw_code, factor, factor_denom, hw_sample_rate, hw_encoding ,hw_precision, and hw_channels. Conversion information is conveyed by stream_filter_list_t. - change the type of sample_rate: u_long -> u_int - add `validbits,' which represents the valid data size in precision bits. It is required in order to distinguish 24/32bit from 24/24bit or 32/32bit. * audio_hw_if - add two parameters to set_params() stream_filter_list_t pfil, stream_filter_list rfil A HW driver should set filter recipes for requested formats - constify audio_params parameters of trigger_output() and trigger_input(). They represent audio formats for the hardware. - make open() and close() optional - add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize() * sw_code is replaced with stream_filter_t. stream_filer_t converts audio data in an input buffer and writes into another output buffer unlike sw_code, which converts data in single buffer. converters in dev/auconv.c, dev/mulaw.c, dev/aurateconv.c, dev/tc/bba.c, dev/ic/msm6258.c, and arch/arm/iomd/vidcaudio.c are reimplemented as stream_filter_t * MI audio - audiosetinfo() builds filter pipelines from stream_filter_list_t filled by audio_hw_if::set_params() - audiosetinfo() returns with EINVAL if mmapped and set_params() requests filters - audio_write(), audio_pint(), and audio_rint() invoke a filter pipeline. - ioctl() for FIONREAD, AUDIO_WSEEK, AUDIO_GETIOFFS, AUDIO_GETOOFFS, and audio_prinfo::{seek,samples} for AUDIO_GETINFO handle values for a buffer nearest to userland. * add `struct device ' parameter to ac97_attach() all of audio HW drivers follow audio_hw_if and ac97 changes
1.25	13-Nov-2004	kent	branches: 1.25.2; adopt auconv_set_converter()
1.24	09-Nov-2004	kent	remove mixer setting code in yds_attach() because ac97_attach() does it.
1.23	08-Nov-2004	kent	ANSIfy and style fixes
1.22	29-Oct-2004	yamt	constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.21	22-Sep-2004	kent	ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR: kern/26973
1.20	09-Jul-2004	mycroft	There's no need to halt the input and output pipes in the close routine -- audio_close() takes care of this.
1.19	23-Apr-2004	itojun	pass string length (= boundary info) to pci_devinfo so that we do not run over the end of memory region
1.18	29-Sep-2003	wiz	branches: 1.18.2; available, not avaliable. From miod@openbsd.
1.17	12-Mar-2003	minoura	branches: 1.17.2; Disable ISA DMA emulation for SoundBlaster module explicitly. This fixes the floppy access problem on some hardware. Problem reported by Miroslav Kure via Jason Wright <jason [at] thought [dot] net> of OpenBSD. Thanks. Fix from ALSA driver; although this setting is undocumented but according to Jaroslav Kysela <perex [at] suse [dot] cz> of ALSA, there's no error reported so far.
1.16	01-Feb-2003	thorpej	Add extensible malloc types, adapted from FreeBSD. This turns malloc types into a structure, a pointer to which is passed around, instead of an int constant. Allow the limit to be adjusted when the malloc type is defined, or with a function call, as suggested by Jonathan Stone.
1.15	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.14	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.13	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.12	10-Jun-2002	someya	Make the yds driver work on big-endian systems.
1.11	10-Jan-2002	someya	branches: 1.11.8; 1.11.10; remove a unnecessary message print
1.10	30-Dec-2001	someya	add powerhook support to resume correctly, fix PR#13079.
1.9	25-Dec-2001	someya	add 600 times dummy read from codec for YMF744 (revision < 2).
1.8	13-Nov-2001	lukem	add RCSID
1.7	03-Oct-2001	augustss	Add a new optional method, dev_ioctl, to the audio hardware driver interface. It is called when an unrecognized ioctl() is performed on a device, thus allowing ioctl()s that frob the hardware driver (like loading microcode).
1.6	07-Jul-2001	thorpej	branches: 1.6.2; 1.6.4; bzero -> memset
1.5	21-May-2001	minoura	Correct an easy mistake in initializing gain registers. This will fix the L-R imbalance. Suggested by Ryo HAYASAKA <ryoh@jaist.ac.jp>.
1.4	15-May-2001	lukem	delint: need ; between goto label and }
1.3	05-Apr-2001	minoura	branches: 1.3.2; 1.3.4; 1.3.6; Make this compile on alpha with DIAGNOSTIC. Thanks to atatat.
1.2	30-Mar-2001	minoura	Update TODO list.
1.1	30-Mar-2001	minoura	Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.3.6.4	29-Jul-2001	he	Apply patch (requested by minoura): Correct argument to pci_intr_map(). Fixes PR#13508.
1.3.6.3	26-May-2001	he	Pull up revision 1.5 (requested by minoura): Correct bug in initializing gain registers. Fixes L-R imbalance.
1.3.6.2	01-May-2001	he	Pull up revisions 1.1-1.3 (new, via patch, requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.3.6.1	05-Apr-2001	he	file yds.c was added on branch netbsd-1-5 on 2001-05-01 11:45:46 +0000
1.3.4.2	21-Apr-2001	bouyer	Sync with HEAD
1.3.4.1	05-Apr-2001	bouyer	file yds.c was added on branch thorpej_scsipi on 2001-04-21 17:49:32 +0000
1.3.2.11	18-Oct-2002	nathanw	Catch up to -current.
1.3.2.10	20-Jun-2002	nathanw	Catch up to -current.
1.3.2.9	28-Feb-2002	nathanw	Catch up to -current.
1.3.2.8	11-Jan-2002	nathanw	More catchup.
1.3.2.7	08-Jan-2002	nathanw	Catch up to -current.
1.3.2.6	14-Nov-2001	nathanw	Catch up to -current.
1.3.2.5	08-Oct-2001	nathanw	Catch up to -current.
1.3.2.4	24-Aug-2001	nathanw	Catch up with -current.
1.3.2.3	21-Jun-2001	nathanw	Catch up to -current.
1.3.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.3.2.1	05-Apr-2001	nathanw	file yds.c was added on branch nathanw_sa on 2001-04-09 01:57:14 +0000
1.6.4.1	11-Oct-2001	fvdl	Catch up with -current. Fix some bogons in the sparc64 kbd/ms attach code. cd18xx conversion provided by mrg.
1.6.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.6.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.6.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.6.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.11.10.1	19-Jun-2003	grant	Pull up revision 1.17 (requested by hubertf in ticket #1322): Disable ISA DMA emulation for SoundBlaster module explicitly. This fixes the floppy access problem on some hardware. Problem reported by Miroslav Kure via Jason Wright <jason [at] thought [dot] net> of OpenBSD. Thanks. Fix from ALSA driver; although this setting is undocumented but according to Jaroslav Kysela <perex [at] suse [dot] cz> of ALSA, there's no error reported so far.
1.11.8.1	20-Jun-2002	gehenna	catch up with -current.
1.17.2.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.17.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.17.2.7	29-Nov-2004	skrll	Sync with HEAD.
1.17.2.6	14-Nov-2004	skrll	Sync with HEAD.
1.17.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.17.2.4	24-Sep-2004	skrll	Sync with HEAD.
1.17.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.17.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.17.2.1	03-Aug-2004	skrll	Sync with HEAD
1.18.2.1	22-Sep-2004	jmc	Pullup rev 1.21 (requested by kent in ticket #867) ac97_host_if::reset() returns non-zero value if codec reset fails, and ac97 is not attached in that case. PR#26973
1.25.2.2	09-Jan-2005	kent	audio_hw_if change * make open() and close() optional * add int (AUMODE_PLAY or AUMODE_RECORD) and audio_params_t parameters to round_blocksize()
1.25.2.1	02-Jan-2005	kent	* adopt the filter pipeline framework * follow ac97_attach() change
1.29.2.5	17-Mar-2008	yamt	sync with head.
1.29.2.4	21-Jan-2008	yamt	sync with head
1.29.2.3	27-Oct-2007	yamt	sync with head.
1.29.2.2	30-Dec-2006	yamt	sync with head.
1.29.2.1	21-Jun-2006	yamt	sync with head.
1.30.14.2	22-Jun-2006	chap	Complete a sync sys/ with head.
1.30.14.1	19-Jun-2006	chap	Sync with head.
1.30.8.3	14-Sep-2006	yamt	sync with head.
1.30.8.2	03-Sep-2006	yamt	sync with head.
1.30.8.1	26-Jun-2006	yamt	sync with head.
1.30.6.1	07-Jun-2006	kardel	Sync with head.
1.30.4.1	09-Sep-2006	rpaulo	sync with head
1.31.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.34.4.2	10-Dec-2006	yamt	sync with head.
1.34.4.1	22-Oct-2006	yamt	sync with head
1.34.2.1	18-Nov-2006	ad	Sync with head.
1.37.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.37.24.3	23-Mar-2008	matt	sync with HEAD
1.37.24.2	09-Jan-2008	matt	sync with HEAD
1.37.24.1	06-Nov-2007	matt	sync with HEAD
1.37.22.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.37.22.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.37.22.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.37.22.1	05-Oct-2007	joerg	Convert yds(4) to PNP power management.
1.37.8.1	23-Oct-2007	ad	Sync with head.
1.37.6.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.38.6.1	11-Dec-2007	yamt	sync with head.
1.38.4.1	26-Dec-2007	ad	Sync with head.
1.39.10.4	17-Jan-2009	mjf	Sync with HEAD.
1.39.10.3	28-Sep-2008	mjf	Sync with HEAD.
1.39.10.2	02-Jun-2008	mjf	Sync with HEAD.
1.39.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.39.6.1	24-Mar-2008	keiichi	sync with head.
1.42.10.1	19-Oct-2008	haad	Sync with HEAD.
1.42.8.1	18-Jul-2008	simonb	Sync with head.
1.42.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.42.4.4	11-Aug-2010	yamt	sync with head.
1.42.4.3	11-Mar-2010	yamt	sync with head
1.42.4.2	16-May-2009	yamt	sync with head
1.42.4.1	04-May-2009	yamt	sync with head.
1.43.6.2	12-Dec-2008	ad	Checkpoint work in progress.
1.43.6.1	11-Dec-2008	ad	Checkpoint work in progress.
1.43.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.45.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.48.2.2	17-Aug-2010	uebayasi	Sync with HEAD.
1.48.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.49.2.1	05-Mar-2011	rmind	sync with head
1.50.10.2	22-Nov-2011	jmcneill	get rid of unused 'powerstate' callback in audio_hw_if
1.50.10.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.50.8.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.50.8.1	17-Apr-2012	yamt	sync with head
1.52.2.1	18-Feb-2012	mrg	merge to -current.
1.53.10.1	18-May-2014	rmind	sync with head
1.53.6.2	03-Dec-2017	jdolecek	update from HEAD
1.53.6.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.56.2.2	28-Aug-2017	skrll	Sync with HEAD
1.56.2.1	05-Oct-2016	skrll	Sync with HEAD
1.59.6.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.59.6.1	10-Jun-2019	christos	Sync with HEAD
1.59.4.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.61.2.4	04-May-2019	isaki	Move dev/audio_if.h -> dev/audio/audio_if.h
1.61.2.3	04-May-2019	isaki	Remove obsoleted methods in audio_hw_if. - drain: is handled in audio upper layer now. - mappage: is handled in audio upper layer now. - setfd: no one uses and it's meaningless now.
1.61.2.2	29-Apr-2019	isaki	Adapt to audio2.
1.61.2.1	21-Apr-2019	isaki	Use C99 style struct initializer to audio_format.
1.64.4.1	29-Feb-2020	ad	Sync with head.
1.65.6.1	03-Apr-2021	thorpej	Sync with HEAD.
1.66.2.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.67.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.68.6.1	14-Jan-2024	martin	Pull up following revision(s) (requested by chs in ticket #542): sys/dev/pci/yds.c: revision 1.69 sys/dev/pci/cmpci.c: revision 1.61 sys/dev/pci/sv.c: revision 1.62 sys/dev/pci/fms.c: revision 1.50 mpu / opl: add an interface attribute to config_found() calls for these specify an interface attribute when attaching mpu and opl devices. the PCI devices that are the parents of these midi devices (cmpci, eso, fms, sv, yds) have two interface attributes: "audiobus", and also one named the same as the parent device, eg. "cmpci" has an interface attribute named "cmpci". we must specify the latter one to attach these children.
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	15-Jan-2005	kent	ansify and KNF
1.4	12-Mar-2003	minoura	branches: 1.4.2; Disable ISA DMA emulation for SoundBlaster module explicitly. This fixes the floppy access problem on some hardware. Problem reported by Miroslav Kure via Jason Wright <jason [at] thought [dot] net> of OpenBSD. Thanks. Fix from ALSA driver; although this setting is undocumented but according to Jaroslav Kysela <perex [at] suse [dot] cz> of ALSA, there's no error reported so far.
1.3	12-Mar-2003	minoura	Typo. Register value symbols YDS_PCI_EX_LEGACY_SBMOD_* -> YDS_PCI_EX_LEGACY_SMOD_*.
1.2	04-Nov-2002	kent	Add power-management register definitions.
1.1	30-Mar-2001	minoura	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.22; Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.1.22.1	19-Jun-2003	grant	Pull up revisions 1.3-1.4 (requested by hubertf in ticket #1322): Typo. Register value symbols YDS_PCI_EX_LEGACY_SBMOD_* -> YDS_PCI_EX_LEGACY_SMOD_*. Disable ISA DMA emulation for SoundBlaster module explicitly. This fixes the floppy access problem on some hardware. Problem reported by Miroslav Kure via Jason Wright <jason [at] thought [dot] net> of OpenBSD. Thanks. Fix from ALSA driver; although this setting is undocumented but according to Jaroslav Kysela <perex [at] suse [dot] cz> of ALSA, there's no error reported so far.
1.1.6.2	01-May-2001	he	Pull up revision 1.1 (new, requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.1.6.1	30-Mar-2001	he	file ydsreg.h was added on branch netbsd-1-5 on 2001-05-01 11:45:26 +0000
1.1.4.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.1	30-Mar-2001	bouyer	file ydsreg.h was added on branch thorpej_scsipi on 2001-04-21 17:49:32 +0000
1.1.2.3	11-Nov-2002	nathanw	Catch up to -current
1.1.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.1.2.1	30-Mar-2001	nathanw	file ydsreg.h was added on branch nathanw_sa on 2001-04-09 01:57:14 +0000
1.4.2.1	17-Jan-2005	skrll	Sync with HEAD.
1.13	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.12	25-Jun-2017	christos	branches: 1.12.6; 1.12.10; PR/52331: ydc driver: sleep-under-spin-mutex bugs in yds_allocmem Don't hold the spin interrupt mutex while calling yds_init from resume. Instead use a flag to short-circuit the interrupt while disabled.
1.11	23-Nov-2011	jmcneill	branches: 1.11.8; 1.11.26; Merge jmcneill-audiomp3 branch, which is derived from ad-audiomp2. From the original ad-audiomp branch notes: Add MP locking to the audio drivers. Making the audio drivers MP safe is necessary before efforts can be made to make the VM system MP safe. The are two locks per device instance, an ISR lock and a character device lock. The ISR lock replaces calls to splaudio()/splx(), and will be held across calls to device methods which were called at splaudio() before (e.g. trigger_output). The character device lock is held across calls to nearly all of the methods, excluding some only used for initialization, e.g. get_locks. Welcome to 5.99.57.
1.10	09-Jul-2008	joerg	branches: 1.10.6; 1.10.28; 1.10.30; - device/softc split
1.9	01-Apr-2008	xtraeme	branches: 1.9.4; 1.9.6; 1.9.8; 1.9.10; Fix fallout from device_t/softc split to mpu(4): always pass the mpu_softc to mpu_intr(). Found by Takeshi Nakayama.
1.8	04-Mar-2007	christos	branches: 1.8.36; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.7	19-Jun-2006	jmcneill	branches: 1.7.10; 1.7.12; Restore yds(4) state on resume.
1.6	11-Dec-2005	christos	branches: 1.6.4; 1.6.8; 1.6.14; 1.6.16; merge ktrace-lwp.
1.5	15-Jan-2005	kent	branches: 1.5.10; ansify and KNF
1.4	13-Nov-2004	kent	adopt auconv_set_converter()
1.3	08-Nov-2004	kent	ANSIfy and style fixes
1.2	25-Dec-2001	someya	branches: 1.2.16; add 600 times dummy read from codec for YMF744 (revision < 2).
1.1	30-Mar-2001	minoura	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Yamaha YMF724/740/744/754 (DS-1) PCI audio driver.
1.1.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.6.2	01-May-2001	he	Pull up revision 1.1 (new, requested by minoura): Add Yamaha YMF724/740/744/745-based sound driver and its subordinates.
1.1.6.1	30-Mar-2001	he	file ydsvar.h was added on branch netbsd-1-5 on 2001-05-01 11:45:27 +0000
1.1.4.2	21-Apr-2001	bouyer	Sync with HEAD
1.1.4.1	30-Mar-2001	bouyer	file ydsvar.h was added on branch thorpej_scsipi on 2001-04-21 17:49:33 +0000
1.1.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.1.2.2	09-Apr-2001	nathanw	Catch up with -current.
1.1.2.1	30-Mar-2001	nathanw	file ydsvar.h was added on branch nathanw_sa on 2001-04-09 01:57:15 +0000
1.2.16.3	17-Jan-2005	skrll	Sync with HEAD.
1.2.16.2	29-Nov-2004	skrll	Sync with HEAD.
1.2.16.1	14-Nov-2004	skrll	Sync with HEAD.
1.5.10.2	03-Sep-2007	yamt	sync with head.
1.5.10.1	21-Jun-2006	yamt	sync with head.
1.6.16.1	13-Jul-2006	gdamore	Merge from HEAD.
1.6.14.1	22-Jun-2006	chap	Complete a sync sys/ with head.
1.6.8.1	26-Jun-2006	yamt	sync with head.
1.6.4.1	09-Sep-2006	rpaulo	sync with head
1.7.12.1	27-Feb-2007	ad	Add MP locking for audio drivers.
1.7.10.1	12-Mar-2007	rmind	Sync with HEAD.
1.8.36.2	28-Sep-2008	mjf	Sync with HEAD.
1.8.36.1	03-Apr-2008	mjf	Sync with HEAD.
1.9.10.1	19-Oct-2008	haad	Sync with HEAD.
1.9.8.1	18-Jul-2008	simonb	Sync with head.
1.9.6.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.9.4.1	04-May-2009	yamt	sync with head.
1.10.30.1	19-Nov-2011	jmcneill	port ad-audiomp2 branch changes to 5.99.56 todo: - port ad-audiomp2 pcppi changes - these drivers need to be fixed: - emuxki - gcscaudio - hdafg - uaudio - umidi - MD audio drivers
1.10.28.1	17-Apr-2012	yamt	sync with head
1.10.6.1	11-Dec-2008	ad	Checkpoint work in progress.
1.11.26.1	28-Aug-2017	skrll	Sync with HEAD
1.11.8.1	03-Dec-2017	jdolecek	update from HEAD
1.12.10.1	29-Apr-2019	isaki	Adapt to audio2.
1.12.6.1	10-Jun-2019	christos	Sync with HEAD
1.4	12-Mar-2003	wiz	Merge first bktr-sf import.
1.3	01-Jul-2000	wiz	branches: 1.3.2; re-add README that got lost in 2.12-to-2.14 merge.
1.2	01-Jul-2000	wiz	resolve conflicts.
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.2	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	01-Jul-2000	bouyer	file README was added on branch thorpej_scsipi on 2000-11-20 22:35:45 +0000
1.7	12-Mar-2003	wiz	Remove some.
1.6	29-Nov-2001	wiz	Update note about ioctl tsleep error with error report.
1.5	28-Oct-2000	wiz	branches: 1.5.2; 1.5.4; 1.5.6; Replace remaining vm_offset_t's in NetBSD code by vaddr_t; remove now unnecessary casts added by thorpej in 1.11 of bktr_core.c. Strike one from the TODO list.
1.4	03-Sep-2000	wiz	remove one
1.3	07-May-2000	wiz	man page is written, remove it from TODO
1.2	07-May-2000	wiz	add one
1.1	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.5.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.5.4.1	08-Jan-2002	nathanw	Catch up to -current.
1.5.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.5.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.2.1	28-Oct-2000	bouyer	file TODO was added on branch thorpej_scsipi on 2000-11-20 22:35:45 +0000
1.7	08-Sep-2024	rillig	normalize wording in import helpers
1.6	30-Apr-2008	martin	branches: 1.6.126; Convert TNF licenses to new 2 clause variant
1.5	21-Mar-2007	isaki	branches: 1.5.30; 1.5.32; 1.5.34; Fix mis-substitution of RCS Id, and avoid it. Reviewed by wiz@.
1.4	12-Mar-2003	wiz	branches: 1.4.18; 1.4.56; 1.4.60; 1.4.62; 1.4.64; Update script for bktr-sf imports.
1.3	28-Oct-2000	wiz	branches: 1.3.2; Delete more files before import.
1.2	07-May-2000	wiz	multiple inclusion protection
1.1	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.3.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	28-Oct-2000	bouyer	file bktr2netbsd was added on branch thorpej_scsipi on 2000-11-20 22:35:45 +0000
1.4.64.1	29-Mar-2007	reinoud	Pullup to -current
1.4.62.1	11-Jul-2007	mjf	Sync with head.
1.4.60.1	15-Jul-2007	ad	Sync with head.
1.4.56.1	24-Mar-2007	yamt	sync with head.
1.4.18.1	03-Sep-2007	yamt	sync with head.
1.5.34.1	16-May-2008	yamt	sync with head.
1.5.32.1	18-May-2008	yamt	sync with head.
1.5.30.1	02-Jun-2008	mjf	Sync with HEAD.
1.6.126.1	02-Aug-2025	perseant	Sync with HEAD
1.20	19-Dec-2021	andvar	fix various typos in bktr driver comments.
1.19	07-Dec-2012	msaitoh	- sprintf() -> snprintf() - Fix typo in comment.
1.18	16-Jan-2008	jmcneill	branches: 1.18.44; 1.18.54; Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.17	19-Oct-2007	ad	branches: 1.17.2; 1.17.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	24-Nov-2006	wiz	branches: 1.16.2; 1.16.8; 1.16.16; 1.16.22; 1.16.24; 1.16.28; s/existance/existence/, from Zafer.
1.15	08-Mar-2006	lukem	branches: 1.15.10; 1.15.12; Use the SI capitalization for "Hz", "kHz", and "MHz" in comments and strings. Add a space between numbers and Hz unit.
1.14	06-Dec-2005	wiz	branches: 1.14.4; 1.14.6; 1.14.8; 1.14.10; Handle MSP3435G the same as MSP3430G. From Scott Presnell in PR 28682.
1.13	12-May-2004	wiz	branches: 1.13.10; 1.13.12; Fix typo in comment; from Robert Nagy via OpenBSD.
1.12	12-Mar-2003	wiz	branches: 1.12.2; Merge some more.
1.11	12-Mar-2003	wiz	Merge first bktr-sf import.
1.10	04-Jan-2003	wiz	Spell output with two ts.
1.9	13-Nov-2001	lukem	add RCSIDs
1.8	30-Dec-2000	wiz	branches: 1.8.2; 1.8.4; Merge 2.18 to trunk.
1.7	28-Oct-2000	wiz	branches: 1.7.2; Merge 2.17 changes to trunk.
1.6	01-Jul-2000	wiz	resolve conflicts.
1.5	21-May-2000	wiz	branches: 1.5.4; Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.4	07-May-2000	veego	It is a bad idea to declare a static bootverbose variable here. Remove it.
1.3	07-May-2000	wiz	remove unnecessary output to console
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.5.4.1	03-Jul-2000	wiz	Pull up revision 1.6 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.7.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.7.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.7.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.7.2.1	28-Oct-2000	bouyer	file bktr_audio.c was added on branch thorpej_scsipi on 2000-11-20 22:35:45 +0000
1.8.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.2	07-Jan-2003	thorpej	Sync with HEAD.
1.8.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.12.2.4	11-Dec-2005	christos	Sync with head.
1.12.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.12.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.12.2.1	03-Aug-2004	skrll	Sync with HEAD
1.13.12.4	21-Jan-2008	yamt	sync with head
1.13.12.3	27-Oct-2007	yamt	sync with head.
1.13.12.2	30-Dec-2006	yamt	sync with head.
1.13.12.1	21-Jun-2006	yamt	sync with head.
1.13.10.1	07-Dec-2005	riz	Pull up following revision(s) (requested by wiz in ticket #1035): sys/dev/pci/bktr/bktr_audio.c: revision 1.14 Handle MSP3435G the same as MSP3430G. From Scott Presnell in PR 28682.
1.14.10.1	19-Apr-2006	elad	sync with head.
1.14.8.1	13-Mar-2006	yamt	sync with head.
1.14.6.1	22-Apr-2006	simonb	Sync with head.
1.14.4.1	09-Sep-2006	rpaulo	sync with head
1.15.12.1	10-Dec-2006	yamt	sync with head.
1.15.10.1	12-Jan-2007	ad	Sync with head.
1.16.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.24.2	23-Mar-2008	matt	sync with HEAD
1.16.24.1	06-Nov-2007	matt	sync with HEAD
1.16.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.16.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.16.8.1	23-Oct-2007	ad	Sync with head.
1.16.2.1	21-Jan-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1046): sys/dev/pci/bktr/bktr_reg.h: revision 1.18 sys/dev/pci/bktr/bktr_audio.c: revision 1.18 sys/dev/pci/bktr/bktr_os.c: revision 1.49 sys/dev/pci/bktr/bktr_core.c: revision 1.46 sys/dev/ic/bt8xx.h: revision 1.9 sys/dev/pci/bktr/bktr_core.h: revision 1.10 Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.17.8.1	19-Jan-2008	bouyer	Sync with HEAD
1.17.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.18.54.1	25-Feb-2013	tls	resync with head
1.18.44.1	16-Jan-2013	yamt	sync with (a bit old) head
1.4	11-Dec-2005	christos	merge ktrace-lwp.
1.3	12-May-2004	wiz	Fix typo in comment; from Robert Nagy via OpenBSD.
1.2	12-Mar-2003	wiz	branches: 1.2.2; Merge first bktr-sf import.
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.2	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.1	07-May-2000	wiz	branches: 1.1.1.1.6; Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.1.1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.1.1.6.1	07-May-2000	bouyer	file bktr_audio.h was added on branch thorpej_scsipi on 2000-11-20 22:35:46 +0000
1.2.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.2.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.2.2.1	03-Aug-2004	skrll	Sync with HEAD
1.25	16-Oct-2013	christos	remove unused variables.
1.24	19-Oct-2007	ad	branches: 1.24.54; 1.24.64; 1.24.70; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.23	16-Nov-2006	christos	branches: 1.23.8; 1.23.22; 1.23.24; 1.23.28; __unused removal on arguments; approved by core.
1.22	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.21	06-Dec-2005	wiz	branches: 1.21.20; 1.21.22; Add support for two models of Terratec TValue cards. From FreeBSD problem report 53383, submitted by Tanja Wittke, updated by Simon Barner. Adapted for NetBSD by Alexander Becher, via port-i386.
1.20	30-May-2005	christos	branches: 1.20.2; - avoid variable shadowing - add const
1.19	24-Jul-2004	mrg	match LeadTek's original PCI vendor ID as well as the 3-alternates when setting CARD_LEADTEK. now with just options BKTR_SYSTEM_DEFAULT=BROOKTREE_PAL i get a shiny happy working card on my macppc box: bktr0 at pci1 dev 18 function 0 bktr0: interrupting at irq 52 bktr0: Leadtek WinFast TV2000/VC100, Philips FR1216 PAL FM tuner.
1.18	12-May-2004	wiz	Add one more unofficial leadtek PCI vendor id; from Robert Nagy via OpenBSD.
1.17	23-Apr-2004	simonb	s/the the/the/ (only in sources that aren't regularly imported from elsewhere).
1.16	30-Oct-2003	simonb	Remove some assigned-to but otherwise unused variables.
1.15	12-Mar-2003	wiz	branches: 1.15.2; Merge first bktr-sf import.
1.14	14-Jan-2003	wiz	Recognise tuner on WinTV-GO-FM. From PR 19844 by Wolfgang S. Rupprecht.
1.13	10-Jan-2003	mjl	Make this work on big endian platforms (mostly from OpenBSD): Uploaded scripts work better if they are little endian, as the card's engine expects, so convert to le first. Also clean up attach routine a bit (use pa_id and PCI_REVISION instead of fetching it ourselves). This makes the driver work on my macppc G4, it can decode and display video and tuner input. Sound does not seem to work, but this may be my wonky formac bktr-for-macintosh card.
1.12	23-Mar-2002	hamajima	add WinFast VC100
1.11	13-Nov-2001	lukem	add RCSIDs
1.10	16-Sep-2001	wiz	Spell 'occurred' with two 'r's.
1.9	30-Dec-2000	wiz	branches: 1.9.2; 1.9.4; 1.9.6; Merge 2.18 to trunk.
1.8	28-Oct-2000	wiz	branches: 1.8.2; Merge 2.17 changes to trunk.
1.7	03-Sep-2000	wiz	Use NetBSD's pcidevs file for vendors/cards as much as possible. Recognize Pinnacle Studio PCTV Rave as Miro card, tested by rh@netbsd.org.
1.6	01-Jul-2000	wiz	resolve conflicts.
1.5	21-May-2000	wiz	branches: 1.5.4; Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.4	20-May-2000	augustss	Add a missing bktr%d: in one of the attach messages. (When is this code going to be KNFed?)
1.3	07-May-2000	veego	It is a bad idea to declare a static bootverbose variable here. Remove it.
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.5.4.1	03-Jul-2000	wiz	Pull up revision 1.6 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.8.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.8.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.8.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.2.1	28-Oct-2000	bouyer	file bktr_card.c was added on branch thorpej_scsipi on 2000-11-20 22:35:46 +0000
1.9.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.9.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.9.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.2.4	15-Jan-2003	thorpej	Sync with HEAD.
1.9.2.3	01-Apr-2002	nathanw	Catch up to -current. (CVS: It's not just a program. It's an adventure!)
1.9.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.9.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.15.2.5	11-Dec-2005	christos	Sync with head.
1.15.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.15.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.15.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.15.2.1	03-Aug-2004	skrll	Sync with HEAD
1.20.2.3	27-Oct-2007	yamt	sync with head.
1.20.2.2	30-Dec-2006	yamt	sync with head.
1.20.2.1	21-Jun-2006	yamt	sync with head.
1.21.22.2	10-Dec-2006	yamt	sync with head.
1.21.22.1	22-Oct-2006	yamt	sync with head
1.21.20.1	18-Nov-2006	ad	Sync with head.
1.23.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.23.24.1	06-Nov-2007	matt	sync with HEAD
1.23.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.23.8.1	23-Oct-2007	ad	Sync with head.
1.24.70.1	18-May-2014	rmind	sync with head
1.24.64.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.24.54.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6	24-Aug-2012	msaitoh	Fix typos
1.5	06-Dec-2005	wiz	branches: 1.5.110; Add support for two models of Terratec TValue cards. From FreeBSD problem report 53383, submitted by Tanja Wittke, updated by Simon Barner. Adapted for NetBSD by Alexander Becher, via port-i386.
1.4	12-Mar-2003	wiz	branches: 1.4.2; 1.4.18; Merge first bktr-sf import.
1.3	01-Jul-2000	wiz	branches: 1.3.2; resolve conflicts.
1.2	07-May-2000	wiz	branches: 1.2.4; Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.3	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.2.4.1	03-Jul-2000	wiz	Pull up revision 1.3 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	01-Jul-2000	bouyer	file bktr_card.h was added on branch thorpej_scsipi on 2000-11-20 22:35:46 +0000
1.4.18.1	21-Jun-2006	yamt	sync with head.
1.4.2.1	11-Dec-2005	christos	Sync with head.
1.5.110.1	30-Oct-2012	yamt	sync with head
1.60	27-Jun-2025	andvar	Grammar and spelling fixes, mainly in comments. A few in documentation, logging, test description, and SCSI ASC/ASCQ assignment descriptions.
1.59	19-Dec-2021	andvar	branches: 1.59.10; fix various typos in bktr driver comments.
1.58	23-May-2020	ad	Move proc_lock into the data segment. It was dynamically allocated because at the time we had mutex_obj_alloc() but not __cacheline_aligned.
1.57	23-Feb-2020	skrll	typo in comment
1.56	03-Feb-2019	mrg	branches: 1.56.6; - add or adjust /* FALLTHROUGH */ where appropriate - add __unreachable() after functions that can return but won't in this case, and thus can't be marked __dead easily
1.55	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.54	14-Dec-2012	joerg	branches: 1.54.36; 1.54.38; bt848_format is only defined in a __FreeBSD__ section of bktr_os.c and overriding the local (static) definition with extern is bogus.
1.53	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.52	30-Jun-2011	wiz	branches: 1.52.2; 1.52.12; dependant -> dependent
1.51	13-Nov-2010	uebayasi	Don't pull in the whole uvm(9) API to access only PAGE_SIZE and some other constants. These are provided by sys/param.h now.
1.50	18-Mar-2009	cegger	branches: 1.50.4; bzero -> memset
1.49	24-Apr-2008	ad	branches: 1.49.2; 1.49.10; 1.49.16; Network protocol interrupts can now block on locks, so merge the globals proclist_mutex and proclist_lock into a single adaptive mutex (proc_lock). Implications: - Inspecting process state requires thread context, so signals can no longer be sent from a hardware interrupt handler. Signal activity must be deferred to a soft interrupt or kthread. - As the proc state locking is simplified, it's now safe to take exit() and wait() out from under kernel_lock. - The system spends less time at IPL_SCHED, and there is less lock activity.
1.48	10-Apr-2008	cegger	branches: 1.48.2; use aprint_*_dev and device_xname
1.47	01-Mar-2008	rmind	Welcome to 4.99.55: - Add a lot of missing selinit() and seldestroy() calls. - Merge selwakeup() and selnotify() calls into a single selnotify(). - Add an additional 'events' argument to selnotify() call. It will indicate which event (POLL_IN, POLL_OUT, etc) happen. If unknown, zero may be used. Note: please pass appropriate value of 'events' where possible. Proposed on: <tech-kern>
1.46	16-Jan-2008	jmcneill	branches: 1.46.2; 1.46.6; Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.45	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.44	19-Oct-2007	ad	branches: 1.44.2; 1.44.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.43	09-Jul-2007	ad	branches: 1.43.6; 1.43.8; 1.43.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.42	04-Mar-2007	christos	branches: 1.42.2; 1.42.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.41	09-Feb-2007	ad	branches: 1.41.2; Merge newlock2 to head.
1.40	16-Nov-2006	christos	branches: 1.40.2; 1.40.4; __unused removal on arguments; approved by core.
1.39	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.38	03-Sep-2006	bouyer	branches: 1.38.2; 1.38.4; stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.37	28-Aug-2006	christos	remove unreachable code.
1.36	11-Dec-2005	christos	branches: 1.36.4; 1.36.8; 1.36.18; merge ktrace-lwp.
1.35	30-May-2005	wiz	branches: 1.35.2; Fix replaco in comment.
1.34	30-May-2005	christos	- avoid variable shadowing - add const
1.33	29-Jun-2003	fvdl	branches: 1.33.2; 1.33.16; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.32	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.31	03-May-2003	wiz	DMA, not dma nor Dma.
1.30	28-Apr-2003	fvdl	Quell a strict-alias warning by applying the same cast to '&target' as already done in all other cases.
1.29	12-Mar-2003	wiz	Merge some more.
1.28	12-Mar-2003	wiz	Merge first bktr-sf import.
1.27	10-Jan-2003	mjl	Make this work on big endian platforms (mostly from OpenBSD): Uploaded scripts work better if they are little endian, as the card's engine expects, so convert to le first. Also clean up attach routine a bit (use pa_id and PCI_REVISION instead of fetching it ourselves). This makes the driver work on my macppc G4, it can decode and display video and tuner input. Sound does not seem to work, but this may be my wonky formac bktr-for-macintosh card.
1.26	25-Dec-2002	toshii	Fix broken programming of VSCALE_HI registers in yuv422_prog.
1.25	25-Dec-2002	toshii	Consistently use BKTR_GEN_IRQ and BKTR_RESYNC instead of numbers. Also, do some whitespace cleanup.
1.24	25-Dec-2002	toshii	Merge FreeBSD rev.1.122: Be consistent about "static" functions: if the function is marked static in its prototype, mark it static at the definition too.
1.23	26-Nov-2002	christos	si_ -> sel_
1.22	28-Aug-2002	toshii	Remove a superfluous interrupt request bit in yuvpack_prog, which confuses the state machine in common_bktr_intr. This makes "mplayer -tv ..." work (but only with NTSC).
1.21	22-May-2002	wiz	bcmp -> memcmp.
1.20	13-Nov-2001	lukem	branches: 1.20.8; add RCSIDs
1.19	28-Sep-2001	chs	don't depend on other headers to include sys/proc.h for us.
1.18	18-Sep-2001	wiz	Give initiali[sz]e all the "i"s it deserves.
1.17	03-Mar-2001	mjl	branches: 1.17.2; 1.17.4; 1.17.6; Correct wrong error return in mmap fn.
1.16	01-Mar-2001	mjl	Fix typo to make this compile even w/ debug defines.
1.15	18-Jan-2001	jdolecek	constify
1.14	30-Dec-2000	wiz	Merge 2.18 to trunk.
1.13	28-Oct-2000	wiz	branches: 1.13.2; Replace remaining vm_offset_t's in NetBSD code by vaddr_t; remove now unnecessary casts added by thorpej in 1.11 of bktr_core.c. Strike one from the TODO list.
1.12	28-Oct-2000	wiz	Merge 2.17 changes to trunk.
1.11	06-Sep-2000	thorpej	Cast the argument to vtophys() to vaddr_t.
1.10	03-Sep-2000	wiz	Use NetBSD's pcidevs file for vendors/cards as much as possible. Recognize Pinnacle Studio PCTV Rave as Miro card, tested by rh@netbsd.org.
1.9	01-Jul-2000	wiz	remove unnecessary #ifdef __NetBSD__'s in #ifdef __FreeBSD__; move one #include <vm/vm.h> into #ifndef __NetBSD__
1.8	01-Jul-2000	wiz	resolve conflicts.
1.7	30-Jun-2000	veego	Move the include of <vm/vm.h> into the ifndef __NetBSD__ part.
1.6	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.5	22-May-2000	wiz	branches: 1.5.4; Replaced overlooked BROOKTREE_SYSTEM_DEFAULT with BKTR_SYSTEM_DEFAULT, as noted by Roger Hardiman.
1.4	21-May-2000	wiz	Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.3	07-May-2000	veego	It is a bad idea to declare a static bootverbose variable here. Remove it.
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.5.4.1	03-Jul-2000	wiz	Pull up revision 1.9 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.13.2.6	12-Mar-2001	bouyer	Sync with HEAD.
1.13.2.5	11-Feb-2001	bouyer	Sync with HEAD.
1.13.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.13.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.13.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.13.2.1	28-Oct-2000	bouyer	file bktr_core.c was added on branch thorpej_scsipi on 2000-11-20 22:35:46 +0000
1.17.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.17.4.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.17.4.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.17.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.17.2.8	15-Jan-2003	thorpej	Sync with HEAD.
1.17.2.7	29-Dec-2002	thorpej	Sync with HEAD.
1.17.2.6	11-Dec-2002	thorpej	Sync with HEAD.
1.17.2.5	17-Sep-2002	nathanw	Catch up to -current.
1.17.2.4	20-Jun-2002	nathanw	Catch up to -current.
1.17.2.3	14-Nov-2001	nathanw	Catch up to -current.
1.17.2.2	08-Oct-2001	nathanw	Catch up to -current.
1.17.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.20.8.2	29-Aug-2002	gehenna	catch up with -current.
1.20.8.1	30-May-2002	gehenna	Catch up with -current.
1.33.16.2	18-Nov-2006	bouyer	Apply patch (requested by aymeric in ticket #1587) fix bug introduced by missing braces around a if() block in ticket #1511
1.33.16.1	16-Sep-2006	ghen	branches: 1.33.16.1.2; Pull up following revision(s) (requested by bouyer in ticket #1511): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.33.16.1.2.1	18-Nov-2006	bouyer	Apply patch (requested by aymeric in ticket #1587) fix bug introduced by missing braces around a if() block in ticket #1511
1.33.2.4	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.33.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.33.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.33.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.35.2.7	17-Mar-2008	yamt	sync with head.
1.35.2.6	21-Jan-2008	yamt	sync with head
1.35.2.5	27-Oct-2007	yamt	sync with head.
1.35.2.4	03-Sep-2007	yamt	sync with head.
1.35.2.3	26-Feb-2007	yamt	sync with head.
1.35.2.2	30-Dec-2006	yamt	sync with head.
1.35.2.1	21-Jun-2006	yamt	sync with head.
1.36.18.1	14-Sep-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #151): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.36.8.2	14-Sep-2006	yamt	sync with head.
1.36.8.1	03-Sep-2006	yamt	sync with head.
1.36.4.1	09-Sep-2006	rpaulo	sync with head
1.38.4.2	10-Dec-2006	yamt	sync with head.
1.38.4.1	22-Oct-2006	yamt	sync with head
1.38.2.2	19-Jan-2007	ad	Acquire proclist_mutex before sending signals.
1.38.2.1	18-Nov-2006	ad	Sync with head.
1.40.4.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.40.2.1	21-Jan-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1046): sys/dev/pci/bktr/bktr_reg.h: revision 1.18 sys/dev/pci/bktr/bktr_audio.c: revision 1.18 sys/dev/pci/bktr/bktr_os.c: revision 1.49 sys/dev/pci/bktr/bktr_core.c: revision 1.46 sys/dev/ic/bt8xx.h: revision 1.9 sys/dev/pci/bktr/bktr_core.h: revision 1.10 Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.41.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.42.4.1	11-Jul-2007	mjf	Sync with head.
1.42.2.2	23-Oct-2007	ad	Sync with head.
1.42.2.1	10-Apr-2007	ad	Changes to select/poll: - Make them MP safe and decouple from the proc locks. - selwakeup: don't call p_find, or traverse per-proc LWP lists (ouch). - selwakeup: don't lock the sleep queue unless we need to.
1.43.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.43.8.3	23-Mar-2008	matt	sync with HEAD
1.43.8.2	09-Jan-2008	matt	sync with HEAD
1.43.8.1	06-Nov-2007	matt	sync with HEAD
1.43.6.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.44.8.2	19-Jan-2008	bouyer	Sync with HEAD
1.44.8.1	08-Jan-2008	bouyer	Sync with HEAD
1.44.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.46.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.46.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.46.2.1	24-Mar-2008	keiichi	sync with head.
1.48.2.1	18-May-2008	yamt	sync with head.
1.49.16.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.49.10.1	28-Apr-2009	skrll	Sync with HEAD.
1.49.2.1	04-May-2009	yamt	sync with head.
1.50.4.1	05-Mar-2011	rmind	sync with head
1.52.12.2	25-Feb-2013	tls	resync with head
1.52.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.52.2.2	23-Jan-2013	yamt	sync with head
1.52.2.1	30-Oct-2012	yamt	sync with head
1.54.38.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.54.38.1	10-Jun-2019	christos	Sync with HEAD
1.54.36.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.56.6.1	29-Feb-2020	ad	Sync with head.
1.59.10.1	02-Aug-2025	perseant	Sync with HEAD
1.12	22-May-2022	andvar	fix various small typos, mainly in comments.
1.11	19-Dec-2021	andvar	fix various typos in bktr driver comments.
1.10	16-Jan-2008	jmcneill	Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.9	04-Mar-2007	christos	branches: 1.9.16; 1.9.22; 1.9.28; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.8	03-Sep-2006	bouyer	branches: 1.8.6; 1.8.8; 1.8.12; stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.7	11-Dec-2005	christos	branches: 1.7.4; 1.7.8; 1.7.18; merge ktrace-lwp.
1.6	29-Jun-2003	fvdl	branches: 1.6.2; 1.6.16; 1.6.18; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.5	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.4	12-Mar-2003	wiz	Merge first bktr-sf import.
1.3	01-Jul-2000	wiz	branches: 1.3.2; resolve conflicts.
1.2	21-May-2000	wiz	branches: 1.2.4; Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.3	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.2.4.1	03-Jul-2000	wiz	Pull up revision 1.3 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.3.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.2.1	01-Jul-2000	bouyer	file bktr_core.h was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.6.18.4	21-Jan-2008	yamt	sync with head
1.6.18.3	03-Sep-2007	yamt	sync with head.
1.6.18.2	30-Dec-2006	yamt	sync with head.
1.6.18.1	21-Jun-2006	yamt	sync with head.
1.6.16.1	16-Sep-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1511): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.6.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.7.18.1	14-Sep-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #151): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.7.8.1	14-Sep-2006	yamt	sync with head.
1.7.4.1	09-Sep-2006	rpaulo	sync with head
1.8.12.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.8.8.1	12-Mar-2007	rmind	Sync with HEAD.
1.8.6.1	21-Jan-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1046): sys/dev/pci/bktr/bktr_reg.h: revision 1.18 sys/dev/pci/bktr/bktr_audio.c: revision 1.18 sys/dev/pci/bktr/bktr_os.c: revision 1.49 sys/dev/pci/bktr/bktr_core.c: revision 1.46 sys/dev/ic/bt8xx.h: revision 1.9 sys/dev/pci/bktr/bktr_core.h: revision 1.10 Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.9.28.1	19-Jan-2008	bouyer	Sync with HEAD
1.9.22.1	18-Feb-2008	mjf	Sync with HEAD.
1.9.16.1	23-Mar-2008	matt	sync with HEAD
1.5	18-Mar-2009	cegger	bzero -> memset
1.4	04-Mar-2007	christos	branches: 1.4.40; 1.4.50; 1.4.56; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.3	12-Mar-2003	wiz	branches: 1.3.18; 1.3.56; Merge first bktr-sf import.
1.2	13-Nov-2001	lukem	add RCSIDs
1.1	28-Oct-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.2	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.1	28-Oct-2000	wiz	branches: 1.1.1.1.2; 1.1.1.1.4; 1.1.1.1.6; Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.1.6.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.1.1.1.4.1	14-Nov-2001	nathanw	Catch up to -current.
1.1.1.1.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.1.1.2.1	28-Oct-2000	bouyer	file bktr_mem.c was added on branch thorpej_scsipi on 2000-11-22 16:04:30 +0000
1.3.56.1	12-Mar-2007	rmind	Sync with HEAD.
1.3.18.1	03-Sep-2007	yamt	sync with head.
1.4.56.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.4.50.1	28-Apr-2009	skrll	Sync with HEAD.
1.4.40.1	04-May-2009	yamt	sync with head.
1.2	12-Mar-2003	wiz	Merge first bktr-sf import.
1.1	28-Oct-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.2	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.1	28-Oct-2000	wiz	branches: 1.1.1.1.2; Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.1.2.2	22-Nov-2000	bouyer	Sync with HEAD.
1.1.1.1.2.1	28-Oct-2000	bouyer	file bktr_mem.h was added on branch thorpej_scsipi on 2000-11-22 16:04:31 +0000
1.69	23-May-2022	rin	Audit unload/unmap v.s. free against DMA buffer for sys/dev/pci; make sure that bus_dmamap_unload(9) [or bus_dmamap_destroy(9)] or bus_dmamem_unmap(9) are preceding to freeing DMA buffer, if it is loaded or mapped, respectively. This is mandatory for some archs. See, e.g.: http://www.nerv.org/netbsd/?q=id:20210511T013030Z.013443cc790088147e4beed43f53dedabeaf9312 http://www.nerv.org/netbsd/?q=id:20220511T172220Z.561179f0b6fcc5b9cd73e274f69d74e2ce9e4c93 XXX XXX XXX Compile test only (for amd64/ALL). Thanks riastradh@ for double check.
1.68	09-Dec-2018	jdolecek	use pci_intr_establish_xname()
1.67	14-Jul-2016	msaitoh	branches: 1.67.16; 1.67.18; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.66	07-Jul-2016	msaitoh	KNF. Remove extra spaces. No functional change.
1.65	25-Jul-2014	dholland	branches: 1.65.4; Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.64	29-Mar-2014	christos	branches: 1.64.2; make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.63	16-Mar-2014	dholland	Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.62	10-Aug-2013	dholland	Redo -r1.53 correctly.
1.61	27-Oct-2012	chs	branches: 1.61.2; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.60	30-Jun-2011	wiz	branches: 1.60.2; 1.60.12; dependant -> dependent
1.59	23-Apr-2010	macallan	don't try to print a bus_space_handle_t now this at least compiles on sparc64
1.58	22-Mar-2010	dyoung	MI code must not printf() a bus_space_tag_t, so don't do that.
1.57	06-May-2009	cegger	branches: 1.57.2; 1.57.4; struct device * -> device_t, no functional changes intended.
1.56	06-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.55	15-Mar-2009	cegger	ansify function definitions
1.54	14-Mar-2009	dsl	Change about 4500 of the K&R function definitions to ANSI ones. There are still about 1600 left, but they have ',' or /* ... */ in the actual variable definitions - which my awk script doesn't handle. There are also many that need () -> (void). (The script does handle misordered arguments.)
1.53	24-Jun-2008	gmcgarry	branches: 1.53.4; 1.53.10; Both legs of a digraph must return the same type. Therefore, replace digraph.
1.52	08-Jun-2008	tsutsui	branches: 1.52.2; Use device_lookup_private() rather than using cd_devs[] directly to get softc. XXX maybe we should change a type of cd_devs[] in struct cfdriver from (void *) to device_t.
1.51	10-Apr-2008	cegger	branches: 1.51.2; 1.51.4; 1.51.6; use aprint_*_dev and device_xname
1.50	01-Mar-2008	rmind	Welcome to 4.99.55: - Add a lot of missing selinit() and seldestroy() calls. - Merge selwakeup() and selnotify() calls into a single selnotify(). - Add an additional 'events' argument to selnotify() call. It will indicate which event (POLL_IN, POLL_OUT, etc) happen. If unknown, zero may be used. Note: please pass appropriate value of 'events' where possible. Proposed on: <tech-kern>
1.49	16-Jan-2008	jmcneill	branches: 1.49.2; 1.49.6; Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.48	19-Oct-2007	ad	branches: 1.48.2; 1.48.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.47	25-Sep-2007	ad	branches: 1.47.2; Use selinit() / seldestroy().
1.46	04-Mar-2007	christos	branches: 1.46.2; 1.46.14; 1.46.16; 1.46.18; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.45	16-Nov-2006	christos	branches: 1.45.2; 1.45.4; 1.45.8; __unused removal on arguments; approved by core.
1.44	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.43	03-Sep-2006	bouyer	branches: 1.43.2; 1.43.4; stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.42	28-Aug-2006	christos	remove unreachable code.
1.41	28-Aug-2006	christos	add missing initializer.
1.40	28-Mar-2006	thorpej	branches: 1.40.6; Use device_unit().
1.39	11-Dec-2005	christos	branches: 1.39.4; 1.39.6; 1.39.8; 1.39.10; 1.39.12; merge ktrace-lwp.
1.38	29-Oct-2004	yamt	branches: 1.38.10; 1.38.12; constify audio_hw_if, midi_hw_if, and radio_hw_if.
1.37	29-Jun-2003	fvdl	branches: 1.37.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.36	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.35	12-Mar-2003	wiz	Merge first bktr-sf import.
1.34	10-Jan-2003	mjl	Make this work on big endian platforms (mostly from OpenBSD): Uploaded scripts work better if they are little endian, as the card's engine expects, so convert to le first. Also clean up attach routine a bit (use pa_id and PCI_REVISION instead of fetching it ourselves). This makes the driver work on my macppc G4, it can decode and display video and tuner input. Sound does not seem to work, but this may be my wonky formac bktr-for-macintosh card.
1.33	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.32	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.31	01-Oct-2002	thorpej	Use CFATTACH_DECL().
1.30	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.29	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.28	07-Jan-2002	jmcneill	branches: 1.28.8; Re-enable radio at bktr, only if NRADIO > 0 and an FM tuner is present.
1.27	07-Jan-2002	tron	Disable broken radio(4) support for bktr(4). The configuration code doesn't even check if the TV card has a radio tuner.
1.26	07-Jan-2002	tron	Clean up preprocess handling of optional radio(4) support a little bit.
1.25	06-Jan-2002	tron	Compile in radio(4) only it is activated in the kernel configuration.
1.24	06-Jan-2002	jmcneill	Add missing definition of STATUSBIT_STEREO.
1.23	06-Jan-2002	jmcneill	Add basic support for the radio(4) interface.
1.22	13-Nov-2001	lukem	add RCSIDs
1.21	18-Oct-2001	wiz	Fix check in device open. Patch by KUROSAWA Takahiro in kern/14289.
1.20	18-Sep-2001	wiz	Give initiali[sz]e all the "i"s it deserves.
1.19	15-Mar-2001	ross	branches: 1.19.2; 1.19.4; LP64 a kprintf(9)
1.18	03-Mar-2001	mjl	branches: 1.18.2; Correct wrong error return in mmap fn.
1.17	30-Dec-2000	wiz	Undo pci_intr_map change in non-NetBSD part of the file.
1.16	30-Dec-2000	wiz	Merge 2.18 to trunk.
1.15	28-Dec-2000	sommerfeld	Change pci_intr_map to get interrupt source information from a "struct pci_attach_args *" instead of from four separate parameters which in all cases were extracted from the same "struct pci_attach_args". This both simplifies the driver api, and allows for alternate PCI interrupt mapping schemes, such as one using the tables described in the Intel Multiprocessor Spec which describe interrupt wirings for devices behind pci-pci bridges based on the device's location rather the bridge's location. Tested on alpha and i386; welcome to 1.5Q
1.14	28-Oct-2000	wiz	branches: 1.14.2; Replace remaining vm_offset_t's in NetBSD code by vaddr_t; remove now unnecessary casts added by thorpej in 1.11 of bktr_core.c. Strike one from the TODO list.
1.13	28-Oct-2000	wiz	Merge 2.17 changes to trunk.
1.12	03-Sep-2000	wiz	Use NetBSD's pcidevs file for vendors/cards as much as possible. Recognize Pinnacle Studio PCTV Rave as Miro card, tested by rh@netbsd.org.
1.11	01-Jul-2000	wiz	remove unnecessary #ifdef __NetBSD__'s in #ifdef __FreeBSD__; move one #include <vm/vm.h> into #ifndef __NetBSD__
1.10	01-Jul-2000	wiz	resolve conflicts.
1.9	30-Jun-2000	veego	Move the include of <vm/vm.h> into the ifndef __NetBSD__ part.
1.8	27-Jun-2000	simonb	Revert some of the previous mmap() changes that were in "#ifdef FreeBSD" parts of this file.
1.7	26-Jun-2000	mrg	remove/move more mach vm header files: <vm/pglist.h> -> <uvm/uvm_pglist.h> <vm/vm_inherit.h> -> <uvm/uvm_inherit.h> <vm/vm_kern.h> -> into <uvm/uvm_extern.h> <vm/vm_object.h> -> nothing <vm/vm_pager.h> -> into <uvm/uvm_pager.h> also includes a bunch of <vm/vm_page.h> include removals (due to redudancy with <vm/vm.h>), and a scattering of other similar headers.
1.6	26-Jun-2000	simonb	Change the kernel mmap interface so that the offset to map is an "off_t" and the return value is a "paddr_t" to allow mappings at offsets past 2^31 bytes. Somewhat inspired by FreeBSD, which only changed the offset to a "vm_offset_t". Includes updates for the i386, pc532 and sh3 mmmmap from Jason Thorpe.
1.5	21-May-2000	wiz	branches: 1.5.4; Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.4	07-May-2000	veego	It is a bad idea to declare a static bootverbose variable here. Remove it.
1.3	07-May-2000	wiz	compilation fix
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.5.4.3	24-Oct-2001	jhawk	Pull up revision 1.21 (requested by lukem): Check unit count correctly (correct fencepost error) on device open.
1.5.4.2	03-Jul-2000	wiz	Pull up revision 1.11 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.5.4.1	30-Jun-2000	simonb	Pull up mmap paddr_t/off_t changes from trunk.
1.14.2.6	27-Mar-2001	bouyer	Sync with HEAD.
1.14.2.5	12-Mar-2001	bouyer	Sync with HEAD.
1.14.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.14.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.14.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.14.2.1	28-Oct-2000	bouyer	file bktr_os.c was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.18.2.10	15-Jan-2003	thorpej	Sync with HEAD.
1.18.2.9	11-Nov-2002	nathanw	Catch up to -current
1.18.2.8	18-Oct-2002	nathanw	Catch up to -current.
1.18.2.7	17-Sep-2002	nathanw	Catch up to -current.
1.18.2.6	28-Feb-2002	nathanw	Catch up to -current.
1.18.2.5	11-Jan-2002	nathanw	More catchup.
1.18.2.4	14-Nov-2001	nathanw	Catch up to -current.
1.18.2.3	22-Oct-2001	nathanw	Catch up to -current.
1.18.2.2	21-Sep-2001	nathanw	Catch up to -current.
1.18.2.1	09-Apr-2001	nathanw	Catch up with -current.
1.19.4.3	01-Oct-2001	fvdl	Catch up with -current.
1.19.4.2	26-Sep-2001	fvdl	* add a VCLONED vnode flag that indicates a vnode representing a cloned device. * rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass to VOP_REVOKE * the revoke system call will revoke all aliases, as before, but not the clones * vdevgone is called when detaching a device, so make it use REVOKECLONE to get rid of all clones as well * clean up all uses of VOP_OPEN wrt. locking. * add a few VOPS to spec_vnops that need to do something when it's a clone vnode (access and getattr) * add a copy of the vnode vattr structure of the original 'master' vnode to the specinfo of a cloned vnode. could possibly redirect getattr to the 'master' vnode, but this has issues with revoke * add a vdev_reassignvp function that disassociates a vnode from its original device, and reassociates it with the specified dev_t. to be used by cloning devices only, in case a new minor is allocated. * change all direct references in drivers to v_devcookie and v_rdev to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes when debugging race conditions that still exist wrt. locking and revoking vnodes. * make the locking state of a vnode consistent when passed to d_open and d_close (unlocked). locked would be better, but has some deadlock issues
1.19.4.1	07-Sep-2001	thorpej	Commit my "devvp" changes to the thorpej-devvp branch. This replaces the use of dev_t in most places with a struct vnode *. This will form the basic infrastructure for real cloning device support (besides being architecurally cleaner -- it'll be good to get away from using numbers to represent objects).
1.19.2.2	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.19.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.28.8.1	16-May-2002	gehenna	Add the character device switch.
1.37.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.37.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.37.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.37.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.38.12.6	17-Mar-2008	yamt	sync with head.
1.38.12.5	21-Jan-2008	yamt	sync with head
1.38.12.4	27-Oct-2007	yamt	sync with head.
1.38.12.3	03-Sep-2007	yamt	sync with head.
1.38.12.2	30-Dec-2006	yamt	sync with head.
1.38.12.1	21-Jun-2006	yamt	sync with head.
1.38.10.1	16-Sep-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1511): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.39.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.39.10.1	19-Apr-2006	elad	sync with head.
1.39.8.3	14-Sep-2006	yamt	sync with head.
1.39.8.2	03-Sep-2006	yamt	sync with head.
1.39.8.1	01-Apr-2006	yamt	sync with head.
1.39.6.1	22-Apr-2006	simonb	Sync with head.
1.39.4.1	09-Sep-2006	rpaulo	sync with head
1.40.6.1	14-Sep-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #151): sys/dev/pci/bktr/bktr_core.c: revision 1.38 sys/dev/pci/bktr/bktr_os.c: revision 1.43 sys/dev/pci/bktr/bktr_core.h: revision 1.8 stop attaching when a bus_dma operation fails, instead of using bogus addresses. Remplace vtophys() with proper bus_dma(9) use. This driver now works properly on Xen
1.43.4.2	10-Dec-2006	yamt	sync with head.
1.43.4.1	22-Oct-2006	yamt	sync with head
1.43.2.1	18-Nov-2006	ad	Sync with head.
1.45.8.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.45.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.45.2.1	21-Jan-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1046): sys/dev/pci/bktr/bktr_reg.h: revision 1.18 sys/dev/pci/bktr/bktr_audio.c: revision 1.18 sys/dev/pci/bktr/bktr_os.c: revision 1.49 sys/dev/pci/bktr/bktr_core.c: revision 1.46 sys/dev/ic/bt8xx.h: revision 1.9 sys/dev/pci/bktr/bktr_core.h: revision 1.10 Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.46.18.1	06-Oct-2007	yamt	sync with head.
1.46.16.2	23-Mar-2008	matt	sync with HEAD
1.46.16.1	06-Nov-2007	matt	sync with HEAD
1.46.14.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.46.14.1	02-Oct-2007	joerg	Sync with HEAD.
1.46.2.2	23-Oct-2007	ad	Sync with head.
1.46.2.1	09-Oct-2007	ad	Sync with head.
1.47.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.48.8.1	19-Jan-2008	bouyer	Sync with HEAD
1.48.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.49.6.4	29-Jun-2008	mjf	Sync with HEAD.
1.49.6.3	02-Jun-2008	mjf	Sync with HEAD.
1.49.6.2	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.49.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.49.2.1	24-Mar-2008	keiichi	sync with head.
1.51.6.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.51.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.51.4.3	11-Aug-2010	yamt	sync with head.
1.51.4.2	16-May-2009	yamt	sync with head
1.51.4.1	04-May-2009	yamt	sync with head.
1.51.2.1	17-Jun-2008	yamt	sync with head.
1.52.2.1	27-Jun-2008	simonb	Sync with head.
1.53.10.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.53.4.1	28-Apr-2009	skrll	Sync with HEAD.
1.57.4.1	30-May-2010	rmind	sync with head
1.57.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.60.12.3	03-Dec-2017	jdolecek	update from HEAD
1.60.12.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.60.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.60.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.60.2.1	30-Oct-2012	yamt	sync with head
1.61.2.2	18-May-2014	rmind	sync with head
1.61.2.1	28-Aug-2013	rmind	sync with head
1.64.2.1	10-Aug-2014	tls	Rebase.
1.65.4.2	05-Oct-2016	skrll	Sync with HEAD
1.65.4.1	09-Jul-2016	skrll	Sync with HEAD
1.67.18.1	10-Jun-2019	christos	Sync with HEAD
1.67.16.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7	30-Jun-2011	wiz	dependant -> dependent
1.6	12-Mar-2003	wiz	Merge first bktr-sf import.
1.5	16-Mar-2001	ross	Sigh. Revert the revert. Naturally, I considered OpenBSD and FreeBSD when I fixed the incorrect use of the spl() interface. The change I made is _required_ for both NetBSD _and_ OpenBSD, or the code won't even COMPILE except on i386, and it is acceptable on FreeBSD. Your revert and mod rebroke it on OpenBSD and tangled things up on NetBSD. It made no difference on FreeBSD. In particular, there are 2,895 uses of splx() within the FreeBSD kernel, and only a mere 21, that's "twenty one" uses of intrmask_t, and those are almost exclusively in the guts of the interrupt implementation, _not_ in the _use_ of the exported spl() functions. It's perfectly OK to `int s = spltty()' in a portable driver in FreeBSD. For that matter, FreeBSD (-current at least) does not even use spl*() any more and stubs them all out with inlines that do _nothing_ except return 0, making intrmask_t vs int _even less_ important there than it already was. I think it's great that you want to start hacking on the kernel, but do note that this is certainly the most simple of the kernel interfaces. It just gets worse from here. Be careful out there!
1.4	16-Mar-2001	wiz	Restore non-NetBSD parts -- we're still sharing this code.
1.3	15-Mar-2001	ross	Do critical sections a better way.
1.2	28-Oct-2000	wiz	branches: 1.2.2; 1.2.4; Replace remaining vm_offset_t's in NetBSD code by vaddr_t; remove now unnecessary casts added by thorpej in 1.11 of bktr_core.c. Strike one from the TODO list.
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.3	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	branches: 1.1.1.1.4; Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.1.1.1.4.1	03-Jul-2000	wiz	Pull up revision 1.1.1.2 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.2.4.1	09-Apr-2001	nathanw	Catch up with -current.
1.2.2.4	27-Mar-2001	bouyer	Sync with HEAD.
1.2.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.2.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.2.2.1	28-Oct-2000	bouyer	file bktr_os.h was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.22	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.21	26-Jan-2011	dyoung	branches: 1.21.4; 1.21.14; unifdef -U__bsdi__ -U__FreeBSD__ -D__NetBSD__. No functional change intended.
1.20	06-May-2009	cegger	branches: 1.20.4; 1.20.6; 1.20.8; struct device * -> device_t, no functional changes intended.
1.19	24-Apr-2008	ad	branches: 1.19.2; 1.19.16; Network protocol interrupts can now block on locks, so merge the globals proclist_mutex and proclist_lock into a single adaptive mutex (proc_lock). Implications: - Inspecting process state requires thread context, so signals can no longer be sent from a hardware interrupt handler. Signal activity must be deferred to a soft interrupt or kthread. - As the proc state locking is simplified, it's now safe to take exit() and wait() out from under kernel_lock. - The system spends less time at IPL_SCHED, and there is less lock activity.
1.18	16-Jan-2008	jmcneill	branches: 1.18.6; 1.18.8; Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.17	19-Oct-2007	ad	branches: 1.17.2; 1.17.8; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.16	30-Dec-2005	mjl	branches: 1.16.24; 1.16.30; 1.16.38; 1.16.44; 1.16.46; 1.16.50; This part got lost during merging with the sourceforge bktr driver: "Also make those variables int instead of char so it can actually hold 454...".
1.15	11-Dec-2005	christos	merge ktrace-lwp.
1.14	30-May-2005	christos	branches: 1.14.2; - avoid variable shadowing - add const
1.13	12-Mar-2003	wiz	branches: 1.13.2; Merge first bktr-sf import.
1.12	10-Jan-2003	mjl	Make this work on big endian platforms (mostly from OpenBSD): Uploaded scripts work better if they are little endian, as the card's engine expects, so convert to le first. Also clean up attach routine a bit (use pa_id and PCI_REVISION instead of fetching it ourselves). This makes the driver work on my macppc G4, it can decode and display video and tuner input. Sound does not seem to work, but this may be my wonky formac bktr-for-macintosh card.
1.11	23-Oct-2002	christos	remove duplicate cpp defs.
1.10	18-Sep-2001	wiz	Give initiali[sz]e all the "i"s it deserves.
1.9	30-Dec-2000	wiz	branches: 1.9.2; 1.9.4; 1.9.6; Merge 2.18 to trunk.
1.8	28-Oct-2000	wiz	branches: 1.8.2; Merge 2.17 changes to trunk.
1.7	24-Sep-2000	jdolecek	use new generic bootverbose instead of local definition
1.6	03-Sep-2000	wiz	Use NetBSD's pcidevs file for vendors/cards as much as possible. Recognize Pinnacle Studio PCTV Rave as Miro card, tested by rh@netbsd.org.
1.5	01-Jul-2000	wiz	resolve conflicts.
1.4	07-May-2000	veego	branches: 1.4.4; Change one 0 for bootverbose to 1 if DEBUG is defined.
1.3	07-May-2000	veego	Add a dummy bootverbose.
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.4.4.1	03-Jul-2000	wiz	Pull up revision 1.5 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.8.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.8.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.8.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.8.2.1	28-Oct-2000	bouyer	file bktr_reg.h was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.9.6.1	01-Oct-2001	fvdl	Catch up with -current.
1.9.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.9.2.3	15-Jan-2003	thorpej	Sync with HEAD.
1.9.2.2	11-Nov-2002	nathanw	Catch up to -current
1.9.2.1	21-Sep-2001	nathanw	Catch up to -current.
1.13.2.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.14.2.3	21-Jan-2008	yamt	sync with head
1.14.2.2	27-Oct-2007	yamt	sync with head.
1.14.2.1	21-Jun-2006	yamt	sync with head.
1.16.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.16.46.2	23-Mar-2008	matt	sync with HEAD
1.16.46.1	06-Nov-2007	matt	sync with HEAD
1.16.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.16.38.1	03-Jun-2008	skrll	Sync with netbsd-4.
1.16.30.1	23-Oct-2007	ad	Sync with head.
1.16.24.1	21-Jan-2008	bouyer	Pull up following revision(s) (requested by jmcneill in ticket #1046): sys/dev/pci/bktr/bktr_reg.h: revision 1.18 sys/dev/pci/bktr/bktr_audio.c: revision 1.18 sys/dev/pci/bktr/bktr_os.c: revision 1.49 sys/dev/pci/bktr/bktr_core.c: revision 1.46 sys/dev/ic/bt8xx.h: revision 1.9 sys/dev/pci/bktr/bktr_core.h: revision 1.10 Don't assume that sizeof(long) == sizeof(int); makes bktr work on amd64.
1.17.8.1	19-Jan-2008	bouyer	Sync with HEAD
1.17.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.18.8.1	18-May-2008	yamt	sync with head.
1.18.6.1	02-Jun-2008	mjf	Sync with HEAD.
1.19.16.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.19.2.1	16-May-2009	yamt	sync with head
1.20.8.1	08-Feb-2011	bouyer	Sync with HEAD
1.20.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.20.4.1	05-Mar-2011	rmind	sync with head
1.21.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.21.4.1	30-Oct-2012	yamt	sync with head
1.13	19-Dec-2021	andvar	fix various typos in bktr driver comments.
1.12	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.11	28-Dec-2005	wiz	branches: 1.11.30; 1.11.44; 1.11.46; 1.11.50; Sync with SourceForge 1.5: revision 1.5 date: 2003/03/13 22:21:04; author: emsi; state: Exp; lines: +17 -18 Fill in band switch crosspoints and use that instead of hardcoded values in the band switching code. For now fill in the formerly hard-coded values 160 and 454 until we can find out the real values for each tuner. Also make those variables int instead of char so it can actually hold 454...
1.10	12-Mar-2003	wiz	branches: 1.10.18; Merge first bktr-sf import.
1.9	13-Nov-2001	lukem	add RCSIDs
1.8	18-Jan-2001	jdolecek	branches: 1.8.2; 1.8.4; constify
1.7	30-Dec-2000	wiz	Merge 2.18 to trunk.
1.6	28-Oct-2000	wiz	branches: 1.6.2; Merge 2.17 changes to trunk.
1.5	01-Jul-2000	wiz	resolve conflicts.
1.4	25-Jun-2000	wiz	printf argument was in wrong position. Noted by Roger Hardiman <roger@cs.strath.ac.uk>. (This code was ifdef'd out by default.)
1.3	21-May-2000	wiz	branches: 1.3.4; Use bktr_softc.bktr_dev.dv_xname instead of "bktr0" when printf'ing as requested by Klaus Klein in kern/10166. Also add device instance name to all printf's that didn't have it.
1.2	07-May-2000	wiz	Some NetBSD specific changes; also TODO, and import script, based on bind2netbsd by Bernd Ernesti
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.5	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.4	30-Dec-2000	wiz	Import release 2.18 of the bktr driver. Changes since 2.17: o Add new ioctls to allow userland programs fuller access to the I2C bus, BT848_MSP_READ BT848_MSP_WRITE and BT848_MSP_RESET. o Add tuner type 0x2a. Remove 0x2c which was incorrect. o Christian Gusenbauer <Christian.Gusenbauer@netway.at> added support for audio on Hauppauge cards without the audio mux. The MSP is used for audio selection. (the 44xxx models)
1.1.1.3	28-Oct-2000	wiz	Import release 2.17 of the bktr driver. Changes: Integration of some NetBSD changes. (FreeBSD version now is also usable as a KLD (~= LKM).)
1.1.1.2	01-Jul-2000	wiz	Import release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types.
1.1.1.1	07-May-2000	wiz	Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.3.4.1	03-Jul-2000	wiz	Pull up revision 1.5 (requested by wiz): Sync with release 2.14 of the bktr driver. Changes: mostly integration of NetBSD changes; support for new Hauppauge tuner types. Approved by thorpej.
1.6.2.5	11-Feb-2001	bouyer	Sync with HEAD.
1.6.2.4	05-Jan-2001	bouyer	Sync with HEAD
1.6.2.3	22-Nov-2000	bouyer	Sync with HEAD.
1.6.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.6.2.1	28-Oct-2000	bouyer	file bktr_tuner.c was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.8.4.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.8.2.1	14-Nov-2001	nathanw	Catch up to -current.
1.10.18.2	27-Oct-2007	yamt	sync with head.
1.10.18.1	21-Jun-2006	yamt	sync with head.
1.11.50.1	25-Oct-2007	bouyer	Sync with HEAD.
1.11.46.1	06-Nov-2007	matt	sync with HEAD
1.11.44.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.11.30.1	23-Oct-2007	ad	Sync with head.
1.2	12-Mar-2003	wiz	Merge first bktr-sf import.
1.1	07-May-2000	wiz	branches: 1.1.1; Initial revision
1.1.1.2	12-Mar-2003	wiz	Import of current state of bktr(4) in sourceforge repository. Changes since last import: . lots of whitespace cleanups . typo fixes (e.g. hz, compatibilty) . fix brightness ioctl return value . wait for int ready using DELAY() instead of tight loop
1.1.1.1	07-May-2000	wiz	branches: 1.1.1.1.6; Import of (finally busspaced) bktr-driver from FreeBSD (part 1). This version is from the location specified in the README, modified by bktr2netbsd, to be committed later.
1.1.1.1.6.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.1.1.1.6.1	07-May-2000	bouyer	file bktr_tuner.h was added on branch thorpej_scsipi on 2000-11-20 22:35:47 +0000
1.5	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.4	19-Apr-2018	christos	branches: 1.4.2; s/static inline/static __inline/g for consistency.
1.3	27-Oct-2012	chs	branches: 1.3.36; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.2	18-May-2011	dyoung	branches: 1.2.4; 1.2.14; MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_adapter.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_adapter.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_adapter.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.2.4.1	30-Oct-2012	yamt	sync with head
1.3.36.1	22-Apr-2018	pgoyette	Sync with HEAD
1.4.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_ael1002.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_ael1002.c was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_ael1002.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.4	29-Jul-2022	andvar	s/collissions/collisions/
1.3	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.2	23-Jan-2013	joerg	branches: 1.2.36; Make cxgb at least compilable. Fix _rt_key usage.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.14; 1.1.24; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.24.1	25-Feb-2013	tls	resync with head
1.1.14.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_common.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_common.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_common.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2.36.1	22-Apr-2018	pgoyette	Sync with HEAD
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_config.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_config.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_config.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_ctl_defs.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_ctl_defs.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_ctl_defs.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2	02-Jun-2024	andvar	Fix various typos, mainly triple letters.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_firmware_exports.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_firmware_exports.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_firmware_exports.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2	11-Apr-2019	kamil	Fix CVS Id usage
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.66; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.66.1	10-Jun-2019	christos	Sync with HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_include.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_include.h was added on branch rmind-uvmplock on 2010-05-30 05:17:39 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_include.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2	06-Sep-2015	dholland	More on PR 41200: headers that declare ioctls should include sys/ioccom.h. This covers (I think) all the MI headers outside of external/ (and dist/).
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.24; 1.1.42; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.42.1	22-Sep-2015	skrll	Sync with HEAD
1.1.24.1	03-Dec-2017	jdolecek	update from HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_ioctl.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_ioctl.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_ioctl.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.3	19-Jun-2025	andvar	s/partilar/particular/ in comment.
1.2	19-Apr-2018	christos	branches: 1.2.40; s/static inline/static __inline/g for consistency.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.64; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.64.1	22-Apr-2018	pgoyette	Sync with HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_jhash.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_jhash.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_jhash.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2.40.1	02-Aug-2025	perseant	Sync with HEAD
1.5	22-Dec-2018	maxv	Replace M_ALIGN and MH_ALIGN by m_align.
1.4	11-Oct-2016	maxv	branches: 1.4.14; 1.4.16; There are two memory leaks here, found by mootja; just add some XXXs.
1.3	25-Mar-2014	christos	branches: 1.3.6; 1.3.10; kill sprintf
1.2	23-Jan-2013	joerg	branches: 1.2.2; Make cxgb at least compilable. Fix _rt_key usage.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.14; 1.1.24; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.24.3	03-Dec-2017	jdolecek	update from HEAD
1.1.24.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.24.1	25-Feb-2013	tls	resync with head
1.1.14.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_l2t.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_l2t.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_l2t.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2.2.1	18-May-2014	rmind	sync with head
1.3.10.1	04-Nov-2016	pgoyette	Sync with HEAD
1.3.6.1	05-Dec-2016	skrll	Sync with HEAD
1.4.16.1	10-Jun-2019	christos	Sync with HEAD
1.4.14.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.2	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.64; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.64.1	22-Apr-2018	pgoyette	Sync with HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_l2t.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_l2t.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_l2t.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2	18-May-2011	dyoung	MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_lro.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_lro.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_lro.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.10	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.9	24-Apr-2021	thorpej	branches: 1.9.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.8	04-Feb-2020	thorpej	branches: 1.8.8; Use ifmedia_fini().
1.7	09-Dec-2018	jdolecek	branches: 1.7.6; use pci_intr_establish_xname() XXX due to 2008-01-17 '10 gigabit Chelsio cards now compile and work, although XXX not very well yet' commit this looks like having worked in past, but since XXX ~2010 received only mechanical updates, and currently it's unreferenced XXX and uncompilable - does it make sense to keep the driver around?
1.6	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.5	26-Jun-2018	msaitoh	branches: 1.5.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.4	23-Jan-2013	joerg	branches: 1.4.36; Make cxgb at least compilable. Fix _rt_key usage.
1.3	27-Oct-2012	chs	split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.2	18-May-2011	dyoung	branches: 1.2.4; 1.2.14; MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_main.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:05 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_main.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_main.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2.14.2	25-Feb-2013	tls	resync with head
1.2.14.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.2.4.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.1	30-Oct-2012	yamt	sync with head
1.4.36.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.4.36.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.4.36.1	28-Jul-2018	pgoyette	Sync with HEAD
1.5.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.5.2.1	10-Jun-2019	christos	Sync with HEAD
1.7.6.1	29-Feb-2020	ad	Sync with head.
1.8.8.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.8.8.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.9.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_mbuf.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_mbuf.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_mbuf.h was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.2	08-Feb-2018	dholland	Typos.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_mc5.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_mc5.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_mc5.c was added on branch uebayasi-xip on 2010-04-30 14:43:44 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_mv88e1xxx.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_mv88e1xxx.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_mv88e1xxx.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.3	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.2	02-Oct-2016	christos	branches: 1.2.14; MFREE -> m_free
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.24; 1.1.42; 1.1.46; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.46.1	04-Nov-2016	pgoyette	Sync with HEAD
1.1.42.1	05-Oct-2016	skrll	Sync with HEAD
1.1.24.1	03-Dec-2017	jdolecek	update from HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_mvec.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_mvec.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_mvec.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2.14.1	22-Apr-2018	pgoyette	Sync with HEAD
1.7	08-Aug-2021	andvar	s/propely/properly/
1.6	27-Dec-2019	msaitoh	s/suport/support/
1.5	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.4	21-Sep-2014	christos	branches: 1.4.18; 1.4.20; comment out code that was never used, but probably intended to be used.
1.3	25-Mar-2014	christos	branches: 1.3.4; kill sprintf
1.2	18-May-2011	dyoung	branches: 1.2.4; 1.2.14; 1.2.18; MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_offload.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_offload.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_offload.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2.18.1	18-May-2014	rmind	sync with head
1.2.14.2	03-Dec-2017	jdolecek	update from HEAD
1.2.14.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.4.1	30-Oct-2014	martin	Pull up following revision(s) (requested by maxv in ticket #164): sys/dev/ieee1394/fwcrom.c: revision 1.15 sys/dev/ieee1394/fwcrom.c: revision 1.16 sys/dev/marvell/if_gfe.c: revision 1.43 sys/dev/usb/stuirda.c: revision 1.16 sys/dev/pci/cxgb/cxgb_offload.c: revision 1.4 sys/dev/pci/twa.c: revision 1.51 sys/dev/pci/twa.c: revision 1.52 Various fixes in dev/: remove dead code and fix two inconsistencies.
1.4.20.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.4.20.1	10-Jun-2019	christos	Sync with HEAD
1.4.18.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.2	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.64; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.64.1	22-Apr-2018	pgoyette	Sync with HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_offload.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_offload.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_offload.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2	18-May-2011	dyoung	MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_osdep.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_osdep.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_osdep.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.7	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.6	19-Apr-2018	christos	branches: 1.6.2; s/static inline/static __inline/g for consistency.
1.5	25-May-2016	ozaki-r	branches: 1.5.16; Use M_GETCTX and M_SETCTX No functional change.
1.4	20-Mar-2014	skrll	branches: 1.4.6; Mechanically replace simplelock with kmutex_t.
1.3	23-Jan-2013	joerg	branches: 1.3.2; Make cxgb at least compilable. Fix _rt_key usage.
1.2	18-May-2011	dyoung	branches: 1.2.4; 1.2.14; MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_osdep.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_osdep.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_osdep.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2.14.3	03-Dec-2017	jdolecek	update from HEAD
1.2.14.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.2.14.1	25-Feb-2013	tls	resync with head
1.2.4.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.2.1	18-May-2014	rmind	sync with head
1.4.6.1	29-May-2016	skrll	Sync with HEAD
1.5.16.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.5.16.1	22-Apr-2018	pgoyette	Sync with HEAD
1.6.2.1	10-Jun-2019	christos	Sync with HEAD
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_regs.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_regs.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_regs.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.8	28-May-2025	andvar	fix few typos in comments.
1.7	22-Dec-2018	maxv	branches: 1.7.36; Replace M_ALIGN and MH_ALIGN by m_align.
1.6	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.5	26-Sep-2017	knakahara	branches: 1.5.2; 1.5.4; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.4	10-Jun-2016	ozaki-r	branches: 1.4.10; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.3	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.2	18-May-2011	dyoung	branches: 1.2.14; 1.2.32; MI code must #include <sys/bus.h>, not <machine/bus.h>.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.8.1	06-Jun-2011	jruoho	Sync with HEAD.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_sge.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.3	31-May-2011	rmind	sync with head
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_sge.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_sge.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2.32.2	09-Jul-2016	skrll	Sync with HEAD
1.2.32.1	19-Mar-2016	skrll	Sync with HEAD
1.2.14.1	03-Dec-2017	jdolecek	update from HEAD
1.4.10.1	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.5.4.1	10-Jun-2019	christos	Sync with HEAD
1.5.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.5.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.7.36.1	02-Aug-2025	perseant	Sync with HEAD
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_sge_defs.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_sge_defs.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_sge_defs.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_t3_cpl.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_t3_cpl.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_t3_cpl.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.9	29-Aug-2024	andvar	s/Perfoms/Performs/ in comment.
1.8	23-May-2024	andvar	branches: 1.8.2; s/separare/separate/ and s/separete/separate/ in comments.
1.7	10-Dec-2021	andvar	s/occured/occurred/ in comments, log messages and man pages.
1.6	10-Nov-2021	msaitoh	s/endianess/endianness/
1.5	02-Aug-2021	andvar	fix various typos in comments and log messages.
1.4	29-Sep-2020	msaitoh	s/occurence/occurrence/
1.3	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.2	08-Feb-2018	dholland	branches: 1.2.2; 1.2.4; Typos.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_t3_hw.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_t3_hw.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_t3_hw.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2.4.1	10-Jun-2019	christos	Sync with HEAD
1.2.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.8.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_tcb.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_tcb.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_tcb.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.2	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.64; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.64.1	22-Apr-2018	pgoyette	Sync with HEAD
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_toedev.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_toedev.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_toedev.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_version.h was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_version.h was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_version.h was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_vsc7323.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_vsc7323.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_vsc7323.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_vsc8211.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:06 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_vsc8211.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_vsc8211.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.3	26-Mar-2023	andvar	fix various typos in documentation, comments and sysctl device description. mainly aion -> ation and inlude -> include.
1.2	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.1	21-Mar-2010	jklos	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.64; 1.1.66; Reorganizing all Chelsio 10 gig files into separate directory.
1.1.66.1	10-Jun-2019	christos	Sync with HEAD
1.1.64.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1.6.2	11-Aug-2010	yamt	sync with head.
1.1.6.1	21-Mar-2010	yamt	file cxgb_xgmac.c was added on branch yamt-nfs-mp on 2010-08-11 22:54:07 +0000
1.1.4.2	30-May-2010	rmind	sync with head
1.1.4.1	21-Mar-2010	rmind	file cxgb_xgmac.c was added on branch rmind-uvmplock on 2010-05-30 05:17:40 +0000
1.1.2.2	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.2.1	21-Mar-2010	uebayasi	file cxgb_xgmac.c was added on branch uebayasi-xip on 2010-04-30 14:43:45 +0000
1.23	21-Oct-2025	pgoyette	cut&paste error - forgot to fix the function name
1.22	21-Oct-2025	pgoyette	It helps to include the sys/module.h header file,
1.21	21-Oct-2025	pgoyette	Add MODULE glue and build infrasructure. XXX We won't actually build or install it due to kern/59716 but XXX it can be manually built and installed if you want.
1.20	01-Jun-2025	rin	igc(4): Merge and clean up igc_{tx_ctx,tso}_setup(), NFC
1.19	01-Jun-2025	rin	igc(4): TSOv[46]: Fix and register into capabilities by syncing TX mbuf after L3/L4 headers are fixed up. Will be pulled up into netbsd-10.
1.18	07-May-2025	rin	igc(4): igc_detach_counters: Correct sizes for kmem_free(9) Now, igc(4) can be successfully detached by `drvctl -d`.
1.17	24-Nov-2024	mlelstv	MBUFTRACE
1.16	05-Jul-2024	rin	sys: Drop redundant NULL check before m_freem(9) m_freem(9) safely has accepted NULL argument at least since 4.2BSD: https://www.tuhs.org/cgi-bin/utree.pl?file=4.2BSD/usr/src/sys/sys/uipc_mbuf.c Compile-tested on amd64/ALL. Suggested by knakahara@
1.15	29-Jun-2024	riastradh	branches: 1.15.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.14	12-Jun-2024	rin	igc: Add missing '\n' to a DPRINTF message
1.13	21-Feb-2024	msaitoh	igc(4): Print EtrackID.
1.12	21-Feb-2024	msaitoh	igc(4): Modify dmesg output of PHY and NVM info. - Print PHY info first and then print NVM info. - Remove debug output.
1.11	08-Feb-2024	msaitoh	igc: Add missing igc_check_for_link() call. It's required to set the collision distance, configure flow control from the negotiated result and set the LTR thresholds. With this change, ifconfig igcN show the flow control status correctly.
1.10	25-Jan-2024	msaitoh	igc(4): Count iqdrops. TODO: RQDPC should be visible via evcnt(9).
1.9	20-Dec-2023	skrll	Remove unnecssary #include
1.8	02-Nov-2023	rin	igc(4): Notify which of 64- or 32-bit DMA is used
1.7	15-Oct-2023	oster	Fix build of the MODULAR kernel, which explicitly excludes vlans.
1.6	12-Oct-2023	rin	igc(4): igc_init: Enable interrupt after everything is set up Inspired by "ixgbe: Enable interrupt after setting IFF_RUNNING.": http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/dev/pci/ixgbe/ixgbe.c#rev1.337 Just for sure, callout_schedule(9) for igc_tick() is also postponed. This is NFC for now, although. XXX Check other drivers.
1.5	11-Oct-2023	riastradh	igc(4): `nvm_ver & ...' has type int, so use %x, not %hx. This happens even though nvm_ver is unsigned short (uint16_t), because of the integer promotions. Should fix clang build. XXX pullup-10
1.4	10-Oct-2023	msaitoh	igc(4): Fix half duplex setting From FreeBSD: 3b8d04f845b416d29a258658b8a48d1afb4a2e81
1.3	04-Oct-2023	rin	branches: 1.3.2; igc(4): XXX: Temporally disable 64-bit DMA for aarch64 Until PR port-arm/57643 is properly addressed.
1.2	04-Oct-2023	rin	igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.3.2.7	01-Jun-2025	martin	Pull up following revision(s) (requested by rin in ticket #1125): sys/dev/pci/igc/if_igc.c: revision 1.19 sys/dev/pci/igc/if_igc.c: revision 1.20 share/man/man4/igc.4: revision 1.3 igc(4): TSOv[46]: Fix and register into capabilities by syncing TX mbuf after L3/L4 headers are fixed up. igc(4): Merge and clean up igc_{tx_ctx,tso}_setup(), NFC igc.4: TSO is supported now. Bump date.
1.3.2.6	09-May-2025	martin	Pull up following revision(s) (requested by rin in ticket #1110): sys/dev/pci/igc/if_igc.c: revision 1.18 igc(4): igc_detach_counters: Correct sizes for kmem_free(9) Now, igc(4) can be successfully detached by `drvctl -d`.
1.3.2.5	21-Jun-2024	martin	Pull up following revision(s) (requested by rin in ticket #713): sys/dev/pci/igc/if_igc.c: revision 1.14 sys/dev/pci/igc/if_igc.c: revision 1.7 sys/dev/pci/igc/if_igc.c: revision 1.8 sys/dev/pci/igc/if_igc.c: revision 1.9 Fix build of the MODULAR kernel, which explicitly excludes vlans. igc(4): Notify which of 64- or 32-bit DMA is used Remove unnecssary #include igc: Add missing '\n' to a DPRINTF message
1.3.2.4	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #605): sys/dev/pci/igc/if_igc.c: revision 1.12 sys/dev/pci/igc/if_igc.c: revision 1.13 sys/dev/pci/igc/igc_defines.h: revision 1.3 sys/dev/pci/igc/if_igc.c: revision 1.10 sys/dev/pci/igc/if_igc.c: revision 1.11 igc(4): Count iqdrops. TODO: RQDPC should be visible via evcnt(9). igc: Add missing igc_check_for_link() call. It's required to set the collision distance, configure flow control from the negotiated result and set the LTR thresholds. With this change, ifconfig igcN show the flow control status correctly. igc(4): Modify dmesg output of PHY and NVM info. - Print PHY info first and then print NVM info. - Remove debug output. igc(4): Print EtrackID.
1.3.2.3	14-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #410): sys/dev/pci/igc/if_igc.c: revision 1.4 sys/dev/pci/igc/if_igc.c: revision 1.5 sys/dev/pci/igc/if_igc.c: revision 1.6 igc(4): Fix half duplex setting From FreeBSD: 3b8d04f845b416d29a258658b8a48d1afb4a2e81 igc(4): `nvm_ver & ...' has type int, so use %x, not %hx. This happens even though nvm_ver is unsigned short (uint16_t), because of the integer promotions. Should fix clang build. igc(4): igc_init: Enable interrupt after everything is set up Inspired by "ixgbe: Enable interrupt after setting IFF_RUNNING.": http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/dev/pci/ixgbe/ixgbe.c#rev1.337 Just for sure, callout_schedule(9) for igc_tick() is also postponed. This is NFC for now, although.
1.3.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.3.2.1	04-Oct-2023	martin	file if_igc.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.15.2.1	02-Aug-2025	perseant	Sync with HEAD
1.3	27-Jun-2024	rin	igc: Use __HAVE_ATOMIC64_LOADSTORE instead of __HAVE_ATOMIC64_OPS to detect atomic_loadstore(9) is applicable to 64-bit integers. Thanks riastradh@ for explanation in PR kern/58340
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.3	27-Jun-2024	martin	Pull up following revision(s) (requested by rin in ticket #729): sys/dev/pci/igc/if_igc.h: revision 1.3 igc: Use __HAVE_ATOMIC64_LOADSTORE instead of __HAVE_ATOMIC64_OPS to detect atomic_loadstore(9) is applicable to 64-bit integers. Thanks riastradh@ for explanation in PR kern/58340
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file if_igc.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_api.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_api.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_base.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_base.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.4	22-Aug-2024	andvar	s/Shfit/Shift/ in comment.
1.3	21-Feb-2024	msaitoh	branches: 1.3.2; igc(4): Print EtrackID.
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.4	12-Sep-2024	martin	Pull up following revision(s) (requested by rin in ticket #853): sys/dev/pci/igc/igc_defines.h: revision 1.4 share/man/man4/igc.4: revision 1.2 Fix typo in email. s/Shfit/Shift/ in comment.
1.2.2.3	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #605): sys/dev/pci/igc/if_igc.c: revision 1.12 sys/dev/pci/igc/if_igc.c: revision 1.13 sys/dev/pci/igc/igc_defines.h: revision 1.3 sys/dev/pci/igc/if_igc.c: revision 1.10 sys/dev/pci/igc/if_igc.c: revision 1.11 igc(4): Count iqdrops. TODO: RQDPC should be visible via evcnt(9). igc: Add missing igc_check_for_link() call. It's required to set the collision distance, configure flow control from the negotiated result and set the LTR thresholds. With this change, ifconfig igcN show the flow control status correctly. igc(4): Modify dmesg output of PHY and NVM info. - Print PHY info first and then print NVM info. - Remove debug output. igc(4): Print EtrackID.
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_defines.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.3.2.1	02-Aug-2025	perseant	Sync with HEAD
1.1	04-Oct-2023	rin	branches: 1.1.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.1.2.1	04-Oct-2023	martin	file igc_evcnt.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_hw.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.3	25-Feb-2025	andvar	Fix a few typos in the word 'offset' in comments and error message.
1.2	04-Oct-2023	rin	branches: 1.2.2; 1.2.8; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.8.1	02-Aug-2025	perseant	Sync with HEAD
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_i225.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_i225.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_mac.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_mac.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_nvm.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_nvm.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_phy.c was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_phy.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	04-Oct-2023	rin	branches: 1.2.2; igc(4): Add support to Intel I225 / I226 series ethernet devices Originally written by kevlo@o for OpenBSD, and ported by knakahara@, msaitoh@, and myself. The driver is EXPERIMENTAL at the moment, as some minor error handling paths are not fully implemented. Hardware VLAN tagging and TSO are not supported yet. Although, we have never observed strange behaviors at least on amd64, aarch64{,eb}, and evbppc (IBM405), except for PR port-arm/57643. We will send pullup request to netbsd-10, after successful snapshot build for -current.
1.1	04-Oct-2023	rin	branches: 1.1.1; Initial revision
1.1.1.1	04-Oct-2023	rin	Import igc(4) from OpenBSD-current as of 2023-10-04
1.2.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by rin in ticket #393): sys/dev/pci/igc/if_igc.c up to 1.3 sys/dev/pci/igc/if_igc.h up to 1.2 sys/dev/pci/igc/igc_api.c up to 1.2 sys/dev/pci/igc/igc_api.h up to 1.2 sys/dev/pci/igc/igc_base.c up to 1.2 sys/dev/pci/igc/igc_base.h up to 1.2 sys/dev/pci/igc/igc_defines.h up to 1.2 sys/dev/pci/igc/igc_evcnt.h up to 1.1 sys/dev/pci/igc/igc_hw.h up to 1.2 sys/dev/pci/igc/igc_i225.c up to 1.2 sys/dev/pci/igc/igc_i225.h up to 1.2 sys/dev/pci/igc/igc_mac.c up to 1.2 sys/dev/pci/igc/igc_mac.h up to 1.2 sys/dev/pci/igc/igc_nvm.c up to 1.2 sys/dev/pci/igc/igc_nvm.h up to 1.2 sys/dev/pci/igc/igc_phy.c up to 1.2 sys/dev/pci/igc/igc_phy.h up to 1.2 sys/dev/pci/igc/igc_regs.h up to 1.2 distrib/sets/lists/man/mi: revision 1.1766 sys/arch/amd64/conf/GENERIC: revision 1.606 sys/arch/evbppc/conf/DHT: revision 1.5 sys/arch/evbarm/conf/GENERIC64: revision 1.213 share/man/man4/Makefile: revision 1.733 sys/arch/amd64/conf/ALL: revision 1.181 share/man/man4/igc.4: revision 1.1 sys/dev/pci/files.pci: revision 1.447 sys/arch/amd64/conf/XEN3_DOM0: revision 1.200 doc/CHANGES (apply patch) Add igc(4) for Intel I225/I226 series ethernet devices
1.2.2.1	04-Oct-2023	martin	file igc_regs.h was added on branch netbsd-10 on 2023-10-08 13:19:34 +0000
1.2	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.1	21-Jan-2014	mlelstv	branches: 1.1.4; 1.1.6; 1.1.10; 1.1.36; 1.1.40; one more file needed
1.1.40.1	05-May-2019	isaki	Remove obsoleted au{,rate,vol}conv and mulaw attributes. audio provides the equivalent of them inseparably.
1.1.36.1	10-Jun-2019	christos	Sync with HEAD
1.1.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.10.1	21-Jan-2014	tls	file files.igma was added on branch tls-maxphys on 2014-08-20 00:03:48 +0000
1.1.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.1	21-Jan-2014	yamt	file files.igma was added on branch yamt-pagecache on 2014-05-22 11:40:34 +0000
1.1.4.2	18-May-2014	rmind	sync with head
1.1.4.1	21-Jan-2014	rmind	file files.igma was added on branch rmind-smpnet on 2014-05-18 17:45:45 +0000
1.5	04-Oct-2025	thorpej	Use device_{get,set}prop_bool() for "is_console".
1.4	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.3	24-Apr-2021	thorpej	branches: 1.3.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.2	24-Feb-2020	rin	branches: 1.2.8; 0x%p --> %p for non-external codes.
1.1	21-Jan-2014	mlelstv	branches: 1.1.4; 1.1.6; 1.1.10; 1.1.36; 1.1.44; wscons driver for Intel Graphics Media Accelerator. Initial commit that already works for a couple of Notebooks based on G35, G45, Sandy Bridge and Ivy Bridge chips. Despite the word 'Accelerator' there is nothing acclerated yet.
1.1.44.1	29-Feb-2020	ad	Sync with head.
1.1.36.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.1.10.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.10.1	21-Jan-2014	tls	file igmafb.c was added on branch tls-maxphys on 2014-08-20 00:03:48 +0000
1.1.6.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.6.1	21-Jan-2014	yamt	file igmafb.c was added on branch yamt-pagecache on 2014-05-22 11:40:34 +0000
1.1.4.2	18-May-2014	rmind	sync with head
1.1.4.1	21-Jan-2014	rmind	file igmafb.c was added on branch rmind-smpnet on 2014-05-18 17:45:45 +0000
1.2.8.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.3.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.5	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.3	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.2	24-Apr-2015	msaitoh	branches: 1.2.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.2	05-Dec-2016	skrll	Sync with HEAD
1.1.30.1	06-Jun-2015	skrll	Sync with HEAD
1.1.28.1	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.2.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.3	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.2	05-Aug-2015	msaitoh	branches: 1.2.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.2	05-Dec-2016	skrll	Sync with HEAD
1.1.30.1	22-Sep-2015	skrll	Sync with HEAD
1.1.28.1	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.2.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.10	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.9	17-Aug-2021	andvar	branches: 1.9.4; fix multiplei repetitive typos in comments, messages and documentation. mainly because copy paste code big amount of files are affected.
1.8	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.7	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.6	12-Aug-2020	msaitoh	branches: 1.6.6; Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different.
1.5	16-Dec-2019	msaitoh	No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.4	04-Apr-2018	msaitoh	branches: 1.4.2; 1.4.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.3	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.2	22-Nov-2017	msaitoh	branches: 1.2.2; 1.2.4; 1.2.6; Fix a bug that bypass adapter's sysctls aren't set. Tested with non-genuine X550-T2 bypass adapter: - Call ixgbe_sysctl_instance() in ixgbe_bypass_init() to get sysctl's top node correctly. - ixgbe_init_device_features() refers adapter->hw.bus.func for bypass adapter. Call set_lan_id() to set adapter->hw.bus.func before calling ixgbe_init_device_features(). Without this, bypass sysctl's are added to both the first and second port. - Initalize node.sysctl_data before calling sysctl_lookup() to read correct value.
1.1	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.2.6.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.2.6.1	15-Mar-2018	pgoyette	Synch with HEAD
1.2.4.8	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.2.4.7	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.2.4.6	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1600): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.2.4.5	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.2.4.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.2.4.3	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.2.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.2.4.1	22-Nov-2017	snj	file if_bypass.c was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.2.2.2	03-Dec-2017	jdolecek	update from HEAD
1.2.2.1	22-Nov-2017	jdolecek	file if_bypass.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.4.8.4	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.4.8.3	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.4.8.2	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1070): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.4.8.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.4.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.6.6.1	13-May-2021	thorpej	Sync with HEAD.
1.9.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.6	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.5	30-Apr-2021	msaitoh	branches: 1.5.14; Add missing __KERNEL_RCSID().
1.4	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.3	07-Sep-2020	msaitoh	branches: 1.3.6; - Remove extra callout_stop() in ixgbe_detach(). Found by knakahara@. - Rename ix{gbe,v}_free_workqueue() to ix{gbe,v}_free_deferred_handlers(). - Add KASSERT() to functions who are called from ixgbe_handle_admin().
1.2	04-Apr-2018	msaitoh	branches: 1.2.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.6.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1.4.5	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file if_fdir.c was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file if_fdir.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.2.8.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.2.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.6.1	13-May-2021	thorpej	Sync with HEAD.
1.5.14.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.18	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.17	24-Dec-2021	msaitoh	branches: 1.17.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.16	24-Dec-2021	msaitoh	Change the first argument of ixgbe_send_vf_msg(). No functional change. - This is a part of FreeBSD ix-3.2.17.
1.15	24-Dec-2021	msaitoh	Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. - Sync with FreeBSD ix-3.3.18. - No functional change.
1.14	10-Dec-2021	msaitoh	Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). From FreeBSD ix-3.3.18. No functional change.
1.13	10-Dec-2021	msaitoh	Change argument. No functional change. This file is not used in NetBSD.
1.12	10-Dec-2021	msaitoh	Remove unused argument. Note that this file is not used in NetBSD.
1.11	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.10	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.9	25-Jan-2021	mrg	branches: 1.9.4; s/boolean_t/bool/. boolean_t is obsolete (from Mach), and this also fixes a new build issue in libkvm on sparc*.
1.8	07-Sep-2020	msaitoh	branches: 1.8.2; - Remove extra callout_stop() in ixgbe_detach(). Found by knakahara@. - Rename ix{gbe,v}_free_workqueue() to ix{gbe,v}_free_deferred_handlers(). - Add KASSERT() to functions who are called from ixgbe_handle_admin().
1.7	25-Jun-2020	msaitoh	Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.6	27-Jun-2019	msaitoh	branches: 1.6.2; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.5	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.4	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.3	04-Apr-2018	msaitoh	branches: 1.3.2; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.2	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.6.4	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1.6.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.1.6.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1.6.1	15-Mar-2018	pgoyette	Synch with HEAD
1.1.4.9	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.8	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.1.4.7	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.6	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.5	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.1.4.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.1.4.3	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file if_sriov.c was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file if_sriov.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.3.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.3.2.1	10-Jun-2019	christos	Sync with HEAD
1.6.2.4	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.6.2.3	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.6.2.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6.2.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.8.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.9.4.1	13-May-2021	thorpej	Sync with HEAD.
1.17.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.117	29-Jun-2024	riastradh	if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.116	30-Dec-2023	msaitoh	ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.115	29-Dec-2023	msaitoh	ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring.
1.114	28-Dec-2023	msaitoh	ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO).
1.113	28-Dec-2023	msaitoh	ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN).
1.112	28-Dec-2023	msaitoh	ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment.
1.111	13-Dec-2023	msaitoh	ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof().
1.110	08-Dec-2023	msaitoh	ixgbe: Update if_opackets outside the loop in ixgbe_txeof().
1.109	08-Dec-2023	msaitoh	ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof().
1.108	16-Nov-2023	msaitoh	ixgbe: Modify for the readability. No functional change.
1.107	14-Nov-2023	msaitoh	ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor.
1.106	14-Nov-2023	msaitoh	ixgbe(4): Move assignment of TXD. NFCI.
1.105	02-Nov-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.104	12-Oct-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.103	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.102	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.101	03-Oct-2023	msaitoh	Remove obsolete comment.
1.100	16-Sep-2022	knakahara	branches: 1.100.4; ixg(4) add an option for Tx to use deferred softint regardless of whether can get txq lock or not. That imporve (7%) and stabilize throughput. But that can cause latency degradation, so off by default. ok'ed by msaitoh@n.o.
1.99	07-Aug-2022	andvar	fix typos in comments.
1.98	11-May-2022	bouyer	bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others.
1.97	25-Apr-2022	msaitoh	Use cached rx_copy_len in ixgbe_rxeof().
1.96	25-Apr-2022	msaitoh	Modify comment for consistency. No functional change.
1.95	25-Jan-2022	msaitoh	Use atomic_{load,store}_relaxed() for event counters.
1.94	08-Sep-2021	msaitoh	Reduce bus_dmamap_sync() cost. - Don't sync whole descriptor ring but limited counts(rx_process_limit) descriptors. - Dont' sync every loop.
1.93	08-Sep-2021	msaitoh	Don't pre-allocate a cluster not to do m_freem() it on RXCOPY case.
1.92	07-Sep-2021	msaitoh	Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet.
1.91	07-Sep-2021	msaitoh	Sprinkle __predict_false() in the RX path.
1.90	03-Sep-2021	msaitoh	Save the discard_multidesc state to not to forget the state by exiting rxeof().
1.89	03-Sep-2021	msaitoh	Rename variable. No functional change.
1.88	26-Aug-2021	msaitoh	Remove "j" or "jumbo" because we always use MCLBYTES sized buffer.
1.87	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.86	19-Aug-2021	msaitoh	Use m_adj(ETHER_ALIGN) more. Tested by me (amd64,aarch64) and rin (alpha). - Align with ETHER_ALIGN everywhere where mbuf is allocated. - Remove extra setting of M_PKTHDR. No functional change. - Add comment.
1.85	19-Aug-2021	msaitoh	KNF. No functional change.
1.84	19-Aug-2021	msaitoh	Don't use fixed value. No functional change.
1.83	19-Aug-2021	msaitoh	Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. If an mbuf is ETHER_ALIGNed, the packet size may be shorter than MCLBYTES. For example, if the max frame size is 2048 by changing the interface's MTU, an mbuf's max length is 2046. So, don't use rx_mbuf_sz for bus_dmamap_sync.
1.82	18-Aug-2021	msaitoh	Refresh unrefreshed descriptors' buffers correctly. - Update next_to_refresh at least before ixgbe_rx_unrefresed() to detect the unrefreshed status correctly in ixgbe_rxeof(). - next_to_refresh points to the previous entry of the first unrefreshed descriptor, so fix a loop variable to point to the correct one in ixgbe_refresh_mbufs(). - Without the above two fixes, RX ring may have some unrefreshed entries which have inconsistent state. On such state, "ifconfig down up" causes panic in bus_dmamap_sync() on aarch64. - Tested on amd64 and aarch64. OK'd by knakahara.
1.81	07-Jul-2021	msaitoh	Add new sysctl "rx_copy_len". ixgbe_rxeof() has an optimization "RX_COPY" to reduce costs of bus_dmamap_load_mbuf() and bus_dmamap_unload() by copying a mbuf cluster's memory to a newly allocated mbuf's MH_databuf[] and recycle the original map. The optimization is used when a length of a packet is smaller than a specific value. The value is calculated based on MHLEN. The size of MHLEN is architecture specific. It's 256 or 512. Make the threshold controllable by adding a new sysctl.
1.80	07-Jul-2021	msaitoh	Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). ixgbe_setup_receive_ring() fully allocates rx buffers. When a descriptor ring is full, rxr->next_to_refresh should point to rxr_next_to_check -1. Before this change, rxr->next_to_refresh is set to 0 and ixgbe_refresh_mbufs() wastefully loops in ixgbe_refresh_mbufs() because it means the ring is empty.
1.79	27-May-2021	msaitoh	Do no_jmbuf++ when ixgbe_getjcl() failed in ixgbe_setup_receive_ring(), too.
1.78	20-May-2021	ryo	fix little-endian dependence
1.77	20-May-2021	msaitoh	Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work.
1.76	20-May-2021	ryo	Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. In ixgbe, TX/RX descriptor rings are configured in 16-byte units. If BUS_DMA_COHERENT is not specified, cpu cache (writeback/invalidate) operations by bus_dmamap_sync() in aarch64 (arm/arm32/bus_dma.c) are done per cache line size (usually 64 bytes). As a result, adjacent descriptors conflict with the DMA operation, resulting in unstable operation. To avoid this, descriptors area should be mapped as non-cache with BUS_DMA_COHERENT. thanks to msaitoh@ for his help in debugging.
1.75	18-May-2021	msaitoh	- Cleanup an rxbuf entry when bus_dmamap_load_mbuf() failed to prevent panic. - Print error number when error occurred.
1.74	14-May-2021	msaitoh	Keep m_len and m_pkthdr.len consistent to prevent panic on arm. Arm's bus_dmamap_load_mbuf() keeps a pointer to the original mbuf and bus_dmamap_sync() refers it. ixgbe_rxeof() modified mbuf's m_len inconsistently with m_pkthdr and dm_mapsize. "ifconfig down up" made panic by referring the inconsistent mbuf length.
1.73	14-May-2021	knakahara	Comment out flow director processing in fast path. ixgbe_xmit() is per-queue fast path. It should reduce access to per-device data (adapter->feat_en).
1.72	11-May-2021	rin	Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). Otherwise, MMU fault occurs for some bus_dma(9) implementations. With this fix, X550-T1 and X540-T1 work fine on alpha (at least DS10 with PCI-PCIe reverse bridge). Discussed with msaitoh. Thanks!
1.71	30-Apr-2021	msaitoh	branches: 1.71.2; Add missing __KERNEL_RCSID().
1.70	31-Mar-2021	msaitoh	branches: 1.70.2; KNF a bit. No functional change.
1.69	12-Mar-2021	knakahara	branches: 1.69.2; Refactor rxr->next_to_check updating.
1.68	12-Mar-2021	knakahara	Remove extra unlock/lock processing around if_percpuq_enqueue(). same as if_wm.c:r1.700
1.67	09-Mar-2021	msaitoh	Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.66	08-Mar-2021	msaitoh	Use no_jmbuf evcnt for the failure case of ixgbe_getjcl().
1.65	02-Mar-2021	msaitoh	Fix jcl's starvation case in ixgbe_rxeof() again. ix_txrx.c rev.1.64 preallocates jcl to prevent starvation but it's not perfect. Don't use ixgbe_rx_discard() and just update the old descriptor for reuse. It's also required for multiple descriptors case to refresh subsequent descriptor(s). Reviewed by knakahara@.
1.64	18-Jan-2021	knakahara	Fix ixg(4) Rx interrupt stall when Rx buffers are exhausted. Current ixgbe_rxeof() implementation calls ixgbe_refresh_mbufs()(and ixgbe_getjcl()) after ixgbe_rx_input()(and if_percpuq_enqueue()). And ixg(4) uses its own m_ext structure. That causes Rx interrupt stall when Rx buffers are exhausted. e.g. TCP iperf3 with large windows size. Furthermore, Rx descriptor problem has occurred after stoppping the load. To fix that problem, ixgbe_rxeof() must call ixgbe_getjcl() before ixgbe_rx_input(), and must discard the new packet when ixgbe_getjcl() fails. By the way, ixg(4) should use MCLGET() instead of own m_ext structure. ok'ed by msaitoh@n.o, thanks.
1.63	17-Apr-2020	msaitoh	branches: 1.63.2; No functional change: - modify comment - whitespace fix
1.62	05-Feb-2020	msaitoh	branches: 1.62.4; No functional change: - Add debug printf()s. - Remove unused macros. - Remove extra newline.
1.61	30-Jan-2020	thorpej	Adopt <net/if_stats.h>.
1.60	21-Jan-2020	msaitoh	Fix the freeing code for some error paths. Found and tested by Patrick Welche.
1.59	20-Jan-2020	msaitoh	Free jumbo mem structure correctly. Found by yamaguchi@ using with LOCKDEBUG.
1.58	16-Dec-2019	msaitoh	branches: 1.58.2; No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.57	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.56	16-Oct-2019	knakahara	Fix missing kpreempt_disable() before softint_schedule() like if_vmx.c:r1.51.
1.55	04-Sep-2019	msaitoh	printf -> device_printf
1.54	04-Jul-2019	msaitoh	branches: 1.54.2; Fix hung queue check when the queue number >= 31.
1.53	27-Jun-2019	msaitoh	Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.52	22-Feb-2019	msaitoh	Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.51	20-Dec-2018	knakahara	Apply the same fix as if_wm.c:r1.606 to ixg(4). Pointed out by msaitoh@n.o, thanks. XXX pullup-8
1.50	06-Sep-2018	msaitoh	Fix a bug that ixgbe_mq_start(an if_transmit function) returned wrong vaue on error. pcq_put returns false on error, so returning it to caller indicated no error. XXX pullup-8
1.49	31-Jul-2018	msaitoh	Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.48	26-Jun-2018	msaitoh	branches: 1.48.2; Implement the BPF direction filter (BIOC[GS]DIRECTION). It provides backward compatibility with BIOC[GS]SEESENT ioctl. The userland interface is the same as FreeBSD. This change also fixes a bug that the direction is misunderstand on some environment by passing the direction to bpf_mtap*() instead of checking m->m_pkthdr.rcvif.
1.47	23-May-2018	msaitoh	Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE.
1.46	23-May-2018	msaitoh	Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922).
1.45	18-May-2018	msaitoh	Revert new watchdog timer commits. The new watchdog timer made stability worse than before. It seems unknown problems exists. http://mail-index.netbsd.org/source-changes/2018/05/08/msg095020.html http://mail-index.netbsd.org/source-changes/2018/05/16/msg095240.html
1.44	16-May-2018	msaitoh	Whitespace fix. No functional change.
1.43	16-May-2018	msaitoh	Fix a problem that the watchdog timer sometimes mistakenly fires. Restore TX sending check in the end of ixgbe_txeof which was wrongly removed in ix_txrx.c rev. 1.42.
1.42	08-May-2018	msaitoh	- Fix broken watchdog timer. This change detects TX device timeout correctly. NOTE: It's supporsed not to be called {ixgbe,ixv}_rearm_queues() in the timer. Those are not required if any chip have no bug. In reality, ixgbe_rearm_queues() is required on 82599 and newer chip AND other than queue 0 to prevent device timeout. When it occured, packet was sent but the descriptor's DD bit wasn't set even though IXGBE_TXD_CMD_EOP and IXGBE_TXD_CMD_RS were set. After forcing interrupt by writing EICS register in ixgbe_rearm_queues(), DD is set. Why? Is this an undocumented errata? It might be possible not call rearm_queues on 82598 or queue 0, we call in any cases in case the problem occurs. On ixv(4), I have not seen this problem yet (though I tested only on X550_X(Xeon D 12xx)'s virtual function), but we do rearm in case TX device timeout happen. - ixv(4): Call callout_stop() earlier in ixv_stop() like ixgbe_stop(). - KNF.
1.41	25-Apr-2018	msaitoh	Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet.
1.40	17-Apr-2018	msaitoh	Remove unused structure entries. No functional change.
1.39	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.38	02-Apr-2018	knakahara	Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o.
1.37	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.36	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.35	09-Mar-2018	msaitoh	Make some event counters MP safe. Now all of the event counters are MP safe.
1.34	02-Mar-2018	knakahara	branches: 1.34.2; ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o.
1.33	26-Feb-2018	knakahara	Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o.
1.32	22-Feb-2018	msaitoh	Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.31	20-Feb-2018	msaitoh	Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8
1.30	04-Dec-2017	msaitoh	- Fixes two problem: 1) RX may accesses freed area. 2) if_init() takes long time on many core machine. Call ixgbe_jcl_reinit() not in ixgbe_setup_receive_ring() but in the biginning of ixgbe_setup_receive_structures(). It was OK for pre multiqueue, but it's not OK now because we support multiqueue. - Fix comment of ixgbe_free_receive_ring().
1.29	26-Sep-2017	knakahara	branches: 1.29.2; VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html XXX need pullup to -8 branch
1.28	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.27	13-Jun-2017	msaitoh	Sync with FreeBSD r316541: > Fix a double free in ixgbe_rxeof() > > Submitted by: rstone > MFC after: 1 week > Differential Revision: https://reviews.freebsd.org/D10255
1.26	13-Jun-2017	msaitoh	Fix the partial chain check in ixgbe_rx_discard(). This bug was addded in rev. 1.33.
1.25	09-Jun-2017	msaitoh	Increase total number of RX buffers on multiqueue.
1.24	18-May-2017	msaitoh	branches: 1.24.2; Fix a bug that number of input packet counted twice. This bug was added since sys/net/if.c rev.1.368. http://mail-index.netbsd.org/source-changes/2016/12/15/msg079882.html
1.23	08-May-2017	msaitoh	Fix a bug that if_obytes and if_omcasts aren't counted. This bug was added in ixgbe.c rev. 1.27 by me. Reported by Uwe Toenjes.
1.22	02-Mar-2017	msaitoh	branches: 1.22.4; Fix a problem that m_defrag() isn't called in if_transmit path. Now both if_start and if_transmit do m_defrag() correctly. This change improves the performance of TSO.
1.21	02-Mar-2017	msaitoh	Fix a bug that TSO doesn't work correctly. This bug was added in ix_txrx.c rev. 1.17.
1.20	13-Feb-2017	msaitoh	Use percpuq.
1.19	10-Feb-2017	msaitoh	- Both mbuf_defrag_failed and m_defrag_failed existed. Remove m_defrag_failed. - rx_irq, dropped_pkts, morerx, moretx and txloops counters are uncounted, so remove them. - Count handleq in ixv_handle_que(). - Detach event counters correctly. - Set some per-queue event names correctly.
1.18	01-Feb-2017	msaitoh	TX multiqueue. If you want to disable it, enable IXGBE_LEGACY_TX in ixgbe_netbsd.h
1.17	30-Jan-2017	msaitoh	- Count tso_err again. It was accidentally removed in ixgbe.c rev. 1.28. - ixgbe.h: Sort entries to reduce diff against FreeBSD.
1.16	19-Jan-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.15	18-Jan-2017	msaitoh	Call ixgbe_free_transmit_buffers instead of ixgbe_free_transmit_structures() in ixgbe_allocate_transmit_buffers()...
1.14	05-Jan-2017	msaitoh	branches: 1.14.2; 1.14.4; Fix INIT_DEBUGOUT() messages.
1.13	16-Dec-2016	msaitoh	The rx_bytes event counter is cleared when if_init() is called, but not for rx_packets. It's inconsistent. It makes rx_packets is bigger than rx_bytes. IMHO, it's not required to clear them. At least the above comment says "/* Setup our descriptor indices */". It's not descriptor indice...
1.12	15-Dec-2016	ozaki-r	Move bpf_mtap and if_ipackets++ on Rx of each driver to percpuq if_input The benefits of the change are: - We can reduce codes - We can provide the same behavior between drivers - Where/When if_ipackets is counted up - Note that some drivers still update packet statistics in their own way (periodical update) - Moved bpf_mtap run in softint - This makes it easy to MP-ify bpf Proposed on tech-kern and tech-net
1.11	14-Dec-2016	msaitoh	Fix ip[46]csum-tx doesn't work other than TCP and UDP.
1.10	05-Dec-2016	msaitoh	branches: 1.10.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.9	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r294578: - Fixup SFP module insertion on the 82599 when insertion happens after the system is booted and running. Add PHY detection logic to ixgbe_handle_mod() and add locking to ixgbe_handle_msf() as well. FreeBSD r293334. - Fix ix advertise value after media change. When ifconfig sets media then the values displayed by the advertise_speed value are invalidated. Fix this by setting the bits correctly including setting advertise to 0 for media = auto. FreeBSD r294578. - Some others (e.g. LRO(not used by NetBSD)).
1.8	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.7	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.6	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.5	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.3	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.2	30-Nov-2016	msaitoh	Extra "buf->map = NULL;"s in ixgbe_txeof() were removed in FreeBSD r282299. This was fixed in NetBSD's ixgbe.c rev. 1.28. Remove our comment.
1.1	28-Nov-2016	msaitoh	FreeBSD r280182 made new file ix_txrx.c and moved ixgbe.c and ixv's common code into it. Before sync with whole of them, just move ixgbe.c and ixv.c's common code into ix_txrx.c from ixgbe.c. In this commit, only ixgbe.c is split into the device dependent part and the common part. ixv.c isn't change to make this commit no functional change. To use ixv.c with ix_txrx.c, it's required to modify the common part's API and functions themselves. This commit is done to make the next change easy to understand.
1.10.2.4	28-Aug-2017	skrll	Sync with HEAD
1.10.2.3	05-Feb-2017	skrll	Sync with HEAD
1.10.2.2	05-Dec-2016	skrll	Sync with HEAD
1.10.2.1	05-Dec-2016	skrll	file ix_txrx.c was added on branch nick-nhusb on 2016-12-05 10:55:16 +0000
1.14.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.14.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.14.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.14.2.1	05-Jan-2017	pgoyette	file ix_txrx.c was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.22.4.2	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.22.4.1	11-May-2017	pgoyette	Sync with HEAD
1.24.2.31	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1933: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.24.2.30	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.24.2.29	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.24.2.28	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.24.2.27	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.24.2.26	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.24.2.25	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.98 sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others. Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.24.2.24	30-May-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1744: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.24.2.23	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1708: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.24.2.22	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.24.2.21	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.24.2.20	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.24.2.19	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.24.2.18	10-Nov-2019	martin	Pull up following revision(s) (requested by knakahara in ticket #1431): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.56 Fix missing kpreempt_disable() before softint_schedule() like if_vmx.c:r1.51.
1.24.2.17	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.24.2.16	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.24.2.15	01-Mar-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1199): sys/dev/pci/ixgbe/ixv.c: revision 1.109 sys/dev/pci/ixgbe/ixgbe.h: revision 1.54 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.174 Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.24.2.14	20-Dec-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #1141): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.51 Apply the same fix as if_wm.c:r1.606 to ixg(4). Pointed out by msaitoh@n.o, thanks. XXX pullup-8
1.24.2.13	07-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1014): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.50 Fix a bug that ixgbe_mq_start(an if_transmit function) returned wrong value on error. pcq_put returns false on error, so returning it to caller indicated no error. XXX pullup-8
1.24.2.12	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.24.2.11	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.24.2.10	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.24.2.9	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.24.2.8	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.24.2.7	13-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/dev/pci/ixgbe/ixgbe.c: revision 1.134 sys/dev/pci/ixgbe/ixgbe.h: revision 1.35 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.35 sys/dev/pci/ixgbe/ixv.c: revision 1.87 Make some event counters MP safe. Now all of the event counters are MP safe.
1.24.2.6	06-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #605): sys/dev/pci/ixgbe/ixgbe.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe.h: revision 1.33 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.34 sys/dev/pci/ixgbe/ixgbe.c: revision 1.128 sys/dev/pci/ixgbe/ixv.c: revision 1.83 sys/dev/pci/ixgbe/ixv.c: revision 1.84 Add hw.ixvM.q[01].{interrupt_rate,[tr]xd_head,[tr]xd_tail} sysctls as ixg(4). ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o. ixv(4) also supports workqueue poll mode, but not enabled by default yet, either. ok by msaitoh@n.o. Move the location of "struct work" as FreeBSD's "struct task" in ixgbe.h. No functional change.
1.24.2.5	01-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #597): sys/dev/pci/ixgbe/ixgbe.h: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.c: revision 1.127 sys/dev/pci/ixgbe/ixv.c: revision 1.82 Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o. Apply ixgbe.c:r1.127 to ixv.c. Pointed out by msaitoh@n.o.
1.24.2.4	26-Feb-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #592): sys/dev/pci/ixgbe/ixv.c: revision 1.79 sys/dev/pci/ixgbe/ixgbe.h: revision 1.30 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.31 sys/dev/pci/ixgbe/ixgbe.c: revision 1.120 sys/dev/pci/ixgbe/ixgbe.c: revision 1.121 sys/dev/pci/ixgbe/ixgbe.c: revision 1.123 sys/dev/pci/ixgbe/ixgbe.c: revision 1.124 sys/dev/pci/ixgbe/ixgbe.c: revision 1.125 sys/dev/pci/ixgbe/ixgbe.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.80 sys/dev/pci/ixgbe/ixv.c: revision 1.81 CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558). Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8 Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv) - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8 Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8 Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o. - Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function. Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.24.2.3	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.24.2.2	24-Oct-2017	snj	Pull up following revision(s) (requested by knakahara in ticket #302): sys/arch/powerpc/booke/dev/pq3etsec.c: 1.30-1.31 sys/arch/x86/pci/if_vmx.c: 1.20 sys/dev/ic/i82557.c: 1.148 sys/dev/ic/rtl8169.c: 1.152 sys/dev/pci/cxgb/cxgb_sge.c: 1.5 sys/dev/pci/if_age.c: 1.51 sys/dev/pci/if_alc.c: 1.25 sys/dev/pci/if_ale.c: 1.23 sys/dev/pci/if_bge.c: 1.311 sys/dev/pci/if_bge.c: 1.312 sys/dev/pci/if_bnx.c: 1.62 sys/dev/pci/if_jme.c: 1.32 sys/dev/pci/if_nfe.c: 1.64 sys/dev/pci/if_sip.c: 1.167 sys/dev/pci/if_stge.c: 1.63-1.64 sys/dev/pci/if_ti.c: 1.102 sys/dev/pci/if_txp.c: 1.48 sys/dev/pci/if_vge.c: 1.61 sys/dev/pci/if_wm.c: 1.538 sys/dev/pci/ixgbe/ix_txrx.c: 1.29 via patch sys/net/agr/if_agrether_hash.c: 1.4 sys/net/if_ether.h: 1.67-1.68 sys/net/if_ethersubr.c: 1.244 sys/net/if_vlan.c: 1.100 sys/net80211/ieee80211_input.c: 1.89 sys/net80211/ieee80211_output.c: 1.59 sys/sys/mbuf.h: 1.171 VLAN ID uses pkthdr instead of mtag now. Contributed by s-yamaguchi@IIJ. I just commit by proxy. Reviewed by joerg@n.o and christos@n.o, thanks. See http://mail-index.netbsd.org/tech-net/2017/09/26/msg006459.html -- only get vtag when we have vtag like the other drivers. -- - only get the vtag if we have it like the other drivers - mask the hardware vlan tag -- - add a constant for the vlan mask. - enforce that we have a tag before we get it. only get vtag when we have vtag like the other drivers. like if_bge.c:1.312 and if_stge.c:1.64. fixed by s-yamaguchi@IIJ, thanks.
1.24.2.1	05-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #175): sys/dev/pci/ixgbe/ix_txrx.c: 1.25-1.27 sys/dev/pci/ixgbe/ixgbe.c: 1.91-1.95 sys/dev/pci/ixgbe/ixgbe_type.h: 1.23-1.25 sys/dev/pci/ixgbe/ixgbe_api.c: 1.16 sync ixg(4) up to ixgbe.c rev. 1.95: - Increase total number of RX buffers on multiqueue. - Fix the partial chain check in ixgbe_rx_discard(). This bug was added in ixgbe.c rev. 1.33. - Fix a double free in ixgbe_rxeof(). - Print verbose output in ixgbe_update_link_status() correctly on 5Gbps and 2.5Gbps (NBASE-T) for X550T[12] and newer. - Change hw.ixgN.ts to hw.ixgN.thermal_test. Same as FreeBSD. - Fix a bug that ifconfig ixgN media 1000baseT and 10Gbase-T advertised additional unwanted speeds. - Print PHY ID. - Remove unused variable.
1.29.2.2	03-Dec-2017	jdolecek	update from HEAD
1.29.2.1	26-Sep-2017	jdolecek	file ix_txrx.c was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.34.2.11	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.34.2.10	30-Sep-2018	pgoyette	Ssync with HEAD
1.34.2.9	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.34.2.8	28-Jul-2018	pgoyette	Sync with HEAD
1.34.2.7	25-Jun-2018	pgoyette	Sync with HEAD
1.34.2.6	21-May-2018	pgoyette	Sync with HEAD
1.34.2.5	02-May-2018	pgoyette	Synch with HEAD
1.34.2.4	22-Apr-2018	pgoyette	Sync with HEAD
1.34.2.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.34.2.2	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.34.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.48.2.4	21-Apr-2020	martin	Sync with HEAD
1.48.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.48.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.48.2.1	10-Jun-2019	christos	Sync with HEAD
1.54.2.15	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1792: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.54.2.14	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.54.2.13	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.54.2.12	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.54.2.11	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.54.2.10	30-May-2022	martin	Pull up the following revisions, requested by msaitoh: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.158,1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.54.2.9	13-May-2022	martin	Pull up following revision(s) (requested by bouyer in ticket #1445): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.98 bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others.
1.54.2.8	02-Feb-2022	martin	Pull up the following revisions (requested by msaitoh in ticket #1424): sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.305 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixv.c 1.178 via patch Use atomic_{load,store}_relaxed() for event counters.
1.54.2.7	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1374: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.54.2.6	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.54.2.5	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.54.2.4	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.54.2.3	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.54.2.2	01-Nov-2019	martin	Pull up following revision(s) (requested by knakahara in ticket #387): sys/net/if_gre.c: revision 1.176 sys/net/if_l2tp.c: revision 1.40 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.56 sys/net/if_tap.c: revision 1.114 Fix missing kpreempt_disable() before softint_schedule() like if_vmx.c:r1.51.
1.54.2.1	05-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #180): sys/dev/pci/ixgbe/ixv.c: revision 1.130 sys/dev/pci/ixgbe/ixgbe.c: revision 1.207 sys/dev/pci/ixgbe/ixgbe.c: revision 1.208 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.55 sys/dev/pci/ixgbe/ixv.c: revision 1.129 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.10 printf -> device_printf Set IFM_1000_BX10 correctly. Use aprint_*() in the attach function.
1.58.2.2	29-Feb-2020	ad	Sync with head.
1.58.2.1	25-Jan-2020	ad	Sync with head.
1.62.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.63.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.69.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.70.2.3	01-Aug-2021	thorpej	Sync with HEAD.
1.70.2.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.70.2.1	13-May-2021	thorpej	Sync with HEAD.
1.71.2.1	31-May-2021	cjep	sync with head
1.100.4.5	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #563): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.110 sys/dev/pci/ixgbe/ixgbe.c: revision 1.345 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.111 sys/dev/pci/ixgbe/ixgbe.c: revision 1.346 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.112 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.62 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.113 sys/dev/pci/ixgbe/ixgbe.c: revision 1.348 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.114 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.115 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.116 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.105 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.106 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.107 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.108 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.109 sys/dev/pci/ixgbe/ixv.c: revision 1.193 sys/dev/pci/ixgbe/ixv.c: revision 1.195 sys/dev/pci/ixgbe/ixv.c: revision 1.196 sys/dev/pci/ixgbe/ixgbe.h: revision 1.94 sys/dev/pci/ixgbe/ixgbe.h: revision 1.95 sys/dev/pci/ixgbe/ixgbe.h: revision 1.96 sys/dev/pci/ixgbe/ixgbe.h: revision 1.97 sys/dev/pci/ixgbe/ixgbe.h: revision 1.98 ixgbe: Fix comment. No functional change. ixgbe: Whitespace. No functional change. ixgbe(4): Move assignment of TXD. NFCI. ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor. ixv(4): Remove unused IFF_OACTIVE. No functional change. ixgbe: Clear the WTHRESH bit field before writing it. ixgbe: Modify for the readability. No functional change. ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof(). ixgbe: Update if_opackets outside the loop in ixgbe_txeof(). ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof(). ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO). ixgbe: Remove unused to reduce the size of struct rx_ring. ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring. ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.100.4.4	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #420): sys/dev/pci/ixgbe/ixgbe.c: revision 1.340 sys/dev/pci/ixgbe/ixgbe.c: revision 1.341 sys/dev/pci/ixgbe/ixgbe.c: revision 1.342 sys/dev/pci/ixgbe/ixgbe.c: revision 1.343 sys/dev/pci/ixgbe/ixgbe.c: revision 1.344 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.61 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.104 sys/dev/pci/ixgbe/ixv.c: revision 1.190 sys/dev/pci/ixgbe/ixv.c: revision 1.191 sys/dev/pci/ixgbe/ixv.c: revision 1.192 sys/dev/pci/ixgbe/ixgbe.h: revision 1.93 ixg(4): Print DEVICE_CAPS register. ixgbe: Whitespace. No functional change. ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl. ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. ixgbe: Whitespace. No functional change.
1.100.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.100.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.100.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.354	10-Jul-2024	msaitoh	ixgbe: Remove ifdef NET_MPSAFE. Make ixgbe MP-scalable by default. Same as wm(4).
1.353	10-Jul-2024	msaitoh	ixgbe: ixgbe is always MP-safe, so remove IXGBE_MPSAFE macro. Before this change, IXGBE_MPSAFE is not defined if NET_MPSAFE is not set. In this case, the PCI_INTR_MPSAFE flag is not set by pci_intr_setattr() and interrupt_distribute() fails on arm. Always enable IXGBE_MPSAFE funciton by removing the macro.
1.352	29-Jun-2024	riastradh	branches: 1.352.2; if_stats(9): Add ifp argument to if_stat..._ref. This will enable us to pass the ifp through to a dtrace probe inside. No functional change intended in this change, but this is an API change visible to modules so it shouldn't be pulled up. PR kern/58377
1.351	30-May-2024	msaitoh	Move RECOVERY_MODE feature test code for readability. No functional change. ixgbe.c rev. 1.169 inserted RECOVERY_MODE feature test in between MSI-X allocation and legacy allocation. To improve code readability, move it to earlier location.
1.350	13-May-2024	msaitoh	Modify comment. The number of queues is not limited to 8. The number of queue is calculated from both the number of CPUs and the number of MSI-X vectors.
1.349	24-Jan-2024	msaitoh	ixgbe: Add QPRDC into iqdrops. A receive packet might drop at two different locations. One is the packet buffer that packets are received into the chip first. If the packet buffer is overflowed, the MPC register is incremented. It's currently added to iqdrops. It's no problem. Another is descriptor ring(s). A packet from the packet buffer is DMA'ed into main memory base on the descriptor ring. If the ring is full, the packet is dropped and the QPRDC register is incremented. It should be added to iqdrops but it was not done. Fix it. Reported by ozaki-r@.
1.348	15-Nov-2023	msaitoh	ixgbe: Clear the WTHRESH bit field before writing it.
1.347	02-Nov-2023	yamaguchi	Use ether_bpf_mtap only when the device supports vlan harware tagging The function is bpf_mtap() for ethernet devices and currently it is just handling VLAN tag stripped by the hardware.
1.346	02-Nov-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.345	30-Oct-2023	msaitoh	ixgbe: Fix comment. No functional change.
1.344	18-Oct-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.343	12-Oct-2023	msaitoh	ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate.
1.342	12-Oct-2023	msaitoh	ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl.
1.341	12-Oct-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.340	11-Oct-2023	msaitoh	ixg(4): Print DEVICE_CAPS register.
1.339	11-Oct-2023	msaitoh	ixg(4): Whitespace. No functional change.
1.338	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.337	06-Oct-2023	msaitoh	ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35.
1.336	06-Oct-2023	msaitoh	ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8.
1.335	06-Oct-2023	msaitoh	ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD.
1.334	06-Oct-2023	msaitoh	ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz.
1.333	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.332	06-Oct-2023	msaitoh	ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.331	14-Sep-2023	msaitoh	ixg(4): Rename some descriptions of flow control related event conters.
1.330	14-Sep-2023	msaitoh	ixg(4): Reorder some flow control related event counters for readability.
1.329	13-Sep-2023	msaitoh	ixg(4): Count Queue Bytes {Transmit, Receive} counter.
1.328	13-Sep-2023	msaitoh	ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based.
1.327	13-Sep-2023	msaitoh	ixgbe: Reorder some event counters for readability.
1.326	15-May-2023	msaitoh	Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.325	03-Feb-2023	msaitoh	Use thermal sensor code for IXGBE_DEV_ID_X550EM_A_10G_T, too. PCI device ID 0x15c8 also use X557-AT PHY, so create the thermal sensor sysctl for it, too.
1.324	28-Oct-2022	msaitoh	branches: 1.324.2; Make three "Unsupported SFP+ module..." messages the same.
1.323	06-Jul-2022	msaitoh	Call txeof first, then rxeof for the consistency. There are three functions where the txeof and rxeof are called. The legacy interrupt function and MSI-X function call rxeof first, then rxeof. For the workqueue function. rxeof is called first. Modify it to match with other two.
1.322	18-Jun-2022	skrll	aprintf_normal -> aprint_normal
1.321	02-Jun-2022	msaitoh	Modify comment. ixgbe_enable_queue() can be used on both MSI-X and legacy intr.
1.320	02-Jun-2022	msaitoh	KNF. Modify comment. No functional change.
1.319	02-Jun-2022	msaitoh	KNF. No functional change.
1.318	01-Jun-2022	msaitoh	Correctly enter the recovery mode. Not tested.
1.317	01-Jun-2022	msaitoh	Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Don't enable queue 1-15 interrupt. - Don't enable queue 0 interrupt when the queue work is being scheduled. - OK'd by knakahara.
1.316	01-Jun-2022	msaitoh	Simplify ixgbe_msix_que(). No functional change.
1.315	30-May-2022	msaitoh	Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857.
1.314	25-Apr-2022	msaitoh	Add missing num_tx_desc sysctl.
1.313	25-Apr-2022	msaitoh	Limit {tx,rx}_process_limit sysctl values from 1 to num_{tx,rx}_desc.
1.312	10-Mar-2022	msaitoh	Print ECC, PHY and temp error log using with ratecheck(). - The ratecheck() is for just in case. All of the interrupts might occur only once, but I don't know whether it's true or not. For the fan failure, it seems it occurs only once. - All of the interval is 60s.
1.311	10-Mar-2022	msaitoh	Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR.
1.310	10-Mar-2022	msaitoh	Move the definition of eicr_mask variable. No functional change.
1.309	10-Mar-2022	msaitoh	Modify comment to clarify EIAC setting.
1.308	10-Mar-2022	msaitoh	Add IFF_RUNNING check in ixgbe_legacy_irq() again. - This change might fix small race between ifconfig down and an interrupt. If the race really exists, txeof() is called and it's not so dangerous. The rxeof is blocked in the beginning of ixgbe_handle_que(). - This change makes consistent with ixgbe_handle_que(). - Found and OK'd by knakahara.
1.307	16-Feb-2022	msaitoh	Print Printed Board Assembly (PBA) number.
1.306	01-Feb-2022	msaitoh	Increment legacy interrupt counter after checking INTx sharing.
1.305	25-Jan-2022	msaitoh	Use atomic_{load,store}_relaxed() for event counters.
1.304	18-Jan-2022	msaitoh	Use 64bit for lxon + lxoff.
1.303	17-Jan-2022	msaitoh	Fix ierror counting again. ixgbe.c rev. 1.298 added total values kept in evcnt(9) instead of incremental values read from registers. Fix it.
1.302	31-Dec-2021	riastradh	sys: Use if_init wrapper function. Exception: Not in kern_pmf.c, for the kind of silly reason that it avoids having kern_pmf.c refer to symbols defined only in net; this avoids a pain in the rump.
1.301	24-Dec-2021	msaitoh	Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.300	16-Dec-2021	msaitoh	Print NVM image version on 82598. FreeBSD ix-3.3.29 added code to decode NVM version. On 82598, the NVM offset is NVM_EEP_OFFSET_82598(== 0x2a). My own three different 82598 cards' value in NVM_EEP_OFFSET_82598 are 0xffff. Instead, the dev starter version (0x29) has the value. Two of them have 0x1070 and another has 0x2090. The 82598 specification update notes about 2.9.0.
1.299	15-Dec-2021	msaitoh	Improve PHY's dmesg output for non-MIIVERBOSE case.
1.298	14-Dec-2021	msaitoh	Add some missing error counters to ierror. - FreeBSD: afb1aa4e6df245d38fd2ba683fa521d5dabe8392 or ix-ix-3.3.22 - Note that the checksum error is not added to the ierror.
1.297	10-Dec-2021	msaitoh	Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). From FreeBSD ix-3.3.18. No functional change.
1.296	10-Dec-2021	msaitoh	No functional change. - Sync with FreeBSD ix-3.3.18. - Rename ixgbe_get_advertise() to ixgbe_get_default_advertise. - Sort lines, modify comment and whitespace to reduce diff against FreeBSD.
1.295	07-Dec-2021	andvar	fix typos in word "instead", mainly in log messages.
1.294	12-Nov-2021	skrll	KNF
1.293	30-Sep-2021	yamaguchi	Remove vlan_ifdetach() from ixgbe_detach() The removed function is called in ether_ifdetach(). And ether_ifdetach is changed to be called before a device is stopped, because vlan_ifdetach called from ether_ifdetach configures VLAN settings.
1.292	16-Sep-2021	msaitoh	Fix a bug that an mbuf chain which has more than 63488bytes MAY fail on TX. - Currently, The TX buffer dmamap's max number of segments is set to 32. MCLBYTES(== 2048) * 32 = 65536 and it's enough for IP_MAXPACKET(65535). If an mbuf chain has more than 32 mbufs, we call m_defrag() to make it lower than equal to 32, but it might not work. The reason is that our m_defrag() don't modify the first mbuf entry of the chain. e.g.: if an mbuf chain contains 63600bytes data and the first mbuf has 100bytes in the m_data, the new chain has 100+2048+2048+...+12 and the total number of the chain is not 32 but 33. It result in 43 TCP packets will drop. - One of the way to fix this problem is to change m_defrag() which add a new mbuf cluster to the first mbuf. It's need discussion. - Another solution is to change the max number of the TX DMA segment. It should be at least 33 to avoid the m_defrag()'s current limitation. The document (82599-X550 DS 7.2.1.1 "Transmit Storage in system Memory") says that a packet can be fragmented into 40 - WTHRESH - 2 (for 82598, the document say nothing and some people and code says it's unlimited). Currently WTHRESH is set to 8. 40 - 8 - 2 = 30. !?!?!? {Net,Free,Open}BSD and Linux use 32. Is that safe? Anyway, we change WTHRESH from 8 to 5 to fit it. The added comment in ixgbe.h is based on DragonFly's though they don't use WTHRESH. - Yet another solution is to use the Tx Head Pointer Write Back function instead of WTHRESH based write back (see 82599-X550 DS 7.2.3.5.2 "Tx Head Pointer Write Back" or 82598 "Transmit Completions Head Write Back" ). DragonFly, illumos and DPDK use it. - Yet yet another solution is to add tso_maxsize entry to struct ifnet and use it in the TCP stack.
1.291	16-Sep-2021	msaitoh	Use IXGBE_TXDCTL_WTHRESH_SHIFT macro. No functional change.
1.290	26-Aug-2021	msaitoh	Remove "j" or "jumbo" because we always use MCLBYTES sized buffer.
1.289	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.288	20-Aug-2021	andvar	fix various typos in comments and log messages.
1.287	15-Jul-2021	msaitoh	Add a new sysctl to read rxr->next_to_refresh.
1.286	07-Jul-2021	msaitoh	Add new sysctl "rx_copy_len". ixgbe_rxeof() has an optimization "RX_COPY" to reduce costs of bus_dmamap_load_mbuf() and bus_dmamap_unload() by copying a mbuf cluster's memory to a newly allocated mbuf's MH_databuf[] and recycle the original map. The optimization is used when a length of a packet is smaller than a specific value. The value is calculated based on MHLEN. The size of MHLEN is architecture specific. It's 256 or 512. Make the threshold controllable by adding a new sysctl.
1.285	29-Jun-2021	pgoyette	Rework the xxxVERBOSE option to share the common module-hook-based verbose mechanism with MIIVERBOSE. This reduces some duplicated code and allows us to once again permit auto-unload of MIIVERBOSE. Change details: * Update dev/devlist2h.awk to accomodate miidevs, including generation of MII_STR_oui_model definitions and use of oui and model rather than vendor and product. This also changes the compressed data in the xxxdevs_data.h files to uint32_t (since mii oui's are up to 6 hex digits long) * Update a couple of phy drivers to use new calls to get verbose data * Regen all of the xxxdevs{,_data}.h files (separate commit, coming very soon) * Update mii/mii_verbose.[ch] and mii/mii_physubr.c to use the various DEV_VERBOSE_xxx macros * Update the pci, usb, and hdaudio code as needed, to #include the xxxdevs.h files (in order to get the proper printf format strings) * Since dev/dev_verbose.c now uses non-literal printf format strings, (to deal with the vendor/product vs oui/model issue), we need to make sure it gets compiled with -Wno-error=format-nonliteral, even in userland's libpci and librumpdev! * Bump kernel version for the change in module interfaces Welcome to 9.99.86! XXX It might be useful in the future to extend the MII_STR_oui_model XXX definitions to PCI as well (and perhaps USB and HDAUDIO). This XXX would allow for a single centralized location for the products' XXX descriptions, rather than being dispersed among individual XXX drivers' xxx_match tables.
1.284	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.283	18-May-2021	msaitoh	- Cleanup an rxbuf entry when bus_dmamap_load_mbuf() failed to prevent panic. - Print error number when error occurred.
1.282	07-May-2021	msaitoh	branches: 1.282.2; Print the error value of ixgbe_reset_hw() for debugging.
1.281	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.280	31-Mar-2021	msaitoh	branches: 1.280.2; KNF a bit. No functional change.
1.279	09-Mar-2021	msaitoh	branches: 1.279.2; Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.278	14-Jan-2021	msaitoh	Add "TX " to "Queue No Descriptor Available" to make it more understandable.
1.277	31-Dec-2020	msaitoh	Reduce code duplication. No functional change. Add new ixgbe_intr_admin_common() and use it in both ixgbe_msix_admin() and ixgbe_legacy_irq().
1.276	26-Dec-2020	msaitoh	Copy & paste some missing part (flow director, ECC and temp sensor) from ixgbe_msix_admin() to ixgbe_legacy_irq(). Now it's ready to make a new function to share the common part. It'll be done in near future.
1.275	26-Dec-2020	msaitoh	Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug.
1.274	26-Dec-2020	msaitoh	Disable some interrupt in ixgbe_{legacy_irq,msix_admin}() to prevent log spam.
1.273	26-Dec-2020	msaitoh	Disable/enable the OTHER interrupts correctly. The OTHER interrupt was not blocked correctly when MSI-X is used. ixgbe.c rev. 1.260 added new mutex to avoid the race but it didn't disable the interrupt itself. Calling ixgbe_enable_intr() enables all interrupts, so it's not good to call it when some interrupt sources should not be enabled (e.g.: calling ixgbe_enable_intr() in ixgbe_handle_admin() enables queue interrupt). IXGBE_REQUEST_TASK_NEED_ACKINTR doesn't work as expected because ixgbe_handle_admin() can't know which task is enqueued from the interrupt context and can't re-enable a specific EIMS bit. Solve the above three problems by the following two changes: - MSI-X: Disable the OTHER interrupts in the biginning of ixgbe_msix_admin(). - Set mask bits correctly at the end of ixgbe_legacy_irq() and ixgbe_msix_admin() using with eim_orig, eims_enable and eims_disable. - Remove IXGBE_REQUEST_TASK_NEED_ACKINTR and add IXGBE_REQUEST_TASK_{MOD,MSF}_WOI.
1.272	26-Dec-2020	msaitoh	Check EICR's queue bit instead of IFF_RUNNING. This change fixes a bug that it might incorrectly enable interrupt when IFF_RUNNING is not set. It also changes that the TX/RX is not processed if a queue interrupt isn't occurred.
1.271	26-Dec-2020	msaitoh	Don't use "more" flag for simplify in ixgbe_legacy_irq(). No functional change intended.
1.270	24-Dec-2020	msaitoh	Simplify setting of EIAC register. No functional change intended.
1.269	24-Dec-2020	msaitoh	Fix a bug that INTx is disabled if the INTx line is shared with other device. ixgbe.c rev. 1.264 was not correct. Restore EIMS before return. To read ECIR, clearing EIMC is required for an errata, so 0) save the original EIMS value 1) clear EIMS 2) read EICR 3) restore with the saved value.
1.268	24-Dec-2020	msaitoh	The EICR register are cleared in the beginning of the ixgbe_legacy_irq(), so it's not required to clear each bit later in the function.
1.267	24-Dec-2020	msaitoh	The EICR register's all OTHER interrupt bits are cleared in the beginning of the ixgbe_msix_admin(), so it's not required to clear each bit later in the function.
1.266	24-Dec-2020	msaitoh	Reorder some code to reduce diff between ixgbe_legacy_irq() and ixgbe_msix_admin. No functional change intended.
1.265	24-Dec-2020	msaitoh	Add some debug printf()s and modify comments.
1.264	24-Dec-2020	msaitoh	Remove strange ixgbe_enable_intr() in ixgbe_legacy_irq(). If the interface is UP and the INTx line is shared with other devices, it result in enabling all interrupt sources even if some of them are disabled for the workqueue. Delete ixgbe_enable_intr() in ixgbe_legacy_irq().
1.263	22-Dec-2020	msaitoh	Correctly ACK the fan failure interrupt in ixgbe_legacy_irq().
1.262	11-Dec-2020	msaitoh	Don't use EIMC_OTHER bit because it's read only other than 82598. Documents say: 82598: All of bit 31(OTHER bit) of EIxx are reserved. In reality, at least EIMS_OTHER and EIMC_OTHER exist and the OTHER interrupt doesn't work without EIMS_OTHER. Other than 82598: EIMS_OTHER is read only and EIMC_OTHER doesn't exist. If one of bit 29..16 is set, EIMS_OTHER is set to 1 (Note that bit 30(TCP timer isn't included)). Even if write bit 31 of EIMC to 1, it's ignored (EIMS_OTHER doesn't set). We introduced new spin mutex in ixgbe.c rev. 1.260, so it's OK to remove EIMC_OTHER stuff. We already set EIMS_OTHER in if_init(), so keep it for 82598. No functional change other than 82598. Another solution is to control bit 30..16 directly to mask/unmask interrupt instead of the mutex. TODO: Some MSI-X interrupt(LSC, module insertion/removal etc.)'s mask/unmask code between ixgbe_msix_admin() and ixgbe_handle_admin() may be wrong. It'll be fixed later.
1.261	30-Nov-2020	msaitoh	Apply ixgbe.c rev. 1.204 which was for ixgbe_msix_admin() to ixgbe_legacy_irq(), too. > An interrupt might not arrive when a module is inserted. When an link > status change interrupt occurred and the driver still regard SFP as > unplugged, issue the module softint before issuing LSC interrupt. TODO: Reduce duplicated code.
1.260	17-Nov-2020	knakahara	Add new spin mutex to avoid race between ixgbe_msix_admin() and ixgbe_handle_admin(). At first, it seems "IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_OTHER)" cannot stop interrupts, because 31th bit is reserved for 82598, 82599, X540 and X550. So, the current following design (1) ixgbe_msix_admin() disables interrupts (2) ixgbe_msix_admin() calls workqueue_enqueue() for ixgbe_handle_admin() (3) ixgbe_handle_admin() does interrupt processing (4) after ixgbe_handle_admin() has done all interrupt processings, ixgbe_handle_admin() enables interrupts does not work correctly, that is, interrupts can be lost while ixgbe_handle_admin() is running. To fix that, add new spin mutex(adapter->admmin_mtx) which protects atomically the following two members. - adapter->admin_pending - adapter->task_requests The unnecessary "IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_OTHER)" code will be removed later. Reviewed and tested by hikaru@n.o and msaitoh@n.o, thanks.
1.259	13-Nov-2020	msaitoh	Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too.
1.258	07-Sep-2020	knakahara	branches: 1.258.2; Fix race between ixgbe_msix_admin() and ixgbe_handle_admin(), pointed out by ozaki-r@n.o. The race is caused by the following. CPU#A processes workqueue, CPU#B processes admin interrupt. (0) one of CPUs already calls ixgbe_schedule_admin_tasklet() such as ixgbe_handle_timer() (1) CPU#A: read adapter->task_requests (2) CPU#B: set adapter->task_requests (3) CPU#B: read(and try to set) adapter->admin_pending but adapter->admin_pending is set, so does not call workqueue_enqueue() (4) CPU#A: clear adapter->admin_pending that is, the tasks set by (2) is not processed as missfire workqueue by (3).
1.257	07-Sep-2020	msaitoh	- Remove extra callout_stop() in ixgbe_detach(). Found by knakahara@. - Rename ix{gbe,v}_free_workqueue() to ix{gbe,v}_free_deferred_handlers(). - Add KASSERT() to functions who are called from ixgbe_handle_admin().
1.256	07-Sep-2020	msaitoh	__predict_false() should be __predict_true() for adapter->osdep.detaching == false check.
1.255	07-Sep-2020	msaitoh	Don't schedule admin work while shutdown like the recovery mode timer. The admin workqueue also runs while !IFF_UP like the recovery mode timer. Apply the same change of ixgbe.c 1.254 to ixgbe_schedule_admin_tasklet() to prevent panic. Found by ozaki-r@.
1.254	01-Sep-2020	msaitoh	Fix a panic on shutdown on a machine which use the recovery mode timer. The recovery mode timer is first issued by the callout and it schedule the workqueue. The workqueue then reschedule the callout. It's hard to stop both of them without race only with callout_stop() and workqueue_wait. To solve this problem. add new "detaching" flag and use it. The situation is almost the same as schedule_wqs_ok for the local_timer's callout and workqueue, but the difference is that the local_timer isn't required to run if the interface is not up. If it's not important to prevent running timer while !IFF_UP, the flag can be integrated into one.
1.253	01-Sep-2020	msaitoh	Call ixgbe_ifstop() instead of ixgbe_stop_locked() in ixgbe_setup_low_power_mode() to stop the timer workqueue.
1.252	31-Aug-2020	msaitoh	Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change.
1.251	31-Aug-2020	msaitoh	If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.250	31-Aug-2020	msaitoh	Fix typo in comment.
1.249	31-Aug-2020	msaitoh	Fix 82598 SFP+ problems. On 82598, SFP+'s MOD_ABS isn't connected to the MAC's GPIO pin, so we can't call ixgbe_sfp_cage_full(). Always issue TASK_MOD from ixgbe_handle_timer() on 82598. Fix ixgbe_identify_sfp_module_generic() for ixgbe_phy_nl. In the driver, hw->phy.type sometimes be compared with ixgbe_phy_nl. In ixgbe_identify_sfp_module_generic(), hw->phy.type may be overridden with another value. For ixgbe_phy_nl, some code don't override phy.type but others were not. Make it consistently keep ixgbe_phy_nl. This change fixes a problem that ixgbe_is_sfp() change the return value true to false when any SFP+ devices are connected to the cage on 82598 and never recover from it. Don't schedule MSF(multi speed fiber) task from ixgbe_handle_mod() on 82598. This task is only for devices which support multi speed fiber and 82598 doesn't support it. Before ixgbe.c rev. 1.237, ixgbe_handle_mod() isn't called on 82598 because 82598 has no SFP+ module insertion/removal interrupt. ixgbe.c rev. 1.237 changed to call the function via timer on 82598. This change fixes a bug that 82598 DA interface's link flaps.
1.248	27-Aug-2020	msaitoh	Limit MA10-ST0's quirk only for on-chip devices.
1.247	27-Aug-2020	msaitoh	Fix compile error.
1.246	27-Aug-2020	msaitoh	Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.245	27-Aug-2020	msaitoh	ADD SFP+ MOD_ABS inversion quirk. On X550 EM, GPIO(SDP) and SFP+'s MOD_ABS is directly connected. It has no inverter. GIGABYTE MA10-ST0 has a inverter, so add new quirk for it.
1.244	24-Aug-2020	msaitoh	Fix handling of IXGBE_REQUEST_TASK_NEED_ACKINTR again...
1.243	24-Aug-2020	msaitoh	The admin workqueue can be used even if the interface is not up. OK'd by thorpej@. Will fixes PR#55534 reported by Shinichi Doyashiki
1.242	24-Aug-2020	msaitoh	Fix ixgbe_sfp_cage_full() on X550EM_A. In ixgbe_handle_mod(): switch (hw->mac.type) { case ixgbe_mac_82599EB: cage_full = IXGBE_READ_REG(hw, IXGBE_ESDP) & IXGBE_ESDP_SDP2; break; case ixgbe_mac_X550EM_x: case ixgbe_mac_X550EM_a: cage_full = IXGBE_READ_REG(hw, IXGBE_ESDP) & IXGBE_ESDP_SDP0; break; default: break; } so I had thought that IXGBE_ESDP_SDP0 bit is 1 on cage is full. In reality, at least, X550EM_A's SFP+ cage is 0 on cage is full. So invert the logic of ixgbe_sfp_cage_full() on X550EM_A
1.241	24-Aug-2020	msaitoh	Fix race in ixgbe_detach() to prevent panic on shutdown.
1.240	24-Aug-2020	msaitoh	Fix handling of IXGBE_REQUEST_TASK_NEED_ACKINTR in ixgbe_handle_admin().
1.239	17-Aug-2020	msaitoh	Simplify SFP+ check. No functional change.
1.238	17-Aug-2020	msaitoh	Re-enabling interrupt is required only when a work is scheduled form the interrput context.
1.237	17-Aug-2020	msaitoh	Fix a bug that the driver sometimes missed module insertion. The ixgbe_sfp_probe() function was only for 82598 and other chips had no way to poll SFP+ cage. The ixgbe_handle_mod() already has function to treat module insertion/removal for all chips, so enqueue the work if the cage status changed. All of ixgbe chips' SFP+ module interrupt is only on the inserstion. This change also detect the removal by the timer.
1.236	17-Aug-2020	msaitoh	Add missing workqueue_wait() for the recovery_mode_timer workqueue.
1.235	13-Aug-2020	msaitoh	Set recovery_mode_timer workqueue's name correctly.
1.234	13-Aug-2020	msaitoh	Use atomic_cas_uint() and atomic_store_relaxed(). Advised by thorpej@. Tested by me. OK'd by knakahara.
1.233	25-Jun-2020	msaitoh	Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.232	18-Jun-2020	msaitoh	Modify a liitle to reduce diff between ixgbe.c and ixv.c. No functional change.
1.231	17-Jun-2020	msaitoh	Add missing callout_stop() for recovery_mode_timer. NOTE: One of the difference between the local_timer and recovery_mode_timer is that local_timer runs only when IFF_UP and recovery_mode_timer always runs (it's enabled on attach).
1.230	12-Jun-2020	msaitoh	- Remove extra kpreempt_disable() -> kpreempt_enable() because ixgbe_handle_msf() was changed from softint to workqueue. - Set schedule_wqs_ok before enabling interrupt to prevent the race. - Fix comment.
1.229	11-Jun-2020	msaitoh	Fix typo. No functional change.
1.228	17-Apr-2020	msaitoh	No functional change: - modify comment - whitespace fix
1.227	15-Mar-2020	thorpej	branches: 1.227.2; Define and implement a locking protocol for the ifmedia / mii layers: - MP-safe drivers provide a mutex to ifmedia that is used to serialize access to media-related structures / hardware regsiters. Converted drivers use the new ifmedia_init_with_lock() function for this. The new name is provided to ease the transition. - Un-converted drivers continue to call ifmedia_init(), which will supply a compatibility lock to be used instead. Several media-related entry points must be aware of this compatibility lock, and are able to acquire it recursively a limited number of times, if needed. This is a SPIN mutex with priority IPL_NET. - This same lock is used to serialize access to PHY registers and other MII-related data structures. The PHY drivers are modified to acquire and release the lock, as needed, and assert the lock is held as a diagnostic aid. The "usbnet" framework has had an overhaul of its internal locking protocols to fit in with the media / mii changes, and the drivers adapted. USB wifi drivers have been changed to provide their own adaptive mutex to the ifmedia later via a new ieee80211_media_init_with_lock() function. This is required because the USB drivers need an adaptive mutex. Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv. mcx also now calls ifmedia_init_with_lock() because it needs to also use an adaptive mutex. The mcx driver still needs to be fully converted to NET_MPSAFE.
1.226	06-Feb-2020	thorpej	Ensure we don't call workqueue_enqueue() if the pluggable optics handler is already pending.
1.225	05-Feb-2020	msaitoh	No functional change: - Add debug printf()s. - Remove unused macros. - Remove extra newline.
1.224	05-Feb-2020	msaitoh	Update comment in ixgbe_update_stats_counters() to clarify why we update if_iqdrops and if_ierrors only. No functional change. OK'd by knakahara and thorpej.
1.223	04-Feb-2020	thorpej	- Fix locking problem with optics module interrupts: ifmedia_add() may block on memory allocation, and so it cannot be safely done from a softint nor can it be done while holding a spin lock. Fix this by using a workqueue rather than a softint, and hold the IFNET_LOCK across the entire handler, and the CORE_LOCK only across the code that needs to serialize access to the hardware state. - Use ifmedia_fini(). Tested in a variety of devices by msaitoh@. (Thanks!)
1.222	01-Feb-2020	thorpej	Adopt <net/if_stats.h>. XXX This driver needs some work in this regard (practually no stats are reported).
1.221	21-Jan-2020	msaitoh	Fix the freeing code for some error paths. Found and tested by Patrick Welche.
1.220	03-Jan-2020	pgoyette	branches: 1.220.2; Fix some typos in comments. From vezhlys on freenode IRC.
1.219	23-Dec-2019	msaitoh	Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.218	23-Dec-2019	msaitoh	Add missing core lock in ixgbe_handle_mod().
1.217	17-Dec-2019	msaitoh	Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@.
1.216	18-Nov-2019	msaitoh	Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.215	10-Nov-2019	chs	in many device attach paths, allocate memory with M_WAITOK instead of M_NOWAIT and remove code to handle failures that can no longer happen.
1.214	16-Oct-2019	msaitoh	Apply FreeBSD r353599: > ixgbe: Disable EEE for backplane X550EM_X > > From Zach: > Intel documentation indicates that backplane X550EM_X KR devices do not > support Energy Efficient Ethernet. Prior to this patch, X552 devices > (device ID 0x15AB) will crash the system when transitioning EEE state > via sysctl. > > Signed-off-by: Zach Vargas <zvargas@xes-inc.com> > > PR: 240320 > Submitted by: Zach Vargas <zvargas@xes-inc.com> > Reviewed by: erj@ > MFC after: 3 days > Differential Revision: https://reviews.freebsd.org/D21673
1.213	18-Sep-2019	msaitoh	s/ixgbe_set_multi/ixgbe_set_rxfilter/. No functional change.
1.212	18-Sep-2019	msaitoh	Remove ixgbe_set_promisc() and use ixgbe_set_multi() to reduce code duplication.
1.211	18-Sep-2019	msaitoh	Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. Changing "<" with "<=" fixes this problem, but it's better to check ETHER_F_ALLMULTI than it because it's straightforward.
1.210	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.209	05-Sep-2019	msaitoh	Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer.
1.208	04-Sep-2019	msaitoh	Set IFM_1000_BX10 correctly.
1.207	04-Sep-2019	msaitoh	printf -> device_printf
1.206	29-Aug-2019	knakahara	Fix panic when ncpu == 1 && sysctl -w hw.ixg0.txrx_workqueue=1. Reported by nonaka@n.o. ok by msaitoh@n.o and nonaka@n.o pullup-8, pullup-9
1.205	29-Aug-2019	knakahara	fix typo in error handling of ixgbe_allocate_legacy(). ok by msaitoh@n.o pullup-8, pullup-9
1.204	28-Aug-2019	msaitoh	An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, link becomes up and the real media type is unknown. e.g: % ifconfig -m ixg0 (snip) media: Ethernet autoselect (autoselect rxpause,txpause) status: active supported Ethernet media: media none media autoselect (snip) To resolve this problem, when an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC interrupt.
1.203	28-Aug-2019	msaitoh	X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too.
1.202	21-Aug-2019	msaitoh	Simplify ix{gbe,v}_[un]register_vlan() API suggestesd by knakahara. The API was the same as FreeBSD's pre-iflib's. They use iflib now and it's not required for us to keep the old API.
1.201	21-Aug-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() suggested by knakahara.
1.200	20-Aug-2019	msaitoh	Fix a bug that VLAN HW "tagging" enable/disable may not refrect correctly. - Always call ec_vlan_cb() if it exists. - Some (or all?) ethernet drivers don't enable HW tagging if no any vlan is attached. ixgbe is one of them. Check the the transition and update VLAN HW tagging function. XXX pullup-9
1.199	30-Jul-2019	msaitoh	branches: 1.199.2; ixgbe_setup_low_power_mode() calls ixgbe_stop() in it. ixgbe_detach() first called ixgbe_stop() and then called ixgbe_setup_low_power_mode() which resulted in calling ixgbe_stop() twice. Call only ixgbe_setup_low_power_mode() first to remove extra ixgbe_stop() call.
1.198	26-Jul-2019	msaitoh	Set IVAR (Interrupt Vector Allocation Register) correctly for 82598 in ixgbe_set_ivar(). Found by KUBSan.
1.197	26-Jul-2019	msaitoh	Simplify code. No functional change intended.
1.196	25-Jul-2019	msaitoh	Fix VLAN hardware filter initialization. Use unsigned. Found by KUBSan.
1.195	25-Jul-2019	msaitoh	Fix VLAN filter setting. Found by KUBSan.
1.194	24-Jul-2019	msaitoh	Fix some problems found by KUBSan: - Set TQSM and RQSM (TX/RX queue statistics mapping) correctly. - Set IVAR (Interrupt Vector Allocation Register) correctly in ixgbe_set_ivar(). - Define IXGBE_KRM_LINK_CTRL_1_TETH_AN_RESTART correctly. This macro is used in ixgbe_restart_an_internal_phy_x550em(). - Define IXGBE_KRM_TX_COEFF_CTRL_1_OVRRD_EN correctly. This macro is used in ixgbe_setup_ixfi_x550em_x(). - Define IXGBE_SB_IOSF_CTRL_BUSY correctly. This macro is used in ixgbe_iosf_wait() (X550EM only).
1.193	17-Jul-2019	msaitoh	Implement VLAN hardware filter function(ETHERCAP_VLAN_HWFILTER). First proposed by jmcneill in 2017 and modified by me. How to use: - Set callback function: ether_set_vlan_cb(struct ethercom , ether_vlancb_t) - Callback. This function is called when a vlan is attached/detached to the parent interface: int (ether_vlancb_t)(struct ethercom *ec, uint16_t vlanid, bool set); - ifconfig(8) ifconfig ixg0 [-]vlan-hwfilter Note that ETHERCAP_VLAN_HWFILTER is set by default on ixg(4) because the PF driver usually enable "all block" filter by default.
1.192	04-Jul-2019	msaitoh	On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing.
1.191	02-Jul-2019	msaitoh	Sync with FreeBSD ix-3.3.10 part 2: - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix().
1.190	02-Jul-2019	msaitoh	Sync with FreeBSD ix-3.3.10 part 1. No functional change in this part: - "(u64)1" -> "1ULL" - Add some not-yet-used register definitions.
1.189	27-Jun-2019	msaitoh	Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.188	04-Jun-2019	msaitoh	Match the follwing devices: 82598_BX (0x1508) 82599_KR (0x1517) 82599_SFP_EM (0x1507) X550EM_X_XFI (0x15b0) X550EM_A_QSFP (0x15ca) X550EM_A_QSFP_N(0x15cc)
1.187	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.186	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.185	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.184	17-May-2019	msaitoh	Use new media types.
1.183	15-May-2019	ozaki-r	Store IFF_ALLMULTI in ec_flags instead of if_flags to avoid data races IFF_ALLMULTI is set/unset to if_flags via if_mcast_op. To avoid data races on if_flags, IFNET_LOCK was added for if_mcast_op. Unfortunately it produces a deadlock so we want to remove added IFNET_LOCK by avoiding the data races by another approach. This fix introduces ec_flags to struct ethercom and stores IFF_ALLMULTI to it. ec_flags is protected by ETHER_LOCK and thus IFNET_LOCK is no longer necessary for if_mcast_op. Note that the fix is applied only to MP-safe drivers that the data races matter. In the kernel, IFF_ALLMULTI is set by a driver and used by the driver itself. So changing the storing place doesn't break anything. One exception is ioctl(SIOCGIFFLAGS); we have to include IFF_ALLMULTI in a result if needed to export the flag as well as before. A upcoming commit will remove IFNET_LOCK. PR kern/54189
1.182	14-May-2019	ozaki-r	Remove unnecessary checks of IFF_ALLMULTI IFF_ALLMULTI is changed by only driver itself, so we don't need to check its change on ec_ifflags_cb. This is part of PR kern/54189. NFCI.
1.181	13-May-2019	msaitoh	Fix a bug that manual setting of 10GBASE-SR or 10GBASE-CX4 didn't work if IFM_ETH_XTYPE was set. We don't use IFM_ETH_XTYPE macro, so this was not a real bug.
1.180	10-May-2019	msaitoh	Fix typo. This code is not used yet.
1.179	18-Mar-2019	msaitoh	s/pakcet/packet/ in comment.
1.178	15-Mar-2019	msaitoh	- Simplily. Suggested by knakahara. - Modify comment. Per queue VLAN enable flags is on 82599 and later. - Fix typo in comment.
1.177	13-Mar-2019	msaitoh	- Fix a bug that the VLAN HW tagging function is not correctly disabled when all vlan is detached. - Fix a bug that VLAN HW tagging function is not correctly controlled on 82598. - Control VLAN HW filter function correctly. - Don't clear IXGBE_VLNCTRL_CFIEN bit When ETHERCAP_VLAN_HWFILTER is set. I think it's not required (and Linux doesn't do it). This change has no effect to NetBSD because ETHERCAP_VLAN_HWFILTER is not supported yet.
1.176	05-Mar-2019	msaitoh	It's not required to calculate unused queues' statistics.
1.175	05-Mar-2019	msaitoh	NetBSD currently uses traffic class 0 only. Other traffic classes aren't used yet. When IXGBE_TC_COUNTER_NUM is set to lower than IXGBE_DCB_MAX_TRAFFIC_CLASS (e.g. 1), other traffic classes' counters are not used. It means we don't generate evcnt for them and don't add the values in ixgbe_update_stats_counters().
1.174	22-Feb-2019	msaitoh	Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.173	23-Jan-2019	msaitoh	- Move the location of calling setup_eee() in ixgbe_attach(). The previous location was too early. The new location is before calling ixgbe_set_phy_power(). - Restore call of setup_eee() in ixgbe_init_locked(). It was removed in rev. 1.99 (FreeBSD r320688). - When calling setup_eee(), pass enable/disable correctly.
1.172	23-Jan-2019	msaitoh	Call ixgbe_init_swfw_semaphore(hw) before calling ixgbe_check_wol_support() because ixgbe_check_wol_support()->ixgbe_check_wol_support()-> ixgbe_get_device_caps() accesses semaphore.
1.171	23-Jan-2019	msaitoh	Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them expilicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.170	08-Dec-2018	msaitoh	Don't call callout_halt() if it's not initialzied.
1.169	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.168	03-Dec-2018	msaitoh	Some counters are not per queue but per traffic class. Make new evcnt group "ixgM tcN" (N = 0..7) and move those counters into it. We are using only traffic class 0, so we will see only tc0's counter are updated.
1.167	27-Sep-2018	msaitoh	Don't set LRO capability flags because we don't support it yet. This doesn't change any behavior other than if_capabilities' visibility (e.g. ifconfig).
1.166	14-Sep-2018	msaitoh	- Fix a bug that Denverton accessed wrong register in ixgbe_setup_low_power_mode(). The location of some registers in Denverton were changed. The GRC register is one of them, so use IXGBE_GRC_BY_MAC(hw) instead of IXGBE_GRC. - ixgbe_check_wol_support() tests the capability of WOL, but the sysctl knobs for WOL were added only for IXGBE_DEV_ID_X550EM_X_10G_T. Add the knobs when adapter->wol_support is not 0. XXX pullup-8
1.165	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.164	29-Aug-2018	msaitoh	Fix a bug that media change may fail. I noticed that ifconfig ixgN media XXX took a 10 or more seconds on a Denverton machie. I occurred by trying take a lock and timed out. The reason was that ixgbe_media_change() didn't take CORE_LOCK. Do it. This problem was from FreeBSD's pre-iflib ixgbe (I don't know whether this problem really occurs on FreeBSD or not). Post-iflib ixgbe has no problem because iflib_media_change() takes a lock. XXX pullup-8
1.163	06-Jul-2018	msaitoh	Add force_10_100_autonego sysctl. The default vaule is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while.
1.162	03-Jul-2018	msaitoh	Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.161	06-Jun-2018	kamil	branches: 1.161.2; Comment out unused functions in order to fix Clang build ixv_rearm_queues() and ixgbe_rearm_queues() are unused. Sponsored by <The NetBSD Foundation>
1.160	04-Jun-2018	msaitoh	Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.159	03-Jun-2018	maxv	Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes).
1.158	30-May-2018	msaitoh	Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console.
1.157	30-May-2018	msaitoh	Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294.
1.156	25-May-2018	ozaki-r	Ensure to call if_register after interface initializations finish
1.155	23-May-2018	msaitoh	Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE.
1.154	23-May-2018	msaitoh	Add rxd_nxck (Receive Descriptor next to check) read only sysctl.
1.153	18-May-2018	msaitoh	Revert new watchdog timer commits. The new watchdog timer made stability worse than before. It seems unknown problems exists. http://mail-index.netbsd.org/source-changes/2018/05/08/msg095020.html http://mail-index.netbsd.org/source-changes/2018/05/16/msg095240.html
1.152	15-May-2018	msaitoh	The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change.
1.151	10-May-2018	msaitoh	Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT].
1.150	08-May-2018	msaitoh	- Fix broken watchdog timer. This change detects TX device timeout correctly. NOTE: It's supporsed not to be called {ixgbe,ixv}_rearm_queues() in the timer. Those are not required if any chip have no bug. In reality, ixgbe_rearm_queues() is required on 82599 and newer chip AND other than queue 0 to prevent device timeout. When it occured, packet was sent but the descriptor's DD bit wasn't set even though IXGBE_TXD_CMD_EOP and IXGBE_TXD_CMD_RS were set. After forcing interrupt by writing EICS register in ixgbe_rearm_queues(), DD is set. Why? Is this an undocumented errata? It might be possible not call rearm_queues on 82598 or queue 0, we call in any cases in case the problem occurs. On ixv(4), I have not seen this problem yet (though I tested only on X550_X(Xeon D 12xx)'s virtual function), but we do rearm in case TX device timeout happen. - ixv(4): Call callout_stop() earlier in ixv_stop() like ixgbe_stop(). - KNF.
1.149	19-Apr-2018	msaitoh	Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init().
1.148	17-Apr-2018	msaitoh	Remove unused structure entries. No functional change.
1.147	17-Apr-2018	knakahara	Fix panic when "sysctl -w hw.ixg0.txrx_workqueue=[01]" while there is traffic. The operation is not supported, however causing panic is problem.
1.146	12-Apr-2018	msaitoh	Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.145	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.144	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.143	04-Apr-2018	msaitoh	Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.142	02-Apr-2018	msaitoh	Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031.
1.141	02-Apr-2018	knakahara	Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o.
1.140	30-Mar-2018	msaitoh	- Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.139	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.138	30-Mar-2018	knakahara	Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8
1.137	26-Mar-2018	msaitoh	- Add link related softint's counter. - Fix indent.
1.136	20-Mar-2018	knakahara	Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me.
1.135	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.134	09-Mar-2018	msaitoh	Make some event counters MP safe. Now all of the event counters are MP safe.
1.133	08-Mar-2018	knakahara	Reduce duplicated code which schedule deferred packet processing. No functional change.
1.132	08-Mar-2018	knakahara	Fix INTx/MSI handler did not schedule workqueue. Pointed out by msaitoh@n.o.
1.131	07-Mar-2018	knakahara	Fix another poll mode assumption breaking. Implemented by msaitoh@n.o, I just commit by proxy. ixgbe_rearm_queues() writes EICS register(s). 82599, X540 and X550 specifications say "Following a write of 1b to any bit in the EICS register (interrupt cause set), its corresponding bit in the EIMS register is auto set as well enabling its interrupt." in "Extended Interrupt Auto Mask Enable (EIAM) Register" section. That is, ixgbe_rearm_queues() causes interrupts regardless of the status managed by ixgbe_enable_queue()/ixgbe_disable_queue(). That can break poll mode assumption. In fact, the problem occurs in the following situation - CPU#A has high load traffic, in contrast, CPU#B has not so high load traffic - CPU#A is occurred interrupt by its NIC queue - CPU#A calls ixgbe_disable_queue() in interrupt handler(ixgbe_msix_que()) - CPU#A kick softint handler(ixgbe_handle_que()) - CPU#A begins softint - CPU#A's NIC queue is set que->txr->busy flag - With some reason, CPU#A can do ixg interrupt handler E.g. when one of CPU#A's softnet handlers sleeps, ipl is lowered - CPU#B starts callout - CPU#B calls ixgbe_local_timer1() - CPU#B writes EICS bit corresponding CPU#A's NIC queue bit - CPU#A's NIC queue causes interrupt whie CPU#A is running in poll mode - CPU#A calls ixgbe_disable_queue() in interrupt handler again - CPU#A has done polling, and then CPU#A calls ixgbe_enable_queue() once - CPU#A's NIC queue interrupt is disabled until ixg is detached as ixgbe_disable_queue() is called twice though ixgbe_disable_queue() is called once only NOTE: 82598 does not say so, but it is treated in the same way because of no harm. By the way, we will refactor ixgbe_local_timer(watchdog processing) later. XXX pullup-8
1.130	07-Mar-2018	msaitoh	Don't increment que->req.ev_count in MSI-X interrupt because it's not reschedule.
1.129	07-Mar-2018	msaitoh	- Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code.
1.128	02-Mar-2018	knakahara	branches: 1.128.2; ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o.
1.127	26-Feb-2018	knakahara	Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o.
1.126	22-Feb-2018	msaitoh	Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.125	20-Feb-2018	knakahara	Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o.
1.124	20-Feb-2018	msaitoh	- Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8
1.123	16-Feb-2018	msaitoh	Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv)
1.122	16-Feb-2018	knakahara	Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o.
1.121	14-Feb-2018	msaitoh	Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8
1.120	26-Jan-2018	msaitoh	CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558).
1.119	28-Dec-2017	msaitoh	Fallback from MSI-X to MSI or INTx if MSI-X setup failed.
1.118	21-Dec-2017	msaitoh	- If MSI can't be used on some environment, fallback to INTx correctly. - Use single vector MSI when number of CPU is 1 to save interrupt slot.
1.117	21-Dec-2017	msaitoh	Fix panic when only link interrupt can't be established.
1.116	20-Dec-2017	msaitoh	Don't panic when resource shortage occured. Fixes PR#52820 reported by kardel@: - Don't use if_free() because ixgbe(4) don't use if_alloc(). - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more).
1.115	06-Dec-2017	msaitoh	Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.114	24-Nov-2017	msaitoh	On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0"
1.113	22-Nov-2017	msaitoh	Fix a bug that bypass adapter's sysctls aren't set. Tested with non-genuine X550-T2 bypass adapter: - Call ixgbe_sysctl_instance() in ixgbe_bypass_init() to get sysctl's top node correctly. - ixgbe_init_device_features() refers adapter->hw.bus.func for bypass adapter. Call set_lan_id() to set adapter->hw.bus.func before calling ixgbe_init_device_features(). Without this, bypass sysctl's are added to both the first and second port. - Initalize node.sysctl_data before calling sysctl_lookup() to read correct value.
1.112	16-Nov-2017	ozaki-r	Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@
1.111	12-Nov-2017	msaitoh	Remove obsoleted comment. This comment should have been removed in rev. 1.97.
1.110	09-Nov-2017	msaitoh	On device which has SFP(+) cage and a module is inserted, hw->phy.id is not MII PHY id but SFF 8024 ID. So checking hw->phy.id with 0 doesn't work. Print PHY ID only for copper PHY.
1.109	02-Nov-2017	msaitoh	- Handle hw.ixgN.advertise_speed correctly. - Fix hw.ixgN.advertise_speed sysctl's description.
1.108	26-Oct-2017	msaitoh	- Remove all half duplex setting. - Remove special handling of IXGBE_DEV_ID_82598AT's 1000BaseT. The PHY is compliant with clause 45, so the media is added in the beginning of ixgbe_add_media_types().
1.107	25-Oct-2017	msaitoh	- For X550 and newer, print NVM Image Version with %u.%02x (e.g. "0.05"). - Fix value check of OEM_NVM_IMAGE_VER.
1.106	23-Oct-2017	msaitoh	If if_initialize() failed in the attach function, free resources and return.
1.105	18-Oct-2017	msaitoh	Protect ec_multi* with mutex like other MP safe Ethernet drivers. See if_ether.h rev. 1.66 and related stuff: http://mail-index.netbsd.org/source-changes/2016/12/28/msg080289.html
1.104	13-Oct-2017	msaitoh	Set baudrate correctly. A device which use mii(4) interface can set if_baudrate atomatically by mii_phy_statusmsg(). ixg(4) and ixv(4) don't use mii(4), so calling ifmedia_baudrate() directly is required. XXX need the same same work for wm(4)'s TBI and SERDEV devices.
1.103	04-Oct-2017	msaitoh	- Remove half-duplex setting for more than 1000Mbps in ixgbe_add_media_types() because those medias are not in the spec. - Add 2.5GBASE-T and 5GBASE-T support. Tested with X550-T1. Example: ixg0: flags=0x8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500 capabilities=fff80<TSO4,IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx> capabilities=fff80<TCP4CSUM_Tx,UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx> capabilities=fff80<TCP6CSUM_Tx,UDP6CSUM_Rx,UDP6CSUM_Tx,TSO6,LRO> enabled=7ff80<TSO4,IP4CSUM_Rx,IP4CSUM_Tx,TCP4CSUM_Rx> enabled=7ff80<TCP4CSUM_Tx,UDP4CSUM_Rx,UDP4CSUM_Tx,TCP6CSUM_Rx> enabled=7ff80<TCP6CSUM_Tx,UDP6CSUM_Rx,UDP6CSUM_Tx,TSO6> ec_capabilities=7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU> ec_enabled=7<VLAN_MTU,VLAN_HWTAGGING,JUMBO_MTU> address: a0:36:9f:b0:f5:b0 media: Ethernet 2.5GBASE-T full-duplex status: active supported Ethernet media: media 10Gbase-T media 10Gbase-T mediaopt full-duplex media 1000baseT media 1000baseT mediaopt full-duplex media 100baseTX media 100baseTX mediaopt full-duplex media 2.5GBASE-T mediaopt full-duplex media 5GBASE-T mediaopt full-duplex media autoselect inet 172.16.0.1/24 broadcast 172.16.0.255 flags 0x0 inet6 fe80::a236:9fff:feb0:f5b0%ixg0/64 flags 0x0 scopeid 0x1
1.102	04-Oct-2017	msaitoh	Fix a bug that X550EM (Denverton) can't force 10BaseT.
1.101	27-Sep-2017	msaitoh	Print device name in ixv_attach().
1.100	31-Aug-2017	msaitoh	Fix 2500 KX media setting in last commit.
1.99	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.98	30-Aug-2017	msaitoh	FreeBSD r320688 sorted a lot of functions. Merging that change by one commit makes difficult to know what was really changed. To make change understandable, commit only function sort and white space changes before real change.
1.97	30-Aug-2017	msaitoh	Don't limit number of queue pair to 8. Take max TX queues and max RX queues into account. 82598 has 32 TX queue and 64 RX queues. Others has 128 TX queues and 128 RX queues. Number of MSI-X table entries is 64, so the max queue pairs is 63 (64 minus the Others (e.g. link status change)).
1.96	24-Aug-2017	msaitoh	Remove unused counters.
1.95	03-Jul-2017	msaitoh	Print PHY ID.
1.94	27-Jun-2017	msaitoh	Fix a bug of ixg(4)'s media setting. Before: ifconfig ixg0 media 100baseTX -> advertise 100Mbps only ifconfig ixg0 media 1000baseT -> advertise 1Gbps and 1000Mbps (NG) ifconfig ixg0 media 10Gbase-T -> advertise all (NG) ifconfig ixg0 media auto -> advertise all After: ifconfig ixg0 media 100baseTX -> advertise 100Mbps only ifconfig ixg0 media 1000baseT -> advertise 1Gbps only ifconfig ixg0 media 10Gbase-T -> advertise 10Gbps only ifconfig ixg0 media auto -> advertise all
1.93	27-Jun-2017	msaitoh	Reduce diff against FreeBSD. change hw.ixgN.ts to hw.ixgN.thermal_test (FreeBSD r282289).
1.92	27-Jun-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.91	23-Jun-2017	msaitoh	Print verbose outout in ixgbe_update_link_status() correctly on 5Gbps and 2.5Gbps (NBASE-T) for X550T[12] and newer.
1.90	23-Jun-2017	msaitoh	Use IFM_1000_KX more.
1.89	12-Jun-2017	msaitoh	Use IFM_1000_KX and IFM_2500_KX.
1.88	02-Jun-2017	msaitoh	branches: 1.88.2; - Print NVM Image Version, PHY FW Revision, NVM Map version and OEM NVM Image version - s/TrackID/ETrackID/
1.87	26-May-2017	msaitoh	Add missing {ixgbe,ixv}_stop() in the detach function. Without this change, shutdown won't finish or panic on heay traffic.
1.86	26-May-2017	msaitoh	Count Bad SFD (Start Frame Delimiter). X550 and newer have this register.
1.85	26-May-2017	msaitoh	Clear all event counters when SIOCZIFDATA(ifconfig -z) is called.
1.84	22-May-2017	msaitoh	Fix a bug that the driver prints "Link is up 1 Gbps" even if the link is 100Mbps. This change is the same as a part of ixv.c rev. 1.42.
1.83	22-May-2017	msaitoh	Document says M[LR]FC are valid when link is up and 10Gbps, so: - Add those values when link is up and 10Gbps - Discard those vaules when link becomes up and the speed is 10Gbps
1.82	18-May-2017	msaitoh	- Count "Total Packets Missed" by evcnt(9) - Call evcnt_attach_dynamic()/detach() for stats->mngpdc (Management Packets Dropped). - Make stats->rnbc (Receive No Buffers) evcnt(9). Only 82598 has this register. - Set RQSMR, TQSM (or TQSMR on 82598) to count qprc, qptc, qbrc, qbtc and qprdc in each queue counter. Without this change, all queue's counts are counted in queue 0's. Even if with this change, only MPC[0] is counted and other MPCs are not counted.
1.81	28-Apr-2017	msaitoh	Fix a bug that reading hw.ixgM.qN.interrupt_rate cause the interrupt throtting rate register to 0. Check sysctl read/write correctly...
1.80	03-Mar-2017	msaitoh	branches: 1.80.4; Set PCI_INTR_MPSAFE, SOFTINT_MPSAFE, CALLOUT_MPSAFE and IFEF_START_MPSAFE if NET_MPSAFE is defined.
1.79	24-Feb-2017	msaitoh	Print mac.type and the TrackID.
1.78	16-Feb-2017	msaitoh	Do if_snd check even if IXGBE_LEGACY_TX is used. It's used by ALTQ.
1.77	16-Feb-2017	msaitoh	txt_si softint is used only for multiqueue (ixgbe_mq_start_locked()), so don't schedule it in ixgbe_legacy_irq().
1.76	13-Feb-2017	msaitoh	Remove more ix_start_locked() from interrupt context like previous commit. > Fix a bug that ix_start_locked() is called in interrput context > (ix_msix_que). The function is called in softint(ix_handle_que()). OK'd by >k-nakahara.
1.75	13-Feb-2017	msaitoh	Use percpuq.
1.74	13-Feb-2017	msaitoh	Fix a bug that ix_start_locked() is called in interrput context (ix_msix_que). The function is called in softint(ix*_handle_que()). OK'd by k-nakahara.
1.73	10-Feb-2017	msaitoh	Change the Interrupt Moderation flag from global variable to per device.
1.72	10-Feb-2017	msaitoh	- Both mbuf_defrag_failed and m_defrag_failed existed. Remove m_defrag_failed. - rx_irq, dropped_pkts, morerx, moretx and txloops counters are uncounted, so remove them. - Count handleq in ixv_handle_que(). - Detach event counters correctly. - Set some per-queue event names correctly.
1.71	08-Feb-2017	msaitoh	Print CPU number correctly.
1.70	08-Feb-2017	msaitoh	Call IFQ_SET_MAXLEN() even if TX multiqueue is used. ALTQ uses it.
1.69	08-Feb-2017	msaitoh	Add queue number into interrupt name.
1.68	08-Feb-2017	msaitoh	Set ifm_active to IFM_NONE when the physical interface's link is down. Now ifconfig shows "media: Ethernet autoselect '(none)'".
1.67	01-Feb-2017	msaitoh	TX multiqueue. If you want to disable it, enable IXGBE_LEGACY_TX in ixgbe_netbsd.h
1.66	25-Jan-2017	msaitoh	Call ixgbe_get_supported_physical_layer() before ixgbe_add_media_types() to set new media types correctly.
1.65	25-Jan-2017	msaitoh	Remove debug printf()s.
1.64	25-Jan-2017	msaitoh	Fix SFP+ insersion or replacement stuff.
1.63	19-Jan-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.62	18-Jan-2017	msaitoh	Now we can use multiqueue. It's default on ixg(4). Not yet for ixv(4).
1.61	18-Jan-2017	msaitoh	Fix sysctl stuff. Make the top sysctl node and keep the pointer.
1.60	18-Jan-2017	msaitoh	Use 64bit DMA tag. If not, a lot of bounce buffer is allocated. Fixes PR#49968 reported by Hauke. XXX pullup-[67].
1.59	05-Jan-2017	msaitoh	branches: 1.59.2; Fix INIT_DEBUGOUT() messages.
1.58	28-Dec-2016	msaitoh	Fix a bug that hw.ixgN.wol_enable=0 doesn't work correctly. Clear the WUC and WUFC register if wol_enable is 0. ixgbe_setup_low_power_mode() calls ixgbe_stop() and it resets the chip. The reset change WUC and WUFC registers to the default, so clearing WUC and WUFC should be done after ixgbe_stop().
1.57	28-Dec-2016	msaitoh	Allow 0 as the defalut of hw.ixgN.advertise_speed sysctl.
1.56	27-Dec-2016	msaitoh	Fix flow control setting sysctl. I don't know why disabling autonego when the flow control parameter is changed. Now the ixgN.fc sysctl works as expected.
1.55	16-Dec-2016	msaitoh	Add missing IFM_FDX.
1.54	16-Dec-2016	msaitoh	Remove extra newline in debug printfs. ixgbe's DEBUGOUT macros automatically add newline.
1.53	05-Dec-2016	msaitoh	- Fix wol_enable, fc and wufc sysctls to work correctly. - Fix RCS Id.
1.52	05-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.51	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r294578: - Fixup SFP module insertion on the 82599 when insertion happens after the system is booted and running. Add PHY detection logic to ixgbe_handle_mod() and add locking to ixgbe_handle_msf() as well. FreeBSD r293334. - Fix ix advertise value after media change. When ifconfig sets media then the values displayed by the advertise_speed value are invalidated. Fix this by setting the bits correctly including setting advertise to 0 for media = auto. FreeBSD r294578. - Some others (e.g. LRO(not used by NetBSD)).
1.50	02-Dec-2016	msaitoh	Fix ix{gbe,v}_set_sysctlvalue().
1.49	02-Dec-2016	msaitoh	- Add missing pmf_device_deregister() in ixgbe_detach() - Remove extra newline from an sysctl description. - Fix dmesg output when the number of PCIe lane can't get.
1.48	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.47	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.46	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.45	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.44	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.43	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.42	28-Nov-2016	msaitoh	FreeBSD r280182 made new file ix_txrx.c and moved ixgbe.c and ixv's common code into it. Before sync with whole of them, just move ixgbe.c and ixv.c's common code into ix_txrx.c from ixgbe.c. In this commit, only ixgbe.c is split into the device dependent part and the common part. ixv.c isn't change to make this commit no functional change. To use ixv.c with ix_txrx.c, it's required to modify the common part's API and functions themselves. This commit is done to make the next change easy to understand.
1.41	25-Nov-2016	msaitoh	Remove #ifdef NETBSD_MSI_OR_MSIX to be simple. ixv(4) isn't attached if __HAVE_PCI_MSI_MSIX isn't set because MSI-X must be used.
1.40	13-Oct-2016	jdolecek	provide intr xname
1.39	11-Jul-2016	knakahara	branches: 1.39.2; pci_intr_type() is required pci_chipset_tag_t argument by other than x86. pointed out by nonaka@n.o.
1.38	10-Jun-2016	ozaki-r	Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.37	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.36	17-Aug-2015	knakahara	Add kernel code to support intrctl(8).
1.35	13-Aug-2015	msaitoh	Reduce diff against FreeBSD r280181.
1.34	13-Aug-2015	msaitoh	- Add MSI/MSI-X support. The multiqueue function is not supported yet. - Make ixv.c compilable. _NOT_TESTED_YET_
1.33	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.32	05-Aug-2015	msaitoh	Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic.
1.31	03-Aug-2015	msaitoh	Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach().
1.30	09-May-2015	christos	remove unreachable code.
1.29	06-May-2015	msaitoh	Fixes link error without vlan. PR#49879.
1.28	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.27	14-Apr-2015	msaitoh	Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work.
1.26	14-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r243716: - A lot of bugfixes. Some of them are realted to multi queue and those have not affected in NetBSD because we have not used it yet. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Improve performance.
1.25	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.24	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.23	25-Mar-2015	msaitoh	Fix a bug that vlan setting may not apply to the parent interface correctly.
1.22	19-Mar-2015	msaitoh	Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes.
1.21	24-Feb-2015	msaitoh	Add 82599EN_SFP and 82599_SFP_SF_QP. Not tested.
1.20	24-Feb-2015	msaitoh	Change macro name IXGBE_DEV_ID_82599_SFP_DELL to IXGBE_DEV_ID_82599_SFP_SF2. Same as OpenBSD and Linux. No binary change.
1.19	07-Feb-2015	christos	disable preemption while scheduling the softint to configure the link.
1.18	04-Feb-2015	msaitoh	Fix mutex related problem reported by Uwe Toenjes in PR#49328: - Revert ixgbe_netbsd.c rev. 1.2 - make CORE_LOCK adaptive - Release spin lock while reinitializing the jumb buffer structure.
1.17	04-Feb-2015	msaitoh	Remove extra IXGBE_TX_LOCK()/IXGBE_TX_UNLOCK() from ixgbe_free_transmit_structures(). This function is called in the end of the detach function. if_stop was called and almost all resources were freed, so it's not required to block the TX stuff.
1.16	04-Feb-2015	msaitoh	Add missing IXGBE_RX_LOCK_DESTROY() for the detach path.
1.15	13-Jan-2015	msaitoh	Fix a legacy interrupt problem. If the INTx line was shared with another device, the ixgbe_legacy_irq() enabled the interrupt even if the interface was down. Check the interface state and call functions appropriately.
1.14	21-Apr-2014	chs	branches: 1.14.2; 1.14.4; only attach/detach stats->mngptc once. remove a low-value printf().
1.13	17-Apr-2014	christos	CID/1008347, CID/1008348, CID/1008349: overflow before widen CID/1203198, CID/1203199: Uninitialized scalar variable
1.12	15-Apr-2014	hannken	Add __diagused.
1.11	11-Apr-2014	joerg	Mark ixgbe_disable_queue as unused.
1.10	08-Apr-2014	christos	Add dell's flavor (from uwe, in current-users@)
1.9	30-Mar-2014	skrll	branches: 1.9.2; Make this compile - Fix pci_intr_string usage - Remove ununsed - define variables with their usage kern/48690: ixg* driver doesn't compile Untested.
1.8	29-Mar-2014	christos	make pci_intr_string and eisa_intr_string take a buffer and a length instead of relying in local static storage.
1.7	25-Feb-2014	pooka	Ensure that the top level sysctl nodes (kern, vfs, net, ...) exist before the sysctl link sets are processed, and remove redundancy. Shaves >13kB off of an amd64 GENERIC, not to mention >1k duplicate lines of code.
1.6	09-Feb-2014	njoly	Remove a few noisy debug printf.
1.5	18-Jun-2012	dsl	branches: 1.5.2; 1.5.4; Missed one of the (void *) casts on the args to sysctl_createv().
1.4	16-Jun-2012	dsl	Add explicit (void *) casts for when sysctl_createv() is passed CTLTYPE_INT but the address of a structure (and a function).
1.3	02-Jun-2012	dsl	Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.2	19-Nov-2011	tls	branches: 1.2.4; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.1	12-Aug-2011	dyoung	branches: 1.1.2; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.2.2	30-Oct-2012	yamt	sync with head
1.1.2.1	17-Apr-2012	yamt	sync with head
1.2.4.1	25-Mar-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #1439): sys/dev/pci/ixgbe/ixgbe.c: revision 1.60 via patch Use 64bit DMA tag. If not, a lot of bounce buffer is allocated. Fixes PR#49968 reported by Hauke.
1.5.4.1	18-May-2014	rmind	sync with head
1.5.2.2	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.9.2.1	10-Aug-2014	tls	Rebase.
1.14.4.9	28-Aug-2017	skrll	Sync with HEAD
1.14.4.8	05-Feb-2017	skrll	Sync with HEAD
1.14.4.7	05-Dec-2016	skrll	Sync with HEAD
1.14.4.6	05-Oct-2016	skrll	Sync with HEAD
1.14.4.5	09-Jul-2016	skrll	Sync with HEAD
1.14.4.4	19-Mar-2016	skrll	Sync with HEAD
1.14.4.3	22-Sep-2015	skrll	Sync with HEAD
1.14.4.2	06-Jun-2015	skrll	Sync with HEAD
1.14.4.1	06-Apr-2015	skrll	Sync with HEAD
1.14.2.7	09-Mar-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #1370): sys/dev/pci/ixgbe/ixgbe.c: revision 1.60 Use 64bit DMA tag. If not, a lot of bounce buffer is allocated. Fixes PR#49968 reported by Hauke.
1.14.2.6	14-Jun-2016	snj	branches: 1.14.2.6.2; Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.14.2.5	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.14.2.4	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.14.2.3	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.14.2.2	24-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #545): sys/dev/pci/ixgbe/ixgbe.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.19 Add missing IXGBE_RX_LOCK_DESTROY() for the detach path. Remove extra IXGBE_TX_LOCK()/IXGBE_TX_UNLOCK() from ixgbe_free_transmit_structures(). This function is called in the end of the detach function. if_stop was called and almost all resources were freed, so it's not required to block the TX stuff. Fix mutex related problem reported by Uwe Toenjes in PR#49328: - Revert ixgbe_netbsd.c rev. 1.2 - make CORE_LOCK adaptive - Release spin lock while reinitializing the jumb buffer structure. disable preemption while scheduling the softint to configure the link.
1.14.2.1	17-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #527): sys/dev/pci/ixgbe/ixgbe.c: revision 1.15 Fix a legacy interrupt problem. If the INTx line was shared with another device, the ixgbe_legacy_irq() enabled the interrupt even if the interface was down. Check the interface state and call functions appropriately.
1.14.2.6.2.1	13-Mar-2017	skrll	Sync with netbsd-7-1-RELEASE
1.39.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.39.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.39.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.59.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.80.4.2	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.80.4.1	02-May-2017	pgoyette	Sync with HEAD - tag prg-localcount2-base1
1.88.2.59	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1933: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.88.2.58	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.88.2.57	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.88.2.56	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.88.2.55	21-Jun-2023	martin	Pullup the following revisions, requested by msaitoh in ticket #1828: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.88.2.54	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.88.2.53	06-Jun-2022	martin	Apply patch, requested by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ixgbe.c (apply patch) Fix the location of "break" statement. No functional change intended.
1.88.2.52	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.88.2.51	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.98 sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others. Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.88.2.50	30-May-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1744: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.88.2.49	01-Feb-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1734: sys/dev/pci/ixgbe/ixgbe.c 1.264,1.269,1.272, 1.306 via patch Four INTx related fixes: - Fix a bug that the all interrupt sources are enabled when the interface is UP and the INTx line is shared with other devices. - Fix a bug that it might incorrectly enable interrupt when IFF_RUNNING is not set. - Don't process TX/RX if a queue interrupt isn't occurred. - Increment legacy interrupt counter after checking INTx sharing.
1.88.2.48	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.88.2.47	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.88.2.46	29-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1724: sys/dev/pci/ixgbe/ixgbe.c 1.298, 1.303 via patch Add some missing error counters to ierror.
1.88.2.45	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1708: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.88.2.44	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.88.2.43	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.88.2.42	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.88.2.41	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1600): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.88.2.40	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.88.2.39	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.88.2.38	26-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1478): sys/dev/pci/ixgbe/ixgbe.c: revision 1.218 Add missing core lock in ixgbe_handle_mod().
1.88.2.37	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1451): sys/dev/pci/ixgbe/ixgbe.c: revision 1.216 sys/dev/pci/ixgbe/ixv.c: revision 1.141 Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.88.2.36	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1408): sys/dev/pci/ixgbe/ixgbe.c: revision 1.214 Apply FreeBSD r353599: > ixgbe: Disable EEE for backplane X550EM_X > > From Zach: > Intel documentation indicates that backplane X550EM_X KR devices do not > support Energy Efficient Ethernet. Prior to this patch, X552 devices > (device ID 0x15AB) will crash the system when transitioning EEE state > via sysctl. > > Signed-off-by: Zach Vargas <zvargas@xes-inc.com> > > PR: 240320 > Submitted by: Zach Vargas <zvargas@xes-inc.com> > Reviewed by: erj@ > MFC after: 3 days > Differential Revision: https://reviews.freebsd.org/D21673
1.88.2.35	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.88.2.34	06-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367: fix mistake in previous.
1.88.2.33	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.88.2.32	29-Aug-2019	martin	Pull up following revision(s) (requested by knakahara in ticket #1347): sys/dev/pci/ixgbe/ixgbe.c: revision 1.205 sys/dev/pci/ixgbe/ixgbe.c: revision 1.206 fix typo in error handling of ixgbe_allocate_legacy(). ok by msaitoh@n.o pullup-8, pullup-9 - Fix panic when ncpu == 1 && sysctl -w hw.ixg0.txrx_workqueue=1. Reported by nonaka@n.o. ok by msaitoh@n.o and nonaka@n.o pullup-8, pullup-9
1.88.2.31	01-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1313: sys/dev/pci/ixgbe/ixgbe_type.h 1.41 sys/dev/pci/ixgbe/ixgbe.c 1.194,1.197-1.199 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.24 sys/dev/pci/ixgbe/ixv.c 1.121,1.124-1.125 - Avoid undefined behavior of interrupt vector setting. - Avoid undefined behavior of TX/RX queue statistics calculation. - Avoid undefined behavior of X550EM's PHY accesses. - Avoid undefined behavior of Flow Control Transmit Timer setting. - Don't call {ixgbe,ixv}_stop() twice while detaching.
1.88.2.30	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.88.2.29	01-Apr-2019	martin	Pull up the following, requested by msaitohin ticket #1225: sys/dev/pci/ixgbe/ixgbe.c 1.175-1.178 sys/dev/pci/ixgbe/ixv.c 1.110-1.111 via patch - NetBSD currently uses traffic class 0 only. Other traffic classes aren't used yet. When IXGBE_TC_COUNTER_NUM is set to lower than IXGBE_DCB_MAX_TRAFFIC_CLASS (e.g. 1), other traffic classes' counters are not used. It means we don't generate evcnt for them and don't add the values in ixgbe_update_stats_counters(). - It's not required to calculate unused queues' statistics. - Fix a bug that the VLAN HW tagging function is not correctly disabled when all vlan is detached. - Fix a bug that VLAN HW tagging function is not correctly controlled on 82598. - Control VLAN HW filter function correctly. Note that currently VLAN HW filter function doesn't work because NetBSD doesn't support it yet. - Don't clear IXGBE_VLNCTRL_CFIEN bit When ETHERCAP_VLAN_HWFILTER is set. I think it's not required (and Linux doesn't do it). This change has no effect to NetBSD because ETHERCAP_VLAN_HWFILTER is not supported yet.
1.88.2.28	01-Mar-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1199): sys/dev/pci/ixgbe/ixv.c: revision 1.109 sys/dev/pci/ixgbe/ixgbe.h: revision 1.54 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.174 Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.88.2.27	28-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1171): sys/dev/pci/ixgbe/ixgbe.c: revision 1.168 sys/dev/pci/ixgbe/ixgbe.h: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.172 sys/dev/pci/ixgbe/ixgbe.c: revision 1.173 Some counters are not per queue but per traffic class. Make new evcnt group "ixgM tcN" (N = 0..7) and move those counters into it. We are using only traffic class 0, so we will see only tc0's counter are updated. - Call ixgbe_init_swfw_semaphore(hw) before calling ixgbe_check_wol_support() because ixgbe_check_wol_support()->ixgbe_check_wol_support()-> ixgbe_get_device_caps() accesses semaphore. - - Move the location of calling setup_eee() in ixgbe_attach(). The previous location was too early. The new location is before calling ixgbe_set_phy_power(). - Restore call of setup_eee() in ixgbe_init_locked(). It was removed in rev. 1.99 (FreeBSD r320688). - When calling setup_eee(), pass enable/disable correctly.
1.88.2.26	27-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1172): sys/dev/pci/nvme_pci.c: revision 1.26 sys/dev/pci/xhci_pci.c: revision 1.21 sys/dev/pci/ixgbe/ixv.c: revision 1.108 sys/dev/pci/ixgbe/ixgbe.c: revision 1.171 sys/dev/pci/if_fxp_pci.c: revision 1.84 sys/dev/pci/if_fxp_pci.c: revision 1.85 sys/dev/pci/xhci_pci.c: revision 1.16 remove #ifdef DEBUG printf, it seems to have outlived it's usefulness - KNF. No functional change. - Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them explicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.88.2.25	08-Nov-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1084): sys/dev/pci/ixgbe/ixv.c: revision 1.107 sys/dev/pci/ixgbe/ixgbe.c: revision 1.167 Don't set LRO capability flag because we don't support it yet. This doesn't change any behavior other than if_capabilities' visibility (e.g. ifconfig).
1.88.2.24	23-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1029): sys/dev/pci/ixgbe/ixgbe.c: revision 1.166 - Fix a bug that Denverton accessed wrong register in ixgbe_setup_low_power_mode(). The location of some registers in Denverton were changed. The GRC register is one of them, so use IXGBE_GRC_BY_MAC(hw) instead of IXGBE_GRC. - ixgbe_check_wol_support() tests the capability of WOL, but the sysctl knobs for WOL were added only for IXGBE_DEV_ID_X550EM_X_10G_T. Add the knobs when adapter->wol_support is not 0. XXX pullup-8
1.88.2.23	05-Sep-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1006): sys/dev/pci/ixgbe/ixgbe.c: revision 1.164 Fix a bug that media change may fail. I noticed that ifconfig ixgN media XXX took a 10 or more seconds on a Denverton machie. I occurred by trying take a lock and timed out. The reason was that ixgbe_media_change() didn't take CORE_LOCK. Do it. This problem was from FreeBSD's pre-iflib ixgbe (I don't know whether this problem really occurs on FreeBSD or not). Post-iflib ixgbe has no problem because iflib_media_change() takes a lock. XXX pullup-8
1.88.2.22	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.88.2.21	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #935): sys/dev/pci/ixgbe/ixgbe.c: revision 1.162 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.35 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.16 Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*). -- Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.88.2.20	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.88.2.19	07-Jun-2018	martin	Pull up following revision(s) (requested by ozaki-r in ticket #843): sys/dev/pci/ixgbe/ixv.c: revision 1.101 sys/net/if_bridge.c: revision 1.156 sys/net/if_pppoe.c: revision 1.138 sys/dev/pci/if_wm.c: revision 1.580 sys/dev/pci/ixgbe/ixgbe.c: revision 1.156 sys/net/if_gif.c: revision 1.142 Ensure to call if_register after interface initializations finish
1.88.2.18	17-Apr-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #769): sys/dev/pci/ixgbe/ixgbe.c: revision 1.147 sys/dev/pci/ixgbe/ixgbe.h: revision 1.42 Fix panic when "sysctl -w hw.ixg0.txrx_workqueue=[01]" while there is traffic. The operation is not supported, however causing panic is problem.
1.88.2.17	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.88.2.16	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.88.2.15	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.88.2.14	13-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/dev/pci/ixgbe/ixgbe.c: revision 1.134 sys/dev/pci/ixgbe/ixgbe.h: revision 1.35 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.35 sys/dev/pci/ixgbe/ixv.c: revision 1.87 Make some event counters MP safe. Now all of the event counters are MP safe.
1.88.2.13	08-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #612): sys/dev/pci/ixgbe/ixgbe.c: revision 1.129-1.133 sys/dev/pci/ixgbe/ixgbe.h: revision 1.34 sys/dev/pci/ixgbe/ixv.c: revision 1.85,1.86 - Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code. Don't increment que->req.ev_count in MSI-X interrupt because it's not reschedule. Fix another poll mode assumption breaking. Implemented by msaitoh@n.o, I just commit by proxy. ixgbe_rearm_queues() writes EICS register(s). 82599, X540 and X550 specifications say "Following a write of 1b to any bit in the EICS register (interrupt cause set), its corresponding bit in the EIMS register is auto set as well enabling its interrupt." in "Extended Interrupt Auto Mask Enable (EIAM) Register" section. That is, ixgbe_rearm_queues() causes interrupts regardless of the status managed by ixgbe_enable_queue()/ixgbe_disable_queue(). That can break poll mode assumption. In fact, the problem occurs in the following situation - CPU#A has high load traffic, in contrast, CPU#B has not so high load traffic - CPU#A is occurred interrupt by its NIC queue - CPU#A calls ixgbe_disable_queue() in interrupt handler(ixgbe_msix_que()) - CPU#A kick softint handler(ixgbe_handle_que()) - CPU#A begins softint - CPU#A's NIC queue is set que->txr->busy flag - With some reason, CPU#A can do ixg interrupt handler E.g. when one of CPU#A's softnet handlers sleeps, ipl is lowered - CPU#B starts callout - CPU#B calls ixgbe_local_timer1() - CPU#B writes EICS bit corresponding CPU#A's NIC queue bit - CPU#A's NIC queue causes interrupt whie CPU#A is running in poll mode - CPU#A calls ixgbe_disable_queue() in interrupt handler again - CPU#A has done polling, and then CPU#A calls ixgbe_enable_queue() once - CPU#A's NIC queue interrupt is disabled until ixg is detached as ixgbe_disable_queue() is called twice though ixgbe_disable_queue() is called once only NOTE: 82598 does not say so, but it is treated in the same way because of no harm. By the way, we will refactor ixgbe_local_timer(watchdog processing) later. Fix INTx/MSI handler did not schedule workqueue. Pointed out by msaitoh@n.o. Reduce duplicated code which schedule deferred packet processing. No functional change.
1.88.2.12	06-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #605): sys/dev/pci/ixgbe/ixgbe.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe.h: revision 1.33 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.34 sys/dev/pci/ixgbe/ixgbe.c: revision 1.128 sys/dev/pci/ixgbe/ixv.c: revision 1.83 sys/dev/pci/ixgbe/ixv.c: revision 1.84 Add hw.ixvM.q[01].{interrupt_rate,[tr]xd_head,[tr]xd_tail} sysctls as ixg(4). ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o. ixv(4) also supports workqueue poll mode, but not enabled by default yet, either. ok by msaitoh@n.o. Move the location of "struct work" as FreeBSD's "struct task" in ixgbe.h. No functional change.
1.88.2.11	01-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #597): sys/dev/pci/ixgbe/ixgbe.h: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.c: revision 1.127 sys/dev/pci/ixgbe/ixv.c: revision 1.82 Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o. Apply ixgbe.c:r1.127 to ixv.c. Pointed out by msaitoh@n.o.
1.88.2.10	26-Feb-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #592): sys/dev/pci/ixgbe/ixv.c: revision 1.79 sys/dev/pci/ixgbe/ixgbe.h: revision 1.30 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.31 sys/dev/pci/ixgbe/ixgbe.c: revision 1.120 sys/dev/pci/ixgbe/ixgbe.c: revision 1.121 sys/dev/pci/ixgbe/ixgbe.c: revision 1.123 sys/dev/pci/ixgbe/ixgbe.c: revision 1.124 sys/dev/pci/ixgbe/ixgbe.c: revision 1.125 sys/dev/pci/ixgbe/ixgbe.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.80 sys/dev/pci/ixgbe/ixv.c: revision 1.81 CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558). Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8 Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv) - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8 Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8 Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o. - Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function. Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.88.2.9	26-Feb-2018	snj	Pull up following revision(s) (requested by knakahara in ticket #567): distrib/sets/lists/comp/mi: 1.2182-1.2183 sys/dev/pci/if_wm.c: 1.564 sys/dev/pci/ixgbe/ixgbe.c: 1.122 sys/dev/pci/ixgbe/ixgbe_rss.h: 1.3 sys/dev/pci/ixgbe/ixv.c: 1.78 sys/net/Makefile: 1.35-1.36 sys/net/files.net: 1.15 sys/net/rss_config.c: 1.1 sys/net/rss_config.h: 1.1 Introduce very simple Receive Side Scaling (RSS) utility. ok by msaitoh@n.o. -- Apply RSS utility to wm(4). ok by msaitoh@n.o. -- Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o. -- Fix build failure, sorry. -- Currently, it is not necessary to install rss_config.h. Pointed out by msaitoh@n.o.
1.88.2.8	13-Jan-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #490): sys/dev/pci/ixgbe/ixgbe.c: revision 1.116-1.119 sys/dev/pci/ixgbe/ixv.c: revision 1.76-1.77 Don't panic when resource shortage occured. Fixes PR#52820 reported by kardel@: - Don't use if_free() because ixgbe(4) don't use if_alloc(). - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more). -- Don't panic when resource shortage occured. Like ixgbe.c rev. 1.116. - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more). -- Fix panic when only link interrupt can't be established. -- - If MSI can't be used on some environment, fallback to INTx correctly. - Use single vector MSI when number of CPU is 1 to save interrupt slot. -- Fallback from MSI-X to MSI or INTx if MSI-X setup failed.
1.88.2.7	02-Jan-2018	snj	Pull up following revision(s) (requested by ozaki-r in ticket #456): sys/arch/arm/sunxi/sunxi_emac.c: 1.9 sys/dev/ic/dwc_gmac.c: 1.43-1.44 sys/dev/pci/if_iwm.c: 1.75 sys/dev/pci/if_wm.c: 1.543 sys/dev/pci/ixgbe/ixgbe.c: 1.112 sys/dev/pci/ixgbe/ixv.c: 1.74 sys/kern/sys_socket.c: 1.75 sys/net/agr/if_agr.c: 1.43 sys/net/bpf.c: 1.219 sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416 sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257 sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146 sys/net/if_etherip.c: 1.40 sys/net/if_ethersubr.c: 1.243, 1.246 sys/net/if_faith.c: 1.57 sys/net/if_gif.c: 1.132 sys/net/if_l2tp.c: 1.15, 1.17 sys/net/if_loop.c: 1.98-1.101 sys/net/if_media.c: 1.35 sys/net/if_pppoe.c: 1.131-1.132 sys/net/if_spppsubr.c: 1.176-1.177 sys/net/if_tun.c: 1.142 sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121 sys/net/npf/npf_ifaddr.c: 1.3 sys/net/npf/npf_os.c: 1.8-1.9 sys/net/rtsock.c: 1.230 sys/netcan/if_canloop.c: 1.3-1.5 sys/netinet/if_arp.c: 1.255 sys/netinet/igmp.c: 1.65 sys/netinet/in.c: 1.210-1.211 sys/netinet/in_pcb.c: 1.180 sys/netinet/ip_carp.c: 1.92, 1.94 sys/netinet/ip_flow.c: 1.81 sys/netinet/ip_input.c: 1.362 sys/netinet/ip_mroute.c: 1.147 sys/netinet/ip_output.c: 1.283, 1.285, 1.287 sys/netinet6/frag6.c: 1.61 sys/netinet6/in6.c: 1.251, 1.255 sys/netinet6/in6_pcb.c: 1.162 sys/netinet6/ip6_flow.c: 1.35 sys/netinet6/ip6_input.c: 1.183 sys/netinet6/ip6_output.c: 1.196 sys/netinet6/mld6.c: 1.90 sys/netinet6/nd6.c: 1.239-1.240 sys/netinet6/nd6_nbr.c: 1.139 sys/netinet6/nd6_rtr.c: 1.136 sys/netipsec/ipsec_output.c: 1.65 sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10 kmem_intr_free kmem_intr_[z]alloced memory the underlying pools are the same but api-wise those should match Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@ Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..." scattered all over the source code and makes it easy to identify remaining KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE. No functional change Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE If IFEF_MPSAFE is set, hold the lock and otherwise don't hold. This change requires additions of KERNEL_LOCK to subsequence functions from if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe components. Proposed on tech-kern@ and tech-net@ Ensure to hold if_ioctl_lock when calling if_flags_set Fix locking against myself on ifpromisc vlan_unconfig_locked could be called with holding if_ioctl_lock. Ensure to not turn on IFF_RUNNING of an interface until its initialization completes And ensure to turn off it before destruction as per IFF_RUNNING's description "resource allocated". (The description is a bit doubtful though, I believe the change is still proper.) Ensure to hold if_ioctl_lock on if_up and if_down One exception for if_down is if_detach; in the case the lock isn't needed because it's guaranteed that no other one can access ifp at that point. Make if_link_queue MP-safe if IFEF_MPSAFE if_link_queue is a queue to store events of link state changes, which is used to pass events from (typically) an interrupt handler to if_link_state_change softint. The queue was protected by KERNEL_LOCK so far, but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it by a spin mutex. Additionally with this change KERNEL_LOCK of if_link_state_change softint is omitted if NET_MPSAFE is enabled. Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of if_timer (see the comment). Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH At that point no other one modifies the list so IFADDR_READER_FOREACH is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though, if we try to detect contract violations of pserialize, using it violates the contract. So avoid using it makes life easy. Ensure to call if_addr_init with holding if_ioctl_lock Get rid of outdated comments Fix build of kernels without ether By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that created a unnecessary dependency from if.c to if_ethersubr.c. PR kern/52790 Rename IFNET_LOCK to IFNET_GLOBAL_LOCK IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then). Wrap if_ioctl_lock with IFNET_* macros (NFC) Also if_ioctl_lock perhaps needs to be renamed to something because it's now not just for ioctl... Reorder some destruction routines in if_detach - Destroy if_ioctl_lock at the end of the if_detach because it's used in various destruction routines - Move psref_target_destroy after pr_purgeif because we want to use psref in pr_purgeif (otherwise destruction procedures can be tricky) Ensure to call if_mcast_op with holding IFNET_LOCK Note that CARP doesn't deal with IFNET_LOCK yet. Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held Describe which lock is used to protect each member variable of struct ifnet Requested by skrll@ Write a guideline for converting an interface to IFEF_MPSAFE Requested by skrll@ Note that IFNET_LOCK must not be held in softint Don't set IFEF_MPSAFE unless NET_MPSAFE at this point Because recent investigations show that interfaces with IFEF_MPSAFE need to follow additional restrictions to work with the flag safely. We should enable it on an interface by default only if the interface surely satisfies the restrictions, which are described in if.h. Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because the network stack is still serialized by the big kernel locks by default.
1.88.2.6	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.88.2.5	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.88.2.4	21-Nov-2017	martin	Revert the pullup of the following revision(s) (requested by msaitoh in ticket #361): sys/dev/pci/ixgbe/ixgbe.c: revision 1.97 Revert a part of ticket #248 because it panics when attaching. It is hard to fix with a small change.
1.88.2.3	31-Aug-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #248): sys/dev/pci/ixgbe/ixgbe.h: revision 1.25 sys/dev/pci/ixgbe/ixgbe.c: revision 1.96 sys/dev/pci/ixgbe/ixgbe.c: revision 1.97 Remove unused counters. Don't limit number of queue pair to 8. Take max TX queues and max RX queues into account. 82598 has 32 TX queue and 64 RX queues. Others has 128 TX queues and 128 RX queues. Number of MSI-X table entries is 64, so the max queue pairs is 63 (64 minus the Others (e.g. link status change)).
1.88.2.2	05-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #175): sys/dev/pci/ixgbe/ix_txrx.c: 1.25-1.27 sys/dev/pci/ixgbe/ixgbe.c: 1.91-1.95 sys/dev/pci/ixgbe/ixgbe_type.h: 1.23-1.25 sys/dev/pci/ixgbe/ixgbe_api.c: 1.16 sync ixg(4) up to ixgbe.c rev. 1.95: - Increase total number of RX buffers on multiqueue. - Fix the partial chain check in ixgbe_rx_discard(). This bug was added in ixgbe.c rev. 1.33. - Fix a double free in ixgbe_rxeof(). - Print verbose output in ixgbe_update_link_status() correctly on 5Gbps and 2.5Gbps (NBASE-T) for X550T[12] and newer. - Change hw.ixgN.ts to hw.ixgN.thermal_test. Same as FreeBSD. - Fix a bug that ifconfig ixgN media 1000baseT and 10Gbase-T advertised additional unwanted speeds. - Print PHY ID. - Remove unused variable.
1.88.2.1	04-Jul-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #81): sys/dev/pci/ixgbe/ixgbe.c: revision 1.89 sys/dev/pci/if_wm.c: revision 1.509 sys/dev/pci/ixgbe/ixgbe.c: revision 1.90 Use IFM_1000_KX and IFM_2500_KX. Use IFM_1000_KX more.
1.128.2.13	26-Jan-2019	pgoyette	Sync with HEAD
1.128.2.12	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.128.2.11	30-Sep-2018	pgoyette	Ssync with HEAD
1.128.2.10	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.128.2.9	28-Jul-2018	pgoyette	Sync with HEAD
1.128.2.8	25-Jun-2018	pgoyette	Sync with HEAD
1.128.2.7	21-May-2018	pgoyette	Sync with HEAD
1.128.2.6	22-Apr-2018	pgoyette	Sync with HEAD
1.128.2.5	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.128.2.4	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.128.2.3	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.128.2.2	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.128.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.161.2.3	21-Apr-2020	martin	Sync with HEAD
1.161.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.161.2.1	10-Jun-2019	christos	Sync with HEAD
1.199.2.30	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1792: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.199.2.29	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.199.2.28	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.199.2.27	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.199.2.26	21-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1647: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.199.2.25	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.199.2.24	03-Jun-2022	snj	Pull up following revision(s) (requested by msaitoh in ticket #1460): sys/dev/pci/ixgbe/ixgbe.c: patch Fix previous to re-enable legacy interrupt correctly. It's a part of ixgbe.c rev. 1.273.
1.199.2.23	02-Jun-2022	martin	Pull up the following revisions, all via patch, requested by msaitoh in ticket #1459: sys/dev/pci/ixgbe/ixgbe.c 1.261,1.265-1.268,1.273,1.275-1.277, 1.312,1.316-1.319 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.182 - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Fix typos in comment. - KNF.
1.199.2.22	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1458): sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.199.2.21	30-May-2022	martin	Pull up the following revisions, requested by msaitoh: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.158,1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.199.2.20	02-Feb-2022	martin	Pull up the following revisions (requested by msaitoh in ticket #1424): sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.305 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixv.c 1.178 via patch Use atomic_{load,store}_relaxed() for event counters.
1.199.2.19	01-Feb-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1420: sys/dev/pci/ixgbe/ixgbe.c 1.264,1.269,1.272, 1.306 via patch Four INTx related fixes: - Fix a bug that the all interrupt sources are enabled when the interface is UP and the INTx line is shared with other devices. - Fix a bug that it might incorrectly enable interrupt when IFF_RUNNING is not set. - Don't process TX/RX if a queue interrupt isn't occurred. - Increment legacy interrupt counter after checking INTx sharing.
1.199.2.18	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.199.2.17	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.199.2.16	29-Jan-2022	martin	Pullup the following revisions, requested by msaitoh: sys/dev/pci/ixgbe/ixgbe.c 1.298, 1.303 via patch Add some missing error counters to ierror.
1.199.2.15	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1374: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.199.2.14	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.199.2.13	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.199.2.12	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1070): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.199.2.11	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.199.2.10	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.199.2.9	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.199.2.8	24-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #580): sys/dev/pci/ixgbe/ixgbe.c: revision 1.217 sys/dev/pci/ixgbe/ixgbe.c: revision 1.218 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.6 sys/dev/pci/ixgbe/ixv.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.25 Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@. Add missing core lock in ixgbe_handle_mod().
1.199.2.7	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #453): sys/dev/pci/ixgbe/ixgbe.c: revision 1.216 sys/dev/pci/ixgbe/ixv.c: revision 1.141 Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.199.2.6	17-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #348): sys/dev/pci/ixgbe/ixgbe.c: revision 1.214 Apply FreeBSD r353599: > ixgbe: Disable EEE for backplane X550EM_X > > From Zach: > Intel documentation indicates that backplane X550EM_X KR devices do not > support Energy Efficient Ethernet. Prior to this patch, X552 devices > (device ID 0x15AB) will crash the system when transitioning EEE state > via sysctl. > > Signed-off-by: Zach Vargas <zvargas@xes-inc.com> > > PR: 240320 > Submitted by: Zach Vargas <zvargas@xes-inc.com> > Reviewed by: erj@ > MFC after: 3 days > Differential Revision: https://reviews.freebsd.org/D21673
1.199.2.5	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.199.2.4	05-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #180): sys/dev/pci/ixgbe/ixv.c: revision 1.130 sys/dev/pci/ixgbe/ixgbe.c: revision 1.207 sys/dev/pci/ixgbe/ixgbe.c: revision 1.208 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.55 sys/dev/pci/ixgbe/ixv.c: revision 1.129 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.10 printf -> device_printf Set IFM_1000_BX10 correctly. Use aprint_*() in the attach function.
1.199.2.3	01-Sep-2019	martin	Pull up following revision(s) (requested by knakahara in ticket #136): sys/dev/pci/ixgbe/ixgbe.c: revision 1.205 sys/dev/pci/ixgbe/ixgbe.c: revision 1.206 fix typo in error handling of ixgbe_allocate_legacy(). ok by msaitoh@n.o pullup-8, pullup-9 - Fix panic when ncpu == 1 && sysctl -w hw.ixg0.txrx_workqueue=1. Reported by nonaka@n.o. ok by msaitoh@n.o and nonaka@n.o pullup-8, pullup-9
1.199.2.2	01-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #134): sys/dev/pci/ixgbe/ixgbe.c: revision 1.202 sys/dev/pci/ixgbe/ixgbe.c: revision 1.203 sys/dev/pci/ixgbe/ixgbe.c: revision 1.204 sys/dev/pci/ixgbe/ixv.c: revision 1.128 Simplify ix{gbe,v}_[un]register_vlan() API suggested by knakahara. The API was the same as FreeBSD's pre-iflib's. They use iflib now and it's not required for us to keep the old API. X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, link becomes up and the real media type is unknown. e.g: % ifconfig -m ixg0 (snip) media: Ethernet autoselect (autoselect rxpause,txpause) status: active supported Ethernet media: media none media autoselect (snip) To resolve this problem, when an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC interrupt.
1.199.2.1	01-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #133): sys/dev/pci/ixgbe/ixgbe.c: revision 1.200 sys/dev/pci/ixgbe/ixgbe.c: revision 1.201 sys/dev/pci/ixgbe/ixv.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.127 sys/net/if_vlan.c: revision 1.142 sys/net/if_vlan.c: revision 1.143 sys/net/if_vlan.c: revision 1.144 sys/net/if_vlan.c: revision 1.145 sys/net/if_vlan.c: revision 1.146 Check ec_capenable instead of ec_capabilities to control TX side of VLAN HW tagging correctly. XXX pullup-9 Add missing IFNET_LOCK() and IFNET_UNLOCK() in vlan_config(). XXX pullup-9 Fix a bug that VLAN HW "tagging" enable/disable may not reflect correctly. - Always call ec_vlan_cb() if it exists. - Some (or all?) ethernet drivers don't enable HW tagging if no any vlan is attached. ixgbe is one of them. Check the the transition and update VLAN HW tagging function. XXX pullup-9 Use ETHER_LOCK()/ETHER_UNLOCK() suggested by knakahara. - kmem_alloc(,KM_SLEEP) never return NULL, so remove NULL check. - VLAN ID is never duplicated, so break the loop when found. Also move kmen_free() outside of ETHER_LOCK(ec)/ETHER_UNLOCK(ec) to reduce the hold time. suggested by ozaki-r. - Whitespace fix.
1.220.2.2	29-Feb-2020	ad	Sync with head.
1.220.2.1	25-Jan-2020	ad	Sync with head.
1.227.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.258.2.3	03-Apr-2021	thorpej	Sync with HEAD.
1.258.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.258.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.279.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.280.2.3	01-Aug-2021	thorpej	Sync with HEAD.
1.280.2.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.280.2.1	13-May-2021	thorpej	Sync with HEAD.
1.282.2.1	31-May-2021	cjep	sync with head
1.324.2.8	23-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #604): sys/dev/pci/ixgbe/ixgbe.c: revision 1.349 ixgbe: Add QPRDC into iqdrops. A receive packet might drop at two different locations. One is the packet buffer that packets are received into the chip first. If the packet buffer is overflowed, the MPC register is incremented. It's currently added to iqdrops. It's no problem. Another is descriptor ring(s). A packet from the packet buffer is DMA'ed into main memory base on the descriptor ring. If the ring is full, the packet is dropped and the QPRDC register is incremented. It should be added to iqdrops but it was not done. Fix it. Reported by ozaki-r@.
1.324.2.7	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #563): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.110 sys/dev/pci/ixgbe/ixgbe.c: revision 1.345 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.111 sys/dev/pci/ixgbe/ixgbe.c: revision 1.346 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.112 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.62 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.113 sys/dev/pci/ixgbe/ixgbe.c: revision 1.348 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.114 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.115 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.116 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.105 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.106 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.107 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.108 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.109 sys/dev/pci/ixgbe/ixv.c: revision 1.193 sys/dev/pci/ixgbe/ixv.c: revision 1.195 sys/dev/pci/ixgbe/ixv.c: revision 1.196 sys/dev/pci/ixgbe/ixgbe.h: revision 1.94 sys/dev/pci/ixgbe/ixgbe.h: revision 1.95 sys/dev/pci/ixgbe/ixgbe.h: revision 1.96 sys/dev/pci/ixgbe/ixgbe.h: revision 1.97 sys/dev/pci/ixgbe/ixgbe.h: revision 1.98 ixgbe: Fix comment. No functional change. ixgbe: Whitespace. No functional change. ixgbe(4): Move assignment of TXD. NFCI. ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor. ixv(4): Remove unused IFF_OACTIVE. No functional change. ixgbe: Clear the WTHRESH bit field before writing it. ixgbe: Modify for the readability. No functional change. ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof(). ixgbe: Update if_opackets outside the loop in ixgbe_txeof(). ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof(). ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO). ixgbe: Remove unused to reduce the size of struct rx_ring. ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring. ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.324.2.6	03-Nov-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #455): sys/dev/pci/ixgbe/ixgbe.c: revision 1.347 sys/net/if_l2tp.c: revision 1.49 tests/net/if_vlan/t_vlan.sh: revision 1.25 sys/net/if_vlan.c: revision 1.171 sys/net/if_ethersubr.c: revision 1.326 sys/dev/pci/ixgbe/ixv.c: revision 1.194 Use ether_bpf_mtap only when the device supports vlan harware tagging The function is bpf_mtap() for ethernet devices and currently it is just handling VLAN tag stripped by the hardware. l2tp(4): use ether_ifattach() to initialize ethercom Support vlan(4) over l2tp(4) Added the test for vlan over l2tp
1.324.2.5	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #420): sys/dev/pci/ixgbe/ixgbe.c: revision 1.340 sys/dev/pci/ixgbe/ixgbe.c: revision 1.341 sys/dev/pci/ixgbe/ixgbe.c: revision 1.342 sys/dev/pci/ixgbe/ixgbe.c: revision 1.343 sys/dev/pci/ixgbe/ixgbe.c: revision 1.344 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.61 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.104 sys/dev/pci/ixgbe/ixv.c: revision 1.190 sys/dev/pci/ixgbe/ixv.c: revision 1.191 sys/dev/pci/ixgbe/ixv.c: revision 1.192 sys/dev/pci/ixgbe/ixgbe.h: revision 1.93 ixg(4): Print DEVICE_CAPS register. ixgbe: Whitespace. No functional change. ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl. ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. ixgbe: Whitespace. No functional change.
1.324.2.4	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.324.2.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.324.2.2	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.324.2.1	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #202): sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.56 sys/dev/pci/ixgbe/ixgbe.c: revision 1.325 sys/dev/pci/ixgbe/ixgbe.c: revision 1.326 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.44 Use thermal sensor code for IXGBE_DEV_ID_X550EM_A_10G_T, too. PCI device ID 0x15c8 also use X557-AT PHY, so create the thermal sensor sysctl for it, too. Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.352.2.1	02-Aug-2025	perseant	Sync with HEAD
1.100	26-Feb-2025	andvar	Fix typos in comments, mainly s/calcurate/calculate/.
1.99	19-Sep-2024	andvar	s/trafic/traffic/ and s/Crrently/Currently/ in comments.
1.98	30-Dec-2023	msaitoh	branches: 1.98.2; ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.97	30-Dec-2023	msaitoh	ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment.
1.96	29-Dec-2023	msaitoh	ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring.
1.95	29-Dec-2023	msaitoh	ixgbe: Remove unused to reduce the size of struct rx_ring.
1.94	28-Dec-2023	msaitoh	ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment.
1.93	12-Oct-2023	msaitoh	ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate.
1.92	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.91	06-Oct-2023	msaitoh	ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD.
1.90	06-Oct-2023	msaitoh	ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32.
1.89	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.88	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.87	13-Sep-2023	msaitoh	ixgbe: Reorder some event counters for readability.
1.86	30-May-2022	msaitoh	branches: 1.86.4; Remove unused adapter->msix_mem.
1.85	10-Mar-2022	msaitoh	Print ECC, PHY and temp error log using with ratecheck(). - The ratecheck() is for just in case. All of the interrupts might occur only once, but I don't know whether it's true or not. For the fan failure, it seems it occurs only once. - All of the interval is 60s.
1.84	10-Dec-2021	msaitoh	No functional change. - Sync with FreeBSD ix-3.3.18. - Rename ixgbe_get_advertise() to ixgbe_get_default_advertise. - Sort lines, modify comment and whitespace to reduce diff against FreeBSD.
1.83	17-Nov-2021	msaitoh	s/dataseet/datasheet/ in comment.
1.82	19-Sep-2021	andvar	fix various typos in comments, messages and documentation.
1.81	16-Sep-2021	msaitoh	Fix a bug that an mbuf chain which has more than 63488bytes MAY fail on TX. - Currently, The TX buffer dmamap's max number of segments is set to 32. MCLBYTES(== 2048) * 32 = 65536 and it's enough for IP_MAXPACKET(65535). If an mbuf chain has more than 32 mbufs, we call m_defrag() to make it lower than equal to 32, but it might not work. The reason is that our m_defrag() don't modify the first mbuf entry of the chain. e.g.: if an mbuf chain contains 63600bytes data and the first mbuf has 100bytes in the m_data, the new chain has 100+2048+2048+...+12 and the total number of the chain is not 32 but 33. It result in 43 TCP packets will drop. - One of the way to fix this problem is to change m_defrag() which add a new mbuf cluster to the first mbuf. It's need discussion. - Another solution is to change the max number of the TX DMA segment. It should be at least 33 to avoid the m_defrag()'s current limitation. The document (82599-X550 DS 7.2.1.1 "Transmit Storage in system Memory") says that a packet can be fragmented into 40 - WTHRESH - 2 (for 82598, the document say nothing and some people and code says it's unlimited). Currently WTHRESH is set to 8. 40 - 8 - 2 = 30. !?!?!? {Net,Free,Open}BSD and Linux use 32. Is that safe? Anyway, we change WTHRESH from 8 to 5 to fit it. The added comment in ixgbe.h is based on DragonFly's though they don't use WTHRESH. - Yet another solution is to use the Tx Head Pointer Write Back function instead of WTHRESH based write back (see 82599-X550 DS 7.2.3.5.2 "Tx Head Pointer Write Back" or 82598 "Transmit Completions Head Write Back" ). DragonFly, illumos and DPDK use it. - Yet yet another solution is to add tso_maxsize entry to struct ifnet and use it in the TCP stack.
1.80	03-Sep-2021	msaitoh	Save the discard_multidesc state to not to forget the state by exiting rxeof().
1.79	26-Aug-2021	msaitoh	Remove "j" or "jumbo" because we always use MCLBYTES sized buffer.
1.78	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.77	19-Aug-2021	msaitoh	Add forgotten commit for previous. Change IXGBE_RX_COPY_LEN_MAX.
1.76	07-Jul-2021	msaitoh	Add new sysctl "rx_copy_len". ixgbe_rxeof() has an optimization "RX_COPY" to reduce costs of bus_dmamap_load_mbuf() and bus_dmamap_unload() by copying a mbuf cluster's memory to a newly allocated mbuf's MH_databuf[] and recycle the original map. The optimization is used when a length of a packet is smaller than a specific value. The value is calculated based on MHLEN. The size of MHLEN is architecture specific. It's 256 or 512. Make the threshold controllable by adding a new sysctl.
1.75	09-Mar-2021	msaitoh	branches: 1.75.4; Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.74	26-Dec-2020	msaitoh	Disable/enable the OTHER interrupts correctly. The OTHER interrupt was not blocked correctly when MSI-X is used. ixgbe.c rev. 1.260 added new mutex to avoid the race but it didn't disable the interrupt itself. Calling ixgbe_enable_intr() enables all interrupts, so it's not good to call it when some interrupt sources should not be enabled (e.g.: calling ixgbe_enable_intr() in ixgbe_handle_admin() enables queue interrupt). IXGBE_REQUEST_TASK_NEED_ACKINTR doesn't work as expected because ixgbe_handle_admin() can't know which task is enqueued from the interrupt context and can't re-enable a specific EIMS bit. Solve the above three problems by the following two changes: - MSI-X: Disable the OTHER interrupts in the biginning of ixgbe_msix_admin(). - Set mask bits correctly at the end of ixgbe_legacy_irq() and ixgbe_msix_admin() using with eim_orig, eims_enable and eims_disable. - Remove IXGBE_REQUEST_TASK_NEED_ACKINTR and add IXGBE_REQUEST_TASK_{MOD,MSF}_WOI.
1.73	19-Nov-2020	msaitoh	Add comment.
1.72	17-Nov-2020	knakahara	Add new spin mutex to avoid race between ixgbe_msix_admin() and ixgbe_handle_admin(). At first, it seems "IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_OTHER)" cannot stop interrupts, because 31th bit is reserved for 82598, 82599, X540 and X550. So, the current following design (1) ixgbe_msix_admin() disables interrupts (2) ixgbe_msix_admin() calls workqueue_enqueue() for ixgbe_handle_admin() (3) ixgbe_handle_admin() does interrupt processing (4) after ixgbe_handle_admin() has done all interrupt processings, ixgbe_handle_admin() enables interrupts does not work correctly, that is, interrupts can be lost while ixgbe_handle_admin() is running. To fix that, add new spin mutex(adapter->admmin_mtx) which protects atomically the following two members. - adapter->admin_pending - adapter->task_requests The unnecessary "IXGBE_WRITE_REG(hw, IXGBE_EIMC, IXGBE_EIMC_OTHER)" code will be removed later. Reviewed and tested by hikaru@n.o and msaitoh@n.o, thanks.
1.71	31-Aug-2020	msaitoh	branches: 1.71.2; If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.70	27-Aug-2020	msaitoh	ADD SFP+ MOD_ABS inversion quirk. On X550 EM, GPIO(SDP) and SFP+'s MOD_ABS is directly connected. It has no inverter. GIGABYTE MA10-ST0 has a inverter, so add new quirk for it.
1.69	17-Aug-2020	msaitoh	Re-enabling interrupt is required only when a work is scheduled form the interrput context.
1.68	17-Aug-2020	msaitoh	Fix a bug that the driver sometimes missed module insertion. The ixgbe_sfp_probe() function was only for 82598 and other chips had no way to poll SFP+ cage. The ixgbe_handle_mod() already has function to treat module insertion/removal for all chips, so enqueue the work if the cage status changed. All of ixgbe chips' SFP+ module interrupt is only on the inserstion. This change also detect the removal by the timer.
1.67	25-Jun-2020	msaitoh	Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.66	23-Jun-2020	msaitoh	KNF. No functional change.
1.65	06-Feb-2020	thorpej	Ensure we don't call workqueue_enqueue() if the pluggable optics handler is already pending.
1.64	05-Feb-2020	msaitoh	No functional change: - Add debug printf()s. - Remove unused macros. - Remove extra newline.
1.63	04-Feb-2020	thorpej	- Fix locking problem with optics module interrupts: ifmedia_add() may block on memory allocation, and so it cannot be safely done from a softint nor can it be done while holding a spin lock. Fix this by using a workqueue rather than a softint, and hold the IFNET_LOCK across the entire handler, and the CORE_LOCK only across the code that needs to serialize access to the hardware state. - Use ifmedia_fini(). Tested in a variety of devices by msaitoh@. (Thanks!)
1.62	21-Jan-2020	msaitoh	Fix the freeing code for some error paths. Found and tested by Patrick Welche.
1.61	20-Jan-2020	msaitoh	Free jumbo mem structure correctly. Found by yamaguchi@ using with LOCKDEBUG.
1.60	16-Dec-2019	msaitoh	branches: 1.60.2; No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.59	30-Oct-2019	msaitoh	G.C. (if_ipackets)
1.58	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.57	12-Sep-2019	msaitoh	- Make SIOCADDMULTI returns ENOSPC when the Ethenet multicast address list exceeds the limit. - State "Ethernet" multicast address. - Use macro.
1.56	17-Jul-2019	msaitoh	branches: 1.56.2; Implement VLAN hardware filter function(ETHERCAP_VLAN_HWFILTER). First proposed by jmcneill in 2017 and modified by me. How to use: - Set callback function: ether_set_vlan_cb(struct ethercom , ether_vlancb_t) - Callback. This function is called when a vlan is attached/detached to the parent interface: int (ether_vlancb_t)(struct ethercom *ec, uint16_t vlanid, bool set); - ifconfig(8) ifconfig ixg0 [-]vlan-hwfilter Note that ETHERCAP_VLAN_HWFILTER is set by default on ixg(4) because the PF driver usually enable "all block" filter by default.
1.55	27-Jun-2019	msaitoh	Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.54	22-Feb-2019	msaitoh	Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.53	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.52	03-Dec-2018	msaitoh	Some counters are not per queue but per traffic class. Make new evcnt group "ixgM tcN" (N = 0..7) and move those counters into it. We are using only traffic class 0, so we will see only tc0's counter are updated.
1.51	31-Jul-2018	msaitoh	Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.50	25-Jun-2018	msaitoh	branches: 1.50.2; Remove duplicated inclusion of net/bpf.h.
1.49	23-May-2018	msaitoh	Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE.
1.48	18-May-2018	msaitoh	Revert new watchdog timer commits. The new watchdog timer made stability worse than before. It seems unknown problems exists. http://mail-index.netbsd.org/source-changes/2018/05/08/msg095020.html http://mail-index.netbsd.org/source-changes/2018/05/16/msg095240.html
1.47	08-May-2018	msaitoh	- Fix broken watchdog timer. This change detects TX device timeout correctly. NOTE: It's supporsed not to be called {ixgbe,ixv}_rearm_queues() in the timer. Those are not required if any chip have no bug. In reality, ixgbe_rearm_queues() is required on 82599 and newer chip AND other than queue 0 to prevent device timeout. When it occured, packet was sent but the descriptor's DD bit wasn't set even though IXGBE_TXD_CMD_EOP and IXGBE_TXD_CMD_RS were set. After forcing interrupt by writing EICS register in ixgbe_rearm_queues(), DD is set. Why? Is this an undocumented errata? It might be possible not call rearm_queues on 82598 or queue 0, we call in any cases in case the problem occurs. On ixv(4), I have not seen this problem yet (though I tested only on X550_X(Xeon D 12xx)'s virtual function), but we do rearm in case TX device timeout happen. - ixv(4): Call callout_stop() earlier in ixv_stop() like ixgbe_stop(). - KNF.
1.46	25-Apr-2018	msaitoh	Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet.
1.45	19-Apr-2018	christos	s/static inline/static __inline/g for consistency.
1.44	19-Apr-2018	msaitoh	Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer.
1.43	17-Apr-2018	msaitoh	Remove unused structure entries. No functional change.
1.42	17-Apr-2018	knakahara	Fix panic when "sysctl -w hw.ixg0.txrx_workqueue=[01]" while there is traffic. The operation is not supported, however causing panic is problem.
1.41	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.40	02-Apr-2018	knakahara	Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o.
1.39	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.38	30-Mar-2018	knakahara	Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8
1.37	26-Mar-2018	msaitoh	- Add link related softint's counter. - Fix indent.
1.36	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.35	09-Mar-2018	msaitoh	Make some event counters MP safe. Now all of the event counters are MP safe.
1.34	07-Mar-2018	msaitoh	- Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code.
1.33	06-Mar-2018	msaitoh	Move the location of "struct work" as FreeBSD's "struct task" in ixgbe.h. No functional change.
1.32	02-Mar-2018	knakahara	branches: 1.32.2; ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o.
1.31	26-Feb-2018	knakahara	Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o.
1.30	22-Feb-2018	msaitoh	Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.29	06-Dec-2017	msaitoh	Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.28	22-Nov-2017	msaitoh	Fix a bug that bypass adapter's sysctls aren't set. Tested with non-genuine X550-T2 bypass adapter: - Call ixgbe_sysctl_instance() in ixgbe_bypass_init() to get sysctl's top node correctly. - ixgbe_init_device_features() refers adapter->hw.bus.func for bypass adapter. Call set_lan_id() to set adapter->hw.bus.func before calling ixgbe_init_device_features(). Without this, bypass sysctl's are added to both the first and second port. - Initalize node.sysctl_data before calling sysctl_lookup() to read correct value.
1.27	02-Nov-2017	msaitoh	- Handle hw.ixgN.advertise_speed correctly. - Fix hw.ixgN.advertise_speed sysctl's description.
1.26	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.25	24-Aug-2017	msaitoh	Remove unused counters.
1.24	13-Feb-2017	msaitoh	branches: 1.24.6; Use percpuq.
1.23	10-Feb-2017	msaitoh	- Both mbuf_defrag_failed and m_defrag_failed existed. Remove m_defrag_failed. - rx_irq, dropped_pkts, morerx, moretx and txloops counters are uncounted, so remove them. - Count handleq in ixv_handle_que(). - Detach event counters correctly. - Set some per-queue event names correctly.
1.22	01-Feb-2017	msaitoh	TX multiqueue. If you want to disable it, enable IXGBE_LEGACY_TX in ixgbe_netbsd.h
1.21	30-Jan-2017	msaitoh	- Count tso_err again. It was accidentally removed in ixgbe.c rev. 1.28. - ixgbe.h: Sort entries to reduce diff against FreeBSD.
1.20	25-Jan-2017	msaitoh	Fix SFP+ insersion or replacement stuff.
1.19	18-Jan-2017	msaitoh	Fix sysctl stuff. Make the top sysctl node and keep the pointer.
1.18	05-Dec-2016	msaitoh	branches: 1.18.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.17	02-Dec-2016	msaitoh	- Add missing pmf_device_deregister() in ixgbe_detach() - Remove extra newline from an sysctl description. - Fix dmesg output when the number of PCIe lane can't get.
1.16	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.15	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.14	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.13	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.12	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.11	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.10	28-Nov-2016	msaitoh	FreeBSD r280182 made new file ix_txrx.c and moved ixgbe.c and ixv's common code into it. Before sync with whole of them, just move ixgbe.c and ixv.c's common code into ix_txrx.c from ixgbe.c. In this commit, only ixgbe.c is split into the device dependent part and the common part. ixv.c isn't change to make this commit no functional change. To use ixv.c with ix_txrx.c, it's required to modify the common part's API and functions themselves. This commit is done to make the next change easy to understand.
1.9	17-Aug-2015	knakahara	branches: 1.9.2; Add kernel code to support intrctl(8).
1.8	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.7	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.6	14-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r243716: - A lot of bugfixes. Some of them are realted to multi queue and those have not affected in NetBSD because we have not used it yet. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Improve performance.
1.5	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.4	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.3	10-Mar-2015	msaitoh	Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.2	04-Feb-2015	msaitoh	Fix mutex related problem reported by Uwe Toenjes in PR#49328: - Revert ixgbe_netbsd.c rev. 1.2 - make CORE_LOCK adaptive - Release spin lock while reinitializing the jumb buffer structure.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.6	28-Aug-2017	skrll	Sync with HEAD
1.1.30.5	05-Feb-2017	skrll	Sync with HEAD
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.5	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.4	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.3	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.28.2	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.1.28.1	24-Feb-2015	martin	Pull up following revision(s) (requested by msaitoh in ticket #545): sys/dev/pci/ixgbe/ixgbe.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.19 Add missing IXGBE_RX_LOCK_DESTROY() for the detach path. Remove extra IXGBE_TX_LOCK()/IXGBE_TX_UNLOCK() from ixgbe_free_transmit_structures(). This function is called in the end of the detach function. if_stop was called and almost all resources were freed, so it's not required to block the TX stuff. Fix mutex related problem reported by Uwe Toenjes in PR#49328: - Revert ixgbe_netbsd.c rev. 1.2 - make CORE_LOCK adaptive - Release spin lock while reinitializing the jumb buffer structure. disable preemption while scheduling the softint to configure the link.
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.9.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.9.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.18.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.24.6.30	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1933: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.24.6.29	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.24.6.28	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.24.6.27	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.24.6.26	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.24.6.25	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.98 sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others. Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.24.6.24	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.24.6.23	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1708: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.24.6.22	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.24.6.21	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.24.6.20	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.24.6.19	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.24.6.18	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.24.6.17	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.24.6.16	01-Mar-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1199): sys/dev/pci/ixgbe/ixv.c: revision 1.109 sys/dev/pci/ixgbe/ixgbe.h: revision 1.54 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.174 Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.24.6.15	28-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1171): sys/dev/pci/ixgbe/ixgbe.c: revision 1.168 sys/dev/pci/ixgbe/ixgbe.h: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.172 sys/dev/pci/ixgbe/ixgbe.c: revision 1.173 Some counters are not per queue but per traffic class. Make new evcnt group "ixgM tcN" (N = 0..7) and move those counters into it. We are using only traffic class 0, so we will see only tc0's counter are updated. - Call ixgbe_init_swfw_semaphore(hw) before calling ixgbe_check_wol_support() because ixgbe_check_wol_support()->ixgbe_check_wol_support()-> ixgbe_get_device_caps() accesses semaphore. - - Move the location of calling setup_eee() in ixgbe_attach(). The previous location was too early. The new location is before calling ixgbe_set_phy_power(). - Restore call of setup_eee() in ixgbe_init_locked(). It was removed in rev. 1.99 (FreeBSD r320688). - When calling setup_eee(), pass enable/disable correctly.
1.24.6.14	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.24.6.13	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #938): sys/arch/acorn32/podulebus/if_ie.c: revision 1.41 sys/arch/amiga/dev/if_es.c: revision 1.58 sys/arch/amiga/dev/if_qn.c: revision 1.45 sys/arch/arm/at91/at91emac.c: revision 1.20 sys/arch/arm/ep93xx/epe.c: revision 1.37 sys/arch/emips/ebus/if_le_ebus.c: revision 1.14 sys/arch/emips/ebus/if_le_ebus.c: revision 1.15 sys/arch/mac68k/dev/if_mc.c: revision 1.46 sys/arch/macppc/dev/am79c950.c: revision 1.39 sys/arch/newsmips/apbus/if_sn.c: revision 1.40 sys/arch/next68k/dev/mb8795.c: revision 1.59 sys/arch/playstation2/dev/if_smap.c: revision 1.25 sys/arch/playstation2/dev/if_smap.c: revision 1.26 sys/arch/sun2/dev/if_ec.c: revision 1.28 sys/arch/sun3/dev/if_ie.c: revision 1.63 sys/arch/x68k/dev/if_ne_intio.c: revision 1.19 sys/arch/xen/xen/if_xennet_xenbus.c: revision 1.75 sys/arch/xen/xen/xennetback_xenbus.c: revision 1.63 sys/dev/bi/if_ni.c: revision 1.45 sys/dev/cadence/if_cemac.c: revision 1.12 sys/dev/ic/am7990.c: revision 1.78 sys/dev/ic/am79900.c: revision 1.27 sys/dev/ic/an.c: revision 1.67 sys/dev/ic/cs89x0.c: revision 1.40 sys/dev/ic/dm9000.c: revision 1.13 sys/dev/ic/dm9000.c: revision 1.14 sys/dev/ic/dp8390.c: revision 1.88 sys/dev/ic/elink3.c: revision 1.141 sys/dev/ic/elinkxl.c: revision 1.122 sys/dev/ic/hme.c: revision 1.98 sys/dev/ic/i82586.c: revision 1.77 sys/dev/ic/lance.c: revision 1.53 sys/dev/ic/mb86950.c: revision 1.27 sys/dev/ic/mb86960.c: revision 1.86 sys/dev/ic/mtd803.c: revision 1.34 sys/dev/ic/pdq_ifsubr.c: revision 1.59 sys/dev/ic/rrunner.c: revision 1.86 sys/dev/ic/seeq8005.c: revision 1.58 sys/dev/ic/sgec.c: revision 1.47 sys/dev/ic/smc90cx6.c: revision 1.72 sys/dev/ic/smc91cxx.c: revision 1.96 sys/dev/ic/tropic.c: revision 1.49 sys/dev/ic/wi.c: revision 1.245 sys/dev/isa/if_eg.c: revision 1.93 sys/dev/isa/if_el.c: revision 1.95 sys/dev/isa/if_iy.c: revision 1.101 sys/dev/ofw/ofnet.c: revision 1.58 sys/dev/pci/if_alc.c: revision 1.27 sys/dev/pci/if_de.c: revision 1.152 sys/dev/pci/if_fpa.c: revision 1.61 sys/dev/pci/if_jme.c: revision 1.34 sys/dev/pci/if_tl.c: revision 1.108 sys/dev/pci/if_vte.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe.h: revision 1.50 sys/dev/pcmcia/if_cnw.c: revision 1.62 sys/dev/pcmcia/if_malo_pcmcia.c: revision 1.17 sys/dev/pcmcia/if_ray.c: revision 1.89 sys/dev/pcmcia/if_xi.c: revision 1.81 sys/dev/pcmcia/mhzc.c: revision 1.51 sys/dev/pcmcia/xirc.c: revision 1.34 sys/dev/qbus/if_de.c: revision 1.33 sys/dev/qbus/if_qe.c: revision 1.78 sys/dev/qbus/if_qt.c: revision 1.22 sys/dev/sbus/be.c: revision 1.87 sys/dev/sbus/qe.c: revision 1.68 sys/dev/scsipi/if_se.c: revision 1.96 sys/dev/usb/if_atu.c: revision 1.59 sys/net/if_l2tp.c: revision 1.28 via patch sys/net/if_ppp.c: revision 1.160 It's not required to include net/bpfdesc.h. Remove it. -- Simplify like other drivers. NULL check of ifp->if_bpf is done in bpf_mtap(), so it's not required to do it here. -- Remove duplicated inclusion of net/bpf.h. -- Remove duplicated inclusion of net/bpf.h. -- Simplify bpf_mtap() call. No functional change.
1.24.6.12	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.24.6.11	17-Apr-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #769): sys/dev/pci/ixgbe/ixgbe.c: revision 1.147 sys/dev/pci/ixgbe/ixgbe.h: revision 1.42 Fix panic when "sysctl -w hw.ixg0.txrx_workqueue=[01]" while there is traffic. The operation is not supported, however causing panic is problem.
1.24.6.10	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.24.6.9	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.24.6.8	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.24.6.7	13-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/dev/pci/ixgbe/ixgbe.c: revision 1.134 sys/dev/pci/ixgbe/ixgbe.h: revision 1.35 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.35 sys/dev/pci/ixgbe/ixv.c: revision 1.87 Make some event counters MP safe. Now all of the event counters are MP safe.
1.24.6.6	08-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #612): sys/dev/pci/ixgbe/ixgbe.c: revision 1.129-1.133 sys/dev/pci/ixgbe/ixgbe.h: revision 1.34 sys/dev/pci/ixgbe/ixv.c: revision 1.85,1.86 - Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code. Don't increment que->req.ev_count in MSI-X interrupt because it's not reschedule. Fix another poll mode assumption breaking. Implemented by msaitoh@n.o, I just commit by proxy. ixgbe_rearm_queues() writes EICS register(s). 82599, X540 and X550 specifications say "Following a write of 1b to any bit in the EICS register (interrupt cause set), its corresponding bit in the EIMS register is auto set as well enabling its interrupt." in "Extended Interrupt Auto Mask Enable (EIAM) Register" section. That is, ixgbe_rearm_queues() causes interrupts regardless of the status managed by ixgbe_enable_queue()/ixgbe_disable_queue(). That can break poll mode assumption. In fact, the problem occurs in the following situation - CPU#A has high load traffic, in contrast, CPU#B has not so high load traffic - CPU#A is occurred interrupt by its NIC queue - CPU#A calls ixgbe_disable_queue() in interrupt handler(ixgbe_msix_que()) - CPU#A kick softint handler(ixgbe_handle_que()) - CPU#A begins softint - CPU#A's NIC queue is set que->txr->busy flag - With some reason, CPU#A can do ixg interrupt handler E.g. when one of CPU#A's softnet handlers sleeps, ipl is lowered - CPU#B starts callout - CPU#B calls ixgbe_local_timer1() - CPU#B writes EICS bit corresponding CPU#A's NIC queue bit - CPU#A's NIC queue causes interrupt whie CPU#A is running in poll mode - CPU#A calls ixgbe_disable_queue() in interrupt handler again - CPU#A has done polling, and then CPU#A calls ixgbe_enable_queue() once - CPU#A's NIC queue interrupt is disabled until ixg is detached as ixgbe_disable_queue() is called twice though ixgbe_disable_queue() is called once only NOTE: 82598 does not say so, but it is treated in the same way because of no harm. By the way, we will refactor ixgbe_local_timer(watchdog processing) later. Fix INTx/MSI handler did not schedule workqueue. Pointed out by msaitoh@n.o. Reduce duplicated code which schedule deferred packet processing. No functional change.
1.24.6.5	06-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #605): sys/dev/pci/ixgbe/ixgbe.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe.h: revision 1.33 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.34 sys/dev/pci/ixgbe/ixgbe.c: revision 1.128 sys/dev/pci/ixgbe/ixv.c: revision 1.83 sys/dev/pci/ixgbe/ixv.c: revision 1.84 Add hw.ixvM.q[01].{interrupt_rate,[tr]xd_head,[tr]xd_tail} sysctls as ixg(4). ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o. ixv(4) also supports workqueue poll mode, but not enabled by default yet, either. ok by msaitoh@n.o. Move the location of "struct work" as FreeBSD's "struct task" in ixgbe.h. No functional change.
1.24.6.4	01-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #597): sys/dev/pci/ixgbe/ixgbe.h: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.c: revision 1.127 sys/dev/pci/ixgbe/ixv.c: revision 1.82 Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o. Apply ixgbe.c:r1.127 to ixv.c. Pointed out by msaitoh@n.o.
1.24.6.3	26-Feb-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #592): sys/dev/pci/ixgbe/ixv.c: revision 1.79 sys/dev/pci/ixgbe/ixgbe.h: revision 1.30 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.31 sys/dev/pci/ixgbe/ixgbe.c: revision 1.120 sys/dev/pci/ixgbe/ixgbe.c: revision 1.121 sys/dev/pci/ixgbe/ixgbe.c: revision 1.123 sys/dev/pci/ixgbe/ixgbe.c: revision 1.124 sys/dev/pci/ixgbe/ixgbe.c: revision 1.125 sys/dev/pci/ixgbe/ixgbe.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.80 sys/dev/pci/ixgbe/ixv.c: revision 1.81 CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558). Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8 Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv) - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8 Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8 Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o. - Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function. Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.24.6.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.24.6.1	31-Aug-2017	martin	Pull up following revision(s) (requested by msaitoh in ticket #248): sys/dev/pci/ixgbe/ixgbe.h: revision 1.25 sys/dev/pci/ixgbe/ixgbe.c: revision 1.96 sys/dev/pci/ixgbe/ixgbe.c: revision 1.97 Remove unused counters. Don't limit number of queue pair to 8. Take max TX queues and max RX queues into account. 82598 has 32 TX queue and 64 RX queues. Others has 128 TX queues and 128 RX queues. Number of MSI-X table entries is 64, so the max queue pairs is 63 (64 minus the Others (e.g. link status change)).
1.32.2.9	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.32.2.8	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.32.2.7	25-Jun-2018	pgoyette	Sync with HEAD
1.32.2.6	21-May-2018	pgoyette	Sync with HEAD
1.32.2.5	02-May-2018	pgoyette	Synch with HEAD
1.32.2.4	22-Apr-2018	pgoyette	Sync with HEAD
1.32.2.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.32.2.2	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.32.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.50.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.50.2.1	10-Jun-2019	christos	Sync with HEAD
1.56.2.13	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1792: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.56.2.12	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.56.2.11	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.56.2.10	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.56.2.9	02-Jun-2022	martin	Pull up the following revisions, all via patch, requested by msaitoh in ticket #1459: sys/dev/pci/ixgbe/ixgbe.c 1.261,1.265-1.268,1.273,1.275-1.277, 1.312,1.316-1.319 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.182 - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Fix typos in comment. - KNF.
1.56.2.8	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1458): sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.56.2.7	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.56.2.6	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1374: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.56.2.5	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.56.2.4	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.56.2.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.56.2.2	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.56.2.1	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.60.2.2	29-Feb-2020	ad	Sync with head.
1.60.2.1	25-Jan-2020	ad	Sync with head.
1.71.2.3	03-Apr-2021	thorpej	Sync with HEAD.
1.71.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.71.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.75.4.1	01-Aug-2021	thorpej	Sync with HEAD.
1.86.4.5	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #563): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.110 sys/dev/pci/ixgbe/ixgbe.c: revision 1.345 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.111 sys/dev/pci/ixgbe/ixgbe.c: revision 1.346 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.112 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.62 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.113 sys/dev/pci/ixgbe/ixgbe.c: revision 1.348 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.114 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.115 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.116 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.105 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.106 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.107 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.108 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.109 sys/dev/pci/ixgbe/ixv.c: revision 1.193 sys/dev/pci/ixgbe/ixv.c: revision 1.195 sys/dev/pci/ixgbe/ixv.c: revision 1.196 sys/dev/pci/ixgbe/ixgbe.h: revision 1.94 sys/dev/pci/ixgbe/ixgbe.h: revision 1.95 sys/dev/pci/ixgbe/ixgbe.h: revision 1.96 sys/dev/pci/ixgbe/ixgbe.h: revision 1.97 sys/dev/pci/ixgbe/ixgbe.h: revision 1.98 ixgbe: Fix comment. No functional change. ixgbe: Whitespace. No functional change. ixgbe(4): Move assignment of TXD. NFCI. ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor. ixv(4): Remove unused IFF_OACTIVE. No functional change. ixgbe: Clear the WTHRESH bit field before writing it. ixgbe: Modify for the readability. No functional change. ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof(). ixgbe: Update if_opackets outside the loop in ixgbe_txeof(). ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof(). ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO). ixgbe: Remove unused to reduce the size of struct rx_ring. ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring. ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.86.4.4	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #420): sys/dev/pci/ixgbe/ixgbe.c: revision 1.340 sys/dev/pci/ixgbe/ixgbe.c: revision 1.341 sys/dev/pci/ixgbe/ixgbe.c: revision 1.342 sys/dev/pci/ixgbe/ixgbe.c: revision 1.343 sys/dev/pci/ixgbe/ixgbe.c: revision 1.344 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.61 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.104 sys/dev/pci/ixgbe/ixv.c: revision 1.190 sys/dev/pci/ixgbe/ixv.c: revision 1.191 sys/dev/pci/ixgbe/ixv.c: revision 1.192 sys/dev/pci/ixgbe/ixgbe.h: revision 1.93 ixg(4): Print DEVICE_CAPS register. ixgbe: Whitespace. No functional change. ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl. ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. ixgbe: Whitespace. No functional change.
1.86.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.86.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.86.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.98.2.1	02-Aug-2025	perseant	Sync with HEAD
1.19	06-Jun-2022	msaitoh	s/filer/filter/ in comment. No functional change.
1.18	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.17	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.16	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.15	25-Jun-2020	msaitoh	branches: 1.15.6; Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). Same as ixgbe_common.c rev. 1.24. Found by KUBSan.
1.14	03-Jan-2020	pgoyette	Fix some typos in comments. From vezhlys on freenode IRC.
1.13	23-Dec-2019	msaitoh	Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.12	04-Apr-2018	msaitoh	branches: 1.12.2; 1.12.8; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.11	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.10	06-Dec-2017	msaitoh	branches: 1.10.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.9	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.8	02-Dec-2016	msaitoh	branches: 1.8.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.7	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.5	05-Aug-2015	msaitoh	branches: 1.5.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.4	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.8.8.8	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.8.8.7	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.8.8.6	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.8.8.5	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.8.8.4	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.8.8.3	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.8.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.8.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.10.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.12.8.6	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.12.8.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.12.8.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.12.8.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.12.8.2	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.12.8.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.12.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.15.6.1	13-May-2021	thorpej	Sync with HEAD.
1.9	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.8	04-Apr-2018	msaitoh	branches: 1.8.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.7	06-Dec-2017	msaitoh	branches: 1.7.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.6	30-Aug-2017	msaitoh	branches: 1.6.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.5	01-Dec-2016	msaitoh	branches: 1.5.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.3	24-Apr-2015	msaitoh	branches: 1.3.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.2	02-Apr-2015	msaitoh	branches: 1.2.2; 1.2.4; Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.1	27-Mar-2015	msaitoh	Add some forgotten files in last commit.
1.2.4.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.4.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.4.1	02-Apr-2015	riz	file ixgbe_82598.h was added on branch netbsd-7 on 2015-04-19 14:42:14 +0000
1.2.2.4	05-Dec-2016	skrll	Sync with HEAD
1.2.2.3	06-Jun-2015	skrll	Sync with HEAD
1.2.2.2	06-Apr-2015	skrll	Sync with HEAD
1.2.2.1	02-Apr-2015	skrll	file ixgbe_82598.h was added on branch nick-nhusb on 2015-04-06 15:18:12 +0000
1.3.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.5.8.3	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.5.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.5.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.6.2.2	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	30-Aug-2017	jdolecek	file ixgbe_82598.h was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.7.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.8.8.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.32	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.31	06-Oct-2023	msaitoh	ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz.
1.30	04-Oct-2023	msaitoh	ixgbe: Fix typo in comment. No functional change.
1.29	24-Dec-2021	msaitoh	branches: 1.29.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.28	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.27	14-Dec-2021	msaitoh	Add IPv6 mask for flow director. - FreeBSD: c4f73d5d6a6c0117e08a03920cce69202865ba13 or ix-3.3.18 DPDK: 21feefa2fcd5899ee26a10be405c17c0a1109860 Write FDIRIP6M register to allow flow director filter to set ipv6 rules without setting ipv6 source/destination address. - No functional change because NetBSD doesn't support flow director.
1.26	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.25	10-Dec-2021	msaitoh	Remove unnecessary return value check. FreeBSD: 3a89005394bc5d82ce9b6baa9e7f8dee362354ae DPDK: 4b0ee6529b7897c2a08dd56669f07ac1f46a8474
1.24	10-Dec-2021	msaitoh	Add typecast for type mismatch. FreeBSD: 994dd6328c66fc277438ad51ed074f3c52096147 DPDK: d8e52b2cf771c31b523b46852fd86225b5a2c721
1.23	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.22	23-Dec-2019	msaitoh	branches: 1.22.12; Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.21	06-Dec-2018	msaitoh	branches: 1.21.4; Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.20	11-Jun-2018	msaitoh	branches: 1.20.2; Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*).
1.19	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.18	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.17	30-Mar-2018	msaitoh	- Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.16	06-Dec-2017	msaitoh	branches: 1.16.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.15	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.14	02-Dec-2016	msaitoh	branches: 1.14.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.13	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.12	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.11	13-Aug-2015	msaitoh	branches: 1.11.2; - Add MSI/MSI-X support. The multiqueue function is not supported yet. - Make ixv.c compilable. _NOT_TESTED_YET_
1.10	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.9	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.8	14-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r243716: - A lot of bugfixes. Some of them are realted to multi queue and those have not affected in NetBSD because we have not used it yet. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Improve performance.
1.7	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.6	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.5	24-Feb-2015	msaitoh	Add 82599EN_SFP and 82599_SFP_SF_QP. Not tested.
1.4	24-Feb-2015	msaitoh	Change macro name IXGBE_DEV_ID_82599_SFP_DELL to IXGBE_DEV_ID_82599_SFP_SF2. Same as OpenBSD and Linux. No binary change.
1.3	17-Apr-2014	christos	branches: 1.3.2; 1.3.4; CID 140077: Add missing break and FALLTHROUGH
1.2	08-Apr-2014	christos	Add dell's flavor (from uwe, in current-users@)
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.4.4	05-Dec-2016	skrll	Sync with HEAD
1.3.4.3	22-Sep-2015	skrll	Sync with HEAD
1.3.4.2	06-Jun-2015	skrll	Sync with HEAD
1.3.4.1	06-Apr-2015	skrll	Sync with HEAD
1.3.2.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.3.2.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.3.2.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.3.2.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.11.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.14.8.11	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.14.8.10	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.14.8.9	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.14.8.8	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.14.8.7	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.14.8.6	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.14.8.5	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.14.8.4	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #935): sys/dev/pci/ixgbe/ixgbe.c: revision 1.162 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.35 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.16 Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*). -- Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.14.8.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.14.8.2	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.14.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.16.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.16.2.2	25-Jun-2018	pgoyette	Sync with HEAD
1.16.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.20.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.20.2.1	10-Jun-2019	christos	Sync with HEAD
1.21.4.6	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.21.4.5	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.21.4.4	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.21.4.3	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.21.4.2	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.21.4.1	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.22.12.1	13-May-2021	thorpej	Sync with HEAD.
1.29.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.29.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.8	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.7	04-Apr-2018	msaitoh	branches: 1.7.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.6	06-Dec-2017	msaitoh	branches: 1.6.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.5	30-Aug-2017	msaitoh	branches: 1.5.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.4	01-Dec-2016	msaitoh	branches: 1.4.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.3	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.2	24-Apr-2015	msaitoh	branches: 1.2.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.1	27-Mar-2015	msaitoh	branches: 1.1.2; 1.1.4; Add some forgotten files in last commit.
1.1.4.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.4.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.4.1	27-Mar-2015	riz	file ixgbe_82599.h was added on branch netbsd-7 on 2015-04-19 14:42:14 +0000
1.1.2.4	05-Dec-2016	skrll	Sync with HEAD
1.1.2.3	06-Jun-2015	skrll	Sync with HEAD
1.1.2.2	06-Apr-2015	skrll	Sync with HEAD
1.1.2.1	27-Mar-2015	skrll	file ixgbe_82599.h was added on branch nick-nhusb on 2015-04-06 15:18:12 +0000
1.2.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.4.8.3	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.4.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.4.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.5.2.2	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	30-Aug-2017	jdolecek	file ixgbe_82599.h was added on branch tls-maxphys on 2017-12-03 11:37:29 +0000
1.6.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.7.8.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.29	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.28	06-Jun-2022	msaitoh	branches: 1.28.4; s/filer/filter/ in comment. No functional change.
1.27	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.26	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.25	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.24	03-Jan-2020	pgoyette	branches: 1.24.12; Fix some typos in comments. From vezhlys on freenode IRC.
1.23	27-Jun-2019	msaitoh	branches: 1.23.2; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.22	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.21	11-Jun-2018	msaitoh	branches: 1.21.2; Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*).
1.20	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.19	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.18	06-Dec-2017	msaitoh	branches: 1.18.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.17	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.16	27-Jun-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.15	05-Dec-2016	msaitoh	branches: 1.15.8; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.14	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.13	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.12	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.11	06-Feb-2016	riastradh	branches: 1.11.2; Avoid shadowing global `min'.
1.10	06-Feb-2016	riastradh	Avoid shadowing global `min'.
1.9	13-Aug-2015	msaitoh	Reduce diff against FreeBSD r280181.
1.8	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.7	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.6	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.5	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.4	24-Feb-2015	msaitoh	Add 82599EN_SFP and 82599_SFP_SF_QP. Not tested.
1.3	24-Feb-2015	msaitoh	Change macro name IXGBE_DEV_ID_82599_SFP_DELL to IXGBE_DEV_ID_82599_SFP_SF2. Same as OpenBSD and Linux. No binary change.
1.2	08-Apr-2014	christos	branches: 1.2.2; 1.2.4; Add dell's flavor (from uwe, in current-users@)
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.7	28-Aug-2017	skrll	Sync with HEAD
1.2.4.6	05-Feb-2017	skrll	Sync with HEAD
1.2.4.5	05-Dec-2016	skrll	Sync with HEAD
1.2.4.4	19-Mar-2016	skrll	Sync with HEAD
1.2.4.3	22-Sep-2015	skrll	Sync with HEAD
1.2.4.2	06-Jun-2015	skrll	Sync with HEAD
1.2.4.1	06-Apr-2015	skrll	Sync with HEAD
1.2.2.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.2.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.2.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.2.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.11.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.15.8.10	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.15.8.9	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.15.8.8	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.15.8.7	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.15.8.6	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.15.8.5	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.15.8.4	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #935): sys/dev/pci/ixgbe/ixgbe.c: revision 1.162 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.35 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.16 Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*). -- Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.15.8.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.15.8.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.15.8.1	05-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #175): sys/dev/pci/ixgbe/ix_txrx.c: 1.25-1.27 sys/dev/pci/ixgbe/ixgbe.c: 1.91-1.95 sys/dev/pci/ixgbe/ixgbe_type.h: 1.23-1.25 sys/dev/pci/ixgbe/ixgbe_api.c: 1.16 sync ixg(4) up to ixgbe.c rev. 1.95: - Increase total number of RX buffers on multiqueue. - Fix the partial chain check in ixgbe_rx_discard(). This bug was added in ixgbe.c rev. 1.33. - Fix a double free in ixgbe_rxeof(). - Print verbose output in ixgbe_update_link_status() correctly on 5Gbps and 2.5Gbps (NBASE-T) for X550T[12] and newer. - Change hw.ixgN.ts to hw.ixgN.thermal_test. Same as FreeBSD. - Fix a bug that ifconfig ixgN media 1000baseT and 10Gbase-T advertised additional unwanted speeds. - Print PHY ID. - Remove unused variable.
1.18.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.18.2.2	25-Jun-2018	pgoyette	Sync with HEAD
1.18.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.21.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.21.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.21.2.1	10-Jun-2019	christos	Sync with HEAD
1.23.2.5	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.23.2.4	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.23.2.3	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.23.2.2	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.23.2.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.24.12.1	13-May-2021	thorpej	Sync with HEAD.
1.28.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.17	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.16	24-Dec-2021	msaitoh	branches: 1.16.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.15	27-Jun-2019	msaitoh	branches: 1.15.2; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.14	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.13	04-Apr-2018	msaitoh	branches: 1.13.2; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.12	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.11	06-Dec-2017	msaitoh	branches: 1.11.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.10	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.9	02-Dec-2016	msaitoh	branches: 1.9.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.8	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.7	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.6	13-Aug-2015	msaitoh	branches: 1.6.2; Reduce diff against FreeBSD r280181.
1.5	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.4	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.9.8.6	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.9.8.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.9.8.4	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.9.8.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.9.8.2	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.9.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.11.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.11.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.11.2.1	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.13.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.13.2.1	10-Jun-2019	christos	Sync with HEAD
1.15.2.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.15.2.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.16.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.5	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.4	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.3	24-Dec-2021	msaitoh	branches: 1.3.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.2	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.14; 1.1.26; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.26.1	13-May-2021	thorpej	Sync with HEAD.
1.1.14.4	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.1.14.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.14.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.14.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.6	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.1.4.5	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_bypass.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_bypass.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.3.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.3.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.48	17-Feb-2025	andvar	s/bye/byte/ in comments.
1.47	06-Oct-2023	msaitoh	branches: 1.47.6; ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.46	06-Oct-2023	msaitoh	ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33.
1.45	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.44	15-May-2023	msaitoh	Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.43	06-Jun-2022	msaitoh	branches: 1.43.4; s/filer/filter/ in comment. No functional change.
1.42	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.41	14-Dec-2021	msaitoh	Wait longer for link after fiber MAC setup. - FreeBSD: aaa3af802f90d93fdffb99100fe56f0bc3dda119 96ef6eb3ae9d622906fb838c82ede3074f864cdc DPDK: 64f1c8539c8ce99214b9eb1fb728a2c6745f3300 73247f1ced303c16987bb366d38a2d8a0fc40db4 After setting up the link on a fiber port, the maximum wait time for the link to come up is 500 ms in ixgbe_setup_mac_link_multispeed_fiber(). On an x550 SFP+ port, this is often not sufficiently long for the link to come up. This can result in never being able to retrieve accurate link status for the port using rte_eth_link_get_nowait(). Increase the maximum wait time in ixgbe_setup_mac_link_multispeed_fiber() to 1 s.
1.40	14-Dec-2021	msaitoh	Use macro. No functional change. Same as FreeBSD.
1.39	14-Dec-2021	msaitoh	Change error level in ixgbe_fc_autoneg(). - FreeBSD: f511cd22586a9c0358b86334a51e3db60ca2db01 or ix-3.3.18 DPDK: ab6ac48d483ef7f906b90f45182f2ddf3254d876 - No functional change in NetBSD.
1.38	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.37	10-Dec-2021	msaitoh	Fix infinite recursion on PCIe link down. - FreeBSD: 8270b7174c48417a4d5f3effa4a4f4588205e687 or ix-3.3.14 - DPDK: 2d04b9e856125197ec8e967471426d56ab7efcf0 In some corner cases the functions ixgbe_clear_rar_generic and ixgbe_clear_vmdq_generic may call one another leading to infinite recursion. When ixgbe_clear_vmdq_generic is called with IXGBE_CLEAR_VMDQ_ALL flag, it's going to clear MPSAR registers, and proceed to call ixgbe_clear_rar_generic, which in turn will clear the RAR registers, and recursively call back ixgbe_clear_vmdq_generic. Normally, the latter would detect that MPSAR registers have already been cleared and terminate the recursion. However, when PCIe link is down, and before the driver has had the opportunity to shut itself down, all register reads return 0xFFFFFFFF, and all register writes fail silently. In such case, because ixgbe_clear_vmdq_generic blindly assumes that clearing MPSAR registers succeeded, it's going to always call ixgbe_clear_rar_generic, which in turn will always call back ixgbe_clear_vmdq_generic, creating infinite recursion. This patch re-reads MPSAR register values after they had been cleared. In case of PCIe link failure, the values read will be non-zero, which will terminate the recursion. On the other hand, under normal circumstances the value read from MPSAR registers is going to be equal to the value previously written, so this patch is expected not to cause any regressions. - Note that NetBSD doesn't support VMDQ.
1.36	10-Dec-2021	msaitoh	Remove debug error message.
1.35	10-Dec-2021	msaitoh	Remove unnecessary return value check. FreeBSD: 3a89005394bc5d82ce9b6baa9e7f8dee362354ae DPDK: 4b0ee6529b7897c2a08dd56669f07ac1f46a8474
1.34	10-Dec-2021	msaitoh	Add typecast for type mismatch. FreeBSD: 994dd6328c66fc277438ad51ed074f3c52096147 DPDK: d8e52b2cf771c31b523b46852fd86225b5a2c721
1.33	02-Aug-2021	andvar	fix various typos in comments and log messages.
1.32	24-Jul-2021	andvar	Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889. Also fixed some additional typos in comments, found on review of same files or typos.
1.31	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.30	31-Aug-2020	msaitoh	branches: 1.30.6; Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change.
1.29	31-Aug-2020	msaitoh	If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.28	27-Aug-2020	msaitoh	Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.27	05-Feb-2020	msaitoh	No functional change: - Add debug printf()s. - Remove unused macros. - Remove extra newline.
1.26	16-Dec-2019	msaitoh	branches: 1.26.2; No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.25	25-Jul-2019	msaitoh	branches: 1.25.2; Fix VLAN filter setting. Found by KUBSan.
1.24	25-Jul-2019	msaitoh	Set FCTTV(Flow Control Transmit Timer Value) register correctly in ixgbe_fc_enable_generic(). Found by KUBSan.
1.23	27-Jun-2019	msaitoh	Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.22	04-Apr-2018	msaitoh	branches: 1.22.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.21	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.20	30-Mar-2018	msaitoh	- Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.19	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.18	16-Mar-2018	msaitoh	- Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+.
1.17	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.16	06-Dec-2017	msaitoh	branches: 1.16.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.15	13-Oct-2017	msaitoh	Show 2.5GBASE-T and 5GBASE-T correctly on ixv(4). Tested on a virtual function on X550 T1.
1.14	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.13	26-May-2017	msaitoh	branches: 1.13.2; Count Bad SFD (Start Frame Delimiter). X550 and newer have this register.
1.12	19-Jan-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.11	05-Dec-2016	msaitoh	branches: 1.11.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.10	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.9	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.8	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.7	06-Feb-2016	riastradh	branches: 1.7.2; Avoid shadowing global `min'.
1.6	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.5	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.4	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.3	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.2	17-Apr-2014	christos	branches: 1.2.2; 1.2.4; CID 141398: Fix incorrect test
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.7	28-Aug-2017	skrll	Sync with HEAD
1.2.4.6	05-Feb-2017	skrll	Sync with HEAD
1.2.4.5	05-Dec-2016	skrll	Sync with HEAD
1.2.4.4	19-Mar-2016	skrll	Sync with HEAD
1.2.4.3	22-Sep-2015	skrll	Sync with HEAD
1.2.4.2	06-Jun-2015	skrll	Sync with HEAD
1.2.4.1	06-Apr-2015	skrll	Sync with HEAD
1.2.2.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.2.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.2.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.7.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.7.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.11.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.13.2.16	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.13.2.15	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.13.2.14	21-Jun-2023	martin	Pullup the following revisions, requested by msaitoh in ticket #1828: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.13.2.13	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.13.2.12	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.13.2.11	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.13.2.10	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1600): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.13.2.9	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.13.2.8	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.13.2.7	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.13.2.6	01-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1313: sys/dev/pci/ixgbe/ixgbe_type.h 1.41 sys/dev/pci/ixgbe/ixgbe.c 1.194,1.197-1.199 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.24 sys/dev/pci/ixgbe/ixv.c 1.121,1.124-1.125 - Avoid undefined behavior of interrupt vector setting. - Avoid undefined behavior of TX/RX queue statistics calculation. - Avoid undefined behavior of X550EM's PHY accesses. - Avoid undefined behavior of Flow Control Transmit Timer setting. - Don't call {ixgbe,ixv}_stop() twice while detaching.
1.13.2.5	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.13.2.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.13.2.3	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.13.2.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.13.2.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.16.2.4	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.16.2.3	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.16.2.2	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.16.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.22.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.22.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.25.2.9	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.25.2.8	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.25.2.7	21-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1647: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.25.2.6	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.25.2.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.25.2.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.25.2.3	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1070): sys/dev/pci/ixgbe/if_bypass.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe.c: revision 1.246 Fix checking return value of atomic_cas_uint(). This change fixes a bug that extra delay() is called only once even if atomic_cas_uint() isn't failed or delay() isn't called when atomic_cas_uint() failed. The reason of this bug was that I simply converted FreeBSD' atomic_cmpset_int() to atomic_cas_uint(). The return value's semantics is different. - Minor change. - Print "X550EM X" instead of "X550EM" for Xeon D devices. - Fix typo in comment. Same as FreeBSD.
1.25.2.2	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.25.2.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.26.2.1	29-Feb-2020	ad	Sync with head.
1.30.6.2	01-Aug-2021	thorpej	Sync with HEAD.
1.30.6.1	13-May-2021	thorpej	Sync with HEAD.
1.43.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.43.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.43.4.1	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #202): sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.56 sys/dev/pci/ixgbe/ixgbe.c: revision 1.325 sys/dev/pci/ixgbe/ixgbe.c: revision 1.326 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.44 Use thermal sensor code for IXGBE_DEV_ID_X550EM_A_10G_T, too. PCI device ID 0x15c8 also use X557-AT PHY, so create the thermal sensor sysctl for it, too. Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.47.6.1	02-Aug-2025	perseant	Sync with HEAD
1.18	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.17	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.16	24-Dec-2021	msaitoh	branches: 1.16.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.15	10-Dec-2021	msaitoh	Remove unnecessary return value check. FreeBSD: 3a89005394bc5d82ce9b6baa9e7f8dee362354ae DPDK: 4b0ee6529b7897c2a08dd56669f07ac1f46a8474
1.14	27-Jun-2019	msaitoh	branches: 1.14.2; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.13	04-Apr-2018	msaitoh	branches: 1.13.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.12	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.11	30-Mar-2018	msaitoh	- Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.10	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.9	06-Dec-2017	msaitoh	branches: 1.9.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.8	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.7	01-Dec-2016	msaitoh	branches: 1.7.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.5	05-Aug-2015	msaitoh	branches: 1.5.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.4	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.7.8.8	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.7.8.7	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.7.8.6	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.7.8.5	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.7.8.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.7.8.3	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.7.8.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.7.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.9.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.9.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.13.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.14.2.3	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.14.2.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.14.2.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.16.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.16.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.15	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.14	06-Oct-2023	msaitoh	ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD.
1.13	24-Dec-2021	msaitoh	branches: 1.13.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.12	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.11	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.10	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.9	03-Sep-2018	riastradh	branches: 1.9.4; 1.9.16; Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.8	04-Apr-2018	msaitoh	branches: 1.8.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.7	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.6	06-Dec-2017	msaitoh	branches: 1.6.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.5	30-Aug-2017	msaitoh	branches: 1.5.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.4	02-Dec-2016	msaitoh	branches: 1.4.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.3	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb.c was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb.c was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.4.8.5	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.4.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.4.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.4.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.4.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.5.2.2	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.6.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.6.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.8.2.1	10-Jun-2019	christos	Sync with HEAD
1.9.16.1	13-May-2021	thorpej	Sync with HEAD.
1.9.4.3	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.9.4.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.9.4.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.13.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.9	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.8	15-Dec-2021	msaitoh	Change DCB credit parameters. No functional change. - From FreeBSD ix-3.3.24. - No functional change because NetBSD doesn't support DCB.
1.7	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.6	04-Apr-2018	msaitoh	branches: 1.6.8; 1.6.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.5	06-Dec-2017	msaitoh	branches: 1.5.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.4	30-Aug-2017	msaitoh	branches: 1.4.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.3	01-Dec-2016	msaitoh	branches: 1.3.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb.h was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb.h was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.3.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.3.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.5.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.6.20.1	13-May-2021	thorpej	Sync with HEAD.
1.6.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.13	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.12	24-Dec-2021	msaitoh	branches: 1.12.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.11	15-Dec-2021	msaitoh	Remove dead code. No functional change. FreeBSD: 0b487fb4547ad1a939be0d523e555557c94b7cd1 DPDK: a6395d471e14e5a7432875dad8fb3533238c5167
1.10	10-Dec-2021	msaitoh	Add typecast for type mismatch. FreeBSD: 994dd6328c66fc277438ad51ed074f3c52096147 DPDK: d8e52b2cf771c31b523b46852fd86225b5a2c721
1.9	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.8	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.7	04-Apr-2018	msaitoh	branches: 1.7.8; 1.7.20; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.6	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.5	06-Dec-2017	msaitoh	branches: 1.5.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.4	30-Aug-2017	msaitoh	branches: 1.4.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.3	01-Dec-2016	msaitoh	branches: 1.3.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb_82598.c was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb_82598.c was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.3.8.5	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.3.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.3.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb_82598.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.5.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.7.20.1	13-May-2021	thorpej	Sync with HEAD.
1.7.8.3	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.7.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.7.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.12.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.8	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.7	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.6	04-Apr-2018	msaitoh	branches: 1.6.8; 1.6.20; 1.6.22; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.5	06-Dec-2017	msaitoh	branches: 1.5.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.4	30-Aug-2017	msaitoh	branches: 1.4.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.3	01-Dec-2016	msaitoh	branches: 1.3.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb_82598.h was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb_82598.h was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.3.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.3.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb_82598.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.5.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.6.22.1	31-May-2021	cjep	sync with head
1.6.20.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.6.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.12	15-Jul-2023	andvar	fix tripple "p" (ppp) to double "p" (pp) in some words in comments.
1.11	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.10	10-Dec-2021	msaitoh	Add typecast for type mismatch. FreeBSD: 994dd6328c66fc277438ad51ed074f3c52096147 DPDK: d8e52b2cf771c31b523b46852fd86225b5a2c721
1.9	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.8	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.7	04-Apr-2018	msaitoh	branches: 1.7.8; 1.7.20; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.6	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.5	06-Dec-2017	msaitoh	branches: 1.5.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.4	30-Aug-2017	msaitoh	branches: 1.4.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.3	01-Dec-2016	msaitoh	branches: 1.3.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb_82599.c was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb_82599.c was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.3.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.3.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb_82599.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.5.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.7.20.1	13-May-2021	thorpej	Sync with HEAD.
1.7.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.7.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.8	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.7	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.6	04-Apr-2018	msaitoh	branches: 1.6.8; 1.6.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.5	06-Dec-2017	msaitoh	branches: 1.5.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.4	30-Aug-2017	msaitoh	branches: 1.4.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.3	01-Dec-2016	msaitoh	branches: 1.3.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.2	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.1	05-Aug-2015	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; Add some files.
1.1.6.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.2	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.1	05-Aug-2015	snj	file ixgbe_dcb_82599.h was added on branch netbsd-7 on 2016-06-14 08:42:34 +0000
1.1.2.3	05-Dec-2016	skrll	Sync with HEAD
1.1.2.2	22-Sep-2015	skrll	Sync with HEAD
1.1.2.1	05-Aug-2015	skrll	file ixgbe_dcb_82599.h was added on branch nick-nhusb on 2015-09-22 12:05:59 +0000
1.3.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.3.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.3.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	30-Aug-2017	jdolecek	file ixgbe_dcb_82599.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.5.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.6.20.1	13-May-2021	thorpej	Sync with HEAD.
1.6.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.5	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.4	24-Dec-2021	msaitoh	branches: 1.4.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.3	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.2	04-Apr-2018	msaitoh	branches: 1.2.8; 1.2.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.6.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1.4.6	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_fdir.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_fdir.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.2.20.1	13-May-2021	thorpej	Sync with HEAD.
1.2.8.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.2.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.2.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.4.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.4	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.3	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.2	06-Dec-2018	msaitoh	branches: 1.2.4; 1.2.16; Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.8.1	10-Jun-2019	christos	Sync with HEAD
1.1.6.1	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.1.4.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_features.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_features.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.2.16.1	13-May-2021	thorpej	Sync with HEAD.
1.2.4.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.2.4.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.19	25-Jan-2022	msaitoh	Use atomic_{load,store}_relaxed() for event counters.
1.18	18-Jan-2022	msaitoh	Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.17	18-Jan-2022	msaitoh	Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf().
1.16	24-Dec-2021	msaitoh	Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.15	24-Dec-2021	msaitoh	Restore some mailbox related functions. No functional change. - Revert some part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. It's also part of FreeBSD 3f66b96d8658f8721e6b1c6f035611bec44675b9. - This change reduces diff against FreeBSD ix-3.3.14.
1.14	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.13	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.12	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.11	27-Jun-2019	msaitoh	branches: 1.11.2; 1.11.14; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.10	04-Apr-2018	msaitoh	branches: 1.10.2; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.9	06-Dec-2017	msaitoh	branches: 1.9.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.8	03-Oct-2017	msaitoh	Make mailbox statistic counters evcnt(9). Example: > ixv0 message TXs 23 0 misc > ixv0 message RXs 2911 0 misc > ixv0 ACKs 23 0 misc > ixv0 REQs 22 0 misc > ixv0 RSTs 0 0 misc
1.7	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.6	05-Dec-2016	msaitoh	branches: 1.6.8; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.5	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.3	02-Apr-2015	msaitoh	branches: 1.3.2; Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.3	05-Feb-2017	skrll	Sync with HEAD
1.1.30.2	05-Dec-2016	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.3.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.6.8.7	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.6.8.6	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.6.8.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6.8.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.6.8.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.6.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.6.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.9.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.10.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.11.14.1	13-May-2021	thorpej	Sync with HEAD.
1.11.2.4	02-Feb-2022	martin	Pull up the following revisions (requested by msaitoh in ticket #1424): sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.305 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixv.c 1.178 via patch Use atomic_{load,store}_relaxed() for event counters.
1.11.2.3	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.11.2.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.11.2.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.20	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.19	24-Dec-2021	msaitoh	branches: 1.19.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.18	24-Dec-2021	msaitoh	Restore some mailbox related functions. No functional change. - Revert some part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. It's also part of FreeBSD 3f66b96d8658f8721e6b1c6f035611bec44675b9. - This change reduces diff against FreeBSD ix-3.3.14.
1.17	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.16	24-Dec-2021	msaitoh	Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. - Sync with FreeBSD ix-3.3.18. - No functional change.
1.15	24-Dec-2021	msaitoh	Move some definitions. No functional change. Part of FreeBSD ix-3.3.18.
1.14	27-Jun-2019	msaitoh	branches: 1.14.2; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.13	04-Apr-2018	msaitoh	branches: 1.13.2; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.12	06-Dec-2017	msaitoh	branches: 1.12.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.11	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.10	02-Dec-2016	msaitoh	branches: 1.10.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.9	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.8	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.7	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.5	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.4	24-Apr-2015	msaitoh	branches: 1.4.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.3	05-Dec-2016	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.10.8.6	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.10.8.5	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.10.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.10.8.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.10.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.10.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.12.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.13.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.14.2.3	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.14.2.2	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.14.2.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.19.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.17	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.16	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.15	25-Jun-2020	msaitoh	branches: 1.15.6; Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.14	17-Apr-2020	msaitoh	No functional change: - modify comment - whitespace fix
1.13	01-Feb-2020	riastradh	branches: 1.13.4; Use atomic_load_acquire for FreeBSDish atomic_load_acq_uint shim.
1.12	21-Jan-2020	msaitoh	Fix the freeing code for some error paths. Found and tested by Patrick Welche.
1.11	20-Jan-2020	msaitoh	Free jumbo mem structure correctly. Found by yamaguchi@ using with LOCKDEBUG.
1.10	04-Sep-2019	msaitoh	branches: 1.10.2; printf -> device_printf
1.9	06-Dec-2018	msaitoh	branches: 1.9.4; Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.8	31-Jul-2018	msaitoh	Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.7	25-Apr-2018	msaitoh	branches: 1.7.2; Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet.
1.6	01-Jun-2017	chs	branches: 1.6.2; 1.6.8; remove checks for failure after memory allocation calls that cannot fail: kmem_alloc() with KM_SLEEP kmem_zalloc() with KM_SLEEP percpu_alloc() pserialize_create() psref_class_create() all of these paths include an assertion that the allocation has not failed, so callers should not assert that again.
1.5	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.3	04-Feb-2015	msaitoh	branches: 1.3.2; Fix mutex related problem reported by Uwe Toenjes in PR#49328: - Revert ixgbe_netbsd.c rev. 1.2 - make CORE_LOCK adaptive - Release spin lock while reinitializing the jumb buffer structure.
1.2	28-Jan-2015	christos	PR/49328: Uwe Toenjes: This driver allocates memory always with NOWAIT, except in the NetBSD-specific code. This happens from ioctl, and usually there is enough memory there since we just freed our resources and we are getting them back. We could even check and not do it again if sizes did not change, but this is how the rest of the code is structured so we follow suit. XXX: pullup-7
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.3	28-Aug-2017	skrll	Sync with HEAD
1.1.30.2	05-Dec-2016	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.3.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.6.8.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.6.8.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.6.8.1	02-May-2018	pgoyette	Synch with HEAD
1.6.2.7	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.6.2.6	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.6.2.5	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.6.2.4	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.6.2.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.6.2.2	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.6.2.1	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.7.2.4	21-Apr-2020	martin	Sync with HEAD
1.7.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.7.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.7.2.1	10-Jun-2019	christos	Sync with HEAD
1.9.4.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.9.4.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.9.4.2	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.9.4.1	05-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #180): sys/dev/pci/ixgbe/ixv.c: revision 1.130 sys/dev/pci/ixgbe/ixgbe.c: revision 1.207 sys/dev/pci/ixgbe/ixgbe.c: revision 1.208 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.55 sys/dev/pci/ixgbe/ixv.c: revision 1.129 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.10 printf -> device_printf Set IFM_1000_BX10 correctly. Use aprint_*() in the attach function.
1.10.2.2	29-Feb-2020	ad	Sync with head.
1.10.2.1	25-Jan-2020	ad	Sync with head.
1.13.4.1	20-Apr-2020	bouyer	Sync with HEAD
1.15.6.1	13-May-2021	thorpej	Sync with HEAD.
1.17	16-Sep-2022	knakahara	ixg(4) add an option for Tx to use deferred softint regardless of whether can get txq lock or not. That imporve (7%) and stabilize throughput. But that can cause latency degradation, so off by default. ok'ed by msaitoh@n.o.
1.16	25-Jan-2022	msaitoh	Fix compile error for non __HAVE_ATOMIC64_LOADSTORE arch.
1.15	25-Jan-2022	msaitoh	Use atomic_{load,store}_relaxed() for event counters.
1.14	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.13	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.12	09-Mar-2021	msaitoh	branches: 1.12.4; 1.12.6; Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.11	05-Mar-2019	msaitoh	branches: 1.11.4; 1.11.12; Centralize ETHER_ALIGN into net/if_ether.h. Note that this commit also changes if_upgt.c's ETHER_ALIGN from 0 to 2.
1.10	21-Dec-2018	msaitoh	Remove ETHERCAP_VLAN_HWFILTER.
1.9	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.8	25-Apr-2018	msaitoh	branches: 1.8.2; Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet.
1.7	08-Feb-2017	msaitoh	branches: 1.7.6; 1.7.12; Enable TX multiqueue by default on both ixg(4) and ixv(4).
1.6	01-Feb-2017	msaitoh	TX multiqueue. If you want to disable it, enable IXGBE_LEGACY_TX in ixgbe_netbsd.h
1.5	01-Dec-2016	msaitoh	branches: 1.5.2; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.3	24-Apr-2015	msaitoh	branches: 1.3.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.2	17-Apr-2014	christos	branches: 1.2.2; 1.2.4; CID/1008347, CID/1008348, CID/1008349: overflow before widen CID/1203198, CID/1203199: Uninitialized scalar variable
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.4	28-Aug-2017	skrll	Sync with HEAD
1.2.4.3	05-Feb-2017	skrll	Sync with HEAD
1.2.4.2	05-Dec-2016	skrll	Sync with HEAD
1.2.4.1	06-Jun-2015	skrll	Sync with HEAD
1.2.2.1	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.3.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.3.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.5.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.7.12.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7.12.1	02-May-2018	pgoyette	Synch with HEAD
1.7.6.6	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.7.6.5	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.7.6.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.7.6.3	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.7.6.2	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.7.6.1	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.8.2.1	10-Jun-2019	christos	Sync with HEAD
1.11.12.1	03-Apr-2021	thorpej	Sync with HEAD.
1.11.4.4	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.11.4.3	02-Feb-2022	martin	Pull up the following revisions (requested by msaitoh in ticket #1424): sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.305 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixv.c 1.178 via patch Use atomic_{load,store}_relaxed() for event counters.
1.11.4.2	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.11.4.1	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.12.6.1	31-May-2021	cjep	sync with head
1.12.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.6	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.5	10-Dec-2021	msaitoh	branches: 1.5.4; ifdef D(ebug)? From FreeBSD ix-3.3.14.
1.4	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.3	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.2	27-Jun-2019	msaitoh	branches: 1.2.2; 1.2.14; Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.8; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.8.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.4.6	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_netmap.c was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_netmap.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.2.14.1	13-May-2021	thorpej	Sync with HEAD.
1.2.2.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.2.2.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.2.2.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.5.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.3	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.2	30-Apr-2021	msaitoh	branches: 1.2.14; Add NetBSD RCS IDs. No functional change.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.14; 1.1.26; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.26.1	13-May-2021	thorpej	Sync with HEAD.
1.1.14.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.14.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.4	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_netmap.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_netmap.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.2.14.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.9	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.8	24-Dec-2021	msaitoh	branches: 1.8.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.7	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.6	17-Dec-2019	msaitoh	branches: 1.6.12; Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@.
1.5	16-Dec-2019	msaitoh	No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.4	04-Apr-2018	msaitoh	branches: 1.4.2; 1.4.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.3	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.2	30-Aug-2017	msaitoh	branches: 1.2.2; 1.2.4; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1	02-Dec-2016	msaitoh	branches: 1.1.2; 1.1.4; 1.1.12; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.1.12.7	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.12.6	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.12.5	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.12.4	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.1.12.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.1.12.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.1.12.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.1.4.1	02-Dec-2016	pgoyette	file ixgbe_osdep.c was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.1.2.2	05-Dec-2016	skrll	Sync with HEAD
1.1.2.1	02-Dec-2016	skrll	file ixgbe_osdep.c was added on branch nick-nhusb on 2016-12-05 10:55:17 +0000
1.2.4.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.2.4.1	15-Mar-2018	pgoyette	Synch with HEAD
1.2.2.2	03-Dec-2017	jdolecek	update from HEAD
1.2.2.1	30-Aug-2017	jdolecek	file ixgbe_osdep.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.4.8.5	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.4.8.4	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.4.8.3	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.4.8.2	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.4.8.1	24-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #580): sys/dev/pci/ixgbe/ixgbe.c: revision 1.217 sys/dev/pci/ixgbe/ixgbe.c: revision 1.218 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.6 sys/dev/pci/ixgbe/ixv.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.25 Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@. Add missing core lock in ixgbe_handle_mod().
1.4.2.1	08-Apr-2020	martin	Merge changes from current as of 20200406
1.6.12.1	13-May-2021	thorpej	Sync with HEAD.
1.8.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.31	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.30	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.29	19-Aug-2021	msaitoh	Use m_adj(ETHER_ALIGN) more. Tested by me (amd64,aarch64) and rin (alpha). - Align with ETHER_ALIGN everywhere where mbuf is allocated. - Remove extra setting of M_PKTHDR. No functional change. - Add comment.
1.28	01-Sep-2020	msaitoh	Fix a panic on shutdown on a machine which use the recovery mode timer. The recovery mode timer is first issued by the callout and it schedule the workqueue. The workqueue then reschedule the callout. It's hard to stop both of them without race only with callout_stop() and workqueue_wait. To solve this problem. add new "detaching" flag and use it. The situation is almost the same as schedule_wqs_ok for the local_timer's callout and workqueue, but the difference is that the local_timer isn't required to run if the interface is not up. If it's not important to prevent running timer while !IFF_UP, the flag can be integrated into one.
1.27	25-Jun-2020	msaitoh	Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.26	11-Jun-2020	msaitoh	Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. Not tested on big endian machine.
1.25	17-Dec-2019	msaitoh	Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@.
1.24	16-Dec-2019	msaitoh	No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.23	31-Jul-2018	msaitoh	branches: 1.23.6; Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.22	25-Apr-2018	msaitoh	branches: 1.22.2; Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet.
1.21	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.20	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.19	06-Dec-2017	msaitoh	branches: 1.19.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.18	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.17	18-Jan-2017	msaitoh	branches: 1.17.6; Now we can use multiqueue. It's default on ixg(4). Not yet for ixv(4).
1.16	05-Dec-2016	msaitoh	branches: 1.16.2; - Fix wol_enable, fc and wufc sysctls to work correctly. - Fix RCS Id.
1.15	05-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.14	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.13	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.12	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.11	25-Nov-2016	msaitoh	Remove #ifdef NETBSD_MSI_OR_MSIX to be simple. ixv(4) isn't attached if __HAVE_PCI_MSI_MSIX isn't set because MSI-X must be used.
1.10	13-Aug-2015	msaitoh	branches: 1.10.2; - Add MSI/MSI-X support. The multiqueue function is not supported yet. - Make ixv.c compilable. _NOT_TESTED_YET_
1.9	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.8	05-Aug-2015	msaitoh	Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic.
1.7	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.6	14-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r243716: - A lot of bugfixes. Some of them are realted to multi queue and those have not affected in NetBSD because we have not used it yet. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Improve performance.
1.5	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.4	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.3	18-Mar-2014	riastradh	branches: 1.3.4; 1.3.6; Merge riastradh-drm2 to HEAD.
1.2	17-Jul-2013	soren	Fix ixb(4) build.
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.24; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.24.1	23-Jul-2013	riastradh	sync with HEAD
1.1.16.1	28-Aug-2013	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.3.6.5	05-Feb-2017	skrll	Sync with HEAD
1.3.6.4	05-Dec-2016	skrll	Sync with HEAD
1.3.6.3	22-Sep-2015	skrll	Sync with HEAD
1.3.6.2	06-Jun-2015	skrll	Sync with HEAD
1.3.6.1	06-Apr-2015	skrll	Sync with HEAD
1.3.4.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.3.4.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.3.4.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.10.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.10.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.16.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.17.6.9	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.17.6.8	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.17.6.7	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.17.6.6	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.17.6.5	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.17.6.4	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.17.6.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.17.6.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.17.6.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.19.2.4	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.19.2.3	02-May-2018	pgoyette	Synch with HEAD
1.19.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.19.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.22.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.22.2.1	10-Jun-2019	christos	Sync with HEAD
1.23.6.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.23.6.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.23.6.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.23.6.2	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.23.6.1	24-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #580): sys/dev/pci/ixgbe/ixgbe.c: revision 1.217 sys/dev/pci/ixgbe/ixgbe.c: revision 1.218 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.6 sys/dev/pci/ixgbe/ixv.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.25 Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@. Add missing core lock in ixgbe_handle_mod().
1.29	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.28	15-Dec-2021	msaitoh	Match X550_PHY_ID correctly on X550. - Sync with FreeBSD ix-3.3.18. - phy_id's revision field is cleared, so use new X550_PHY_ID. - Before this commit, phy_id was set to ixgbe_phy_cu_unknown on X550. Now it's set to ixgbe_phy_aq.
1.27	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.26	10-Dec-2021	msaitoh	Move increments after evaluations. FreeBSD: dc11ba4eb3fe5cce615f361de83e85e07005ca24 or ix-3.3.14 DPDK: 390445ec30b4c52a3d2887c3d2a202d9cf37ea8e The retry variable was being incremented before it was evaluated by the subsequent conditional against the maximum retries to figure out which message to print. So we'll move the increment op to the end.
1.25	10-Dec-2021	msaitoh	Remove unnecessary return value check. FreeBSD: 3a89005394bc5d82ce9b6baa9e7f8dee362354ae DPDK: 4b0ee6529b7897c2a08dd56669f07ac1f46a8474
1.24	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.23	31-Aug-2020	msaitoh	branches: 1.23.6; If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.22	31-Aug-2020	msaitoh	Fix 82598 SFP+ problems. On 82598, SFP+'s MOD_ABS isn't connected to the MAC's GPIO pin, so we can't call ixgbe_sfp_cage_full(). Always issue TASK_MOD from ixgbe_handle_timer() on 82598. Fix ixgbe_identify_sfp_module_generic() for ixgbe_phy_nl. In the driver, hw->phy.type sometimes be compared with ixgbe_phy_nl. In ixgbe_identify_sfp_module_generic(), hw->phy.type may be overridden with another value. For ixgbe_phy_nl, some code don't override phy.type but others were not. Make it consistently keep ixgbe_phy_nl. This change fixes a problem that ixgbe_is_sfp() change the return value true to false when any SFP+ devices are connected to the cage on 82598 and never recover from it. Don't schedule MSF(multi speed fiber) task from ixgbe_handle_mod() on 82598. This task is only for devices which support multi speed fiber and 82598 doesn't support it. Before ixgbe.c rev. 1.237, ixgbe_handle_mod() isn't called on 82598 because 82598 has no SFP+ module insertion/removal interrupt. ixgbe.c rev. 1.237 changed to call the function via timer on 82598. This change fixes a bug that 82598 DA interface's link flaps.
1.21	17-Apr-2020	msaitoh	No functional change: - modify comment - whitespace fix
1.20	23-Dec-2019	msaitoh	branches: 1.20.6; Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.19	16-Dec-2019	msaitoh	No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.18	06-Dec-2018	msaitoh	branches: 1.18.4; Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.17	04-Apr-2018	msaitoh	branches: 1.17.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.16	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.15	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.14	06-Dec-2017	msaitoh	branches: 1.14.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.13	07-Nov-2017	msaitoh	Fix a bug that X550T(1) didn't linkup if it forces 100BaseTX-FDX. Popular switches and OSes don't use auto-negotiation if the media is 100BASE-TX (and 10BASE-T). Do the same thig. Another fix is required for Denverton's _T device which use firmware.
1.12	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.11	19-Jan-2017	msaitoh	branches: 1.11.6; Reduce diff against FreeBSD. No functional change.
1.10	05-Dec-2016	msaitoh	branches: 1.10.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.9	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.8	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.7	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.6	05-Aug-2015	msaitoh	branches: 1.6.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.5	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.4	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.3	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.2	17-Apr-2014	christos	branches: 1.2.2; 1.2.4; CIT 744275: Increase max_retry so we retry once.
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.5	05-Feb-2017	skrll	Sync with HEAD
1.2.4.4	05-Dec-2016	skrll	Sync with HEAD
1.2.4.3	22-Sep-2015	skrll	Sync with HEAD
1.2.4.2	06-Jun-2015	skrll	Sync with HEAD
1.2.4.1	06-Apr-2015	skrll	Sync with HEAD
1.2.2.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.2.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.2.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.6.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.6.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.10.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.11.6.9	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.11.6.8	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.11.6.7	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.11.6.6	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.11.6.5	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.11.6.4	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.11.6.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.11.6.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.11.6.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.14.2.3	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.14.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.14.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.17.2.3	21-Apr-2020	martin	Sync with HEAD
1.17.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.17.2.1	10-Jun-2019	christos	Sync with HEAD
1.18.4.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.18.4.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.18.4.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.18.4.2	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.18.4.1	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.20.6.1	20-Apr-2020	bouyer	Sync with HEAD
1.23.6.1	13-May-2021	thorpej	Sync with HEAD.
1.13	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.12	31-Aug-2020	msaitoh	If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.11	04-Apr-2018	msaitoh	branches: 1.11.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.10	06-Dec-2017	msaitoh	branches: 1.10.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.9	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.8	02-Dec-2016	msaitoh	branches: 1.8.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.7	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.5	05-Aug-2015	msaitoh	branches: 1.5.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.4	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.4	05-Dec-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.8.8.3	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.8.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.8.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.10.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.11.8.1	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.6	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.5	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.4	04-Apr-2018	msaitoh	branches: 1.4.8; 1.4.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.3	16-Feb-2018	knakahara	branches: 1.3.2; Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o.
1.2	06-Dec-2017	msaitoh	branches: 1.2.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_rss.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.2.2.6	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.2.2.5	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.2.2.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.2.2.3	26-Feb-2018	snj	Pull up following revision(s) (requested by knakahara in ticket #567): distrib/sets/lists/comp/mi: 1.2182-1.2183 sys/dev/pci/if_wm.c: 1.564 sys/dev/pci/ixgbe/ixgbe.c: 1.122 sys/dev/pci/ixgbe/ixgbe_rss.h: 1.3 sys/dev/pci/ixgbe/ixv.c: 1.78 sys/net/Makefile: 1.35-1.36 sys/net/files.net: 1.15 sys/net/rss_config.c: 1.1 sys/net/rss_config.h: 1.1 Introduce very simple Receive Side Scaling (RSS) utility. ok by msaitoh@n.o. -- Apply RSS utility to wm(4). ok by msaitoh@n.o. -- Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o. -- Fix build failure, sorry. -- Currently, it is not necessary to install rss_config.h. Pointed out by msaitoh@n.o.
1.2.2.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.2.2.1	06-Dec-2017	snj	file ixgbe_rss.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.3.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.4.20.1	13-May-2021	thorpej	Sync with HEAD.
1.4.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.4.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.6	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.5	24-Dec-2021	msaitoh	branches: 1.5.4; Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.4	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.3	04-Apr-2018	msaitoh	branches: 1.3.8; 1.3.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.2	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.1	30-Aug-2017	msaitoh	branches: 1.1.2; 1.1.4; 1.1.6; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.1.6.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.1.6.1	15-Mar-2018	pgoyette	Synch with HEAD
1.1.4.7	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.1.4.6	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.1.4.5	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.1.4.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.1.4.3	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.1.4.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.1.4.1	30-Aug-2017	snj	file ixgbe_sriov.h was added on branch netbsd-8 on 2017-12-21 19:28:54 +0000
1.1.2.2	03-Dec-2017	jdolecek	update from HEAD
1.1.2.1	30-Aug-2017	jdolecek	file ixgbe_sriov.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.3.20.1	13-May-2021	thorpej	Sync with HEAD.
1.3.8.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.3.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.3.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.5.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.62	15-Nov-2023	msaitoh	ixgbe: Clear the WTHRESH bit field before writing it.
1.61	11-Oct-2023	msaitoh	ixg(4): Print DEVICE_CAPS register.
1.60	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.59	06-Oct-2023	msaitoh	ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz.
1.58	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.57	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.56	15-May-2023	msaitoh	Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.55	24-Dec-2021	msaitoh	branches: 1.55.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.54	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.53	24-Dec-2021	msaitoh	Move some definitions. No functional change. Part of FreeBSD ix-3.3.18.
1.52	15-Dec-2021	msaitoh	Match X550_PHY_ID correctly on X550. - Sync with FreeBSD ix-3.3.18. - phy_id's revision field is cleared, so use new X550_PHY_ID. - Before this commit, phy_id was set to ixgbe_phy_cu_unknown on X550. Now it's set to ixgbe_phy_aq.
1.51	10-Dec-2021	msaitoh	Add some unused macros. Sync with FreeBSD ix-3.3.14.
1.50	01-Nov-2021	andvar	fix typos, mainly in words minimum and maximum, but also few others.
1.49	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.48	26-Dec-2020	msaitoh	branches: 1.48.4; 1.48.6; Disable some interrupt in ixgbe_{legacy_irq,msix_admin}() to prevent log spam.
1.47	26-Dec-2020	msaitoh	Disable/enable the OTHER interrupts correctly. The OTHER interrupt was not blocked correctly when MSI-X is used. ixgbe.c rev. 1.260 added new mutex to avoid the race but it didn't disable the interrupt itself. Calling ixgbe_enable_intr() enables all interrupts, so it's not good to call it when some interrupt sources should not be enabled (e.g.: calling ixgbe_enable_intr() in ixgbe_handle_admin() enables queue interrupt). IXGBE_REQUEST_TASK_NEED_ACKINTR doesn't work as expected because ixgbe_handle_admin() can't know which task is enqueued from the interrupt context and can't re-enable a specific EIMS bit. Solve the above three problems by the following two changes: - MSI-X: Disable the OTHER interrupts in the biginning of ixgbe_msix_admin(). - Set mask bits correctly at the end of ixgbe_legacy_irq() and ixgbe_msix_admin() using with eim_orig, eims_enable and eims_disable. - Remove IXGBE_REQUEST_TASK_NEED_ACKINTR and add IXGBE_REQUEST_TASK_{MOD,MSF}_WOI.
1.46	11-Dec-2020	msaitoh	Don't use EIMC_OTHER bit because it's read only other than 82598. Documents say: 82598: All of bit 31(OTHER bit) of EIxx are reserved. In reality, at least EIMS_OTHER and EIMC_OTHER exist and the OTHER interrupt doesn't work without EIMS_OTHER. Other than 82598: EIMS_OTHER is read only and EIMC_OTHER doesn't exist. If one of bit 29..16 is set, EIMS_OTHER is set to 1 (Note that bit 30(TCP timer isn't included)). Even if write bit 31 of EIMC to 1, it's ignored (EIMS_OTHER doesn't set). We introduced new spin mutex in ixgbe.c rev. 1.260, so it's OK to remove EIMC_OTHER stuff. We already set EIMS_OTHER in if_init(), so keep it for 82598. No functional change other than 82598. Another solution is to control bit 30..16 directly to mask/unmask interrupt instead of the mutex. TODO: Some MSI-X interrupt(LSC, module insertion/removal etc.)'s mask/unmask code between ixgbe_msix_admin() and ixgbe_handle_admin() may be wrong. It'll be fixed later.
1.45	31-Aug-2020	msaitoh	branches: 1.45.2; If an SFP+ module is not inserted, don't try to access SFP+ EEPROM. This change eliminate long timeout. Reduce code duplication using with ixgbe_sfp_cage_full(hw).
1.44	23-Dec-2019	msaitoh	Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.43	20-Sep-2019	msaitoh	- Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - If a multicast entry has range, use ALLMULTI like others. - Remove ixv_set_promisc() and use ixv_set_multi(). And then, rename _set_multi() to _set_rxfilter(). Same as ixgbe.c. - The promisc mode can't be enabled if the PF is not in promisc mode. Identify that state and report it as "the PF may not in promisc mode" (though it might not be perfect).
1.42	12-Sep-2019	msaitoh	Add support ALLMULTI on ixv(4): - Negotiate API version up to 1.3. - On linux's PF driver implementation, the PF replies VF's XCAST_MODE_ALLMULTI message not with NACK but with ACK even if the virtual function is NOT marked "trust" and act as XCAST_MODE_"MULTI". If ixv(4) simply check the return vaule of update_xcast_mode(XCAST_MODE_ALLMULTI), SIOCSADDMULTI success and the user may have trouble with some addresses. Fortunately, the Linux's PF driver's "ACK" message has not XCAST_MODE_"ALL"MULTI but XCAST_MODE_MULTI, so we can check this state by checking if the send message's argument and the reply message's argument are different. - Noy yet for PROMISC.
1.41	24-Jul-2019	msaitoh	branches: 1.41.2; Fix some problems found by KUBSan: - Set TQSM and RQSM (TX/RX queue statistics mapping) correctly. - Set IVAR (Interrupt Vector Allocation Register) correctly in ixgbe_set_ivar(). - Define IXGBE_KRM_LINK_CTRL_1_TETH_AN_RESTART correctly. This macro is used in ixgbe_restart_an_internal_phy_x550em(). - Define IXGBE_KRM_TX_COEFF_CTRL_1_OVRRD_EN correctly. This macro is used in ixgbe_setup_ixfi_x550em_x(). - Define IXGBE_SB_IOSF_CTRL_BUSY correctly. This macro is used in ixgbe_iosf_wait() (X550EM only).
1.40	02-Jul-2019	msaitoh	Sync with FreeBSD ix-3.3.10 part 1. No functional change in this part: - "(u64)1" -> "1ULL" - Add some not-yet-used register definitions.
1.39	27-Jun-2019	msaitoh	Sync with FreeBSD ix-3.3.8 part 1. No functional change in this part: - Move ixgbe_toggle_txdctl() to ixgbe_common.c and modify a bit. No functional change because this function is currently used for SR-IOV and it's not used in NetBSD. - Some modification to match the latest netmap API. - Modify ixgbe_hic_unlocked(). No functional change because neither IXGBE_HOST_INTERFACE_APPLY_UPDATE_CMD(0x38) nor IXGBE_HOST_INTERFACE_SHADOW_RAM_READ_CMD(0x31) are used. - Add ixgbe_clear_mbx(). No functional change because this function is not used yet. - Add some not-yet-used register definitions. - Whitespace fixes.
1.38	05-Mar-2019	msaitoh	NetBSD currently uses traffic class 0 only. Other traffic classes aren't used yet. When IXGBE_TC_COUNTER_NUM is set to lower than IXGBE_DCB_MAX_TRAFFIC_CLASS (e.g. 1), other traffic classes' counters are not used. It means we don't generate evcnt for them and don't add the values in ixgbe_update_stats_counters().
1.37	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.36	06-Jul-2018	msaitoh	Add force_10_100_autonego sysctl. The default vaule is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while.
1.35	11-Jun-2018	msaitoh	branches: 1.35.2; Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*).
1.34	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.33	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.32	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.31	20-Feb-2018	msaitoh	branches: 1.31.2; - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8
1.30	06-Dec-2017	msaitoh	Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.29	04-Oct-2017	msaitoh	Add 2.5GBASE-T and 5GBASE-T support. (forgotten in the last commit)
1.28	03-Oct-2017	msaitoh	Make mailbox statistic counters evcnt(9). Example: > ixv0 message TXs 23 0 misc > ixv0 message RXs 2911 0 misc > ixv0 ACKs 23 0 misc > ixv0 REQs 22 0 misc > ixv0 RSTs 0 0 misc
1.27	14-Sep-2017	msaitoh	Quick hack to fix ixv(4)'s hardware counter. ixgbe_rx_checksum() passes "struct ixgbe_hw_stats *" even if its a virtual function. The layout between ixgbe_hw_stats and ixgbevf_hw_stats are different, so ixgbe_rx_checksum() wrote wrong area if a device is a virtual function. To fix this problem with small change, move the location of ipcs, ipcs_bad, l4cs and l4cs_bad of struct ixgbe(vf)_hw_stats to the same location at the biggining of the struct.
1.26	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.25	28-Jun-2017	msaitoh	Add IXGBE_TIPG_IPGT_MASK and IXGBE_PAP_PACE_MASK for packet pacing.
1.24	27-Jun-2017	msaitoh	Add some register definitions for X550 and newer.
1.23	23-Jun-2017	msaitoh	Remove unused "cached_autoc". It was removed in ixgbe_82599.c rev.1.12 (FreeBSD r280182).
1.22	02-Jun-2017	msaitoh	branches: 1.22.2; - Print NVM Image Version, PHY FW Revision, NVM Map version and OEM NVM Image version - s/TrackID/ETrackID/
1.21	26-May-2017	msaitoh	Count Bad SFD (Start Frame Delimiter). X550 and newer have this register.
1.20	18-May-2017	msaitoh	- Count "Total Packets Missed" by evcnt(9) - Call evcnt_attach_dynamic()/detach() for stats->mngpdc (Management Packets Dropped). - Make stats->rnbc (Receive No Buffers) evcnt(9). Only 82598 has this register. - Set RQSMR, TQSM (or TQSMR on 82598) to count qprc, qptc, qbrc, qbtc and qprdc in each queue counter. Without this change, all queue's counts are counted in queue 0's. Even if with this change, only MPC[0] is counted and other MPCs are not counted.
1.19	24-Feb-2017	msaitoh	branches: 1.19.4; Print mac.type and the TrackID.
1.18	19-Jan-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.17	05-Dec-2016	msaitoh	branches: 1.17.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.16	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.15	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.14	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.13	13-Aug-2015	msaitoh	branches: 1.13.2; Reduce diff against FreeBSD r280181.
1.12	13-Aug-2015	msaitoh	- Add MSI/MSI-X support. The multiqueue function is not supported yet. - Make ixv.c compilable. _NOT_TESTED_YET_
1.11	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.10	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.9	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.8	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.7	24-Feb-2015	msaitoh	Add 82599EN_SFP and 82599_SFP_SF_QP. Not tested.
1.6	24-Feb-2015	msaitoh	Change macro name IXGBE_DEV_ID_82599_SFP_DELL to IXGBE_DEV_ID_82599_SFP_SF2. Same as OpenBSD and Linux. No binary change.
1.5	24-Feb-2015	msaitoh	Enclose macro argument correctly. Same as *BSDs. No binary change.
1.4	24-Feb-2015	msaitoh	Modify comments to make difference among *BSDs small. No binary change.
1.3	24-Feb-2015	msaitoh	Change tabs and spaces to make difference among *BSDs small. No binary change.
1.2	08-Apr-2014	christos	branches: 1.2.2; 1.2.4; Add dell's flavor (from uwe, in current-users@)
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; 1.1.26; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.26.1	10-Aug-2014	tls	Rebase.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.4.6	28-Aug-2017	skrll	Sync with HEAD
1.2.4.5	05-Feb-2017	skrll	Sync with HEAD
1.2.4.4	05-Dec-2016	skrll	Sync with HEAD
1.2.4.3	22-Sep-2015	skrll	Sync with HEAD
1.2.4.2	06-Jun-2015	skrll	Sync with HEAD
1.2.4.1	06-Apr-2015	skrll	Sync with HEAD
1.2.2.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.2.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.2.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.2.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.13.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.13.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.17.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.19.4.1	19-May-2017	pgoyette	Resolve conflicts from previous merge (all resulting from $NetBSD keywork expansion)
1.22.2.22	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1933: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.22.2.21	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.22.2.20	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.22.2.19	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.22.2.18	21-Jun-2023	martin	Pullup the following revisions, requested by msaitoh in ticket #1828: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.22.2.17	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.22.2.16	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.22.2.15	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.22.2.14	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1708: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.22.2.13	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.22.2.12	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.22.2.11	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.22.2.10	01-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1313: sys/dev/pci/ixgbe/ixgbe_type.h 1.41 sys/dev/pci/ixgbe/ixgbe.c 1.194,1.197-1.199 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.24 sys/dev/pci/ixgbe/ixv.c 1.121,1.124-1.125 - Avoid undefined behavior of interrupt vector setting. - Avoid undefined behavior of TX/RX queue statistics calculation. - Avoid undefined behavior of X550EM's PHY accesses. - Avoid undefined behavior of Flow Control Transmit Timer setting. - Don't call {ixgbe,ixv}_stop() twice while detaching.
1.22.2.9	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.22.2.8	01-Apr-2019	martin	Pull up the following, requested by msaitohin ticket #1225: sys/dev/pci/ixgbe/ixgbe.c 1.175-1.178 sys/dev/pci/ixgbe/ixv.c 1.110-1.111 via patch - NetBSD currently uses traffic class 0 only. Other traffic classes aren't used yet. When IXGBE_TC_COUNTER_NUM is set to lower than IXGBE_DCB_MAX_TRAFFIC_CLASS (e.g. 1), other traffic classes' counters are not used. It means we don't generate evcnt for them and don't add the values in ixgbe_update_stats_counters(). - It's not required to calculate unused queues' statistics. - Fix a bug that the VLAN HW tagging function is not correctly disabled when all vlan is detached. - Fix a bug that VLAN HW tagging function is not correctly controlled on 82598. - Control VLAN HW filter function correctly. Note that currently VLAN HW filter function doesn't work because NetBSD doesn't support it yet. - Don't clear IXGBE_VLNCTRL_CFIEN bit When ETHERCAP_VLAN_HWFILTER is set. I think it's not required (and Linux doesn't do it). This change has no effect to NetBSD because ETHERCAP_VLAN_HWFILTER is not supported yet.
1.22.2.7	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.22.2.6	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #935): sys/dev/pci/ixgbe/ixgbe.c: revision 1.162 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.35 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.16 Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*). -- Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.22.2.5	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.22.2.4	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.22.2.3	26-Feb-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #592): sys/dev/pci/ixgbe/ixv.c: revision 1.79 sys/dev/pci/ixgbe/ixgbe.h: revision 1.30 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.31 sys/dev/pci/ixgbe/ixgbe.c: revision 1.120 sys/dev/pci/ixgbe/ixgbe.c: revision 1.121 sys/dev/pci/ixgbe/ixgbe.c: revision 1.123 sys/dev/pci/ixgbe/ixgbe.c: revision 1.124 sys/dev/pci/ixgbe/ixgbe.c: revision 1.125 sys/dev/pci/ixgbe/ixgbe.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.80 sys/dev/pci/ixgbe/ixv.c: revision 1.81 CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558). Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8 Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv) - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8 Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8 Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o. - Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function. Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.22.2.2	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.22.2.1	05-Aug-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #175): sys/dev/pci/ixgbe/ix_txrx.c: 1.25-1.27 sys/dev/pci/ixgbe/ixgbe.c: 1.91-1.95 sys/dev/pci/ixgbe/ixgbe_type.h: 1.23-1.25 sys/dev/pci/ixgbe/ixgbe_api.c: 1.16 sync ixg(4) up to ixgbe.c rev. 1.95: - Increase total number of RX buffers on multiqueue. - Fix the partial chain check in ixgbe_rx_discard(). This bug was added in ixgbe.c rev. 1.33. - Fix a double free in ixgbe_rxeof(). - Print verbose output in ixgbe_update_link_status() correctly on 5Gbps and 2.5Gbps (NBASE-T) for X550T[12] and newer. - Change hw.ixgN.ts to hw.ixgN.thermal_test. Same as FreeBSD. - Fix a bug that ifconfig ixgN media 1000baseT and 10Gbase-T advertised additional unwanted speeds. - Print PHY ID. - Remove unused variable.
1.31.2.5	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.31.2.4	28-Jul-2018	pgoyette	Sync with HEAD
1.31.2.3	25-Jun-2018	pgoyette	Sync with HEAD
1.31.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.31.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.35.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.35.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.35.2.1	10-Jun-2019	christos	Sync with HEAD
1.41.2.12	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1792: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.41.2.11	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.41.2.10	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.41.2.9	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.41.2.8	21-Jun-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1647: sys/dev/pci/ixgbe/ixgbe.c 1.325-1.326 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.44 sys/dev/pci/ixgbe/ixgbe_type.h 1.56 - PCI device ID 0x15c8 also uses X557-AT PHY, so create the thermal sensor sysctl for it, too. - Count the number of link down events in the MAC using with LINK_DN_CNT register.
1.41.2.7	02-Jun-2022	martin	Pull up the following revisions, all via patch, requested by msaitoh in ticket #1459: sys/dev/pci/ixgbe/ixgbe.c 1.261,1.265-1.268,1.273,1.275-1.277, 1.312,1.316-1.319 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.182 - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Fix typos in comment. - KNF.
1.41.2.6	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.41.2.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.41.2.4	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1374: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.41.2.3	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.41.2.2	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.41.2.1	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.45.2.2	03-Jan-2021	thorpej	Sync w/ HEAD.
1.45.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.48.6.1	31-May-2021	cjep	sync with head
1.48.4.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.55.4.6	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #563): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.110 sys/dev/pci/ixgbe/ixgbe.c: revision 1.345 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.111 sys/dev/pci/ixgbe/ixgbe.c: revision 1.346 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.112 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.62 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.113 sys/dev/pci/ixgbe/ixgbe.c: revision 1.348 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.114 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.115 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.116 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.105 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.106 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.107 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.108 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.109 sys/dev/pci/ixgbe/ixv.c: revision 1.193 sys/dev/pci/ixgbe/ixv.c: revision 1.195 sys/dev/pci/ixgbe/ixv.c: revision 1.196 sys/dev/pci/ixgbe/ixgbe.h: revision 1.94 sys/dev/pci/ixgbe/ixgbe.h: revision 1.95 sys/dev/pci/ixgbe/ixgbe.h: revision 1.96 sys/dev/pci/ixgbe/ixgbe.h: revision 1.97 sys/dev/pci/ixgbe/ixgbe.h: revision 1.98 ixgbe: Fix comment. No functional change. ixgbe: Whitespace. No functional change. ixgbe(4): Move assignment of TXD. NFCI. ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor. ixv(4): Remove unused IFF_OACTIVE. No functional change. ixgbe: Clear the WTHRESH bit field before writing it. ixgbe: Modify for the readability. No functional change. ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof(). ixgbe: Update if_opackets outside the loop in ixgbe_txeof(). ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof(). ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO). ixgbe: Remove unused to reduce the size of struct rx_ring. ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring. ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.55.4.5	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #420): sys/dev/pci/ixgbe/ixgbe.c: revision 1.340 sys/dev/pci/ixgbe/ixgbe.c: revision 1.341 sys/dev/pci/ixgbe/ixgbe.c: revision 1.342 sys/dev/pci/ixgbe/ixgbe.c: revision 1.343 sys/dev/pci/ixgbe/ixgbe.c: revision 1.344 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.61 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.104 sys/dev/pci/ixgbe/ixv.c: revision 1.190 sys/dev/pci/ixgbe/ixv.c: revision 1.191 sys/dev/pci/ixgbe/ixv.c: revision 1.192 sys/dev/pci/ixgbe/ixgbe.h: revision 1.93 ixg(4): Print DEVICE_CAPS register. ixgbe: Whitespace. No functional change. ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl. ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. ixgbe: Whitespace. No functional change.
1.55.4.4	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.55.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.55.4.2	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.55.4.1	21-Jun-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #202): sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.56 sys/dev/pci/ixgbe/ixgbe.c: revision 1.325 sys/dev/pci/ixgbe/ixgbe.c: revision 1.326 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.44 Use thermal sensor code for IXGBE_DEV_ID_X550EM_A_10G_T, too. PCI device ID 0x15c8 also use X557-AT PHY, so create the thermal sensor sysctl for it, too. Count the number of link down events in the MAC using with LINK_DN_CNT. - Add new event counter "link_dn_cnt" to count the number of link down events in the MAC. - The LINK_DN_CNT register (at 0x0403c) is described only in the Denverton's datasheet, so use it only on ixgbe_mac_X550EM_a.
1.33	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.32	06-Oct-2023	msaitoh	ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27.
1.31	24-Dec-2021	msaitoh	branches: 1.31.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.30	24-Dec-2021	msaitoh	Whitespace fix. Remove old comment. No functional change. - From FreeBSD ixv-1.5.22.
1.29	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.28	24-Dec-2021	msaitoh	Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. - Sync with FreeBSD ix-3.3.18. - No functional change.
1.27	30-Apr-2021	msaitoh	Add missing __KERNEL_RCSID().
1.26	20-Jan-2021	msaitoh	branches: 1.26.4; The max number of queue(pair) is not 7 but 8. Inspired by DPDK.
1.25	30-Nov-2020	msaitoh	s/ we we / we /
1.24	22-Jul-2020	msaitoh	branches: 1.24.2; s/reseting/resetting/
1.23	05-Feb-2020	msaitoh	No functional change: - Fix typos. - Remove extra newline.
1.22	20-Sep-2019	msaitoh	branches: 1.22.2; - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - If a multicast entry has range, use ALLMULTI like others. - Remove ixv_set_promisc() and use ixv_set_multi(). And then, rename _set_multi() to _set_rxfilter(). Same as ixgbe.c. - The promisc mode can't be enabled if the PF is not in promisc mode. Identify that state and report it as "the PF may not in promisc mode" (though it might not be perfect).
1.21	12-Sep-2019	msaitoh	Add support ALLMULTI on ixv(4): - Negotiate API version up to 1.3. - On linux's PF driver implementation, the PF replies VF's XCAST_MODE_ALLMULTI message not with NACK but with ACK even if the virtual function is NOT marked "trust" and act as XCAST_MODE_"MULTI". If ixv(4) simply check the return vaule of update_xcast_mode(XCAST_MODE_ALLMULTI), SIOCSADDMULTI success and the user may have trouble with some addresses. Fortunately, the Linux's PF driver's "ACK" message has not XCAST_MODE_"ALL"MULTI but XCAST_MODE_MULTI, so we can check this state by checking if the send message's argument and the reply message's argument are different. - Noy yet for PROMISC.
1.20	12-Sep-2019	msaitoh	- Make SIOCADDMULTI returns ENOSPC when the Ethenet multicast address list exceeds the limit. - State "Ethernet" multicast address. - Use macro.
1.19	11-Sep-2019	msaitoh	Print message when the number of multicast addresses exceeded the limit (30).
1.18	02-Jul-2019	msaitoh	branches: 1.18.2; Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf().
1.17	04-Apr-2018	msaitoh	branches: 1.17.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.16	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.15	06-Dec-2017	msaitoh	branches: 1.15.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.14	13-Oct-2017	msaitoh	Show 2.5GBASE-T and 5GBASE-T correctly on ixv(4). Tested on a virtual function on X550 T1.
1.13	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.12	02-Dec-2016	msaitoh	branches: 1.12.8; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.11	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.10	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.9	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.8	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.7	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.6	06-Feb-2016	riastradh	branches: 1.6.2; Convert bool to target type before shift.
1.5	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.4	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.3	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.2	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.5	05-Dec-2016	skrll	Sync with HEAD
1.1.30.4	19-Mar-2016	skrll	Sync with HEAD
1.1.30.3	22-Sep-2015	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.28.2	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.1	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.12.8.10	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.12.8.9	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.12.8.8	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.12.8.7	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.12.8.6	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.12.8.5	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.12.8.4	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.12.8.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.12.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.12.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.15.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.17.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.18.2.7	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.18.2.6	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.18.2.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.18.2.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.18.2.3	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.18.2.2	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.18.2.1	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.22.2.1	29-Feb-2020	ad	Sync with head.
1.24.2.2	03-Apr-2021	thorpej	Sync with HEAD.
1.24.2.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.26.4.1	13-May-2021	thorpej	Sync with HEAD.
1.31.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.18	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.17	24-Jan-2022	msaitoh	branches: 1.17.4; Use 32bit for 32bit counter.
1.16	19-Jan-2022	msaitoh	Improve ixv(4)'s some event counters. - The virtual function's packet counter registers are not cleared on read. To solve this problem, <REGNAME>, base_<REGNAME>, last_<REGNAME> and saved_reset_<REGNAME> are in the struct ixgbevf_hw_stats and some functions use them. However, saved_reset_<REGNAME> is set but never referenced. base_<REGNAME> is set and it's only used for saved_reset_<REGNAME>, so it's also unused in reality. THERE ARE A LOT OF GARBAGE. Remove them. - Call ixv_init_stats() in ixv_clear_evcnt() to make ifconfig -z ixvN work correctly.
1.15	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.14	12-Sep-2019	msaitoh	- Make SIOCADDMULTI returns ENOSPC when the Ethenet multicast address list exceeds the limit. - State "Ethernet" multicast address. - Use macro.
1.13	04-Apr-2018	msaitoh	branches: 1.13.2; 1.13.8; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.12	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.11	06-Dec-2017	msaitoh	branches: 1.11.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.10	14-Sep-2017	msaitoh	Quick hack to fix ixv(4)'s hardware counter. ixgbe_rx_checksum() passes "struct ixgbe_hw_stats *" even if its a virtual function. The layout between ixgbe_hw_stats and ixgbevf_hw_stats are different, so ixgbe_rx_checksum() wrote wrong area if a device is a virtual function. To fix this problem with small change, move the location of ipcs, ipcs_bad, l4cs and l4cs_bad of struct ixgbe(vf)_hw_stats to the same location at the biggining of the struct.
1.9	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.8	10-Feb-2017	msaitoh	branches: 1.8.6; - Both mbuf_defrag_failed and m_defrag_failed existed. Remove m_defrag_failed. - rx_irq, dropped_pkts, morerx, moretx and txloops counters are uncounted, so remove them. - Count handleq in ixv_handle_que(). - Detach event counters correctly. - Set some per-queue event names correctly.
1.7	01-Dec-2016	msaitoh	branches: 1.7.2; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.5	24-Apr-2015	msaitoh	branches: 1.5.2; Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.4	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.3	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.2	10-Mar-2015	msaitoh	Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X.
1.1	12-Aug-2011	dyoung	branches: 1.1.12; 1.1.28; 1.1.30; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.30.4	28-Aug-2017	skrll	Sync with HEAD
1.1.30.3	05-Dec-2016	skrll	Sync with HEAD
1.1.30.2	06-Jun-2015	skrll	Sync with HEAD
1.1.30.1	06-Apr-2015	skrll	Sync with HEAD
1.1.28.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.28.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.28.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.1.12.1	03-Dec-2017	jdolecek	update from HEAD
1.5.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.5.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.7.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.8.6.7	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.8.6.6	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.8.6.5	29-Jan-2022	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1725: sys/dev/pci/ixgbe/ixgbe_vf.h 1.16-1.17 sys/dev/pci/ixgbe/ixv.c 1.176-1.177 Make ifconfig -z ixvN clear event counter.
1.8.6.4	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.8.6.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.8.6.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.8.6.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.11.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.11.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.13.8.4	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.13.8.3	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.13.8.2	29-Jan-2022	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1408: sys/dev/pci/ixgbe/ixgbe_vf.h 1.16-1.17 sys/dev/pci/ixgbe/ixv.c 1.176-1.177 Make ifconfig -z ixvN clear event counter.
1.13.8.1	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.13.2.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.17.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.24	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.23	24-Dec-2021	msaitoh	branches: 1.23.4; Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.22	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.21	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.20	10-Dec-2021	msaitoh	Remove unnecessary return value check. FreeBSD: 3a89005394bc5d82ce9b6baa9e7f8dee362354ae DPDK: 4b0ee6529b7897c2a08dd56669f07ac1f46a8474
1.19	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.18	30-Apr-2021	msaitoh	branches: 1.18.2; Add missing __KERNEL_RCSID().
1.17	13-Nov-2020	msaitoh	branches: 1.17.4; Fix typo in a debug message.
1.16	11-Jun-2018	msaitoh	branches: 1.16.8; 1.16.16; Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*).
1.15	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.14	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.13	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.12	06-Dec-2017	msaitoh	branches: 1.12.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.11	04-Oct-2017	msaitoh	branches: 1.11.2; Add 2.5GBASE-T and 5GBASE-T support. (forgotten in the last commit)
1.10	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.9	19-Jan-2017	msaitoh	branches: 1.9.6; Reduce diff against FreeBSD. No functional change.
1.8	05-Dec-2016	msaitoh	branches: 1.8.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.7	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.6	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.5	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.4	05-Aug-2015	msaitoh	branches: 1.4.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.3	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.2	02-Apr-2015	msaitoh	branches: 1.2.2; 1.2.4; Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.1	27-Mar-2015	msaitoh	Add some forgotten files in last commit.
1.2.4.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.4.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.4.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.4.1	02-Apr-2015	riz	file ixgbe_x540.c was added on branch netbsd-7 on 2015-04-19 14:42:14 +0000
1.2.2.6	05-Feb-2017	skrll	Sync with HEAD
1.2.2.5	05-Dec-2016	skrll	Sync with HEAD
1.2.2.4	22-Sep-2015	skrll	Sync with HEAD
1.2.2.3	06-Jun-2015	skrll	Sync with HEAD
1.2.2.2	06-Apr-2015	skrll	Sync with HEAD
1.2.2.1	02-Apr-2015	skrll	file ixgbe_x540.c was added on branch nick-nhusb on 2015-04-06 15:18:12 +0000
1.4.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.4.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.8.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.9.6.9	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.9.6.8	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.9.6.7	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.9.6.6	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.9.6.5	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.9.6.4	26-Jul-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #935): sys/dev/pci/ixgbe/ixgbe.c: revision 1.162 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.35 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.16 Apply changes of FreeBSD ix-3.2.18.tgz: - Add IXGBE_DEV_ID_82599_LS(0x154f) support. - Increase timeout for newer than ixgbe_mac_X550 (i.e. Xeon-D and Denverton) in ixgbe_acquire_swfw_sync_X540(). - Add some unused macros (IXGBE_FW_LESM_*). -- Remove nmbclusters check. We don't use the mbuf cluster. The old code also had a bug that ixgbe_total_ports adds two every port and never decrement in the detach path. Found by hikaru@. The code was removed in FreeBSD when it switched to use iflib and OpenBSD removed the code 8 years ago.
1.9.6.3	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.9.6.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.9.6.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.11.2.2	03-Dec-2017	jdolecek	update from HEAD
1.11.2.1	04-Oct-2017	jdolecek	file ixgbe_x540.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.12.2.3	25-Jun-2018	pgoyette	Sync with HEAD
1.12.2.2	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.12.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.16.16.1	14-Dec-2020	thorpej	Sync w/ HEAD.
1.16.8.5	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.16.8.4	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.16.8.3	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.16.8.2	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.16.8.1	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.17.4.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.17.4.1	13-May-2021	thorpej	Sync with HEAD.
1.18.2.1	31-May-2021	cjep	sync with head
1.23.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.10	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.9	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.8	04-Apr-2018	msaitoh	branches: 1.8.8; 1.8.20; Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.7	06-Dec-2017	msaitoh	branches: 1.7.2; Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.6	30-Aug-2017	msaitoh	branches: 1.6.2; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.5	01-Dec-2016	msaitoh	branches: 1.5.8; Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.4	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.3	05-Aug-2015	msaitoh	branches: 1.3.2; Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.2	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.1	27-Mar-2015	msaitoh	branches: 1.1.2; 1.1.4; Add some forgotten files in last commit.
1.1.4.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.1.4.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.1.4.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.1.4.1	27-Mar-2015	riz	file ixgbe_x540.h was added on branch netbsd-7 on 2015-04-19 14:42:15 +0000
1.1.2.5	05-Dec-2016	skrll	Sync with HEAD
1.1.2.4	22-Sep-2015	skrll	Sync with HEAD
1.1.2.3	06-Jun-2015	skrll	Sync with HEAD
1.1.2.2	06-Apr-2015	skrll	Sync with HEAD
1.1.2.1	27-Mar-2015	skrll	file ixgbe_x540.h was added on branch nick-nhusb on 2015-04-06 15:18:12 +0000
1.3.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.5.8.4	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.5.8.3	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.5.8.2	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.5.8.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.6.2.2	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	30-Aug-2017	jdolecek	file ixgbe_x540.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.7.2.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.8.20.1	13-May-2021	thorpej	Sync with HEAD.
1.8.8.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.8.8.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.28	06-Oct-2023	msaitoh	ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.27	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.26	31-May-2022	andvar	branches: 1.26.4; s/disbale/disable/ and s/enbale/enable/ in comments. also one more typo fix.
1.25	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.24	15-Dec-2021	msaitoh	Check host interface return status. No functional change on NetBSD. - FreeBSD: c1a56b6f5ffd6f3180a654d058c1783ccb808e8b or ix-3.3.18 DPDK: db18e37090a3b9af47d6a6886248520f6b220bf9 Writing to read-only fields returns a non-OK Return Status for shadow RAM write command for X550. This information was previously discarded. - No functional change on NetBSD because our ixg has no interface to write the NVM.
1.23	10-Dec-2021	msaitoh	Whitespace fix. Sync with FreeBSD ix-3.3.14.
1.22	10-Dec-2021	msaitoh	Add typecast for type mismatch. FreeBSD: 994dd6328c66fc277438ad51ed074f3c52096147 DPDK: d8e52b2cf771c31b523b46852fd86225b5a2c721
1.21	10-Dec-2021	msaitoh	Initialize data field in struct buffer. FreeBSD: b3ebe337ffa06b0f1f460bf8f1e42fb55db77d0b or ix-3.3.14 DPDK: 40543be5376ca415b2a7e196315d0555725b8bdf While sending request using ixgbe_hic_unlocked() the data field in buffer struct is not used. It is set when the struct is overwritten by FW to deliver the response. To not pass random data to FW the whole structure should be zeroed before use.
1.20	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.19	30-Apr-2021	msaitoh	branches: 1.19.2; Add missing __KERNEL_RCSID().
1.18	05-Feb-2020	msaitoh	branches: 1.18.10; No functional change: - Fix typos. - Remove extra newline.
1.17	23-Dec-2019	msaitoh	branches: 1.17.2; Add recovery code for unsupported SFP+. Before this commit: If an unsuppored SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching drvier (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.16	06-Aug-2019	msaitoh	Fix typo in comment. Found by Wataru Ashihara.
1.15	27-Jun-2019	msaitoh	branches: 1.15.2; Sync with FreeBSD ix-3.3.8 part 2: - Add missing XFI support into ixgbe_get_link_capabilities_X550em().
1.14	06-Dec-2018	msaitoh	Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.13	06-Jul-2018	msaitoh	Add force_10_100_autonego sysctl. The default vaule is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while.
1.12	04-Apr-2018	msaitoh	branches: 1.12.2; Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.11	04-Apr-2018	msaitoh	Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.10	30-Mar-2018	msaitoh	- Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.9	16-Mar-2018	msaitoh	- Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+.
1.8	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.7	08-Nov-2017	msaitoh	branches: 1.7.2; 1.7.4; Fix a bug that Denverton which uses firmware don't linkup if the media is forced to 100baseTX-FDX or 10baseT-FDX. As I wrote in ixgbe_phy.c rev. 1.13, popular switches and OSes don't use auto-negotiation if the media is forced to 100BASE-TX or 10BASE-T. Do the same thing. But, if we don't set FW_PHY_ACT_SETUP_LINK_AN in ixgbe_setup_fw_link(), the firmware wrongly set BMCR register. Two problems are observed: a) FDX may not be set. b) BMCR_SPEED1 (bit 6) is always cleard. + -------+------+-----------+-----+ \|request \| BMCR \| BMCR spd \| BMCR \| \| \| (HEX)\| (in bits)\| FDX \| +--------+------+----------+------+ \| 10M \| 0000 \| 10M(00) \| 0 \| \| 10M \| 2100 \| 100M(01) \| 1 \| \| 100M \| 0000 \| 10M(00) \| 0 \| \| 100M \| 0100 \| 10M(00) \| 1 \| +--------------------------+------+ To avoid this problem, after sending request to firmware, check BMCR register and fix the setting if it's required. Before this change: +------------------+---------------------------------------------+ \| \| denverton \| \| +---------+--------+---------+----------------+ \| \| auto \| 1G FDX \| 100 FDX \| 10 FDX \| +---------+--------+---------+--------+---------+----------------+ \| \| auto \| 1G FDX \| 1G FDX \| 100 FDX \| 10FDX/down(NG) \| \| +--------+---------+--------+---------+----------------+ \| \| 1G FDX \| 1G FDX \| 1G FDX \| down \| down \| \| link +--------+---------+--------+---------+----------------+ \| partner \|100 FDX \| down(1)\| down \| down(NG)\| down \| \| +--------+---------+--------+---------+----------------+ \| \| 10 FDX \| down(1)\| down \| down \| down(NG) \| +---------+--------+---------+--------+---------+----------------+ (Observed on: NVM Image Version 0.05 ID 0x8, NVM Map version 1.16, OEM NVM Image version 0.06, ETrackID 8000087c) After this change: +------------------+---------------------------------------------+ \| \| denverton \| \| +---------+--------+---------+----------------+ \| \| auto \| 1G FDX \| 100 FDX \| 10 FDX \| +---------+--------+---------+--------+---------+----------------+ \| \| auto \| 1G FDX \| 1G FDX \| 100 FDX \| 10 FDX \| \| +--------+---------+--------+---------+----------------+ \| \| 1G FDX \| 1G FDX \| 1G FDX \| down \| down \| \| link +--------+---------+--------+---------+----------------+ \| partner \|100 FDX \| down(1)\| down \| 100 FDX \| down \| \| +--------+---------+--------+---------+----------------+ \| \| 10 FDX \| down(1)\| down \| down \| 10 FDX \| +---------+--------+---------+--------+---------+----------------+ *1): may be correct because ixg doesn't support half duplex.
1.6	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.5	19-Jan-2017	msaitoh	branches: 1.5.6; Reduce diff against FreeBSD. No functional change.
1.4	27-Dec-2016	msaitoh	branches: 1.4.2; 1.4.4; Fix DEBUGFUNC() output in ixgbe_disable_rx_x550().
1.3	05-Dec-2016	msaitoh	branches: 1.3.2; Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.2	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.1	01-Dec-2016	msaitoh	Add forgotten ixgbe_x550.c and ixgbe_x550.h.
1.3.2.3	05-Feb-2017	skrll	Sync with HEAD
1.3.2.2	05-Dec-2016	skrll	Sync with HEAD
1.3.2.1	05-Dec-2016	skrll	file ixgbe_x550.c was added on branch nick-nhusb on 2016-12-05 10:55:17 +0000
1.4.4.1	21-Apr-2017	bouyer	Sync with HEAD
1.4.2.3	20-Mar-2017	pgoyette	Sync with HEAD
1.4.2.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.4.2.1	27-Dec-2016	pgoyette	file ixgbe_x550.c was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.5.6.14	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.5.6.13	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.5.6.12	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.5.6.11	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.5.6.10	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.5.6.9	02-Sep-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #1601): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.5.6.8	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.5.6.7	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.5.6.6	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.5.6.5	07-Aug-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #961): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.h: revision 1.51 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.23 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe.c: revision 1.163 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.49 Add force_10_100_autonego sysctl. The default value is 0(false). This sysctl is only for 550EM_a with PHY firmware for a while. Make jcl allocation per queue to reduce mutex spin. Tested by me and knakahara.
1.5.6.4	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.5.6.3	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.5.6.2	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.5.6.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.7.4.5	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.7.4.4	28-Jul-2018	pgoyette	Sync with HEAD
1.7.4.3	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.7.4.2	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.7.4.1	15-Mar-2018	pgoyette	Synch with HEAD
1.7.2.2	03-Dec-2017	jdolecek	update from HEAD
1.7.2.1	08-Nov-2017	jdolecek	file ixgbe_x550.c was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.12.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.12.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.12.2.1	10-Jun-2019	christos	Sync with HEAD
1.15.2.8	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.15.2.7	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.15.2.6	02-Jun-2022	martin	Pull up the following revisions, all via patch, requested by msaitoh in ticket #1459: sys/dev/pci/ixgbe/ixgbe.c 1.261,1.265-1.268,1.273,1.275-1.277, 1.312,1.316-1.319 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.182 - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Fix typos in comment. - KNF.
1.15.2.5	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.15.2.4	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.15.2.3	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.15.2.2	28-Jan-2020	martin	Pull up following revision(s) (requested by msaitoh in ticket #664): sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe.c: revision 1.219 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.44 Add recovery code for unsupported SFP+. Before this commit: If an unsupported SFP module is inserted before booting, the driver attach failed and there was no way to recover form it without rebooting or detaching/reattaching driver (drvctl -d && drvctl -r pciN). After this commit: We can automatically recover any time by replacing it with a supported module.
1.15.2.1	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.17.2.1	29-Feb-2020	ad	Sync with head.
1.18.10.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.18.10.1	13-May-2021	thorpej	Sync with HEAD.
1.19.2.1	31-May-2021	cjep	sync with head
1.26.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.26.4.1	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.7	24-Dec-2021	msaitoh	Update copyright to 2020. FreeBSD: 8455e365f77f5b66ac9521dbcd690f79345ce147
1.6	30-Apr-2021	msaitoh	Add NetBSD RCS IDs. No functional change.
1.5	06-Dec-2018	msaitoh	branches: 1.5.4; 1.5.16; Apply FreeBSD ix-3.3.6.tar.gz's change to NetBSD. Tested on C3000 and X550-T1, but not tested on Xeon D: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM (C3000): - FreeBSD always set IXGBE_FEATURE_RECOVERY_MODE without checking the NVM image version. We compare it against 2.0 to not to make new callout and not to call extra atomic operations. - In some FreeBSD's sysctl functions, atomic_load_acq_int() is called before a null pointer check. We call it after null pointer check. - Before calling atomic_load_acq_uint(), check adapter->feat_en flags to save atomic operation call. - We don't check recovery_mode in ixgbe_set_sysctl_value() because this function doesn't touch any hardware register. - NetBSD don't have FreeBSD's atomic_load_acq_int()-like function, so do it with membar_sync(). Thanks riastradh@ for the advice. - FreeBSD's ix-3.3.6 changed ixgbe_enable_aim from TRUE to FALSE, but we will keep it as TRUE because we have already fixed some bugs. - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and removed in ix-3.3.6.tar.gz. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - if_sriov.c's change doesn't affect to NetBSD because we don't support SR-IOV PF function.
1.4	30-Mar-2018	msaitoh	branches: 1.4.2; - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static.
1.3	30-Aug-2017	msaitoh	branches: 1.3.2; 1.3.4; - Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.2	02-Dec-2016	msaitoh	branches: 1.2.2; 1.2.4; 1.2.12; Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.1	01-Dec-2016	msaitoh	Add forgotten ixgbe_x550.c and ixgbe_x550.h.
1.2.12.5	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.2.12.4	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.2.12.3	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.2.12.2	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.2.12.1	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.2.4.2	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.2.4.1	02-Dec-2016	pgoyette	file ixgbe_x550.h was added on branch pgoyette-localcount on 2017-01-07 08:56:40 +0000
1.2.2.2	05-Dec-2016	skrll	Sync with HEAD
1.2.2.1	02-Dec-2016	skrll	file ixgbe_x550.h was added on branch nick-nhusb on 2016-12-05 10:55:17 +0000
1.3.4.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.3.4.1	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.3.2.2	03-Dec-2017	jdolecek	update from HEAD
1.3.2.1	30-Aug-2017	jdolecek	file ixgbe_x550.h was added on branch tls-maxphys on 2017-12-03 11:37:30 +0000
1.4.2.1	10-Jun-2019	christos	Sync with HEAD
1.5.16.1	13-May-2021	thorpej	Sync with HEAD.
1.5.4.2	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.5.4.1	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.198	10-Jul-2024	msaitoh	ixgbe: Remove ifdef NET_MPSAFE. Make ixgbe MP-scalable by default. Same as wm(4).
1.197	10-Jul-2024	msaitoh	ixgbe: ixgbe is always MP-safe, so remove IXGBE_MPSAFE macro. Before this change, IXGBE_MPSAFE is not defined if NET_MPSAFE is not set. In this case, the PCI_INTR_MPSAFE flag is not set by pci_intr_setattr() and interrupt_distribute() fails on arm. Always enable IXGBE_MPSAFE funciton by removing the macro.
1.196	15-Nov-2023	msaitoh	branches: 1.196.2; ixgbe: Clear the WTHRESH bit field before writing it.
1.195	15-Nov-2023	msaitoh	ixv(4): Remove unused IFF_OACTIVE. No functional change.
1.194	02-Nov-2023	yamaguchi	Use ether_bpf_mtap only when the device supports vlan harware tagging The function is bpf_mtap() for ethernet devices and currently it is just handling VLAN tag stripped by the hardware.
1.193	02-Nov-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.192	18-Oct-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.191	12-Oct-2023	msaitoh	ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate.
1.190	12-Oct-2023	msaitoh	ixgbe: Whitespace. No functional change.
1.189	06-Oct-2023	msaitoh	ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35.
1.188	06-Oct-2023	msaitoh	ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8.
1.187	06-Oct-2023	msaitoh	ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD.
1.186	06-Oct-2023	msaitoh	ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block
1.185	06-Oct-2023	msaitoh	ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30.
1.184	13-Sep-2023	msaitoh	ixgbe: Reorder some event counters for readability.
1.183	06-Jul-2022	msaitoh	branches: 1.183.4; Call txeof first, then rxeof for the consistency. There are three functions where the txeof and rxeof are called. The legacy interrupt function and MSI-X function call rxeof first, then rxeof. For the workqueue function. rxeof is called first. Modify it to match with other two.
1.182	02-Jun-2022	msaitoh	KNF. No functional change.
1.181	30-May-2022	msaitoh	Remove unused adapter->msix_mem.
1.180	25-Apr-2022	msaitoh	Add missing num_{tx,rx}_desc sysctl to sync with ixgbe.c.
1.179	25-Apr-2022	msaitoh	Limit {tx,rx}_process_limit sysctl values from 1 to num_{tx,rx}_desc.
1.178	25-Jan-2022	msaitoh	Use atomic_{load,store}_relaxed() for event counters.
1.177	24-Jan-2022	msaitoh	Fix ixv.c rev. 1.176. Calculate 36bit counter correctly.
1.176	19-Jan-2022	msaitoh	Improve ixv(4)'s some event counters. - The virtual function's packet counter registers are not cleared on read. To solve this problem, <REGNAME>, base_<REGNAME>, last_<REGNAME> and saved_reset_<REGNAME> are in the struct ixgbevf_hw_stats and some functions use them. However, saved_reset_<REGNAME> is set but never referenced. base_<REGNAME> is set and it's only used for saved_reset_<REGNAME>, so it's also unused in reality. THERE ARE A LOT OF GARBAGE. Remove them. - Call ixv_init_stats() in ixv_clear_evcnt() to make ifconfig -z ixvN work correctly.
1.175	18-Jan-2022	msaitoh	Update ixv_print_debug_info() to print similar data to ixg(4). The old hw.ixvN.debug=1 printed evcnt(9) data. The same values are printed by vmstat -e, so it's not worth to print. Remove all of the old output and make it similar to ixgbe_print_debug_info().
1.174	18-Jan-2022	msaitoh	Don't expose garbage data of hw.ixvN.debug.
1.173	24-Dec-2021	msaitoh	Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0.
1.172	24-Dec-2021	msaitoh	Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24.
1.171	24-Dec-2021	msaitoh	Whitespace fix. Remove old comment. No functional change. - From FreeBSD ixv-1.5.22.
1.170	14-Dec-2021	msaitoh	Use adapter->mta for the multicast array memory instead of the on-stack array. - Same as ixv-1.5.22. - Same as ixg(4).
1.169	30-Sep-2021	yamaguchi	Remove vlan_ifdetach() from ixv_detach() The removed function is called in ether_ifdetach().
1.168	16-Sep-2021	msaitoh	Fix a bug that an mbuf chain which has more than 63488bytes MAY fail on TX. - Currently, The TX buffer dmamap's max number of segments is set to 32. MCLBYTES(== 2048) * 32 = 65536 and it's enough for IP_MAXPACKET(65535). If an mbuf chain has more than 32 mbufs, we call m_defrag() to make it lower than equal to 32, but it might not work. The reason is that our m_defrag() don't modify the first mbuf entry of the chain. e.g.: if an mbuf chain contains 63600bytes data and the first mbuf has 100bytes in the m_data, the new chain has 100+2048+2048+...+12 and the total number of the chain is not 32 but 33. It result in 43 TCP packets will drop. - One of the way to fix this problem is to change m_defrag() which add a new mbuf cluster to the first mbuf. It's need discussion. - Another solution is to change the max number of the TX DMA segment. It should be at least 33 to avoid the m_defrag()'s current limitation. The document (82599-X550 DS 7.2.1.1 "Transmit Storage in system Memory") says that a packet can be fragmented into 40 - WTHRESH - 2 (for 82598, the document say nothing and some people and code says it's unlimited). Currently WTHRESH is set to 8. 40 - 8 - 2 = 30. !?!?!? {Net,Free,Open}BSD and Linux use 32. Is that safe? Anyway, we change WTHRESH from 8 to 5 to fit it. The added comment in ixgbe.h is based on DragonFly's though they don't use WTHRESH. - Yet another solution is to use the Tx Head Pointer Write Back function instead of WTHRESH based write back (see 82599-X550 DS 7.2.3.5.2 "Tx Head Pointer Write Back" or 82598 "Transmit Completions Head Write Back" ). DragonFly, illumos and DPDK use it. - Yet yet another solution is to add tso_maxsize entry to struct ifnet and use it in the TCP stack.
1.167	16-Sep-2021	msaitoh	Use IXGBE_TXDCTL_WTHRESH_SHIFT macro. No functional change.
1.166	26-Aug-2021	msaitoh	Remove "j" or "jumbo" because we always use MCLBYTES sized buffer.
1.165	25-Aug-2021	msaitoh	Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. - Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Reviewed by knakahara and ryo.
1.164	15-Jul-2021	msaitoh	Add a new sysctl to read rxr->next_to_refresh.
1.163	07-Jul-2021	msaitoh	Add new sysctl "rx_copy_len". ixgbe_rxeof() has an optimization "RX_COPY" to reduce costs of bus_dmamap_load_mbuf() and bus_dmamap_unload() by copying a mbuf cluster's memory to a newly allocated mbuf's MH_databuf[] and recycle the original map. The optimization is used when a length of a packet is smaller than a specific value. The value is calculated based on MHLEN. The size of MHLEN is architecture specific. It's 256 or 512. Make the threshold controllable by adding a new sysctl.
1.162	16-Jun-2021	riastradh	if_attach and if_initialize cannot fail, don't test return value These were originally made failable back in 2017 when if_initialize allocated a softint in every interface for link state changes, so that it could fail gracefully instead of panicking: https://mail-index.NetBSD.org/source-changes/2017/10/23/msg089053.html However, this spawned many seldom- or never-tested error branches, which are risky to have around. And that softint in every interface has since been replaced by a single global workqueue, because link state changes require thread context but not low latency or high throughput: https://mail-index.NetBSD.org/source-changes/2020/02/06/msg113759.html So there is no longer any reason for if_initialize to fail. (The subroutine if_stats_init can't fail because percpu_alloc can't fail either.) There is a snag: the softint_establish in if_percpuq_create could fail, potentially leading to bad consequences later on trying to use the softint. This change doesn't introduce any new bugs because of the snag -- if_percpuq_attach was already broken. However, the snag can be better addressed without spawning error branches, either by using a single softint or making softints less scarce. (Separate commit will change the signatures of if_attach and if_initialize to return void, scheduled to ride whatever is the next convenient kernel bump.) Patch and testing on amd64 and evbmips64-eb by maya@; commit message soliloquy, and compile-testing on evbppc/i386/earmv7hf, by me.
1.161	19-May-2021	msaitoh	No functional change: - Add NetBSD RCS IDs. - KNF.
1.160	18-May-2021	msaitoh	- Cleanup an rxbuf entry when bus_dmamap_load_mbuf() failed to prevent panic. - Print error number when error occurred.
1.159	30-Apr-2021	msaitoh	branches: 1.159.2; Add missing __KERNEL_RCSID().
1.158	31-Mar-2021	msaitoh	branches: 1.158.2; KNF a bit. No functional change.
1.157	31-Mar-2021	msaitoh	Modify error message to sync with ixgbe.c
1.156	11-Mar-2021	msaitoh	branches: 1.156.2; Add "TX " to "Queue No Descriptor Available" to make it more understandable. Same as ixgbe.c rev. 1.278.
1.155	09-Mar-2021	msaitoh	Modify some parameters to reduce packet dropping. - Background: ixgbe doesn't use common MCLGET() interface and use the driver specific cluster allocation mechanism (jcl). The cluster is pre-allocated with a fixed number and the current number per queue is num_rx_desc * 2 (20482=4096). It's too small. It also has a problem that the max length of the pcq which is used in the TX path is big (4096). Example: 100M <----- [ixg0 ixg1] <----- 1G 2048 TX descs <--- 4096 pcqs <---- 2048 RX descs If a machine forwards a traffic from 1G interface to 100M interface, It would require 2048+4096+2048=8192 descriptors, but the current number is 20482=4096. It's too small. Even if the both interface's link speed is the same and only small number of packet is queued in the pcq, 4096 jcl is small because 2048(RX)+TX(2048)=4096. If jcl is exhausted, not only forwarding from ixg1 to ixg0 is dropped, but also another forwarding path from ixg1 to another interface(e.g. wm0) is also dropped. Sockets also queue packets, so if a lot of sockets are used and/or a socket buffer size is changed to bigger one, it'll also become a problem. If the jcl is exhausted, evcnt(9) counter "ixgX qY Rx no jumbo mbuf" is incremented. Example: vmstat -ev \| grep ixg1 \| grep "no jumbo" ixg1 q0 Rx no jumbo mbuf 0 0 misc ixg1 q1 Rx no jumbo mbuf 0 0 misc ixg1 q2 Rx no jumbo mbuf 141326 0 misc ixg1 q3 Rx no jumbo mbuf 0 0 misc - To solve this problem: - Add new config parameter IXGBE_JCLNUM_MULTI and set the default to 3 (2048 * 3). The minimum number is 2. The total number of jcl per queue is available with hw.ixgN.num_jcl_per_queue sysctl. - Reduce the max length of the pcq() which is used in the TX path from 4096 to 2048. - Reviewed by knakahara@ and ozaki-r@. - TODO: Use MCLGET().
1.154	07-Sep-2020	msaitoh	branches: 1.154.2; - Remove extra callout_stop() in ixgbe_detach(). Found by knakahara@. - Rename ix{gbe,v}_free_workqueue() to ix{gbe,v}_free_deferred_handlers(). - Add KASSERT() to functions who are called from ixgbe_handle_admin().
1.153	31-Aug-2020	msaitoh	Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change.
1.152	13-Aug-2020	msaitoh	Use atomic_cas_uint() and atomic_store_relaxed(). Advised by thorpej@. Tested by me. OK'd by knakahara.
1.151	25-Jun-2020	msaitoh	Reduce ixgbe's busy loop using with workqueue and kpause. - Use workqueue instead of softint to make some functions sleepable. - Use new workqueue and enqueue it in ixgbe_local_timer() and ixgbe_recovery_mode_timer() to make them sleepable. - Make new ixgbe_delay() and use it. This functions sleeps if the time is more than equals 1 tick. If it's not, do delay().
1.150	18-Jun-2020	msaitoh	Modify a liitle to reduce diff between ixgbe.c and ixv.c. No functional change.
1.149	11-Jun-2020	msaitoh	Fix typo. No functional change.
1.148	17-Apr-2020	msaitoh	No functional change: - modify comment - whitespace fix
1.147	15-Mar-2020	thorpej	branches: 1.147.2; Define and implement a locking protocol for the ifmedia / mii layers: - MP-safe drivers provide a mutex to ifmedia that is used to serialize access to media-related structures / hardware regsiters. Converted drivers use the new ifmedia_init_with_lock() function for this. The new name is provided to ease the transition. - Un-converted drivers continue to call ifmedia_init(), which will supply a compatibility lock to be used instead. Several media-related entry points must be aware of this compatibility lock, and are able to acquire it recursively a limited number of times, if needed. This is a SPIN mutex with priority IPL_NET. - This same lock is used to serialize access to PHY registers and other MII-related data structures. The PHY drivers are modified to acquire and release the lock, as needed, and assert the lock is held as a diagnostic aid. The "usbnet" framework has had an overhaul of its internal locking protocols to fit in with the media / mii changes, and the drivers adapted. USB wifi drivers have been changed to provide their own adaptive mutex to the ifmedia later via a new ieee80211_media_init_with_lock() function. This is required because the USB drivers need an adaptive mutex. Besised "usbnet", a few other drivers are converted: vmx, wm, ixgbe / ixv. mcx also now calls ifmedia_init_with_lock() because it needs to also use an adaptive mutex. The mcx driver still needs to be fully converted to NET_MPSAFE.
1.146	05-Feb-2020	msaitoh	Modify note in ixv_update_stats(). VF doesn't count errors by hardware.
1.145	04-Feb-2020	thorpej	Use ifmedia_fini().
1.144	21-Jan-2020	msaitoh	Fix the freeing code for some error paths. Found and tested by Patrick Welche.
1.143	17-Dec-2019	msaitoh	branches: 1.143.2; Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@.
1.142	16-Dec-2019	msaitoh	No functional change: - Remove unused code. - Remove extra spaces. - KNF.
1.141	18-Nov-2019	msaitoh	Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.140	06-Nov-2019	msaitoh	ixv(4): disable RSS configuration on 82599 and X540 VFs. Those VFs share their RSS configuration with PF and, thus, they cannot be configured independently. From FreeBSD r354349.
1.139	01-Oct-2019	msaitoh	Set MTU correctly if mtu > 1500.
1.138	20-Sep-2019	msaitoh	- Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - If a multicast entry has range, use ALLMULTI like others. - Remove ixv_set_promisc() and use ixv_set_multi(). And then, rename _set_multi() to _set_rxfilter(). Same as ixgbe.c. - The promisc mode can't be enabled if the PF is not in promisc mode. Identify that state and report it as "the PF may not in promisc mode" (though it might not be perfect).
1.137	13-Sep-2019	msaitoh	Add PROMISC support.
1.136	13-Sep-2019	msaitoh	if_flags is neither int nor short. It's unsigned short.
1.135	13-Sep-2019	msaitoh	- Return EPERM when the interface is not trusted. - Treat IXGBE_ERR_FEATURE_NOT_SUPPORTED.
1.134	12-Sep-2019	msaitoh	Add support ALLMULTI on ixv(4): - Negotiate API version up to 1.3. - On linux's PF driver implementation, the PF replies VF's XCAST_MODE_ALLMULTI message not with NACK but with ACK even if the virtual function is NOT marked "trust" and act as XCAST_MODE_"MULTI". If ixv(4) simply check the return vaule of update_xcast_mode(XCAST_MODE_ALLMULTI), SIOCSADDMULTI success and the user may have trouble with some addresses. Fortunately, the Linux's PF driver's "ACK" message has not XCAST_MODE_"ALL"MULTI but XCAST_MODE_MULTI, so we can check this state by checking if the send message's argument and the reply message's argument are different. - Noy yet for PROMISC.
1.133	12-Sep-2019	msaitoh	Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30).
1.132	12-Sep-2019	msaitoh	- Make SIOCADDMULTI returns ENOSPC when the Ethenet multicast address list exceeds the limit. - State "Ethernet" multicast address. - Use macro.
1.131	05-Sep-2019	msaitoh	Add support SIOCZIFDATA as ixgbe.c. BUGS The following event counters are not cleared by SIOCZIFDATA because the corresponding registers are read only and not cleared on read: Good Packets Received Good Octets Received Multicast Packets Received Good Packets Transmitted Good Octets Transmitted
1.130	04-Sep-2019	msaitoh	Use aprint_*() in the attach function.
1.129	04-Sep-2019	msaitoh	printf -> device_printf
1.128	21-Aug-2019	msaitoh	Simplify ix{gbe,v}_[un]register_vlan() API suggestesd by knakahara. The API was the same as FreeBSD's pre-iflib's. They use iflib now and it's not required for us to keep the old API.
1.127	21-Aug-2019	msaitoh	Use ETHER_LOCK()/ETHER_UNLOCK() suggested by knakahara.
1.126	20-Aug-2019	msaitoh	Fix a bug that VLAN HW "tagging" enable/disable may not refrect correctly. - Always call ec_vlan_cb() if it exists. - Some (or all?) ethernet drivers don't enable HW tagging if no any vlan is attached. ixgbe is one of them. Check the the transition and update VLAN HW tagging function. XXX pullup-9
1.125	30-Jul-2019	msaitoh	branches: 1.125.2; It's not required to call ixv_stop() twice in ixv_detach().
1.124	26-Jul-2019	msaitoh	Simplify and avoid undefined behavior. This is not a real bug because the max number of queue is 2.
1.123	25-Jul-2019	msaitoh	Fix VLAN hardware filter initialization. Use unsigned. Found by KUBSan.
1.122	25-Jul-2019	msaitoh	Fix VLAN filter setting. Found by KUBSan.
1.121	25-Jul-2019	msaitoh	Set IVAR (Interrupt Vector Allocation Register) correctly in ixgbe_set_ivar(). Found by KUBSan.
1.120	17-Jul-2019	msaitoh	Implement VLAN hardware filter function(ETHERCAP_VLAN_HWFILTER). First proposed by jmcneill in 2017 and modified by me. How to use: - Set callback function: ether_set_vlan_cb(struct ethercom , ether_vlancb_t) - Callback. This function is called when a vlan is attached/detached to the parent interface: int (ether_vlancb_t)(struct ethercom *ec, uint16_t vlanid, bool set); - ifconfig(8) ifconfig ixg0 [-]vlan-hwfilter Note that ETHERCAP_VLAN_HWFILTER is set by default on ixg(4) because the PF driver usually enable "all block" filter by default.
1.119	17-Jul-2019	msaitoh	KNF. No functional change.
1.118	04-Jul-2019	msaitoh	On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing.
1.117	02-Jul-2019	msaitoh	Sync with FreeBSD's ixv-1.5.17: - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Modify to match the newer NETMAP interface. No functional change.
1.116	27-Jun-2019	msaitoh	Don't call set_vfta() if any VLAN is attached. XXX pullup-8.
1.115	29-May-2019	msaitoh	Even if we don't use MII(4), use the common path of SIOC[GS]IFMEDIA in sys/net/if_ethersubr.c if we can. - Add ec_ifmedia into struct ethercom. - ec_mii in struct ethercom is kept and used as it is. It might be used in future. Note that some Ethernet drivers which _DOESN'T_ use mii(4) use ec_mii for keeping the if_media. Those should be changed in future.
1.114	23-May-2019	msaitoh	Whitespace fix (mainly tabify).
1.113	23-May-2019	msaitoh	-No functional change: - KNF - u_int_t -> uint_t.
1.112	18-Mar-2019	msaitoh	s/pakcet/packet/ in comment.
1.111	15-Mar-2019	msaitoh	- Simplily. Suggested by knakahara. - Modify comment. Per queue VLAN enable flags is on 82599 and later. - Fix typo in comment.
1.110	13-Mar-2019	msaitoh	Fix a bug that the VLAN HW tagging function is not correctly disabled when all vlan is detached. Part of ixgbe.c rev. 1.177.
1.109	22-Feb-2019	msaitoh	Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.108	23-Jan-2019	msaitoh	Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them expilicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.107	27-Sep-2018	msaitoh	Don't set LRO capability flags because we don't support it yet. This doesn't change any behavior other than if_capabilities' visibility (e.g. ifconfig).
1.106	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.105	06-Jun-2018	kamil	branches: 1.105.2; Comment out unused functions in order to fix Clang build ixv_rearm_queues() and ixgbe_rearm_queues() are unused. Sponsored by <The NetBSD Foundation>
1.104	04-Jun-2018	msaitoh	Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.103	03-Jun-2018	maxv	Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes).
1.102	30-May-2018	msaitoh	Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294.
1.101	25-May-2018	ozaki-r	Ensure to call if_register after interface initializations finish
1.100	23-May-2018	msaitoh	Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE.
1.99	23-May-2018	msaitoh	Add rxd_nxck (Receive Descriptor next to check) read only sysctl.
1.98	18-May-2018	msaitoh	Revert new watchdog timer commits. The new watchdog timer made stability worse than before. It seems unknown problems exists. http://mail-index.netbsd.org/source-changes/2018/05/08/msg095020.html http://mail-index.netbsd.org/source-changes/2018/05/16/msg095240.html
1.97	14-May-2018	msaitoh	Fix panic or hangup when "sysctl -w hw.ixgN.debug=1". XXX pullup-8
1.96	08-May-2018	msaitoh	- Fix broken watchdog timer. This change detects TX device timeout correctly. NOTE: It's supporsed not to be called {ixgbe,ixv}_rearm_queues() in the timer. Those are not required if any chip have no bug. In reality, ixgbe_rearm_queues() is required on 82599 and newer chip AND other than queue 0 to prevent device timeout. When it occured, packet was sent but the descriptor's DD bit wasn't set even though IXGBE_TXD_CMD_EOP and IXGBE_TXD_CMD_RS were set. After forcing interrupt by writing EICS register in ixgbe_rearm_queues(), DD is set. Why? Is this an undocumented errata? It might be possible not call rearm_queues on 82598 or queue 0, we call in any cases in case the problem occurs. On ixv(4), I have not seen this problem yet (though I tested only on X550_X(Xeon D 12xx)'s virtual function), but we do rearm in case TX device timeout happen. - ixv(4): Call callout_stop() earlier in ixv_stop() like ixgbe_stop(). - KNF.
1.95	19-Apr-2018	msaitoh	Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init().
1.94	17-Apr-2018	msaitoh	Remove unused structure entries. No functional change.
1.93	12-Apr-2018	msaitoh	Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.92	04-Apr-2018	msaitoh	Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment.
1.91	04-Apr-2018	msaitoh	Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace.
1.90	30-Mar-2018	knakahara	Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8
1.89	20-Mar-2018	knakahara	Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o.
1.88	15-Mar-2018	msaitoh	Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix.
1.87	09-Mar-2018	msaitoh	Make some event counters MP safe. Now all of the event counters are MP safe.
1.86	07-Mar-2018	msaitoh	Don't increment que->req.ev_count in MSI-X interrupt because it's not reschedule.
1.85	07-Mar-2018	msaitoh	- Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code.
1.84	02-Mar-2018	knakahara	branches: 1.84.2; ixv(4) also supports workqueue poll mode, but not enabled by default yet, either. ok by msaitoh@n.o.
1.83	27-Feb-2018	msaitoh	Add hw.ixvM.q[01].{interrupt_rate,[tr]xd_head,[tr]xd_tail} sysctls as ixg(4).
1.82	26-Feb-2018	knakahara	Apply ixgbe.c:r1.127 to ixv.c. Pointed out by msaitoh@n.o.
1.81	22-Feb-2018	msaitoh	Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.80	22-Feb-2018	msaitoh	- Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function.
1.79	16-Feb-2018	msaitoh	Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv)
1.78	16-Feb-2018	knakahara	Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o.
1.77	21-Dec-2017	msaitoh	Fix panic when only link interrupt can't be established.
1.76	21-Dec-2017	msaitoh	Don't panic when resource shortage occured. Like ixgbe.c rev. 1.116. - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more).
1.75	06-Dec-2017	msaitoh	Sync with FreeBSD's r326022. All of the following changes have no influence to netbsd: - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Fix for netmap module. - Remove never defined UDP_IPV4_EX - Add SPDX-License-Identifier
1.74	16-Nov-2017	ozaki-r	Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@
1.73	23-Oct-2017	msaitoh	If if_initialize() failed in the attach function, free resources and return.
1.72	18-Oct-2017	msaitoh	Protect ec_multi* with mutex like other MP safe Ethernet drivers. See if_ether.h rev. 1.66 and related stuff: http://mail-index.netbsd.org/source-changes/2016/12/28/msg080289.html
1.71	13-Oct-2017	msaitoh	Show 2.5GBASE-T and 5GBASE-T correctly on ixv(4). Tested on a virtual function on X550 T1.
1.70	13-Oct-2017	msaitoh	Set baudrate correctly. A device which use mii(4) interface can set if_baudrate atomatically by mii_phy_statusmsg(). ixg(4) and ixv(4) don't use mii(4), so calling ifmedia_baudrate() directly is required. XXX need the same same work for wm(4)'s TBI and SERDEV devices.
1.69	04-Oct-2017	msaitoh	- ixv_enableintr(): Fix a bug that ixv_enable_queue() is called with incorrect argument. This bug was added in rev. 1.68. - Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. A bit is set on very low probability and the bit location is incorrect. Is this an errata?
1.68	03-Oct-2017	msaitoh	Fix a problem that mailbox interrupt never occured. Bit definitions of PF's interrupt related registers and VF's interrupt registers a completely different. VF's registers (VF{EICR, EICS, EIMS, EIMC, EIAC, EIAM})'s bits are just MSI-X vector bitmask, so read/write correctly. Read VTEICR instead of VTEICS in ixv_msix_mbx (note that "PF" has a errata that EICS is required to read the cause). Don't write IXGBE_VTEICR in ixv_msix_mbx() because we use auto-clear for mailbox interrupt. We have gotten link status change by timer instead of interrupt before this fix...
1.67	03-Oct-2017	msaitoh	Make mailbox statistic counters evcnt(9). Example: > ixv0 message TXs 23 0 misc > ixv0 message RXs 2911 0 misc > ixv0 ACKs 23 0 misc > ixv0 REQs 22 0 misc > ixv0 RSTs 0 0 misc
1.66	27-Sep-2017	msaitoh	Print device name in ixv_attach().
1.65	15-Sep-2017	msaitoh	xv(4) VLAN fixes: - Dirty hack for VID mask bits. On ixg(4), VLAN filter is disabled. So any vlan ID's packet passes RX filter. On ixv(4), usually, PF driver uses VLAN filter enabled. There is no way to disable PF's VLAN filter function itself. NetBSD's Ethernet driver has currently no API to know which VLAN ID should be accepted. To avoid this proble, enable all VIDs... Yes, I know this is dirty hack. We should rethink. - Call ixv_setup_vlan_support() in ixv_ifflags_cb(). - Don't use global ixv_shadow_vfta[]. - Use local variable (rxr->) to reduce diff against ixgbe.c. No functional change.
1.64	15-Sep-2017	msaitoh	Make ixv(4) TX/RX descriptors size the same as ixg(4).
1.63	13-Sep-2017	msaitoh	- Print driver feature capabilities and enable bits when verbose boot as ixgbe.c rev. 1.99. - Whitespace fix.
1.62	13-Sep-2017	msaitoh	Set get_link_status to true in ixv_init_locked() to get the first link status correctly.
1.61	12-Sep-2017	msaitoh	Fix panic caused by rev. 1.58's change. Don't if_start from interrupt context. The restart code exists in ixv_handle_que().
1.60	11-Sep-2017	msaitoh	Print Mailbox API version.
1.59	11-Sep-2017	msaitoh	Use cprng_strong64() insread of cprng_fast64() to prevent panic.
1.58	30-Aug-2017	msaitoh	- Sync with FreeBSD r320688 (and up to r322865): - Add C3000(Denvertion(-NS)) support. - Add bypass function support for bypass adapters. Sysctls are provided. Not tested because I have no any bypass adapter. - ixv(4): set RSS mapping. - Change EEE sysctl. - ixv(4): Add TSOv6. - ixv(4): Mailbox API 1.2 and more are implemented and comment says it negotiate with 1.2 but it really does 1.1... - Remove thermal test sysctl. - Fix unknown bugs. - Print driver feature capabilities and enable bits when verbose boot.
1.57	30-Aug-2017	msaitoh	FreeBSD r320688 sorted a lot of functions. Merging that change by one commit makes difficult to know what was really changed. To make change understandable, commit only function sort and white space changes before real change.
1.56	26-May-2017	msaitoh	branches: 1.56.2; Add missing {ixgbe,ixv}_stop() in the detach function. Without this change, shutdown won't finish or panic on heay traffic.
1.55	03-Mar-2017	msaitoh	Set PCI_INTR_MPSAFE, SOFTINT_MPSAFE, CALLOUT_MPSAFE and IFEF_START_MPSAFE if NET_MPSAFE is defined.
1.54	16-Feb-2017	msaitoh	Do if_snd check even if IXGBE_LEGACY_TX is used. It's used by ALTQ.
1.53	13-Feb-2017	msaitoh	Remove more ix_start_locked() from interrupt context like previous commit. > Fix a bug that ix_start_locked() is called in interrput context > (ix_msix_que). The function is called in softint(ix_handle_que()). OK'd by >k-nakahara.
1.52	13-Feb-2017	msaitoh	Use percpuq.
1.51	13-Feb-2017	msaitoh	Fix a bug that ix_start_locked() is called in interrput context (ix_msix_que). The function is called in softint(ix*_handle_que()). OK'd by k-nakahara.
1.50	10-Feb-2017	msaitoh	Change the Interrupt Moderation flag from global variable to per device.
1.49	10-Feb-2017	msaitoh	- Both mbuf_defrag_failed and m_defrag_failed existed. Remove m_defrag_failed. - rx_irq, dropped_pkts, morerx, moretx and txloops counters are uncounted, so remove them. - Count handleq in ixv_handle_que(). - Detach event counters correctly. - Set some per-queue event names correctly.
1.48	10-Feb-2017	msaitoh	No functional change: - Rename ixv_sysctl_attach() to ixv_add_device_sysctls(). - Move sysct and evcnt related functions to match ixgbe.c.
1.47	08-Feb-2017	msaitoh	Print CPU number correctly.
1.46	08-Feb-2017	msaitoh	- Fix aprint_*()s in ixv_allocate_msix(). - Style fix. Reduce diff against ixgbe.c. No functional change.
1.45	08-Feb-2017	msaitoh	Call IFQ_SET_MAXLEN() even if TX multiqueue is used. ALTQ uses it.
1.44	08-Feb-2017	msaitoh	Enable multiqueue.
1.43	08-Feb-2017	msaitoh	Use dmat64 if available. Same as ixgbe.c rev. 1.60.
1.42	08-Feb-2017	msaitoh	Fix a bug that 1Gbps is printed even if the phyiscal media is at 100Mbps.
1.41	08-Feb-2017	msaitoh	Fix detach path: - Call softint_disestablish() and pci_intr_disestablish() for all queues correctly. - Add missing sysctl_teardown() to delete sysctl entries. - Call pci_intr_release(). Now NETBSD_MSI_OR_MSIX is completely deleted.
1.40	08-Feb-2017	msaitoh	Add queue number into interrupt name.
1.39	08-Feb-2017	msaitoh	Set ifm_active to IFM_NONE when the physical interface's link is down. Now ifconfig shows "media: Ethernet autoselect '(none)'".
1.38	07-Feb-2017	msaitoh	Set IFM_10G_T on 10G...
1.37	07-Feb-2017	msaitoh	Fix link status stuff: - Set get_link_status in ixv_init_locked() to get link status correctly. - Don't discard checked speed in ixv_config_link()...
1.36	07-Feb-2017	msaitoh	Fix device attach: - Even if ixgbevf_negotiate_api_version() failed, continue attaching. Same as Linux. Tested on ESXi 6.0. - Fix argument of pci_msix_alloc_exact(). - Use different intr_xname for each interrupt. - Add missing pci_intr_release() on error. - Add missign kcpuset_destroy(). - Calculate MSI-X vectors correctly.
1.35	01-Feb-2017	msaitoh	TX multiqueue. If you want to disable it, enable IXGBE_LEGACY_TX in ixgbe_netbsd.h
1.34	30-Jan-2017	msaitoh	Fix to use softint-based if_input in ixv(4). See: http://mail-index.netbsd.org/source-changes/2016/02/09/msg072521.html http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html
1.33	19-Jan-2017	msaitoh	Reduce diff against FreeBSD. No functional change.
1.32	18-Jan-2017	msaitoh	Now we can use multiqueue. It's default on ixg(4). Not yet for ixv(4).
1.31	05-Jan-2017	msaitoh	branches: 1.31.2; Fix INIT_DEBUGOUT() messages.
1.30	16-Dec-2016	msaitoh	Remove extra newline in debug printfs. ixgbe's DEBUGOUT macros automatically add newline.
1.29	05-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r303890: - Configure ixgbe phy & gbic power. Setup phy and gbic power as per Linux 4.3.13 driver. This fixes link not detected on X540-AT2 after booting to Linux which turns the phy power off on detach. (FreeBSD r295093) - Fixup DA cable detection routine. (FreeBSD r303032) - Some minor changes
1.28	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r294578: - Fixup SFP module insertion on the 82599 when insertion happens after the system is booted and running. Add PHY detection logic to ixgbe_handle_mod() and add locking to ixgbe_handle_msf() as well. FreeBSD r293334. - Fix ix advertise value after media change. When ifconfig sets media then the values displayed by the advertise_speed value are invalidated. Fix this by setting the bits correctly including setting advertise to 0 for media = auto. FreeBSD r294578. - Some others (e.g. LRO(not used by NetBSD)).
1.27	02-Dec-2016	msaitoh	Fix ix{gbe,v}_set_sysctlvalue().
1.26	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r292674: - Add X552 SFP+ and X550T single port. - Bug fixes.
1.25	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r289238: - Add support for sysctl knobs to live tune the per interrupt rx/tx packet processing limits in ixgbe(4) - Some others (netmap, etc.)
1.24	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r285590: - Fix igxbe SRIOV VF (if_ixv) initialization bugs. The MAC address for an if_ixv instance can now set at creation time, and the receive ring tail pointer is correctly initialized (previously, things still worked because the receive ring tail pointer was being fixed up as a side effect of other activity).
1.23	02-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r283881: - SRIOV support (not enabled because NetBSD doesn't support it).
1.22	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r282299: - Add support for X55x. - ADD EEE support (not tested). - Add WOL support (not tested). - Add suspend/resume support (not testd). - Add LPLU support (not tested). - Add DMA Coalescing (note verified). - Not tested well for sysctls. - Fix ixgbe_set_advertise() a bit. At least, FreeBSD r294578 is required to work hw.ixg0.advertise_speed sysctl correctly.
1.21	01-Dec-2016	msaitoh	Update ixg(4) and ixv(4) up to FreeBSD r280197: - Add support for 82599_QSFP_SF_QP and X540T1. - Add partial support for X55x. It's required to sync with FreeBSD r282299 to support X55x. - Bugfixes.
1.20	28-Nov-2016	msaitoh	Fix vlan check bug in ixv_setup_vlan_support().
1.19	28-Nov-2016	knakahara	fix build of ixv enabled NO_PCI_MSI_MSIX option.
1.18	25-Nov-2016	msaitoh	Remove #ifdef NETBSD_MSI_OR_MSIX to be simple. ixv(4) isn't attached if __HAVE_PCI_MSI_MSIX isn't set because MSI-X must be used.
1.17	10-Jun-2016	ozaki-r	branches: 1.17.2; Introduce m_set_rcvif and m_reset_rcvif The API is used to set (or reset) a received interface of a mbuf. They are counterpart of m_get_rcvif, which will come in another commit, hide internal of rcvif operation, and reduce the diff of the upcoming change. No functional change.
1.16	09-Feb-2016	ozaki-r	Introduce softint-based if_input This change intends to run the whole network stack in softint context (or normal LWP), not hardware interrupt context. Note that the work is still incomplete by this change; to that end, we also have to softint-ify if_link_state_change (and bpf) which can still run in hardware interrupt. This change softint-ifies at ifp->if_input that is called from each device driver (and ieee80211_input) to ensure Layer 2 runs in softint (e.g., ether_input and bridge_input). To this end, we provide a framework (called percpuq) that utlizes softint(9) and percpu ifqueues. With this patch, rxintr of most drivers just queues received packets and schedules a softint, and the softint dequeues packets and does rest packet processing. To minimize changes to each driver, percpuq is allocated in struct ifnet for now and that is initialized by default (in if_attach). We probably have to move percpuq to softc of each driver, but it's future work. At this point, only wm(4) has percpuq in its softc as a reference implementation. Additional information including performance numbers can be found in the thread at tech-kern@ and tech-net@: http://mail-index.netbsd.org/tech-kern/2016/01/14/msg019997.html Acknowledgment: riastradh@ greatly helped this work. Thank you very much!
1.15	17-Aug-2015	knakahara	Add kernel code to support intrctl(8).
1.14	14-Aug-2015	martin	Make it compilable without DIAGNOSTIC
1.13	14-Aug-2015	msaitoh	- Fix uninitialized data. - include vlan.h and check NVLAN.
1.12	13-Aug-2015	msaitoh	Reduce diff against FreeBSD r280181.
1.11	13-Aug-2015	msaitoh	- Add MSI/MSI-X support. The multiqueue function is not supported yet. - Make ixv.c compilable. _NOT_TESTED_YET_
1.10	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.9	21-May-2015	rtr	remove duplicated assignment of mh->m_len
1.8	24-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes.
1.7	14-Apr-2015	msaitoh	Sync ixg(4) up to FreeBSD r243716: - A lot of bugfixes. Some of them are realted to multi queue and those have not affected in NetBSD because we have not used it yet. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Improve performance.
1.6	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.5	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.4	19-Mar-2015	msaitoh	Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes.
1.3	10-Mar-2015	msaitoh	Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X.
1.2	25-Mar-2014	christos	branches: 1.2.4; 1.2.6; kill sprintf
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; 1.1.16; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.16.1	18-May-2014	rmind	sync with head
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.2.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.2.6.8	28-Aug-2017	skrll	Sync with HEAD
1.2.6.7	05-Feb-2017	skrll	Sync with HEAD
1.2.6.6	05-Dec-2016	skrll	Sync with HEAD
1.2.6.5	09-Jul-2016	skrll	Sync with HEAD
1.2.6.4	19-Mar-2016	skrll	Sync with HEAD
1.2.6.3	22-Sep-2015	skrll	Sync with HEAD
1.2.6.2	06-Jun-2015	skrll	Sync with HEAD
1.2.6.1	06-Apr-2015	skrll	Sync with HEAD
1.2.4.4	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.4.3	06-May-2015	riz	Pull up following revisions (requested by msaitoh in ticket #734): sys/dev/pci/ixgbe/LICENSE 1.2 sys/dev/pci/ixgbe/ixgbe.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe.h 1.7 sys/dev/pci/ixgbe/ixgbe_82598.c 1.4 sys/dev/pci/ixgbe/ixgbe_82598.h 1.3 sys/dev/pci/ixgbe/ixgbe_82599.c 1.9 sys/dev/pci/ixgbe/ixgbe_82599.h 1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.7 sys/dev/pci/ixgbe/ixgbe_api.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.5 sys/dev/pci/ixgbe/ixgbe_common.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.4 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.3 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.7 sys/dev/pci/ixgbe/ixgbe_phy.c 1.5 sys/dev/pci/ixgbe/ixgbe_phy.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.10 sys/dev/pci/ixgbe/ixgbe_vf.c 1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.5 sys/dev/pci/ixgbe/ixgbe_x540.c 1.3 sys/dev/pci/ixgbe/ixgbe_x540.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.8 - Sync ixg(4) up to FreeBSD r250108: - Cleanup some unused counters and some unused code. - Improve performance. - Fix flow control - don't override user value on re-init - Fix to make 1G optics work correctly - Change to interrupt enabling - some bits were incorrect for certain hardware. - Certain stats fixes, remove a duplicate increment of ierror, thanks to Scott Long for pointing these out. - Fix the setting of RX which related to multicast. - Some netmap related fixes. - Fixes link error without vlan. PR#49879 reported by Hauke Fath. [msaitoh, ticket #734]
1.2.4.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.4.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.17.2.2	20-Mar-2017	pgoyette	Sync with HEAD
1.17.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.31.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.56.2.44	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1933: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.56.2.43	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1914: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 via patch - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.56.2.42	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1912): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.56.2.41	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe.c 1.327-1.332 via patch sys/dev/pci/ixgbe/ixgbe.h 1.87-1.88 sys/dev/pci/ixgbe/ixv.c 1.184-1.185 sys/dev/pci/ixgbe/ix_txrx.c 1.101 sys/dev/pci/ixgbe/ixgbe_82599.c 1.30 sys/dev/pci/ixgbe/ixgbe_vf.c 1.32-1.33 sys/dev/pci/ixgbe/ixgbe_vf.h 1.18 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.20 sys/dev/pci/ixgbe/ixgbe_type.h 1.57 - Reorder some event counters for readability. - Rename some descriptions of event counters. - Count Queue Bytes {Transmit, Receive} counter. - Improve error check in ixgbe_check_mac_link_vf(). - Add new IXGBE_VF_GET_LINK_STATE message support. The VF's link state can be forced to down by PF. - Update FCTRL after writing multicast filter. - Update comments.
1.56.2.40	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1796: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.56.2.39	03-Jun-2022	martin	Pull up the following revisions, requestes by msaitoh in ticket #1746: sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.261,1.263, 1.265-1.268,1.273, 1.275-1.277,1.305, 1.312, 1.316-1.321 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.178,1.182 via patch - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Use macro. - Fix typos in comment. - KNF.
1.56.2.38	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.98 sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 bus_dmamem_unmap() before bus_dmamem_free(), otherwise we may give back meomry which is still (and will stay) mapped. Fixes one instance of "panic: HYPERVISOR_mmu_update failed" on Xen. There may be others. Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.56.2.37	30-May-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1744: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.56.2.36	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1730): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.56.2.35	30-Jan-2022	martin	Pull up the following, requested by msaitoh in ticket #1729: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.56.2.34	29-Jan-2022	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1725: sys/dev/pci/ixgbe/ixgbe_vf.h 1.16-1.17 sys/dev/pci/ixgbe/ixv.c 1.176-1.177 Make ifconfig -z ixvN clear event counter.
1.56.2.33	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1708: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.56.2.32	15-Sep-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1696: sysdev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sysdev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sysdev/pci/ixgbe/ix_txrx.c 1.68-1.93 sysdev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sysdev/pci/ixgbe/if_bypass.c 1.7-1.9 sysdev/pci/ixgbe/if_fdir.c 1.4-1.5 sysdev/pci/ixgbe/if_sriov.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_82598.c 1.16 sysdev/pci/ixgbe/ixgbe_82599.c 1.23 sysdev/pci/ixgbe/ixgbe_api.c 1.25 sysdev/pci/ixgbe/ixgbe_bypass.h 1.2 sysdev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sysdev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sysdev/pci/ixgbe/ixgbe_dcb.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sysdev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sysdev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sysdev/pci/ixgbe/ixgbe_fdir.h 1.3 sysdev/pci/ixgbe/ixgbe_features.h 1.3 sysdev/pci/ixgbe/ixgbe_mbx.c 1.12 sysdev/pci/ixgbe/ixgbe_netbsd.c 1.16-1.17 sysdev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sysdev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sysdev/pci/ixgbe/ixgbe_netmap.h 1.2 sysdev/pci/ixgbe/ixgbe_osdep.c 1.7 sysdev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sysdev/pci/ixgbe/ixgbe_phy.c 1.24 sysdev/pci/ixgbe/ixgbe_rss.h 1.5 sysdev/pci/ixgbe/ixgbe_sriov.h 1.4 sysdev/pci/ixgbe/ixgbe_type.h 1.49 sysdev/pci/ixgbe/ixgbe_vf.c 1.27 sysdev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sysdev/pci/ixgbe/ixgbe_x540.h 1.9 sysdev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sysdev/pci/ixgbe/ixgbe_x550.h 1.6 sysdev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.56.2.31	11-Mar-2021	martin	Pull up the following (all via patch), requested by msaitoh in ticket #1663: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.3, 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.56.2.30	05-Aug-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #1587: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.56.2.29	24-Jan-2020	martin	Pull up the following, requested by msaitoh in ticket #1490: sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.195,1.220-1.221 via patch sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.122,1.142,1.144 via patch - Use unsigned to avoid undefined behavior in ix{gbe,v}_[un]register_vlan(). - Free RX structure correctly when detaching. - Remove unused code. - Remove extra spaces. - Fix some typos in comment. - KNF.
1.56.2.28	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1451): sys/dev/pci/ixgbe/ixgbe.c: revision 1.216 sys/dev/pci/ixgbe/ixv.c: revision 1.141 Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.56.2.27	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1435): sys/dev/pci/ixgbe/ixv.c: revision 1.140 ixv(4): disable RSS configuration on 82599 and X540 VFs. Those VFs share their RSS configuration with PF and, thus, they cannot be configured independently. From FreeBSD r354349.
1.56.2.26	08-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1403): sys/dev/pci/ixgbe/ixv.c: revision 1.139 Set MTU correctly if mtu > 1500.
1.56.2.25	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1389: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 via patch sys/dev/pci/ixgbe/ixv.c 1.131-1.138 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 via patch share/man/man4/ixv.4 1.5 via patch - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others.
1.56.2.24	05-Sep-2019	martin	Apply patch, requested by msaitoh in ticket #1367, to pull up the following revisions: sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixgbe.c 1.180,1.203-1.204, 1.207-1.208 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.55 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.10 sys/dev/pci/ixgbe/ixgbe_common.c 1.25 sys/dev/pci/ixgbe/ixv.c 1.129-1.130 - X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. - An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC softint. - Add 10000BASE-LX media if it's 1000BASE-BX. - printf -> device_printf - Avoid undefined behavior of VLAN filter setting. - Simplify code. - Fix typo in unused code. - Fix typo in comment. Found by Wataru Ashihara.
1.56.2.23	01-Aug-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #1313: sys/dev/pci/ixgbe/ixgbe_type.h 1.41 sys/dev/pci/ixgbe/ixgbe.c 1.194,1.197-1.199 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.24 sys/dev/pci/ixgbe/ixv.c 1.121,1.124-1.125 - Avoid undefined behavior of interrupt vector setting. - Avoid undefined behavior of TX/RX queue statistics calculation. - Avoid undefined behavior of X550EM's PHY accesses. - Avoid undefined behavior of Flow Control Transmit Timer setting. - Don't call {ixgbe,ixv}_stop() twice while detaching.
1.56.2.22	22-Jul-2019	martin	Pull up the following revisions (via patch), requested by msaitoh in ticket #1301: sys/dev/pci/ixgbe/if_sriov.c 1.5-1.6 sys/dev/pci/ixgbe/ix_txrx.c 1.53-1.54 sys/dev/pci/ixgbe/ixgbe_x550.h 1.5 sys/dev/pci/ixgbe/ixgbe.c 1.169-1.170,1.176,1.179,1.181,1.185-1.186,1.188-1.192 via patch sys/dev/pci/ixgbe/ixgbe.h 1.53,1.55 sys/dev/pci/ixgbe/ixgbe_82599.c 1.21 sys/dev/pci/ixgbe/ixgbe_api.c 1.22-1.23 sys/dev/pci/ixgbe/ixgbe_api.h 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_common.c 1.23 sys/dev/pci/ixgbe/ixgbe_common.h 1.14 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.11 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.2 sys/dev/pci/ixgbe/ixgbe_features.h 1.2 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.9 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.9-1.10 sys/dev/pci/ixgbe/ixgbe_phy.c 1.18 sys/dev/pci/ixgbe/ixgbe_type.h 1.37,1.39-1.40 sys/dev/pci/ixgbe/ixgbe_vf.c 1.18 sys/dev/pci/ixgbe/ixv.c 1.112-1.114,1.117,1.119 via patch Sync ixgbe up to 20190717 except ETHERCAP or ixv's VLAN stuff: - Add firmware recovery mode for X550, X550A(Xeon D) and X550EM(C3000). - Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. - On X550EMU, use ixgbe_identify_sfp_module_X550em() instead of ixgbe_identify_module_generic(). ixgbe_identify_sfp_module_X550em() has extra check (e.g. exclude 1G copper). - It's not required to calculate unused queues' statistics. - Remove ETHERCAP_VLAN_HWFILTER's definition. - Match 82598_BX(0x1508), 82599_KR(0x1517), 82599_SFP_EM(0x1507), X550EM_X_XFI(0x15b0), X550EM_A_QSFP(0x15ca) and X550EM_A_QSFP_N(0x15cc) - Add missing XFI support into ixgbe_get_link_capabilities_X550em(). - Other than IXGBE_VF_RESET should wait ACK, so use ixgbevf_write_msg_read_ack() instead of write_posted() in ixgbe_update_mc_addr_list_vf(). - When ixv_check_link() failed in the watchdog function, reset the interface. - Remove RXCSUM register modification in ixv_initialize_receive_units(). It seems it's not required. - Remove some debug printf in ixv_print_debug_info(). - Calculate vector's bit location correctly when the vector >= 31 in ixgbe_allocate_msix(). - Fix hung queue check when the queue number >= 31. - On ENETRESET case, not continue and quit the ifflags_cb() function because if_init() will do the same thing. - Fix bugs in unused code. - Fix typo in comment. - Fix typo in unused code. - Whitespace fixes. KNF.
1.56.2.21	01-Apr-2019	martin	Pull up the following, requested by msaitohin ticket #1225: sys/dev/pci/ixgbe/ixgbe.c 1.175-1.178 sys/dev/pci/ixgbe/ixv.c 1.110-1.111 via patch - NetBSD currently uses traffic class 0 only. Other traffic classes aren't used yet. When IXGBE_TC_COUNTER_NUM is set to lower than IXGBE_DCB_MAX_TRAFFIC_CLASS (e.g. 1), other traffic classes' counters are not used. It means we don't generate evcnt for them and don't add the values in ixgbe_update_stats_counters(). - It's not required to calculate unused queues' statistics. - Fix a bug that the VLAN HW tagging function is not correctly disabled when all vlan is detached. - Fix a bug that VLAN HW tagging function is not correctly controlled on 82598. - Control VLAN HW filter function correctly. Note that currently VLAN HW filter function doesn't work because NetBSD doesn't support it yet. - Don't clear IXGBE_VLNCTRL_CFIEN bit When ETHERCAP_VLAN_HWFILTER is set. I think it's not required (and Linux doesn't do it). This change has no effect to NetBSD because ETHERCAP_VLAN_HWFILTER is not supported yet.
1.56.2.20	01-Mar-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1199): sys/dev/pci/ixgbe/ixv.c: revision 1.109 sys/dev/pci/ixgbe/ixgbe.h: revision 1.54 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.52 sys/dev/pci/ixgbe/ixgbe.c: revision 1.174 Fix a bug that if_link_state_change(ifp, LINK_STATE_DOWN) isn't called unless link goes up at least one time. Without this change, never linkup-ed media keeps LINK_STATE_UNKNOWN instead of LINK_STATE_DOWN.
1.56.2.19	27-Jan-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #1172): sys/dev/pci/nvme_pci.c: revision 1.26 sys/dev/pci/xhci_pci.c: revision 1.21 sys/dev/pci/ixgbe/ixv.c: revision 1.108 sys/dev/pci/ixgbe/ixgbe.c: revision 1.171 sys/dev/pci/if_fxp_pci.c: revision 1.84 sys/dev/pci/if_fxp_pci.c: revision 1.85 sys/dev/pci/xhci_pci.c: revision 1.16 remove #ifdef DEBUG printf, it seems to have outlived it's usefulness - KNF. No functional change. - Nowadays some UEFI BIOSes don't enable some PCI devices' address decoding. To resolve this problem, pci_map.c rev. 1.34-1.36 changed the pci_mapreg_(sub)map()'s to set the decode bit if it's not set. It's good for almost all drivers, but some other drivers don't use pci_mapreg_map(). In drivers which don't use pci_mapreg_map(), some of them explicitly enable decoding but others don't. Add code to enable decoding to them. See also the following discussion: http://mail-index.netbsd.org/tech-kern/2017/03/22/msg021678.html
1.56.2.18	08-Nov-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #1084): sys/dev/pci/ixgbe/ixv.c: revision 1.107 sys/dev/pci/ixgbe/ixgbe.c: revision 1.167 Don't set LRO capability flag because we don't support it yet. This doesn't change any behavior other than if_capabilities' visibility (e.g. ifconfig).
1.56.2.17	09-Jun-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #864): sys/dev/pci/ixgbe/ix_txrx.c 1.40-1.47 (patch) sys/dev/pci/ixgbe/ixgbe.c 1.148,1.149,1.151, 1.152,1.154, 1.155,1.157-1.160 (patch) sys/dev/pci/ixgbe/ixgbe.h 1.43,1.44,1.46,1.49 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.7 (patch) sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.8 (patch) sys/dev/pci/ixgbe/ixgbe_osdep.h 1.22 (patch) sys/dev/pci/ixgbe/ixv.c 1.100-1.104 (patch) sys/dev/pci/ixgbe/ixv.c 1.94,1.95,1.99 (patch) Remove unused structure entries. No functional change. - Remove unused IXGBE_FC_HI and IXGBE_FC_LO. The watermark of the flow control is automatically calculated from the size of the packet buffer. - Use ixgbe_eitr_write() when writing the EITR for the link interrupt like queue's EITR to write the register safely. This change is not relatively so important than queue's EITR because link's EITR is written in if_init(). - Don't free and reallocate bus_dmamem when it's not required. Currently, the watchdog timer is completely broken and never fire (it's from FreeBSD (pre iflib)). If the problem is fixed and watchdog fired, ixgbe_init() always calls ixgbe_jcl_reinit() and it causes panic. The reason is that ixgbe_local_timer1(it includes watchdog function) is softint and xgbe_jcl_reinit() calls bus_dmamem() functions. bus_dmamem() can't be called from interrupt context. One of the way to prevent panic is use worqueue for the timer, but it's not a small change. (I'll do it in future). Another way is not reallocate dmamem if it's not required. If both the MTU (rx_mbuf_sz in reality) and the number of RX descriptors are not changed, it's not required to call bus_dmamem_{unmap,free}(). Even if we use workque, this change save time of ixgbe_init(). I have a code to fix broken watchdog timer but it sometime causes watchdog timeout, so I don't commit it yet. - Count some register correctly: - QPRDC register is only for 82599 and newer. - Count IXGBE_QPRDC, PX{ON,OFF}{T,R}XC[NT]. The TQSMR register is not for receiving but for transmitting, so move the initialization from ixgbe_initialize_receive_units() to ixgbe_initialize_transmit_units(). No functional change. - Whitespace fix. No functional change. - Add rxd_nxck (Receive Descriptor next to check) read only sysctl. Don't check IFF_RUNNING in ixgbe_rxeof(). Doing break and leaving a deacriptor with DD bit is worse than just processing the entry. And it's also racy to check IFF_RUNNING in rxeof(). If you'd like to strictly obey IFF_RUNNING, it would be better to do it in the upper layer. Same as DragonFly (a part of 79251f5ebe4cf9dd2f3e6aed590e09d756d39922). Add "bool txr_no_space" for TX descriptor shortage. Use it like IFF_OACTIVE. Clear que->disabled_count in {ixgbe,ixv}_init_locked(). Without this, interrupt mask state and EIMS may mismatch and if_init doesn't recover from TX/RX stall problem. This change itself doesn't fix PR#53294. - Add hw.ixgN.debug sysctl. "sysctl -w hw.ixgN.debug=1" dumps some registers to console. - Constify several variables in ixgbe/ so that they land in .rodata (1038 bytes). - Don't call ixgbe_rearm_queues() in ixgbe_local_timer1(). ixgbe_enable_queue() and ixgbe_disable_queue() try to enable/disable queue interrupt safely. It has the internal counter. When a queue's MSI-X is received, ixgbe_msix_que() is called (IPL_NET). This function disable the queue's interrupt by ixgbe_disable_queue() and issues an softint. ixgbe_handle() queue is called by the softint (IPL_SOFTNET), process TX, RX and call ixgbe_enable_queue() at the end. ixgbe_local_timer1() is a callout and run always on CPU 0 (IPL_SOFTCLOCK). When ixgbe_rearm_queues() called, an MSI-X interrupt is issued for a specific queue. It may not CPU 0. If this interrupt's ixgbe_msix_que() is called and sofint_schedule() is called before the last sofint's softint_execute() is not called, the softint_schedule() fails because of SOFTINT_PENDING. It result in breaking ixgbe_{enable,disable}_queue()'s internal counter. ixgbe_local_timer1() is written not to call ixgbe_rearm_queues() if the interrupt is disabled, but it's called because of unknown bug or a race. One solution to avoid this problem is to not to use the internal counter, but it's little difficult. Another solution is stop using ixgbe_rearm_queues() at all. Essentially, ixgbe_rearm_queues() is not required (it was added in ixgbe.c rev. 1.43 (2016/12/01)). ixgbe_rearm_queues() helps for lost interrupt problem but I've never seen it other than ixgbe_rearm_queues() problem. XXX pullup-8.
1.56.2.16	07-Jun-2018	martin	Pull up following revision(s) (requested by ozaki-r in ticket #843): sys/dev/pci/ixgbe/ixv.c: revision 1.101 sys/net/if_bridge.c: revision 1.156 sys/net/if_pppoe.c: revision 1.138 sys/dev/pci/if_wm.c: revision 1.580 sys/dev/pci/ixgbe/ixgbe.c: revision 1.156 sys/net/if_gif.c: revision 1.142 Ensure to call if_register after interface initializations finish
1.56.2.15	15-May-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #824): sys/dev/pci/ixgbe/ixv.c: revision 1.97 Fix panic or hangup when "sysctl -w hw.ixgN.debug=1". XXX pullup-8
1.56.2.14	14-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #750): sys/dev/pci/ixgbe/ixgbe_x540.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82598.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: revision 1.6 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.18 sys/dev/pci/ixgbe/if_sriov.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.7 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_dcb.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.91 sys/dev/pci/ixgbe/ixgbe_rss.h: revision 1.4 sys/dev/pci/ixgbe/ixv.c: revision 1.92 sys/dev/pci/ixgbe/ixv.c: revision 1.93 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.4 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.12 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.33 sys/dev/pci/ixgbe/if_fdir.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.34 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.39 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.13 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.21 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.22 sys/dev/pci/ixgbe/ixgbe_82599.h: revision 1.7 sys/dev/pci/ixgbe/ixgbe.c: revision 1.144 sys/dev/pci/ixgbe/ixgbe.c: revision 1.145 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe.c: revision 1.146 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_82598.h: revision 1.8 sys/dev/pci/ixgbe/ixgbe.h: revision 1.41 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_phy.h: revision 1.11 Sync with the remaining part of FreeBSD r328265 except sfp_reinit stuff: - Always schedule module intterrupt in ixgbe_config_link() when a device is SFP+ based. - Use not loop index but txr->me in ixv_initialize_{transmit,receive}_units(). It's required for VMDQ but NetBSD doesn't use it, so it's not a bug in NetBSD. - Simplify ixgbe_bp_wd_set(). No functional change. - Whitespace. Sync with FreeBSD r331224 except ixv_if_update_admin_status()'s change: - Fix length of reading buffer when the command is Read Flash in ixgbe_host_interface_command(). - Add missing start_hw() call in ixv_init_locked(). This is not a real bug because start_hw just set hw->adapter_stopped to false and anyone refer it. - Style change. - Update comment. Read sc_if_flags after taking core lock. Same as if_wm.c rev. 1.418.
1.56.2.13	04-Apr-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #690): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe.c: revision 1.138 sys/dev/pci/ixgbe/ixv.c: revision 1.90 sys/dev/pci/ixgbe/ixgbe.c: revision 1.139 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.19 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.11 sys/dev/pci/ixgbe/ixgbe.h: revision 1.38 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe.h: revision 1.39 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.11 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.38 sys/dev/pci/ixgbe/ixgbe.c: revision 1.140 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.141 sys/dev/pci/ixgbe/ixgbe.c: revision 1.142 sys/dev/pci/ixgbe/ixgbe.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe.h: revision 1.40 sys/dev/pci/ixgbe/ixgbe_x550.h: revision 1.4 Fix the problem between eitr and link_speed. In ixgbe_msix_que(), que->eitr_setting is limited to IXGBE_MIN_RSC_EITR_10G1G when link_speed is 1Gbps or 10Gbps. However, que->eitr_setting is set to EITR register in the next Tx/Rx interrupt. If link_speed changes from 100Mbps to 1Gbps ro 10Gbps, que->eitr_setting which is not limited can be set to EITR register, that is, the problem fixed by ixgbe.c:r1.124 can occur in this case. To fix this case, que->eitr_setting should be clear when link_speed is changed or link state is changed. Furthermore, expand the variants used for AIM (txr->bytes, txr->packets, rxr->bytes and rxr->packets) from u32 to u64 to avoid wraparound which causes que->eitr_setting calculation mistake. XXX pullup-8 Don't write EIMC directly. It is required to manage with struct ix_queue status. XXX pullup-8 - Add missing IFM_NONE support. If a interface support linkdown, "ifconfig ixgN media none" drpos link. Not all interface can do link down. Tested: 82598 AT2 (T) 92599 SF+(SFI) (X520-DA2) X540 X550-T1 X550EM_x (X10SDV-8C-TLN4F) X550EM_a (A2SDi-H-TP4F port 0, 1 (T)) Doesn't work: X550EM_a (A2SDi-H-TP4F port 2, 3 (SFP+ (KR))) X550EM_a (MA10-ST0 port 2, 3 (SFP+ (SFI))) (Denverton SFP+ can't force link down because SFP+'s TX_DISABLE pin is pull down. Is there a way to shutdown SFP+ cage's power?) Not tested: 82598 fiber. - Change some functions static. Avoid issues caused by sending old packets at next link-up time. This modification consists by the following two parts. - drain packets in if_snd queue or corresponding txr->txr_interq when link_active == false in ifp->if_start(), ifp->if_transmit(), and deferred Tx processing - drain packets in if_snd queue and all of txr->txr_interq's at link-down time ok by msaitoh@n.o. Fix long standing bug that 82598 SFP+ panics in ixgbe_handle_mod() because hw->mac.ops.setup_sfp is NULL. This change is a part of FreeBSD r327031. Fix a bug that "ifconfig ixgN down up" forgot IFM_NONE setting. - Don't assume autoneg == 0 is the first call of ixgbe_config_link(). Check ifm_media, too. - Don't override autoneg_advertised in ixgbe_get_phy_id_fw() to not to be inconsistent with if_media value.
1.56.2.12	30-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #673): sys/dev/pci/ixgbe/if_sriov.c: revision 1.2 sys/dev/pci/ixgbe/ixgbe.c: revision 1.135 sys/dev/pci/ixgbe/ixgbe.c: revision 1.136 sys/dev/pci/ixgbe/ixgbe.c: revision 1.137 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.17 sys/dev/pci/ixgbe/if_bypass.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.10 sys/dev/pci/ixgbe/ixgbe.h: revision 1.36 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.3 sys/dev/pci/ixgbe/ixgbe.h: revision 1.37 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.36 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.12 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.20 sys/dev/pci/ixgbe/ixv.c: revision 1.88 sys/dev/pci/ixgbe/ixgbe_phy.c: revision 1.15 sys/dev/pci/ixgbe/ixv.c: revision 1.89 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.8 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.9 Add some changes from ix-3.2.17.tar.gz and r328265. Not fully synchronized. Some others (e.g. sfp cage interrupt and bypass adapter stuff) will be merged later: - Initialize firmware command buffer correctly in ixgbe_read_ee_hostif_X550() and ixgbe_read_ee_hostif_buffer_X550(). These functions are used when reading NVM. - Fix a bug that ixgbe_mng_present() misunderstand management capability on X550 and newer on some environment. X550 changed FWSM bit definition. See X540 document and X550's document and compare them. - Fix checksum calculation in ixgbe_set_fw_drv_ver_generic(). This function is not used in NetBSD. - Add some unused funtions. - Whitespace fix. - Check offset correctly in ixgbe_get_oem_prod_version(). Note that this function is not used. - Set PHY correctly in ixgbe_setup_mac_link_sfp_x550a() if a device is a C3000 KR SFP+. Fix race about writing adapter->link_active for ixg(4). adapter->link_active is updated by ixgbe_update_link_status() only. The function is called from the following four functions. - ixgbe_media_status() - ixgbe_local_timer1() - ixgbe_stop() - ixgbe_handle_link() The functions other than ixgbe_handle_link() call ixgbe_update_link_status() with holding IXGBE_CORE_LOCK, however ixgbe_handle_link() calls it without holding IXGBE_CORE_LOCK. That can cause race. So, add IXGBE_CORE_LOCK to ixgbe_handle_link(). Tested by msaitoh@n.o and me. Fix race about writing adapter->link_active for ixv(4). adapter->link_active is updated by ixv_update_link_status() only. The function is called from the following two functions. - ixv_media_status() - ixv_handle_link() ixv_media_status() calls ixv_update_link_status() with holding IXGBE_CORE_LOCK, however ixv_handle_link() calls it without holding IXGBE_CORE_LOCK, the same as ixg(4). ok by msaitoh@n.o. - Add link related softint's counter. - Fix indent.
1.56.2.11	13-Mar-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #626): sys/dev/pci/ixgbe/ixgbe.c: revision 1.134 sys/dev/pci/ixgbe/ixgbe.h: revision 1.35 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.35 sys/dev/pci/ixgbe/ixv.c: revision 1.87 Make some event counters MP safe. Now all of the event counters are MP safe.
1.56.2.10	08-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #612): sys/dev/pci/ixgbe/ixgbe.c: revision 1.129-1.133 sys/dev/pci/ixgbe/ixgbe.h: revision 1.34 sys/dev/pci/ixgbe/ixv.c: revision 1.85,1.86 - Make "Handled queue in softint" and "Requeued in softint" evcnt(9) per queue and count them correctly. - Remove #if 0'ed code. Don't increment que->req.ev_count in MSI-X interrupt because it's not reschedule. Fix another poll mode assumption breaking. Implemented by msaitoh@n.o, I just commit by proxy. ixgbe_rearm_queues() writes EICS register(s). 82599, X540 and X550 specifications say "Following a write of 1b to any bit in the EICS register (interrupt cause set), its corresponding bit in the EIMS register is auto set as well enabling its interrupt." in "Extended Interrupt Auto Mask Enable (EIAM) Register" section. That is, ixgbe_rearm_queues() causes interrupts regardless of the status managed by ixgbe_enable_queue()/ixgbe_disable_queue(). That can break poll mode assumption. In fact, the problem occurs in the following situation - CPU#A has high load traffic, in contrast, CPU#B has not so high load traffic - CPU#A is occurred interrupt by its NIC queue - CPU#A calls ixgbe_disable_queue() in interrupt handler(ixgbe_msix_que()) - CPU#A kick softint handler(ixgbe_handle_que()) - CPU#A begins softint - CPU#A's NIC queue is set que->txr->busy flag - With some reason, CPU#A can do ixg interrupt handler E.g. when one of CPU#A's softnet handlers sleeps, ipl is lowered - CPU#B starts callout - CPU#B calls ixgbe_local_timer1() - CPU#B writes EICS bit corresponding CPU#A's NIC queue bit - CPU#A's NIC queue causes interrupt whie CPU#A is running in poll mode - CPU#A calls ixgbe_disable_queue() in interrupt handler again - CPU#A has done polling, and then CPU#A calls ixgbe_enable_queue() once - CPU#A's NIC queue interrupt is disabled until ixg is detached as ixgbe_disable_queue() is called twice though ixgbe_disable_queue() is called once only NOTE: 82598 does not say so, but it is treated in the same way because of no harm. By the way, we will refactor ixgbe_local_timer(watchdog processing) later. Fix INTx/MSI handler did not schedule workqueue. Pointed out by msaitoh@n.o. Reduce duplicated code which schedule deferred packet processing. No functional change.
1.56.2.9	06-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #605): sys/dev/pci/ixgbe/ixgbe.h: revision 1.32 sys/dev/pci/ixgbe/ixgbe.h: revision 1.33 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.34 sys/dev/pci/ixgbe/ixgbe.c: revision 1.128 sys/dev/pci/ixgbe/ixv.c: revision 1.83 sys/dev/pci/ixgbe/ixv.c: revision 1.84 Add hw.ixvM.q[01].{interrupt_rate,[tr]xd_head,[tr]xd_tail} sysctls as ixg(4). ixg(4) supports workqueue poll mode, but not enabled by default yet. (that is, the default behavior is not changed) At the time of high load near the wire rate, the turnaround time of update/delete processing such as "ifconfig ixg0 inet XXX" or "ifconfig ixg0 delete" is very long. The main reason is CPU starvation caused by ixg(4)'s softint poll mode. ixg(4) uses workqueue poll mode instead of softint poll mode, so that this problem will be fix. This change may cause performance issues, so it is not enabled by default yet. Although there are that issues, if you want to use workqueue poll mode, do "sysctl -w hw.ixgXX.txrx_workqueue=1" while there is no traffic on the ixgXX. ok by msaitoh@n.o. ixv(4) also supports workqueue poll mode, but not enabled by default yet, either. ok by msaitoh@n.o. Move the location of "struct work" as FreeBSD's "struct task" in ixgbe.h. No functional change.
1.56.2.8	01-Mar-2018	martin	Pull up following revision(s) (requested by knakahara in ticket #597): sys/dev/pci/ixgbe/ixgbe.h: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.c: revision 1.127 sys/dev/pci/ixgbe/ixv.c: revision 1.82 Fix poll mode assumption breaking. ixgbe_{enable,disable}_intr() forcibly enable/disable all interrupts regardless of current state. That can break poll mode assumption, that is, queue interrupts must not occur while polling Tx/Rx rings. E.g. "ifconfig ixg0 delete && ifconfig ixg0 192.168.0.1" on heavy load traffic can causes this issue. This fix may have 1% or 2% performance impact at short packets. XXX ixgbe_rearm_queues() which is called only via watchdog can also break this poll mode assumption because writing EICS casues interrupts immediately when interrupt auto mask enabled. We will fix it with other issues about watchdog later. ok by msaitoh@n.o. Apply ixgbe.c:r1.127 to ixv.c. Pointed out by msaitoh@n.o.
1.56.2.7	26-Feb-2018	martin	Pull up following revision(s) (requested by msaitoh in ticket #592): sys/dev/pci/ixgbe/ixv.c: revision 1.79 sys/dev/pci/ixgbe/ixgbe.h: revision 1.30 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.31 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.31 sys/dev/pci/ixgbe/ixgbe.c: revision 1.120 sys/dev/pci/ixgbe/ixgbe.c: revision 1.121 sys/dev/pci/ixgbe/ixgbe.c: revision 1.123 sys/dev/pci/ixgbe/ixgbe.c: revision 1.124 sys/dev/pci/ixgbe/ixgbe.c: revision 1.125 sys/dev/pci/ixgbe/ixgbe.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.80 sys/dev/pci/ixgbe/ixv.c: revision 1.81 CID-1427719: Integer handling issues (BAD_SHIFT). Print bus/slot info correctly on 82599_SFP_SF_QP(DID 0x154a) and 82599_QSFP_SF_QP(DID 0x1558). Fix a bug that RX may stall on heavy load on ixg(4). ixgbe_rxeof() has loop limit and the function returns true if a packet are still in the RX ring. ixgbe_handle_que() didn't check the return value. Check the return vaule and issue a softint. This bug is derived from FreeBSD and ixv(4) has no this bug. XXX pullup-8 Fix a bug that the first call of ifflags_cb() causes linkdown. The first call of ix(gbe\|v)_ifflags_cb() refered uninitialized adapter->if_flags. adapter->if_flags should be initialized in the end of xxx_init(). XXX pullup-[678] (ixgbe) XXX pullup-8 (ixv) - Fix a bug that RX may stall on heavy load on ixg(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. When I use a machine as a NFS client, sometimes one of queue pairs doesn't get any interrupt other than every second tick via ixgbe_local_timer1(). When the problem occured, the queue pair's hw.ixgM.qN.interrupt_rate is always 500000. When this problem occuring, set hw.ixgM.qN.interrupt_rate lower than 166667 recover from stall. i.e.: sysctl -w hw.ixgM.qN.interrupt_rate=166667 (don't revocer) sysctl -w hw.ixgM.qN.interrupt_rate=166666 (recover) Relatios between the interrupt_rate and EICR's ITR_INTERVAL field is as follows: int_rate \| EICR[11:0] \| interval in us \| recover \| \|(ITR_INTERVAL)\| (10G and 1G) \| \| ---------+--------------+----------------+---------+ 500000 \| 0x008(0) \| 2 \| not \| 166667 \| 0x010(1) \| 4 \| not \| 166666 \| 0x018(2) \| 6 \| recover \| The reason why int_rate becomes 500000 is that xgbe_tx_eof() doesn't increment rxr->packets(1). Even if we fix rxr->packets' bug, interrupt_rate might become greater than 166666 and it might cause stall. While reading datasheets, knakahara noticed a section titled with "ITR Affect on RSC Functionality". It says "When RSC is enabled on specific RX queues, the associated ITR interval with these queus must be enabled and must be larger (in time uints) than RSC delay". Currently, RSC_DELAY field in the GPIE register is 0 and it means 4us for 10G and 1G. The greater ITR_INTERVAL value of 4us is 6us == 166666. Yes, BINGO! This description is noted in 82599 and newer datasheets and not in 82598 datasheet. I don't know if 82598 has this limitation but, I apply this limitation all of chips. (1) Note that this bug is going to be fixed in the next commit to distinct between two different bugs. - The bitfield of EITR register is different between 82598 and others. Only ixgbe_msix_que() taken care of it. Make new function ixgbe_eitr_write() and use it in all of functions which modify ITR_INTERVAL. XXX pullup-8 Increment rxr->packets correctly in ixgbe_rxeof() to calculate ITR value of AIM (Auto Interrupt Moderation) correctly. See also ixgbe.c rev. 1.124. XXX pullup-8 Improve a comment about reading EICS register defined write-only by spec. It seems that is workaround for silicon errata. ok by msaitoh@n.o. - Apply ixgbe.c rev. 1.124 to ixv.c. Fix a bug that RX may stall on heavy load on ixv(4) derived from FreeBSD's AIM (Auto Interrupt Moderation) bug. ITR_INTERVAL value must be larger than 4us. - The bitfield of EITR register is different between 82598 and others. ixv.c had a bug that it accessed 82598's way even though only 82599 and newer support virtual function. Fix it using with new ixv_eitr_write() function. Fix a potential bug that TX/RX might stall when TX rate limit reached. This change is almost the same as the RX rate limit bug fix in ixgbe.c rev. 1.121. I've never got any stall, but this must be a bug.
1.56.2.6	26-Feb-2018	snj	Pull up following revision(s) (requested by knakahara in ticket #567): distrib/sets/lists/comp/mi: 1.2182-1.2183 sys/dev/pci/if_wm.c: 1.564 sys/dev/pci/ixgbe/ixgbe.c: 1.122 sys/dev/pci/ixgbe/ixgbe_rss.h: 1.3 sys/dev/pci/ixgbe/ixv.c: 1.78 sys/net/Makefile: 1.35-1.36 sys/net/files.net: 1.15 sys/net/rss_config.c: 1.1 sys/net/rss_config.h: 1.1 Introduce very simple Receive Side Scaling (RSS) utility. ok by msaitoh@n.o. -- Apply RSS utility to wm(4). ok by msaitoh@n.o. -- Apply RSS utility to ixg(4) and ixv(4). ok by msaitoh@n.o. -- Fix build failure, sorry. -- Currently, it is not necessary to install rss_config.h. Pointed out by msaitoh@n.o.
1.56.2.5	13-Jan-2018	snj	Pull up following revision(s) (requested by msaitoh in ticket #490): sys/dev/pci/ixgbe/ixgbe.c: revision 1.116-1.119 sys/dev/pci/ixgbe/ixv.c: revision 1.76-1.77 Don't panic when resource shortage occured. Fixes PR#52820 reported by kardel@: - Don't use if_free() because ixgbe(4) don't use if_alloc(). - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more). -- Don't panic when resource shortage occured. Like ixgbe.c rev. 1.116. - Move location of {ixgbe,ixv}_setup_interface() call at a position that any error don't occur. One of the reason is that it should be. Another reason is that it's hard to call ether_ifdetach() and if_detach() when cold == 1 (because of pserialize_perform, xc_wait, timing of domaininit and maybe more). -- Fix panic when only link interrupt can't be established. -- - If MSI can't be used on some environment, fallback to INTx correctly. - Use single vector MSI when number of CPU is 1 to save interrupt slot. -- Fallback from MSI-X to MSI or INTx if MSI-X setup failed.
1.56.2.4	02-Jan-2018	snj	Pull up following revision(s) (requested by ozaki-r in ticket #456): sys/arch/arm/sunxi/sunxi_emac.c: 1.9 sys/dev/ic/dwc_gmac.c: 1.43-1.44 sys/dev/pci/if_iwm.c: 1.75 sys/dev/pci/if_wm.c: 1.543 sys/dev/pci/ixgbe/ixgbe.c: 1.112 sys/dev/pci/ixgbe/ixv.c: 1.74 sys/kern/sys_socket.c: 1.75 sys/net/agr/if_agr.c: 1.43 sys/net/bpf.c: 1.219 sys/net/if.c: 1.397, 1.399, 1.401-1.403, 1.406-1.410, 1.412-1.416 sys/net/if.h: 1.242-1.247, 1.250, 1.252-1.257 sys/net/if_bridge.c: 1.140 via patch, 1.142-1.146 sys/net/if_etherip.c: 1.40 sys/net/if_ethersubr.c: 1.243, 1.246 sys/net/if_faith.c: 1.57 sys/net/if_gif.c: 1.132 sys/net/if_l2tp.c: 1.15, 1.17 sys/net/if_loop.c: 1.98-1.101 sys/net/if_media.c: 1.35 sys/net/if_pppoe.c: 1.131-1.132 sys/net/if_spppsubr.c: 1.176-1.177 sys/net/if_tun.c: 1.142 sys/net/if_vlan.c: 1.107, 1.109, 1.114-1.121 sys/net/npf/npf_ifaddr.c: 1.3 sys/net/npf/npf_os.c: 1.8-1.9 sys/net/rtsock.c: 1.230 sys/netcan/if_canloop.c: 1.3-1.5 sys/netinet/if_arp.c: 1.255 sys/netinet/igmp.c: 1.65 sys/netinet/in.c: 1.210-1.211 sys/netinet/in_pcb.c: 1.180 sys/netinet/ip_carp.c: 1.92, 1.94 sys/netinet/ip_flow.c: 1.81 sys/netinet/ip_input.c: 1.362 sys/netinet/ip_mroute.c: 1.147 sys/netinet/ip_output.c: 1.283, 1.285, 1.287 sys/netinet6/frag6.c: 1.61 sys/netinet6/in6.c: 1.251, 1.255 sys/netinet6/in6_pcb.c: 1.162 sys/netinet6/ip6_flow.c: 1.35 sys/netinet6/ip6_input.c: 1.183 sys/netinet6/ip6_output.c: 1.196 sys/netinet6/mld6.c: 1.90 sys/netinet6/nd6.c: 1.239-1.240 sys/netinet6/nd6_nbr.c: 1.139 sys/netinet6/nd6_rtr.c: 1.136 sys/netipsec/ipsec_output.c: 1.65 sys/rump/net/lib/libnetinet/netinet_component.c: 1.9-1.10 kmem_intr_free kmem_intr_[z]alloced memory the underlying pools are the same but api-wise those should match Unify IFEF__MPSAFE into IFEF_MPSAFE There are already two flags for if_output and if_start, however, it seems such MPSAFE flags are eventually needed for all if_XXX operations. Having discrete flags for each operation is wasteful of if_extflags bits. So let's unify the flags into one: IFEF_MPSAFE. Fortunately IFEF__MPSAFE flags have never been included in any releases, so we can change them without breaking backward compatibility of the releases (though the kernel version of -current should be bumped). Note that if an interface have both MP-safe and non-MP-safe operations at a time, we have to set the IFEF_MPSAFE flag and let callees of non-MP-safe opeartions take the kernel lock. Proposed on tech-kern@ and tech-net@ Provide macros for softnet_lock and KERNEL_LOCK hiding NET_MPSAFE switch It reduces C&P codes such as "#ifndef NET_MPSAFE KERNEL_LOCK(1, NULL); ..." scattered all over the source code and makes it easy to identify remaining KERNEL_LOCK and/or softnet_lock that are held even if NET_MPSAFE. No functional change Hold KERNEL_LOCK on if_ioctl selectively based on IFEF_MPSAFE If IFEF_MPSAFE is set, hold the lock and otherwise don't hold. This change requires additions of KERNEL_LOCK to subsequence functions from if_ioctl such as ifmedia_ioctl and ifioctl_common to protect non-MP-safe components. Proposed on tech-kern@ and tech-net@ Ensure to hold if_ioctl_lock when calling if_flags_set Fix locking against myself on ifpromisc vlan_unconfig_locked could be called with holding if_ioctl_lock. Ensure to not turn on IFF_RUNNING of an interface until its initialization completes And ensure to turn off it before destruction as per IFF_RUNNING's description "resource allocated". (The description is a bit doubtful though, I believe the change is still proper.) Ensure to hold if_ioctl_lock on if_up and if_down One exception for if_down is if_detach; in the case the lock isn't needed because it's guaranteed that no other one can access ifp at that point. Make if_link_queue MP-safe if IFEF_MPSAFE if_link_queue is a queue to store events of link state changes, which is used to pass events from (typically) an interrupt handler to if_link_state_change softint. The queue was protected by KERNEL_LOCK so far, but if IFEF_MPSAFE is enabled, it becomes unsafe because (perhaps) an interrupt handler of an interface with IFEF_MPSAFE doesn't take KERNEL_LOCK. Protect it by a spin mutex. Additionally with this change KERNEL_LOCK of if_link_state_change softint is omitted if NET_MPSAFE is enabled. Note that the spin mutex is now ifp->if_snd.ifq_lock as well as the case of if_timer (see the comment). Use IFADDR_WRITER_FOREACH instead of IFADDR_READER_FOREACH At that point no other one modifies the list so IFADDR_READER_FOREACH is unnecessary. Use of IFADDR_READER_FOREACH is harmless in general though, if we try to detect contract violations of pserialize, using it violates the contract. So avoid using it makes life easy. Ensure to call if_addr_init with holding if_ioctl_lock Get rid of outdated comments Fix build of kernels without ether By throwing out if_enable_vlan_mtu and if_disable_vlan_mtu that created a unnecessary dependency from if.c to if_ethersubr.c. PR kern/52790 Rename IFNET_LOCK to IFNET_GLOBAL_LOCK IFNET_LOCK will be used in another lock, if_ioctl_lock (might be renamed then). Wrap if_ioctl_lock with IFNET_* macros (NFC) Also if_ioctl_lock perhaps needs to be renamed to something because it's now not just for ioctl... Reorder some destruction routines in if_detach - Destroy if_ioctl_lock at the end of the if_detach because it's used in various destruction routines - Move psref_target_destroy after pr_purgeif because we want to use psref in pr_purgeif (otherwise destruction procedures can be tricky) Ensure to call if_mcast_op with holding IFNET_LOCK Note that CARP doesn't deal with IFNET_LOCK yet. Remove IFNET_GLOBAL_LOCK where it's unnecessary because IFNET_LOCK is held Describe which lock is used to protect each member variable of struct ifnet Requested by skrll@ Write a guideline for converting an interface to IFEF_MPSAFE Requested by skrll@ Note that IFNET_LOCK must not be held in softint Don't set IFEF_MPSAFE unless NET_MPSAFE at this point Because recent investigations show that interfaces with IFEF_MPSAFE need to follow additional restrictions to work with the flag safely. We should enable it on an interface by default only if the interface surely satisfies the restrictions, which are described in if.h. Note that enabling IFEF_MPSAFE solely gains a few benefit on performance because the network stack is still serialized by the big kernel locks by default.
1.56.2.3	21-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #441): sys/dev/pci/files.pci: 1.390 sys/dev/pci/ixgbe/if_bypass.c: new sys/dev/pci/ixgbe/if_fdir.c: new sys/dev/pci/ixgbe/if_sriov.c: new sys/dev/pci/ixgbe/ixgbe_bypass.h: new sys/dev/pci/ixgbe/ixgbe_fdir.h: new sys/dev/pci/ixgbe/ixgbe_features.h: new sys/dev/pci/ixgbe/ixgbe_netmap.c: new sys/dev/pci/ixgbe/ixgbe_netmap.h: new sys/dev/pci/ixgbe/ixgbe_rss.h: new sys/dev/pci/ixgbe/ixgbe_sriov.h: new sys/dev/pci/ixgbe/ix_txrx.c: 1.28, 1.30 via patch sys/dev/pci/ixgbe/ixgbe.c: 1.97-1.105, 1.107-1.111, 1.113-1.115 sys/dev/pci/ixgbe/ixgbe.h: 1.26-1.29 via patch sys/dev/pci/ixgbe/ixv.c: 1.57-1.58, 1.60-1.61, 1.63-1.64, 1.66-1.72, 1.75 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.17-1.18 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.26-1.30 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_82598.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c: 1.15-1.16 via patch sys/dev/pci/ixgbe/ixgbe_82599.h: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_api.h: 1.10-1.11 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.14-1.16 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.5-1.6 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_mbx.c: 1.7-1.9 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h: 1.11-1.12 via patch sys/dev/pci/ixgbe/ixgbe_osdep.c: 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.18-1.19 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.12-1.14 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.9-1.10 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.13-1.15 via patch sys/dev/pci/ixgbe/ixgbe_vf.h: 1.9-1.11 via patch sys/dev/pci/ixgbe/ixgbe_x540.c: 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_x540.h: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.c: 1.6-1.7 sys/dev/pci/ixgbe/ixgbe_x550.h: 1.3 Sync with FreeBSD's up to r326022, fix bugs and some improvements: - Don't limit number of queue pair to 8. Take chip's max TX queues and max RX queues into account. - Add C3000 (Denverton) support. - Add bypass function support for bypass adapters. - Change EEE sysctl. - Remove thermal test sysctl. - ixv(4): set RSS mapping. - ixv(4): Add TSOv6. - ixv(4): Fix hardware counter. - ixv(4): Make TX/RX descriptors size the same as ixg(4). - ixv(4): Print device name in ixv_attach(). - ixv(4): Make mailbox statistic counters evcnt(9). - ixv(4): Fix a problem that mailbox interrupt never occurred. - ixv(4): Don't check neither VTEICR nor VTEICS same as Linux. It seems that both registers can't be used to check which MSI-X vector is triggered. - ixv(4): Print Mailbox API version. - Print driver feature capabilities and enable bits when verbose boot. - Fix a bug that X550EM_A (Denverton) can't force 10BaseT. - Remove all half-duplex setting because of the chip specification. - Add 2.5GBASE-T and 5GBASE-T support. - Set ifm_baudrate correctly. - Protect ec_multi* with mutex like other MP safe Ethernet drivers. - On X550 and newer, print NVM Image Version with %u.%02x. - On X540, print PHY FW Revision with %u.%x. 0x4030 will be printed as "Revision 4.3 ID 0x0" - Fix value check of OEM_NVM_IMAGE_VER. - Print PHY ID only for copper PHY. - Fix a bug that X550 and newer didn't linkup if it forces 100BaseTX-FDX or 10BaseT-FDX. - Fix a bug that RX may accesses freed area. It also fixes a problem that if_init() takes long time on many core machine. - Check ETHERCAP_VLAN_HWTAGGING in ixgbe_setup_vlan_hw_support(). This change has no influence to netbsd because it's enabled by default and NetBSD has no API to disable it. - Whitespace fix.
1.56.2.2	10-Dec-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #427): sys/arch/amiga/dev/if_bah_zbus.c: 1.17 sys/arch/arm/broadcom/bcm53xx_eth.c: 1.30 sys/arch/powerpc/booke/dev/pq3etsec.c: 1.32 sys/arch/usermode/dev/if_veth.c: 1.9 sys/dev/ic/an.c: 1.66 sys/dev/ic/athn.c: 1.17 sys/dev/ic/atw.c: 1.162 sys/dev/ic/bwi.c: 1.33 sys/dev/ic/dwc_gmac.c: 1.41-1.42 sys/dev/ic/malo.c: 1.10 sys/dev/ic/rt2560.c: 1.31 sys/dev/ic/rt2661.c: 1.36 sys/dev/ic/rt2860.c: 1.29 sys/dev/ic/rtw.c: 1.127 sys/dev/ic/rtwvar.h: 1.46 sys/dev/ic/smc90cx6.c: 1.71 sys/dev/ic/smc90cx6var.h: 1.12 sys/dev/ic/wi.c: 1.244 sys/dev/pci/if_ipw.c: 1.66 sys/dev/pci/if_iwi.c: 1.104 sys/dev/pci/if_iwm.c: 1.76 sys/dev/pci/if_iwn.c: 1.86 sys/dev/pci/if_rtwn.c: 1.13 sys/dev/pci/if_wm.c: 1.541 sys/dev/pci/if_wpi.c: 1.79 sys/dev/pci/ixgbe/ixgbe.c: 1.106 sys/dev/pci/ixgbe/ixv.c: 1.73 via patch sys/dev/pcmcia/if_malo_pcmcia.c: 1.15 sys/dev/scsipi/if_se.c: 1.95 sys/dev/usb/if_upl.c: 1.60 sys/net/if.c: 1.396 sys/net/if.h: 1.241 sys/net/if_arc.h: 1.23 sys/net/if_arcsubr.c: 1.78 sys/net/if_bridge.c: 1.136-1.137 sys/net/if_etherip.c: 1.39 sys/net/if_faith.c: 1.56 sys/net/if_gif.c: 1.131 sys/net/if_loop.c: 1.96 sys/net/if_mpls.c: 1.30 sys/net/if_pppoe.c: 1.129 sys/net/if_srt.c: 1.27 sys/net/if_stf.c: 1.102 sys/net/if_tap.c: 1.100 sys/net/if_vlan.c: 1.105 sys/netinet/ip_carp.c: 1.91 sys/rump/net/lib/libshmif/if_shmem.c: 1.73-1.74 sys/rump/net/lib/libvirtif/if_virt.c: 1.55-1.56 if_initalize() and if_attach() failed when resource allocation failed (e.g. allocating softint). Without this change, it panics. It's bad because resource shortage really occured when a lot of pseudo interface is created. To avoid this problem, don't panic and change return value of if_initialize() and if_attach() to int. Caller fanction will be recover from error cleanly by checking the return value. Return if bah_attach_subr() failed. If if_attach() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add some missing frees in bridge_clone_destroy(). - KNF If error occured in bcmeth_ccb_attach(), free resources and return. If error occured in pq3etsec_attach(), free resources and return. If error occured in the attach function, free resources and return. - If if_initialize() failed in athn_attach(), free resources and return. - Add missing pmf_event_deregister() in athn_detach(). - Free resources correctly on some errors in atw_attach(). - Use apint*() insread of printf() in the attach function. If if_initialize() failed in the attach function, return. - If if_initialize() failed in the attach function, free resources and return. - Add missing dwc_gmac_free_dma_rings() and mutex_destroy() when attach failed. - If if_initialize() failed in the attach function, free resources and return. - ifp is always not NULL in iwi_detach(). Check correctly with ifp->if_softc. - If if_initialize() failed in the attach function, free resources and return. - Fix error path in the attach function correctly. If if_initialize() failed in the attach function, free resources and return. If if_attach() failed in the attach function, free resources and return. - If if_initialize() failed in the attach function, free resources and return. - KNF - If if_attach() failed in the attach function, free resources and return. - KNF Fix compile error. Fix compile error. We don't need '&mii', but just 'mii' for mii_detach(). Don't free sc_rthash twice
1.56.2.1	23-Sep-2017	snj	Pull up following revision(s) (requested by msaitoh in ticket #288): sys/dev/pci/ixgbe/ixv.c: 1.59 and 1.62 via patch - Use cprng_strong64() instead of cprng_fast64() to prevent panic. - Set get_link_status to true before calling ixv_config_link() in ixv_init_locked() to get the first link status correctly.
1.84.2.11	26-Jan-2019	pgoyette	Sync with HEAD
1.84.2.10	30-Sep-2018	pgoyette	Ssync with HEAD
1.84.2.9	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.84.2.8	25-Jun-2018	pgoyette	Sync with HEAD
1.84.2.7	21-May-2018	pgoyette	Sync with HEAD
1.84.2.6	22-Apr-2018	pgoyette	Sync with HEAD
1.84.2.5	16-Apr-2018	pgoyette	Sync with HEAD, resolve some conflicts
1.84.2.4	07-Apr-2018	pgoyette	Sync with HEAD. 77 conflicts resolved - all of them $NetBSD$
1.84.2.3	30-Mar-2018	pgoyette	Resolve conflicts between branch and HEAD
1.84.2.2	22-Mar-2018	pgoyette	Synch with HEAD, resolve conflicts
1.84.2.1	15-Mar-2018	pgoyette	Synch with HEAD
1.105.2.3	21-Apr-2020	martin	Sync with HEAD
1.105.2.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.105.2.1	10-Jun-2019	christos	Sync with HEAD
1.125.2.25	03-Feb-2024	martin	Pull up the following, requested by msaitoh in ticket #1792: sys/dev/pci/ixgbe/ix_txrx.c 1.105-1.116 via patch sys/dev/pci/ixgbe/ixgbe.c 1.345-1.346,1.349 via patch sys/dev/pci/ixgbe/ixgbe.h 1.94-1.98 sys/dev/pci/ixgbe/ixgbe_type.h 1.62 sys/dev/pci/ixgbe/ixv.c 1.193,1.195-1.196 - Clear the WTHRESH bit field before writing it. - Optimize ixgbe_txeof(). - Use kmem_zalloc() instead of malloc(,M_ZERO). - Add QPRDC(Queue Packet Receive Drop Count) into iqdrops. - No functional change - Move assignment of TXD. - ixv(4): Remove unused IFF_OACTIVE. - Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. - The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). - Use #ifdef LRO more to reduce the size of struct rx_ring. - Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. - Th RSC (hardware receive side coalescing) feature has been disabled all along, so enclose the code with #ifdef RSC. - Remove unused. - Modify for the readability. - Modify comment. - Fix comment. Whitespace.
1.125.2.24	18-Oct-2023	martin	Pull up the following, requested by msaitoh in ticket #1753: sys/dev/pci/ixgbe/ix_txrx.c 1.103-1.104 sys/dev/pci/ixgbe/ixgbe.c 1.334-1.338, 1.341-1.344 via patch sys/dev/pci/ixgbe/ixgbe.h 1.90-1.93 sys/dev/pci/ixgbe/ixgbe_82599.c 1.31-1.32 sys/dev/pci/ixgbe/ixgbe_api.c 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.5 sys/dev/pci/ixgbe/ixgbe_common.c 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_common.h 1.18 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.14-1.15 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.13 sys/dev/pci/ixgbe/ixgbe_type.h 1.59-1.61 sys/dev/pci/ixgbe/ixgbe_x540.c 1.24 sys/dev/pci/ixgbe/ixgbe_x550.c 1.28 sys/dev/pci/ixgbe/ixv.c 1.187-1.192 - ixg(4): Add 82599 LS support once again. - ixg(4): Filter out spurious link up indication more. - ixg(4): Print DEVICE_CAPS register. - ixg(4): Fix a bug that the number of queues is unintentionally limited to a small number or wrong error message may be printed when two devices' number of MSI-X vectors are different. - Modify error message of wrong TX/RX descriptor size. - Enable interrupt after setting IFF_RUNNING. - Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. - Cleanup the code.
1.125.2.23	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1749): sys/dev/pci/ixgbe/if_bypass.c 1.10 sys/dev/pci/ixgbe/if_fdir.c 1.6 via patch sys/dev/pci/ixgbe/if_sriov.c 1.18 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.102 via patch sys/dev/pci/ixgbe/ixgbe.c 1.333,1.339 via patch sys/dev/pci/ixgbe/ixgbe.h 1.89 sys/dev/pci/ixgbe/ixgbe_api.h 1.17 sys/dev/pci/ixgbe/ixgbe_common.h 1.17 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.4 sys/dev/pci/ixgbe/ixgbe_common.c 1.45 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.5 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.6 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.9 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.6 sys/dev/pci/ixgbe/ixgbe_type.h 1.58 sys/dev/pci/ixgbe/ixgbe_x550.c 1.27 sys/dev/pci/ixgbe/ixv.c 1.186 via patch ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.125.2.22	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1745): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.125.2.21	23-Jan-2023	martin	Pull up the following revisions, requested by msaitoh in ticket #1579: sys/dev/pci/files.pci 1.442 sys/dev/pci/ixgbe/ix_txrx.c 1.99-1.100 sys/dev/pci/ixgbe/ixgbe.c 1.320-1.324 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.19 sys/dev/pci/ixgbe/ixgbe_api.c 1.28 sys/dev/pci/ixgbe/ixgbe_common.c 1.43 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.17 sys/dev/pci/ixgbe/ixv.c 1.183 - Add an option for Tx to use deferred softint regardless of whether can get txq lock or not. It's off by default. - Call txeof first, then rxeof for the consistency. - Make three "Unsupported SFP+ module..." messages the same. - KNF. Modify comment. Fix typo.
1.125.2.20	02-Jun-2022	martin	Pull up the following revisions, all via patch, requested by msaitoh in ticket #1459: sys/dev/pci/ixgbe/ixgbe.c 1.261,1.265-1.268,1.273,1.275-1.277, 1.312,1.316-1.319 via patch sys/dev/pci/ixgbe/ixgbe.h 1.85 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.46-1.47 sys/dev/pci/ixgbe/ixgbe_x550.c 1.26 sys/dev/pci/ixgbe/ixv.c 1.182 - Reduce code duplication between ixgbe_msix_admin() and ixgbe_legacy_irq(). - Add missing code which was not in ixgbe_msix_admin() from ixgbe_legacy_irq() and vice versa. - Reorder code. - Disable/enable the OTHER interrupts correctly. - Don't return in the middle of ixgbe_msix_admin() when an flow director reinit failed. NetBSD currently doesn't support flow director, so this is not a real bug. - Print ECC, PHY and temp error log using with ratecheck(). - Correctly re-enable queue interrupt in ixgbe_legacy_irq(). - Correctly enter the recovery mode. - No functional change: - Add some debug printf()s. - Don't use "more" flag for simplify. - Fix typos in comment. - KNF.
1.125.2.19	31-May-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1458): sys/dev/pci/ixgbe/ixv.c: revision 1.181 sys/dev/pci/ixgbe/ixgbe.c: revision 1.315 sys/dev/pci/ixgbe/ixgbe.h: revision 1.86 Fix a bug that the legacy interrupt doesn't work when MSI-X allocation failed. Fixes PR kern/56857. Remove unused adapter->msix_mem.
1.125.2.18	30-May-2022	martin	Pull up the following revisions, requested by msaitoh: sys/dev/pci/ixgbe/ixgbe.c 1.270,1.280,1.307-1.311, 1.313-1.314 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.96-1.97 sys/dev/pci/ixgbe/ixv.c 1.158,1.179-1.180 via patch - ixg(4): Print Printed Board Assembly (PBA) number. - ixg(4): Add IFF_RUNNING check in ixgbe_legacy_irq() again. this might fix small race but it's not so dangerous. - Add value check for {tx,rx}_process_limit sysctl to avoid setting wrong value. - Add missing num_tx_desc sysctl. - No functional change: - KNF a bit. - Simplify setting of EIAC register. - Move the definition of eicr_mask variable. - Enclose flow director stuff in ixgbe_intr_admin_common() with IXGBE_FIR which is not defined in NetBSD. - Modify comment for consistency. - Use cached rx_copy_len in ixgbe_rxeof().
1.125.2.17	02-Feb-2022	martin	Pull up the following revisions (requested by msaitoh in ticket #1424): sys/dev/pci/ixgbe/ix_txrx.c 1.95 sys/dev/pci/ixgbe/ixgbe.c 1.305 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.19 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.15-1.16 sys/dev/pci/ixgbe/ixv.c 1.178 via patch Use atomic_{load,store}_relaxed() for event counters.
1.125.2.16	31-Jan-2022	martin	Pull up following revision(s) (requested by msaitoh in ticket #1416): sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.19 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.23 sys/dev/pci/ixgbe/if_sriov.c: revision 1.17 sys/dev/pci/ixgbe/ixv.c: revision 1.172 sys/dev/pci/ixgbe/ixv.c: revision 1.173 sys/dev/pci/ixgbe/ixgbe.c: revision 1.301 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.55 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.16 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.17 sys/dev/pci/ixgbe/ixgbe_mbx.c: revision 1.18 Add code to support API version 1.5. No functional change. - This change adds almost all code to support API 1.5 except real negotiation and upgrade mailbox functions. - From ix-3.3.18, ix-3.3.22 and ixv-1.5.24. Enable mailbox API 1.5 support. Tested on ESXi with ixgben 1.10.3.0. Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_vf(). Don't clear mailbox related counters in ixgbe_upgrade_mbx_params_pf().
1.125.2.15	30-Jan-2022	martin	Pull up the following revisions, requested by msaitoh in ticket #1414: sys/dev/pci/ixgbe/if_sriov.c 1.12-1.16 sys/dev/pci/ixgbe/ixgbe.c 1.295-1.297,1.300,1.304 via patch sys/dev/pci/ixgbe/ixgbe.h 1.84 sys/dev/pci/ixgbe/ixgbe_82598.c 1.17-1.18 sys/dev/pci/ixgbe/ixgbe_82598.h 1.9 sys/dev/pci/ixgbe/ixgbe_82599.c 1.24-1.28 sys/dev/pci/ixgbe/ixgbe_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_api.c 1.26-1.27 sys/dev/pci/ixgbe/ixgbe_api.h 1.16 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.3 sys/dev/pci/ixgbe/ixgbe_common.c 1.34-1.42 sys/dev/pci/ixgbe/ixgbe_common.h 1.15-1.16 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.12-1.13 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.10-1.12 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.8 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.8 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.4 sys/dev/pci/ixgbe/ixgbe_features.h 1.4 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.13-1.15 sys/dev/pci/ixgbe/ixgbe_mbx.h 1.15-1.18 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.8 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.31 sys/dev/pci/ixgbe/ixgbe_phy.c 1.25-1.29 sys/dev/pci/ixgbe/ixgbe_phy.h 1.13 sys/dev/pci/ixgbe/ixgbe_rss.h 1.6 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.5 sys/dev/pci/ixgbe/ixgbe_type.h 1.51-1.54 sys/dev/pci/ixgbe/ixgbe_vf.c 1.28-1.29 sys/dev/pci/ixgbe/ixgbe_vf.h 1.15 sys/dev/pci/ixgbe/ixgbe_x540.c 1.20-1.22 sys/dev/pci/ixgbe/ixgbe_x540.h 1.10 sys/dev/pci/ixgbe/ixgbe_x550.c 1.21-1.25 sys/dev/pci/ixgbe/ixgbe_x550.h 1.7 sys/dev/pci/ixgbe/ixv.c 1.170, 1.174-1.175 via patch - Add typecast for type mismatch. - Fix retry count calculation of I2C read/write. - Wait longer for link after fiber MAC setup. - ixv(4): Use adapter->mta for the multicast array memory instead of the on-stack array. - Match X550_PHY_ID correctly on X550. - Print NVM image version on 82598. - Use 64bit for lxon + lxoff. - Don't expose garbage data of hw.ixvN.debug. - Some NetBSD unrelated changes: - Fix infinite recursion on PCIe link down if VMDQ is used. - Move PF mailbox initialization from ixgbe_attach() to ixgbe_init_iov(). - Add IPv6 mask for flow director. - Change error level in ixgbe_fc_autoneg(). - Check host interface return status when writing NVM. - Change DCB credit parameters. - Restore some mailbox related functions. Revert part of ixgbe_mbx.c rev. 1.7 and ixgbe_mbx.h rev. 1.11. No functional change. - Rename IXGBE_VT_MSGTYPE_{ACK,NACK} to IXGBE_VT_MSGTYPE_{SUCCESS,FAILURE}. No functional change. - Remove unused argument. Change argument. - Remove unnecessary return value check. - Remove debug error message. - Remove dead code. - Add some unused macros. - Fix typo in comment. - Rename some functions. - Sort lines, modify comment. - Whitespace fix.
1.125.2.14	29-Jan-2022	martin	Pull up the following revisions (all via patch), requested by msaitoh in ticket #1408: sys/dev/pci/ixgbe/ixgbe_vf.h 1.16-1.17 sys/dev/pci/ixgbe/ixv.c 1.176-1.177 Make ifconfig -z ixvN clear event counter.
1.125.2.13	20-Nov-2021	martin	Pull up the following, requested by msaitoh in ticket #1374: sys/dev/pci/ixgbe/ixgbe.h 1.81-1.83 sys/dev/pci/ixgbe/ixgbe.c 1.291-1.292 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.50 sys/dev/pci/ixgbe/ixv.c 1.167-1.168 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.94 - Fix a bug that a near 64KB TSO segment can't send. - Reduce bus_dmamap_sync() cost. - Use macro. Fix typos in comment.
1.125.2.12	15-Sep-2021	martin	Pull up the following (via patch), requested by msaitoh in ticket #1346: sys/dev/pci/ixgbe/ixgbe.c 1.252, 1.280-1.283, 1.286-1.287, 1.289-1.290 via patch sys/dev/pci/ixgbe/ixgbe.h 1.73, 1.76-1.80 via patch sys/dev/pci/ixgbe/ix_txrx.c 1.68-1.93 sys/dev/pci/ixgbe/ixv.c 1.153, 1.157-1.161, 1.163-1.166 via patch sys/dev/pci/ixgbe/if_bypass.c 1.7-1.9 sys/dev/pci/ixgbe/if_fdir.c 1.4-1.5 sys/dev/pci/ixgbe/if_sriov.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_82598.c 1.16 sys/dev/pci/ixgbe/ixgbe_82599.c 1.23 sys/dev/pci/ixgbe/ixgbe_api.c 1.25 sys/dev/pci/ixgbe/ixgbe_bypass.h 1.2 sys/dev/pci/ixgbe/ixgbe_common.c 1.30-1.33 sys/dev/pci/ixgbe/ixgbe_dcb.c 1.10-1.11 sys/dev/pci/ixgbe/ixgbe_dcb.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h 1.7 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c 1.8-1.9 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h 1.7 sys/dev/pci/ixgbe/ixgbe_fdir.h 1.3 sys/dev/pci/ixgbe/ixgbe_features.h 1.3 sys/dev/pci/ixgbe/ixgbe_mbx.c 1.12 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.13, 1.16-1.17 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.13-1.14 sys/dev/pci/ixgbe/ixgbe_netmap.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h 1.2 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.7 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.29-1.30 sys/dev/pci/ixgbe/ixgbe_phy.c 1.24 sys/dev/pci/ixgbe/ixgbe_rss.h 1.5 sys/dev/pci/ixgbe/ixgbe_sriov.h 1.4 sys/dev/pci/ixgbe/ixgbe_type.h 1.49 sys/dev/pci/ixgbe/ixgbe_vf.c 1.27 sys/dev/pci/ixgbe/ixgbe_x540.c 1.18-1.19 sys/dev/pci/ixgbe/ixgbe_x540.h 1.9 sys/dev/pci/ixgbe/ixgbe_x550.c 1.19-1.20 sys/dev/pci/ixgbe/ixgbe_x550.h 1.6 sys/dev/pci/files.pci 1.438 share/man/man4/ixg.4 1.15 share/man/man4/ixv.4 1.8 - Use MCLGET() instead of homegrown cluster (jcl) allocation mechanism. Before this commit, resource shortage was easily occurred because the total number of the clusters is small. - Improve performance: - Use m_adj(ETHER_ALIGN) more. - Sprinkle __predict_false() in the RX path. - Don't pre-allocate a cluster for RXCOPY case to improve short packet's performance. - Call bus_dmamap_unload(9) via ixgbe_dmamap_unload(), before freeing DMA buffer. Also, when the buffer is already freed, do not call bus_dmamap_unload(9) (no resource leaks with this change). This change is required to make ixg(4) work on alpha. - Keep m_len and m_pkthdr.len consistent to prevent panic on arm. - Fix panic when bus_dmamap_load_mbuf() failed in ixgbe_setup_receive_ring(). - Added BUS_DMA_COHERENT flag to bus_dmamem_map() to improve stability on aarch64. - Use uint64_t instead of bus_addr_t for the TX descriptor's buffer address. At least, this change is required for macppc (sizeof(bus_addr_t) == 4) to make TX work. - Fix little-endian dependence. - Set rxr->next_to_refresh correctly in ixgbe_setup_receive_ring(). - Refresh unrefreshed descriptors' buffers correctly. - Don't call bus_dmamap_sync with rx_mbuf_sz(== MCLBYTES) to prevent panic. - Save the discard_multidesc state to not to forget the state by exiting rxeof(). - Add missing increment of no_mbuf error counter. - Don't increment no_mbuf evcnt(9) when discarding multi-descriptor packet. - ixv: Modify error message to sync with ixgbe.c - Print the error value of ixgbe_reset_hw() for debugging. - Remove extra unlock/lock processing around if_percpuq_enqueue(). - Refactor rxr->next_to_check updating. - Add new sysctl "rx_copy_len". - Add a new sysctl to read rxr->next_to_refresh. - Print error number when error occurred. - Rename ix{gbe,v}_stop() with ix{gbe,v}_stop_locked(). No functional change. - Don't use fixed value. - Comment out flow director processing in fast path. - Add missing NetBSD RCS IDs and __KERNEL_RCSID()s. - KNF. - Fix typos.
1.125.2.11	11-Mar-2021	martin	Pull up the following (all via patch) requested by msaitoh in ticket #1231: sys/dev/pci/ixgbe/ixgbe.c 1.259, 1.278-1.279 sys/dev/pci/ixgbe/ixgbe.h 1.75 sys/dev/pci/ixgbe/ixgbe_netbsd.h 1.12 sys/dev/pci/ixgbe/ixgbe_vf.c 1.24-1.26 sys/dev/pci/ixgbe/ixgbe_x550.c 1.17 sys/dev/pci/ixgbe/ixv.c 1.155-1.156 sys/dev/pci/ixgbe/ix_txrx.c 1.64-67 sys/dev/pci/files.pci 1.436 share/man/man4/ixg.4 1.13-1.14 share/man/man4/ixv.4 1.6-1.7 - Fix a problem that the RX path stalled when the mbuf cluster is exhausted. - Modify some parameters to reduce packet dropping. See also the manual's OPTIONS section for the detail. - ixv(4): The max number of queue(pair) is not 7 but 8. Correctly reset the hardware. - Add "TX " to "Queue No Descriptor Available" evcnt(9) name to make it more understandable. - Fix a bug that some advertise speeds can't be set with hw.ixgN.advertise_speed if both 2.5G and 5G are set. Fix the error message, too. - Fix typo in comment or debug message.
1.125.2.10	10-Jul-2020	martin	Pull up the following revisions, requested by msaitoh in ticket #997: sys/dev/pci/ixgbe/ix_txrx.c 1.62-1.63 via patch sys/dev/pci/ixgbe/ixgbe.c 1.225, 1.228-1.229, 1.232 via patch sys/dev/pci/ixgbe/ixgbe.h 1.64, 1.66 sys/dev/pci/ixgbe/ixv.c 1.146, 1.148-1.150 sys/dev/pci/ixgbe/ixgbe_common.c 1.27 sys/dev/pci/ixgbe/ixgbe_vf.c 1.23 sys/dev/pci/ixgbe/ixgbe_82598.c 1.15 sys/dev/pci/ixgbe/ixgbe_x550.c 1.18 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.14 sys/dev/pci/ixgbe/ixgbe_phy.c 1.21 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.26 - Fix IXGBE_LE32_TO_CPUS() macro for big endian machine. This problem was only on X550*. - Add debug printf()s. - Use unsigned to avoid undefined behavior in ixgbe_fc_enable_generic(). - Modify a little to reduce diff between ixgbe.c and ixv.c. No functional change. - Modify comment. - Remove unused macros. - Whitespace fix. - Fix typos.
1.125.2.9	26-Jan-2020	martin	Pull up the following (via patch), requested by msaitoh in ticket #648 sys/dev/pci/ixgbe/if_bypass.c 1.5 sys/dev/pci/ixgbe/ixgbe_osdep.c 1.5 sys/dev/pci/ixgbe/ix_txrx.c 1.58-1.60 sys/dev/pci/ixgbe/ixgbe.c 1.220-1.221 sys/dev/pci/ixgbe/ixgbe.h 1.60-1.2 sys/dev/pci/ixgbe/ixgbe_api.c 1.24 sys/dev/pci/ixgbe/ixgbe_common.c 1.26 sys/dev/pci/ixgbe/ixgbe_netbsd.c 1.11-1.12 sys/dev/pci/ixgbe/ixgbe_osdep.h 1.24 sys/dev/pci/ixgbe/ixgbe_phy.c 1.19 sys/dev/pci/ixgbe/ixgbe_82598.c 1.14 sys/dev/pci/ixgbe/ixv.c 1.142,1.144 - Free RX structure correctly when detaching. - Remove unused code. - Fix some typos in comment. - Remove extra spaces. - KNF.
1.125.2.8	24-Dec-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #580): sys/dev/pci/ixgbe/ixgbe.c: revision 1.217 sys/dev/pci/ixgbe/ixgbe.c: revision 1.218 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.6 sys/dev/pci/ixgbe/ixv.c: revision 1.143 sys/dev/pci/ixgbe/ixgbe_osdep.h: revision 1.25 Use bus_space_barrier() instead of x86 specific *fence instruction. Written by riastradh@. Add missing core lock in ixgbe_handle_mod().
1.125.2.7	19-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #453): sys/dev/pci/ixgbe/ixgbe.c: revision 1.216 sys/dev/pci/ixgbe/ixv.c: revision 1.141 Print MAC address. XXX Should we move such type of printf() to ether_ifattach?
1.125.2.6	14-Nov-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #422): sys/dev/pci/ixgbe/ixv.c: revision 1.140 ixv(4): disable RSS configuration on 82599 and X540 VFs. Those VFs share their RSS configuration with PF and, thus, they cannot be configured independently. From FreeBSD r354349.
1.125.2.5	08-Oct-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #293): sys/dev/pci/ixgbe/ixv.c: revision 1.139 Set MTU correctly if mtu > 1500.
1.125.2.4	26-Sep-2019	martin	Pull up the following revisions, requested by msaitoh in ticket #246: sys/dev/pci/ixgbe/ixgbe_type.h 1.42-1.43 sys/dev/pci/ixgbe/ixgbe.c 1.209-1.213 sys/dev/pci/ixgbe/ixgbe_x550.c 1.16 sys/dev/pci/ixgbe/ixv.c 1.131-1.138 sys/dev/pci/ixgbe/ixgbe_vf.c 1.19-1.22 sys/dev/pci/ixgbe/ixgbe_vf.h 1.14 sys/dev/pci/ixgbe/ixgbe.h 1.57-1.58 share/man/man4/ixv.4 1.5 - Fix a bug that MBSDC (Bad SFD Count) isn't counted on X550EM_X and X550EM_A. The register is for X550 and newer. - ixv(4): Make SIOCADDMULTI returns ENOSPC and print error message when the Ethernet multicast address list exceeds the limit(30) and can't be ALLMULTI. - ixv(4): SIOCZIFDATA clear the event counters as ixgbe.c. - Reduce ixv(4)'s multicast table array size in ixv_set_multi from MAX_NUM_MULTICAST_ADDRESSES(128) to IXGBE_MAX_VF_MC(30). - ixv(4): Add support ALLMULTI and PROMISC. - if_flags is neither int nor short. It's unsigned short. - ixg(4): Fix a bug that the multicast filter isn't correctly initialized when the total number of the Ethernet multicast addresses is just 128. - Make ixv_set_multi() work correctly (especially for PROMISC) when the function is called from if_init(). - Remove _set_promisc() and use _set_multi(). And then, rename _set_multi() to _set_rxfilter(). - ixv(4): If a multicast entry has range, use ALLMULTI like others. - Fix typo in comment. Found by Wataru Ashihara.
1.125.2.3	05-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #180): sys/dev/pci/ixgbe/ixv.c: revision 1.130 sys/dev/pci/ixgbe/ixgbe.c: revision 1.207 sys/dev/pci/ixgbe/ixgbe.c: revision 1.208 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.55 sys/dev/pci/ixgbe/ixv.c: revision 1.129 sys/dev/pci/ixgbe/ixgbe_netbsd.c: revision 1.10 printf -> device_printf Set IFM_1000_BX10 correctly. Use aprint_*() in the attach function.
1.125.2.2	01-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #134): sys/dev/pci/ixgbe/ixgbe.c: revision 1.202 sys/dev/pci/ixgbe/ixgbe.c: revision 1.203 sys/dev/pci/ixgbe/ixgbe.c: revision 1.204 sys/dev/pci/ixgbe/ixv.c: revision 1.128 Simplify ix{gbe,v}_[un]register_vlan() API suggested by knakahara. The API was the same as FreeBSD's pre-iflib's. They use iflib now and it's not required for us to keep the old API. X550EM supports QSFP, so check ixgbe_media_type_fiber_qsfp too. An interrupt might not arrive when a module is inserted. When an link status change interrupt occurred and the driver still regard SFP as unplugged, link becomes up and the real media type is unknown. e.g: % ifconfig -m ixg0 (snip) media: Ethernet autoselect (autoselect rxpause,txpause) status: active supported Ethernet media: media none media autoselect (snip) To resolve this problem, when an link status change interrupt occurred and the driver still regard SFP as unplugged, issue the module softint before issuing LSC interrupt.
1.125.2.1	01-Sep-2019	martin	Pull up following revision(s) (requested by msaitoh in ticket #133): sys/dev/pci/ixgbe/ixgbe.c: revision 1.200 sys/dev/pci/ixgbe/ixgbe.c: revision 1.201 sys/dev/pci/ixgbe/ixv.c: revision 1.126 sys/dev/pci/ixgbe/ixv.c: revision 1.127 sys/net/if_vlan.c: revision 1.142 sys/net/if_vlan.c: revision 1.143 sys/net/if_vlan.c: revision 1.144 sys/net/if_vlan.c: revision 1.145 sys/net/if_vlan.c: revision 1.146 Check ec_capenable instead of ec_capabilities to control TX side of VLAN HW tagging correctly. XXX pullup-9 Add missing IFNET_LOCK() and IFNET_UNLOCK() in vlan_config(). XXX pullup-9 Fix a bug that VLAN HW "tagging" enable/disable may not reflect correctly. - Always call ec_vlan_cb() if it exists. - Some (or all?) ethernet drivers don't enable HW tagging if no any vlan is attached. ixgbe is one of them. Check the the transition and update VLAN HW tagging function. XXX pullup-9 Use ETHER_LOCK()/ETHER_UNLOCK() suggested by knakahara. - kmem_alloc(,KM_SLEEP) never return NULL, so remove NULL check. - VLAN ID is never duplicated, so break the loop when found. Also move kmen_free() outside of ETHER_LOCK(ec)/ETHER_UNLOCK(ec) to reduce the hold time. suggested by ozaki-r. - Whitespace fix.
1.143.2.2	29-Feb-2020	ad	Sync with head.
1.143.2.1	25-Jan-2020	ad	Sync with head.
1.147.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.154.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.156.2.1	03-Apr-2021	thorpej	Sync with HEAD.
1.158.2.3	01-Aug-2021	thorpej	Sync with HEAD.
1.158.2.2	17-Jun-2021	thorpej	Sync w/ HEAD.
1.158.2.1	13-May-2021	thorpej	Sync with HEAD.
1.159.2.1	31-May-2021	cjep	sync with head
1.183.4.6	03-Feb-2024	martin	Pull up following revision(s) (requested by msaitoh in ticket #563): sys/dev/pci/ixgbe/ix_txrx.c: revision 1.110 sys/dev/pci/ixgbe/ixgbe.c: revision 1.345 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.111 sys/dev/pci/ixgbe/ixgbe.c: revision 1.346 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.112 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.62 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.113 sys/dev/pci/ixgbe/ixgbe.c: revision 1.348 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.114 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.115 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.116 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.105 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.106 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.107 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.108 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.109 sys/dev/pci/ixgbe/ixv.c: revision 1.193 sys/dev/pci/ixgbe/ixv.c: revision 1.195 sys/dev/pci/ixgbe/ixv.c: revision 1.196 sys/dev/pci/ixgbe/ixgbe.h: revision 1.94 sys/dev/pci/ixgbe/ixgbe.h: revision 1.95 sys/dev/pci/ixgbe/ixgbe.h: revision 1.96 sys/dev/pci/ixgbe/ixgbe.h: revision 1.97 sys/dev/pci/ixgbe/ixgbe.h: revision 1.98 ixgbe: Fix comment. No functional change. ixgbe: Whitespace. No functional change. ixgbe(4): Move assignment of TXD. NFCI. ixgbe(4): Modify comment. No functional change. ixgbe_tx_ctx_setup() may or may not consume one TX descriptor. ixv(4): Remove unused IFF_OACTIVE. No functional change. ixgbe: Clear the WTHRESH bit field before writing it. ixgbe: Modify for the readability. No functional change. ixgbe: micro-optimize ixgbe_txeof() Update txr->tx_avail and txr->txr_no_space outside the loop in ixgbe_txeof(). ixgbe: Update if_opackets outside the loop in ixgbe_txeof(). ixgbe: micro-optimize ixgbe_txeof() Update txr->packets outside the loop in ixgbe_txeof(). ixgbe: Use #ifdef IXGBE_FDIR more Don't include the Flow Director related members to reduce the size of struct tx_ring. On amd64 and aarch64, the real size is not changed because of the alignment. ixgbe: Simplify. No functional change. The descriptor ring size and the alignment are tested in the attach function, so it's not required to use roundup2(size, DBA_ALIGN). ixgbe: Use kmem_zalloc() instead of malloc(,M_ZERO). ixgbe: Remove unused to reduce the size of struct rx_ring. ixgbe: Use #ifdef LRO more to reduce the size of struct rx_ring. ixgbe: Change "me" from 32bit to 8bit because the max is 128. This commit doesn't change the real size of ix_queue, tx_ring and rx_ring because of the alignment. ixgbe: Use #ifdef RSC This feature (hardware receive side coalescing) has been disabled all along, so enclose the code with #ifdef RSC.
1.183.4.5	03-Nov-2023	martin	Pull up following revision(s) (requested by yamaguchi in ticket #455): sys/dev/pci/ixgbe/ixgbe.c: revision 1.347 sys/net/if_l2tp.c: revision 1.49 tests/net/if_vlan/t_vlan.sh: revision 1.25 sys/net/if_vlan.c: revision 1.171 sys/net/if_ethersubr.c: revision 1.326 sys/dev/pci/ixgbe/ixv.c: revision 1.194 Use ether_bpf_mtap only when the device supports vlan harware tagging The function is bpf_mtap() for ethernet devices and currently it is just handling VLAN tag stripped by the hardware. l2tp(4): use ether_ifattach() to initialize ethercom Support vlan(4) over l2tp(4) Added the test for vlan over l2tp
1.183.4.4	18-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #420): sys/dev/pci/ixgbe/ixgbe.c: revision 1.340 sys/dev/pci/ixgbe/ixgbe.c: revision 1.341 sys/dev/pci/ixgbe/ixgbe.c: revision 1.342 sys/dev/pci/ixgbe/ixgbe.c: revision 1.343 sys/dev/pci/ixgbe/ixgbe.c: revision 1.344 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.61 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.104 sys/dev/pci/ixgbe/ixv.c: revision 1.190 sys/dev/pci/ixgbe/ixv.c: revision 1.191 sys/dev/pci/ixgbe/ixv.c: revision 1.192 sys/dev/pci/ixgbe/ixgbe.h: revision 1.93 ixg(4): Print DEVICE_CAPS register. ixgbe: Whitespace. No functional change. ixg(4): Don't print wrong error message about ixgbe_num_queues. Don't override the ixgbe_num_queues global variable. It's the default value of the number of queues and should not override it because it will be referenced by later device attach. For example, the number of MSI-X vector is 64 on X540 and 18 on 82599. When both cards are inserted to a machine that the number of CPU is 24 and X540 is probed earlier, ixgbe_num_queues is overridden to 24 and the following error message is printed when attaching 82599: ixg2: autoconfiguration error: ixgbe_num_queues (24) is too large, using reduced amount (17). Note that the number of queues is in sc->num_queuss and referenced by hw.ixgN.num_queues sysctl. ixgbe: Don't override the {ixgbe,ixv}_max_interrupt_rate global variable. Fix a bug that changing hw.ix[gv]X.qY.interrupt_rate would change all devices all queues default interrupt rate. ixgbe: Whitespace. No functional change.
1.183.4.3	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #407): sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.47 sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.28 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.60 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.31 sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.32 sys/dev/pci/ixgbe/ixv.c: revision 1.187 sys/dev/pci/ixgbe/ixv.c: revision 1.188 sys/dev/pci/ixgbe/ixv.c: revision 1.189 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.18 sys/dev/pci/ixgbe/ixgbe_x540.c: revision 1.24 sys/dev/pci/ixgbe/ixgbe.c: revision 1.334 sys/dev/pci/ixgbe/ixgbe.c: revision 1.335 sys/dev/pci/ixgbe/ixgbe.c: revision 1.336 sys/dev/pci/ixgbe/ixgbe.c: revision 1.337 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.103 sys/dev/pci/ixgbe/ixgbe.c: revision 1.338 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.59 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: revision 1.13 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.14 sys/dev/pci/ixgbe/ixgbe_dcb.c: revision 1.15 sys/dev/pci/ixgbe/ixgbe_api.c: revision 1.29 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.5 sys/dev/pci/ixgbe/ixgbe.h: revision 1.90 sys/dev/pci/ixgbe/ixgbe.h: revision 1.91 sys/dev/pci/ixgbe/ixgbe.h: revision 1.92 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.46 ixg(4): Remove unused and old function. No functional change. - From FreeBSD's ix-3.3.32. ixg(4): Add 82599 LS once again. - From: FreeBSD: 9228ac3a69c4c7401a743e6465e118101a2beeb0 DPDK: 549ccd3dc01539e060597b503f2b65b272de3347 - This was removed 5 years ago. From the commit message: > Remove IXGBE_DEV_ID_82599_LS(0x154f) support again. I don't know why. This > was added in ix-3.2.18.tar.gz(NetBSD: ixgbe_82599.c rev. 1.20) and > removed in ix-3.3.6.tar.gz. ixg(4): Filter out spurious link up indication - Extend SFP+ cage crosstalk fix by re-checking link state after 5ms delay to filter out spurious link up indication by transceiver with no fiber cable connected. - From FreeBSD: In-tree: 04a5e6d7cadd06b10169c3c3a560649e7dc7444c Out of tree: ix-3.3.33. ixgbe: Simplify definitions. No functional change. Both DEFAULT_{TX,RX}D and PERFORM_{TX,RX}D are 2048. Use DEFAULT_{TX,RX}D. Same as FreeBSD. ixgbe: Modify error message of wrong TX/RX descriptor size. - Based from FreeBSD ix-3.3.35. I think ix-3.3.35's RING_INCREMENT(== 32) is wrong. It should be 8(DBA_ALIGN / sizeof(union ixgbe_adv_[tr]x_desc)). Linux also uses 8. ixgbe: Remove NO_82599_SUPPORT and NO_X540_SUPPORT support. NFCI. From FreeBSD ix-3.3.35. Note that this file is not used in NetBSD. ixgbe: Enable interrupt after setting IFF_RUNNING. Same as FreeBSD x-3.3.35. ixgbe: Cleanup. No fucntional change. Remove unused code, fix whitespace and modify comment to reduce against FreeBSD(mainly from ix-3.3.35).
1.183.4.2	13-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #405): sys/dev/pci/ixgbe/ixgbe_x550.c: revision 1.27 sys/dev/pci/ixgbe/ixgbe_fdir.h: revision 1.5 sys/dev/pci/ixgbe/ixv.c: revision 1.186 sys/dev/pci/ixgbe/ixgbe_osdep.c: revision 1.9 sys/dev/pci/ixgbe/ixgbe_common.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe_api.h: revision 1.17 sys/dev/pci/ixgbe/ixgbe.h: revision 1.89 sys/dev/pci/ixgbe/if_fdir.c: revision 1.6 sys/dev/pci/ixgbe/if_sriov.c: revision 1.18 sys/dev/pci/ixgbe/ixgbe.c: revision 1.333 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.102 sys/dev/pci/ixgbe/ixgbe.c: revision 1.339 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.58 sys/dev/pci/ixgbe/ixgbe_sriov.h: revision 1.6 sys/dev/pci/ixgbe/if_bypass.c: revision 1.10 sys/dev/pci/ixgbe/ixgbe_bypass.h: revision 1.4 sys/dev/pci/ixgbe/ixgbe_netmap.h: revision 1.3 sys/dev/pci/ixgbe/ixgbe_netmap.c: revision 1.6 sys/dev/pci/ixgbe/ixgbe_common.c: revision 1.45 ixgbe: Rename some definitions, modify comment. No functional change. Apply changes from FreeBSD's ix-3.3.31 and ixv-1.5.32. - struct adapter adapter -> struct ixgbe_softc sc - master -> primary - black -> block ixg(4): Whitespace. No functional change.
1.183.4.1	08-Oct-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #395): sys/dev/pci/ixgbe/ixgbe_82599.c: revision 1.30 sys/dev/pci/ixgbe/ixv.c: revision 1.184 sys/dev/pci/ixgbe/ixv.c: revision 1.185 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.32 sys/dev/pci/ixgbe/ixgbe_vf.c: revision 1.33 sys/dev/pci/ixgbe/ixgbe.h: revision 1.87 sys/dev/pci/ixgbe/ixgbe.h: revision 1.88 sys/dev/pci/ixgbe/ixgbe.c: revision 1.330 sys/dev/pci/ixgbe/ixgbe.c: revision 1.331 sys/dev/pci/ixgbe/ixgbe.c: revision 1.332 sys/dev/pci/ixgbe/ixgbe_vf.h: revision 1.18 sys/dev/pci/ixgbe/ix_txrx.c: revision 1.101 sys/dev/pci/ixgbe/ixgbe_type.h: revision 1.57 sys/dev/pci/ixgbe/ixgbe_mbx.h: revision 1.20 sys/dev/pci/ixgbe/ixgbe.c: revision 1.327 sys/dev/pci/ixgbe/ixgbe.c: revision 1.328 sys/dev/pci/ixgbe/ixgbe.c: revision 1.329 ixgbe: Reorder some event counters for readability. ixg(4): Rename some descriptions of event counters. - Rename some descriptions from register name to the meaning. - For the same meaning's counters, add "(soft)" or "(reg)". The former is for a software level counter and the latter is for a statistics counter register based. ixg(4): Count Queue Bytes {Transmit, Receive} counter. ixg(4): Reorder some flow control related event counters for readability. ixg(4): Rename some descriptions of flow control related event conters. Remove obsolete comment. ixgbe: Fix typo in comment. No functional change. ixv(4): Improve error check. ixgbe_vf.c rev. 1.31 changed the behavior of the ixgbe_check_mac_link_vf() function. It was from FreeBSD's ixv-1.5.25 to resolve mailbox collision problem. The change had a problem that error checks have not done at all if the API version >= 1.5. Fix it. From FreeBSD ixv-1.5.27. ixv(4): Add new IXGBE_VF_GET_LINK_STATE message support. PF can control vf's link state by this change. Note that Linux's PF driver can't control the link to force up (i.e. ip link set XXX vf Y state enable). From FreeBSD ixv-1.5.30. ixg(4): Update FCTRL after writing multicast filter. Same as other OSes. From FreeBSD 395cc55d896654b8f75071e71e856b22aed87da5.
1.196.2.1	02-Aug-2025	perseant	Sync with HEAD
1.8	01-Dec-2016	msaitoh	ixv.h is no more.
1.7	17-Aug-2015	knakahara	branches: 1.7.2; Add kernel code to support intrctl(8).
1.6	05-Aug-2015	msaitoh	Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. TODO: - Merge r280182 and newer. It's required to support X55x. - MSI/MSI-X support.
1.5	02-Apr-2015	msaitoh	Update our ixg(4) driver up to FreeBSD r238149: - Add TSO6 support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Fix a lot of bugs. - Improve performance.
1.4	27-Mar-2015	msaitoh	Synchronize our ixg(4) driver up to FreeBSD r230775: - Add X540 support. - Add 100BaseTX support. - Fix a lot of bugs. - Improve performance.
1.3	10-Mar-2015	msaitoh	Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X.
1.2	27-Oct-2012	chs	branches: 1.2.12; 1.2.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.1	12-Aug-2011	dyoung	branches: 1.1.2; 1.1.12; Add sources for ixgbe(4), a Intel 82599 10-gigabit ethernet driver ported from FreeBSD.
1.1.12.2	03-Dec-2017	jdolecek	update from HEAD
1.1.12.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.1.2.1	30-Oct-2012	yamt	sync with head
1.2.14.3	05-Dec-2016	skrll	Sync with HEAD
1.2.14.2	22-Sep-2015	skrll	Sync with HEAD
1.2.14.1	06-Apr-2015	skrll	Sync with HEAD
1.2.12.3	14-Jun-2016	snj	Pull up following revision(s) (requested by msaitoh in ticket #1174): sys/dev/pci/ixgbe/README: 1.2 sys/dev/pci/ixgbe/ixgbe.c: 1.30-1.33, 1.35 via patch sys/dev/pci/ixgbe/ixgbe.h: 1.8 via patch sys/dev/pci/ixgbe/ixgbe_82598.c: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_82599.c: 1.10 via patch sys/dev/pci/ixgbe/ixgbe_api.c: 1.8-1.11 via patch sys/dev/pci/ixgbe/ixgbe_api.h: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_common.c: 1.6-1.7 via patch sys/dev/pci/ixgbe/ixgbe_common.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h: 1.9 via patch sys/dev/pci/ixgbe/ixgbe_phy.c: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.h: 1.5 via patch sys/dev/pci/ixgbe/ixgbe_type.h: 1.11, 1.13 via patch sys/dev/pci/ixgbe/ixgbe_vf.c: 1.5-1.6 via patch sys/dev/pci/ixgbe/ixv.c: 1.9-1.10, 1.12-1.14 via patch sys/dev/pci/ixgbe/ixv.h: 1.6 via patch sys/dev/pci/ixgbe/ixgbe_dcb.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82598.h: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.c: 1.1 sys/dev/pci/ixgbe/ixgbe_dcb_82599.h: 1.1 Sync ixg(4) up to ixgbe.c rev. 1.35 except MSI/MSI-X and NET_MPSAFE: - Sync ixg(4) up to FreeBSD r279805 (or r280181) which include some bugfixes. - Add missing evcnt_detach(&rxr->rx_copies) in ixgbe_detach(). - Fix error path in ixgbe_attach() and ixgbe_detatch() to prevent panic. - Remove unreachable code. - Remove duplicated assignment of mh->m_len. - Reduce the diff against FreeBSD (tabs, spaces and comments) - ixv.c: Fix unintialized data. Include vlan.h and check NVLAN. - Avoid shadowing global `min'.
1.2.12.2	19-Apr-2015	riz	Apply patch (requested by msaitoh in ticket #697): sys/dev/pci/files.pci 1.375 via patch sys/dev/pci/ixgbe/ixgbe.c 1.24-1.27 via patch sys/dev/pci/ixgbe/ixgbe.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_82598.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_82598.h 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_82599.c 1.6-1.8 via patch sys/dev/pci/ixgbe/ixgbe_82599.h 1.1 via patch sys/dev/pci/ixgbe/ixgbe_api.c 1.5-1.6 via patch sys/dev/pci/ixgbe/ixgbe_api.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_common.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_common.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_mbx.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_osdep.h 1.4-1.6 via patch sys/dev/pci/ixgbe/ixgbe_phy.c 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_phy.h 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_type.h 1.8-1.9 via patch sys/dev/pci/ixgbe/ixgbe_vf.c 1.2-1.3 via patch sys/dev/pci/ixgbe/ixgbe_vf.h 1.3-1.4 via patch sys/dev/pci/ixgbe/ixgbe_x540.c 1.1-1.2 via patch sys/dev/pci/ixgbe/ixgbe_x540.h 1.1 via patch sys/dev/pci/ixgbe/ixv.c 1.5-1.7 via patch sys/dev/pci/ixgbe/ixv.h 1.4-1.5 via patch Synchronize our ixg(4) driver up to FreeBSD r243716: - Add X540 support. - Add TSO6 support. - Add 100BaseTX support. - The max size in dma tag is changed from 65535 to 262140 (IXGBE_TSO_SIZE). The value is the same as other *BSDs. The change might cause a address space shortage (ixgbe_dmamap_create() might fail) on some machines. - Show 1000Base-SX correctly. - Fix if_baudrate from 1G to 10G. - Fix a bug that ifconfig -z (SOICZIFDATA) doesn't work. - Fix a lot of bugs. - Improve performance. [msaitoh, ticket #697]
1.2.12.1	26-Mar-2015	martin	Pull up the following revisions (via patch), requested by msaitoh in #641: sys/dev/pci/files.pci 1.374 sys/dev/pci/ixgbe/ixgbe.h 1.3 sys/dev/pci/ixgbe/ixgbe.c 1.20-1.23 sys/dev/pci/ixgbe/ixgbe_type.h 1.3-1.7 sys/dev/pci/ixgbe/ixgbe_82599.c 1.4-1.5 sys/dev/pci/ixgbe/ixgbe_api.c 1.3-1.4 sys/dev/pci/ixgbe/ixgbe_vf.h 1.2 sys/dev/pci/ixgbe/ixv.c 1.3-1.4 sys/dev/pci/ixgbe/ixv.h 1.3 - Sync ixg(4) up to FreeBSD r230572. - Fix in the interrupt handler to make sure the stack TX queue is processed. (FreeBSD r222588) - The maximum read size of incoming packets is done in 1024-byte increments. The current code was rounding down the maximum frame size instead of routing up, resulting in a read size of 1024 bytes, in the non-jumbo frame case, and splitting the packets across multiple mbufs. (FreeBSD r225045) - Consequently the above problem exposed another issue, which is when packets were splitted across multiple mbufs, and all of the mbufs in the chain have the M_PKTHDR flag set. (FreeBSD r225045) - Use the correct constant for conversion between interrupt rate and EITR values (the previous values were off by a factor of 2) (FreeBSD r230572) - Make dev.ix.N.queueM.interrupt_rate a RW sysctl variable. Changing individual values affects the queue immediately, and propagates to all interfaces at the next reinit. (FreeBSD r230572) - Add dev.ix.N.queueM.irqs rdonly sysctl, to export the actual interrupt counts. (FreeBSD r230572) - Some netmap related changes. - Fix a bug that vlan setting may not apply to the parent interface correctly. - Add 82599EN_SFP and 82599_SFP_SF_QP. - Modify to make Intel Intel 10G Ethernet (ixg(4)) virtual function ixv(4) compilable. Not completed yet. It's required to use MSI-X. - Change comments, tabs and spaces to make difference among BSDs small. - Enclose macro argument correctly. Same as BSDs. No binary change. - Add missing parenthesis in IXGBE_CORE_LOCK_ASSERT() macro. No binary change.
1.7.2.1	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.8	22-May-2022	riastradh	opencrypto: Make freesession callback return void. No functional change intended: all drivers already return zero unconditionally.
1.7	06-Nov-2021	msaitoh	s/allign/align/
1.6	14-Jun-2020	riastradh	qat(4): Simplify iv generation logic with cprng_fast.
1.5	05-Mar-2020	msaitoh	branches: 1.5.4; s/Intialize/Initialize/ in comment.
1.4	01-Feb-2020	riastradh	opencrypto orders new_session -> process for us. No need for membar_producer. Discussed with hikaru@.
1.3	04-Dec-2019	hikaru	branches: 1.3.2; Fix macro error if QAT_DUMP is defined for debugging.
1.2	02-Dec-2019	msaitoh	Use PCI_MSIX_"TBL"BIR_MASK instead of PCI_MSIX_"PBA"BIR_MASK for MSI-X table. This is not a real bug because both macros have the same value.
1.1	20-Nov-2019	hikaru	Add opencrypto driver for Intel QuickAssist.
1.3.2.1	29-Feb-2020	ad	Sync with head.
1.5.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.5.4.1	05-Mar-2020	martin	file qat.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.2	05-Dec-2021	msaitoh	s/recusive/recursive/ in comment.
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_ae.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_aevar.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.2	24-May-2022	knakahara	Workaround for Atom C2000 series (Rangeley) qat, ok'ed by hikaru@n.o. C2000 qat has two engines, however it doesn't have arbiter. If we enable both engines, qat interrupts twice wrongly (and second interrupt notice invalid data). So, we enable only one engine.
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c2xxx.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c2xxxreg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c3xxx.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c3xxxreg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c62x.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_c62xreg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_d15xx.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_d15xxreg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw15.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw15reg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw15var.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw17.c was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw17reg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qat_hw17var.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.3	08-Apr-2022	andvar	s/postion/position/
1.2	31-Dec-2021	andvar	fix few typos in comments, mainly in word "parameter".
1.1	20-Nov-2019	hikaru	branches: 1.1.8; Add opencrypto driver for Intel QuickAssist.
1.1.8.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.1.8.1	20-Nov-2019	martin	file qatreg.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.3	06-Jul-2022	andvar	fix various typos in comments.
1.2	14-Mar-2020	ad	branches: 1.2.4; - Hide the details of SPCF_SHOULDYIELD and related behind a couple of small functions: preempt_point() and preempt_needed(). - preempt(): if the LWP has exceeded its timeslice in kernel, strip it of any priority boost gained earlier from blocking.
1.1	20-Nov-2019	hikaru	Add opencrypto driver for Intel QuickAssist.
1.2.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.2.4.1	14-Mar-2020	martin	file qatvar.h was added on branch phil-wifi on 2020-04-13 08:04:46 +0000
1.7	08-May-2019	isaki	Merge isaki-audio2 branch, the overhaul of audio subsystem. - Interrupt-oriented system rather than thread-oriented. - Improve stability, quality and performance. - Split playback and record cleanly. Improve halfduplex support. - Many bugs are fixed including deadlocks, resource leaks, abuses, etc. - Simplify audio filter mechanism. The encoding/channels/frequency conversions are completely handled in the upper layer. So the hard- ware driver only converts its hardware encoding (if necessary). - audio_hw_if changes: - Obsoletes query_encoding and add query_format instead. - Obsoletes set_params and add set_format instead. - Remove drain, setfd, mappage. - The call sequences are changed. - ioctl AUDIO_GETFD/SETFD, AUDIO_GETCHAN/SETCHAN are obsoleted. - ioctl AUDIO_{QUERY,GET,SET}FORMAT are introduced. - cleanup config attributes: au*conv and mulaw. - All hardware drivers should follow it (I've done as much as possible). Some file paths are changed: - dev/audio.c -> dev/audio/audio.c (rewritten) - dev/audiovar.h -> dev/audio/audiovar.h - dev/audio_dai.h -> dev/audio/audio_dai.h - dev/audio_if.h -> dev/audio/audio_if.h - dev/audiobell.c -> dev/audio/audiobell.c - dev/audiobellvar.h -> dev/audio/audiobellvar.h - dev/mulaw.[ch] -> dev/audio/mulaw.[ch] + dev/audio/alaw.c
1.6	13-Mar-2013	macallan	branches: 1.6.38; 1.6.42; defflag VOYAGERFB_DEPTH_32
1.5	21-Feb-2012	macallan	branches: 1.5.2; use glyphcache for anti-aliased fonts
1.4	22-Dec-2011	macallan	support running in 32bit colour with anti-aliased fonts enable with options VOYAGERFB_ANTIALIAS
1.3	13-Dec-2011	macallan	define and use PWMCLOCK_DEBUG
1.2	13-Dec-2011	macallan	an alternative clock implementation for Loongson2F and SM502, mostly Gdium-specific. This supports Loongson2F frequency scaling. The problem is that Gdium has no CPU clock independent high resolution timer we can use as a timecounter so we use one of the SM502's PWMs to generate a 100Hz timer interrupt, use the cp0 counter to measure time and adjust for frequency changes. Other Loongson-based machines will need something similar but hopefully less hackish.
1.1	31-Aug-2011	macallan	branches: 1.1.2; 1.1.6; split up voyagerfb into the framebuffer portion and a pseudo bus that we can attach all the other sub-devices to while there attach an i2c bus
1.1.6.2	24-Feb-2012	mrg	sync to -current.
1.1.6.1	18-Feb-2012	mrg	merge to -current.
1.1.2.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.2.1	17-Apr-2012	yamt	sync with head
1.5.2.1	23-Jun-2013	tls	resync from head
1.6.42.1	05-May-2019	isaki	Remove obsoleted au{,rate,vol}conv and mulaw attributes. audio provides the equivalent of them inseparably.
1.6.38.1	10-Jun-2019	christos	Sync with HEAD
1.12	29-May-2020	rin	For struct timecounter, use C99 initializers. Compile tested. No functional changes intended.
1.11	26-Aug-2016	skrll	Adjust evbmips_iointr to pass a clockframe pointer and use it for pwmclock @ voyager. Suggested by matt@ Hi macallan!
1.10	14-May-2013	macallan	branches: 1.10.10; found the missing commit so put this back am I the only one here with a Gdium?
1.9	13-May-2013	christos	make this compile again; macallan, please commit the right fix!
1.8	16-Apr-2013	macallan	unscrew this again
1.7	15-Apr-2013	christos	make this compile again; macallan, please commit the right fix!
1.6	02-Jun-2012	dsl	branches: 1.6.2; Add some pre-processor magic to verify that the type of the data item passed to sysctl_createv() actually matches the declared type for the item itself. In the places where the caller specifies a function and a structure address (typically the 'softc') an explicit (void ) cast is now needed. Fixes bugs in sys/dev/acpi/asus_acpi.c sys/dev/bluetooth/bcsp.c sys/kern/vfs_bio.c sys/miscfs/syncfs/sync_subr.c and setting AcpiGbl_EnableAmlDebugObject. (mostly passing the address of a uint64_t when typed as CTLTYPE_INT). I've test built quite a few kernels, but there may be some unfixed MD fallout. Most likely passing &char[] to char . Also add CTLFLAG_UNSIGNED for unsiged decimals - not set yet.
1.5	23-Feb-2012	macallan	branches: 1.5.2; break long lines
1.4	23-Feb-2012	macallan	set the CPU clock back to full speed on shutdown this is necessary because PMON does not touch the clock scaling register on reboot but still reports full CPU speed which confuses the clock code
1.3	14-Feb-2012	macallan	branches: 1.3.4; another blah() -> blah(void) ok riz
1.2	13-Dec-2011	macallan	define and use PWMCLOCK_DEBUG
1.1	13-Dec-2011	macallan	an alternative clock implementation for Loongson2F and SM502, mostly Gdium-specific. This supports Loongson2F frequency scaling. The problem is that Gdium has no CPU clock independent high resolution timer we can use as a timecounter so we use one of the SM502's PWMs to generate a 100Hz timer interrupt, use the cp0 counter to measure time and adjust for frequency changes. Other Loongson-based machines will need something similar but hopefully less hackish.
1.3.4.3	24-Feb-2012	mrg	sync to -current.
1.3.4.2	18-Feb-2012	mrg	merge to -current.
1.3.4.1	14-Feb-2012	mrg	file pwmclock.c was added on branch jmcneill-usbmp on 2012-02-18 07:34:55 +0000
1.5.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.5.2.3	30-Oct-2012	yamt	sync with head
1.5.2.2	17-Apr-2012	yamt	sync with head
1.5.2.1	23-Feb-2012	yamt	file pwmclock.c was added on branch yamt-pagecache on 2012-04-17 00:07:59 +0000
1.6.2.2	03-Dec-2017	jdolecek	update from HEAD
1.6.2.1	23-Jun-2013	tls	resync from head
1.10.10.1	05-Oct-2016	skrll	Sync with HEAD
1.34	20-Dec-2023	thorpej	Remove unnecessary <sys/malloc.h>.
1.33	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.32	21-May-2021	macallan	branches: 1.32.4; fix wsdisplay attachment in the New Order Of Things
1.31	21-May-2021	macallan	use probed VRAM size to determine where to put the cursor instead of blindly assuming 16MB also make sure the glyph cache can't overlap with the cursor
1.30	24-Apr-2021	thorpej	branches: 1.30.2; 1.30.4; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.29	19-Jan-2018	macallan	branches: 1.29.20; enable font loading
1.28	13-Jan-2016	macallan	some trivial fixes: - rearrange dmesg output so all values are printed properly - use aprint_* - support WSDISPLAYIO_GET_FBINFO
1.27	11-Mar-2014	mrg	branches: 1.27.6; avoid set but unused variables. move variables under their usage #ifdef.
1.26	30-Jul-2013	macallan	avoid uninitialized use of defattr
1.25	19-Mar-2013	macallan	branches: 1.25.6; rearrange things a little bit to avoid showing garbled VRAM content while switching colour depth
1.24	13-Mar-2013	macallan	break a bunch of long lines
1.23	13-Mar-2013	macallan	support anti-aliased fonts in 8bit while there, use the DRAM config register to figure out how much video memory we have instead of guessing
1.22	30-May-2012	macallan	branches: 1.22.2; fix some comments
1.21	23-May-2012	macallan	first step to fix PR 46461 - don't jump through a NULL pointer when we're not the console but the firmware set up the graphics hardware anyway
1.20	19-Apr-2012	macallan	glyphcache_init() works best with its parameters in the right order. Doh. While there request an RGB devcmap.
1.19	20-Mar-2012	macallan	split voyagerfb_putchar() into one method for mono fonts and one for anti-aliased fonts
1.18	13-Mar-2012	elad	Replace the remaining KAUTH_GENERIC_ISSUSER authorization calls with something meaningful. All relevant documentation has been updated or written. Most of these changes were brought up in the following messages: http://mail-index.netbsd.org/tech-kern/2012/01/18/msg012490.html http://mail-index.netbsd.org/tech-kern/2012/01/19/msg012502.html http://mail-index.netbsd.org/tech-kern/2012/02/17/msg012728.html Thanks to christos, manu, njoly, and jmmv for input. Huge thanks to pgoyette for spinning these changes through some build cycles and ATF.
1.17	21-Feb-2012	macallan	use glyphcache for anti-aliased fonts
1.16	11-Jan-2012	macallan	use rasops_init(0,0), WSFONT_GLYPH(), FONT_IS_ALPHA()
1.15	28-Dec-2011	macallan	make anti-aliased fonts with odd widths work The drawing engine needs lines padded to 64bit when doing colour host blits which is of course not mentioned anywhere in the manual.
1.14	27-Dec-2011	macallan	add fastpaths for alpha == 0 and alpha == 255 - no need to do the six multiplications dance there, just use background or foreground colour
1.13	24-Dec-2011	macallan	don't put alpha values in a variable named alpha since alpha doesn't like variables conflicting with platform names ( just in case someone figures out how to stick an sm50x into an alpha )
1.12	22-Dec-2011	macallan	don't probe colour depth on attach - we're going to change it anyway
1.11	22-Dec-2011	macallan	don't switch to 8 bit for WSDISPLAYIO_MODE_EMUL if VOYAGERFB_ANTIALIAS is defined
1.10	22-Dec-2011	macallan	support running in 32bit colour with anti-aliased fonts enable with options VOYAGERFB_ANTIALIAS
1.9	08-Nov-2011	macallan	branches: 1.9.4; switch to 32bit colour for X, switch back to 8 bit for console emulation
1.8	08-Nov-2011	macallan	run the console in 8 bit colour for additional speed ( on Gdium this scrolls more than twice as fast as the 16bit default mode )
1.7	18-Oct-2011	macallan	branches: 1.7.2; add backlight control via sm502 PWM
1.6	28-Sep-2011	macallan	support a hardware cursor - now X with wsfb on gdium is a little less annoying
1.5	22-Sep-2011	macallan	use BUS_SPACE_MAP_PREFETCHABLE
1.4	20-Sep-2011	macallan	support ioctl(WSDISPLAYIO_SVIDEO) and friends so wsfb can turn the backlight off
1.3	06-Sep-2011	macallan	add simple backlight control support, only on/off for now
1.2	06-Sep-2011	macallan	wsdisplayio_busid_pci() needs out parent since we don't attach directly to a PCI bus anymore
1.1	31-Aug-2011	macallan	split up voyagerfb into the framebuffer portion and a pseudo bus that we can attach all the other sub-devices to while there attach an i2c bus
1.7.2.5	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.7.2.4	30-Oct-2012	yamt	sync with head
1.7.2.3	23-May-2012	yamt	sync with head.
1.7.2.2	17-Apr-2012	yamt	sync with head
1.7.2.1	10-Nov-2011	yamt	sync with head
1.9.4.5	02-Jun-2012	mrg	sync to latest -current.
1.9.4.4	29-Apr-2012	mrg	sync to latest -current.
1.9.4.3	05-Apr-2012	mrg	sync to latest -current.
1.9.4.2	24-Feb-2012	mrg	sync to -current.
1.9.4.1	18-Feb-2012	mrg	merge to -current.
1.22.2.3	03-Dec-2017	jdolecek	update from HEAD
1.22.2.2	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.22.2.1	23-Jun-2013	tls	resync from head
1.25.6.2	18-May-2014	rmind	sync with head
1.25.6.1	28-Aug-2013	rmind	sync with head
1.27.6.1	19-Mar-2016	skrll	Sync with HEAD
1.29.20.1	21-Mar-2021	thorpej	Give config_found() the same variadic arguments treatment as config_search(). This commit only adds the CFARG_EOL sentinel to the existing config_found() calls. Conversion of config_found_sm_loc() and config_found_ia() call sites will be in subsequent commits.
1.30.4.1	31-May-2021	cjep	sync with head
1.30.2.1	17-Jun-2021	thorpej	Sync w/ HEAD.
1.32.4.1	04-Aug-2021	thorpej	Adapt to CFARGS().

OpenGrok