Cross Reference: /src/sys/dev/ata

History log of /src/sys/dev/ata
Revision	Date	Author	Comments
1.6	11-Dec-2005	christos	merge ktrace-lwp.
1.5	14-Dec-2003	thorpej	Serial ATA register definitions.
1.4	05-Aug-2002	soren	branches: 1.4.6; Add ATA SMART reporting support from Ben Collver in kern/12787.
1.3	11-Apr-2001	jdolecek	branches: 1.3.2; 1.3.14; Put back all header files defining ioctls
1.2	12-Oct-1998	bouyer	branches: 1.2.12; 1.2.24; Adjust include list.
1.1	12-Jun-1998	cgd	Rework the way kernel include files are installed. In the new method, as with user-land programs, include files are installed by each directory in the tree that has includes to install. (This allows more flexibility as to what gets installed, makes 'partial installs' easier, and gives us more options as to which machines' includes get installed at any given time.) The old SYS_INCLUDES={symlinks,copies} behaviours are _both_ still supported, though at least one bug in the 'symlinks' case is fixed by this change. Include files can't be build before installation, so directories that have includes as targets (e.g. dev/pci) have to move those targets into a different Makefile.
1.2.24.2	13-Aug-2002	nathanw	Catch up to -current.
1.2.24.1	21-Jun-2001	nathanw	Catch up to -current.
1.2.12.1	21-Apr-2001	bouyer	Sync with HEAD
1.3.14.1	29-Aug-2002	gehenna	catch up with -current.
1.3.2.1	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.4.6.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.6.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.6.1	03-Aug-2004	skrll	Sync with HEAD
1.13	24-Oct-2018	jdolecek	remove TODO.ncq; the FIS note was added to atastart(), and the note about dump to unopened wd(4) device was attached to PR kern/16789, and nothing else remains
1.12	24-Oct-2018	jdolecek	move the comment (and expand) about NCQ TRIM to wd_trim()
1.11	24-Oct-2018	jdolecek	detach the controller itself on shutdown; adjust to not detach already detached atabus/channel
1.10	23-Oct-2018	jdolecek	add support for DIOCCACHESYNC (!), and DIOCGCACHE; code adapted from ccd(4) tested with Intel ATA Raid
1.9	22-Oct-2018	jdolecek	remove the note for dump, moral equivalent is done and biodone() most certainly should not be done
1.8	22-Oct-2018	jdolecek	remove the wd* at umass? entry (comment moved to umass_isdata.c and manpage) remove MSI for ahcisata(4), siisata(4) - these were done
1.7	22-Oct-2018	jdolecek	move mvsata(4) related items from ata/TODO.ncq to dev/ic/mvsata.c, it's unlikely this will ever be done, but still good to note somewhere
1.6	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.5	14-Oct-2018	jdolecek	remove M_CANFAIL flag for malloc(9) - it was completely ignored, so had actually no effect
1.4	31-Aug-2018	jdolecek	branches: 1.4.2; one less
1.3	01-Nov-2017	jdolecek	branches: 1.3.2; 1.3.4; 1.3.6; one down - mlelstv converted wd(4) to dksubr
1.2	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.1	19-Apr-2017	jdolecek	branches: 1.1.2; file TODO.ncq was initially added on branch jdolecek-ncq.
1.1.2.47	29-Sep-2017	jdolecek	introduce ATA_BSIZE and use it instead of DEV_BSIZE for get params and recovery, where they are by spec 512 bytes and not negotiable
1.1.2.46	28-Sep-2017	jdolecek	add sysctls to control if NCQ is being used, and how many max tags; I have a drive which is significantly slower with NCQ than non-NCQ, and it's generally useful to have this easily overridable while here, also move the frequency settings for WD_CHAOS_MONKEY to a sysctl and make it per-drive
1.1.2.45	28-Sep-2017	jdolecek	add note about DEV_BSIZE use
1.1.2.44	21-Sep-2017	jdolecek	wddone() doesn't need more locking
1.1.2.43	20-Sep-2017	jdolecek	MVSATA_WITHOUTDMA seems to work fine
1.1.2.42	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.1.2.41	13-Sep-2017	jdolecek	two more - mvsata MVSATA_WITHOUTDMA seems to be broken, and latest channel lock changes introduced panic in mi_switch() on code paths which kpause/tsleep
1.1.2.40	11-Sep-2017	jdolecek	doesn't seem the freeze/thaw in error recovery can cause the thread to panic, all commands within are executed as polled and hence thread is not invoked
1.1.2.39	11-Sep-2017	jdolecek	atastart() rechecked - it's okay to call it even in error path in those several cases we do, it's just optimization to skip the call; I'm not even very convinced it's useful to have this conditional, but keeping for now until proven harmful
1.1.2.38	10-Sep-2017	jdolecek	refactor code so that xfer c_start() hook is called with channel mutex held, and hence the controller submit code no longer relies on spl tested all the affected drivers - wdc (via piixide), ahci, mvsata, siisata, both disk and atapi I/O
1.1.2.37	29-Aug-2017	jdolecek	some notes around c_start/atastart()
1.1.2.36	13-Aug-2017	jdolecek	move the dump note to post-merge, it's not new siisata seems to be fine, no longer holds the merge remove the kill active transfers after software drive reset - not relevant now only the wd* at umass? stays as new item, but I won't hold the merge for this, as I don't have the hardware and it is contained enough to be resolved on HEAD
1.1.2.35	12-Aug-2017	jdolecek	add NCQ error recovery for mvsata(4) fix wait flags for mvsata_bio_start(), AT_* constants are only valid for cmd change the debug logging to similar form as ahcisata/siisata, so it's easier to selectively show
1.1.2.34	12-Aug-2017	jdolecek	convert the atabus thread to use the channel lock and a condvar, adjust code which sets the relevant channel flags to take the lock while doing so
1.1.2.33	12-Aug-2017	jdolecek	remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status, so that there is less hidden state shared by commands; primary intent is to make the NCQ and non-NCQ paths more similar, and remove possibility of incorrect handling for the NCQ commands tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4), ahcisata(4), mvsata(4) on real hw
1.1.2.32	01-Aug-2017	jdolecek	another one down
1.1.2.31	31-Jul-2017	jdolecek	some more notes
1.1.2.30	19-Jul-2017	jdolecek	update to note remaining work move some stuff to 'after-merge'
1.1.2.29	03-Jul-2017	jdolecek	expand what needs to be done with error handling
1.1.2.28	28-Jun-2017	jdolecek	ATAPI on siisata(4), ahcisata(4) tested
1.1.2.27	27-Jun-2017	jdolecek	note ATAPI on siisata, ahcisata needs to be tested
1.1.2.26	24-Jun-2017	jdolecek	fix dump on mvsata - there was misplaced mvsata_bio_intr() call causing diagnostic panic, and we also need to switch controller to appropriate DMA mode (NCQ vs. non-NCQ DMA) while here, also reduce delay for the polled command wait to 100 us, same as siisata(4) and ahcisata(4)
1.1.2.25	24-Jun-2017	jdolecek	add detach code for umass_isdata; compile-tested only
1.1.2.24	24-Jun-2017	jdolecek	only limit the openings for I/O xfer if the drive actually supports NCQ; if it's non-NCQ drive, the tag is not going to be used, so we can use any xfer
1.1.2.23	23-Jun-2017	jdolecek	fix dump for siisata(4) to work - need to call drv_done() callback in order to not leak xfers while here, reduce the DELAY() for polled bio to make it go much faster, same value as used in ahcisata(4) needs also rev. 1.252 src/sys/arch/x86/x86/pmap.c to not trigger the assertion, the pmap.c fix committed only on HEAD
1.1.2.22	23-Jun-2017	jdolecek	restart I/O processing after freeing xfer, i.e. now even after commands like cache flush or standby; the command handling no longer use on-stack xfer, hence use queue slot and compete with normal I/O for the xfers the restart give change to all drives attached to the same channel in round-robin fashion, for fair usage and to recover from situation when disk is idle due to all xfers being consumed by other drives make special concession for flush cache - ignore any new I/O requests on the particular disk while the flush cache is waiting for xfer, so that I/O queue won't starve the flush cache and the flush cache would be done ASAP tested on piixide(4), ahci(4), siisata(4)
1.1.2.21	21-Jun-2017	jdolecek	hold channel lock for ata_exec_xfer() and most of atastart(), convert C_WAITACT handling to use condvar add comment for the code block with C_FREE and ata_free_xfer() explaining why it's not abstraction layer violation
1.1.2.20	20-Jun-2017	jdolecek	some more notes for dump on unopened device (PR kern/16789)
1.1.2.19	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.1.2.18	19-Jun-2017	jdolecek	add ata_channel lock, use it to protect queue manipulation (only that for now); add ata_channel_detach() to destroy the locks change ata_get_xfer() so that it can wait for xfer, convert all on-stack xfer code to use the blocking variant fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(), convert drive probe code also over from slot0 XXX to ata_get_xfer() drive reset and PMP now works on siisata(4) too; changes tested also on piixide(4), ahci(4), mvsata(4)
1.1.2.17	17-Jun-2017	jdolecek	more notes, esp. PMP on siisata doesn't work + more problems with downsize
1.1.2.16	17-Jun-2017	jdolecek	make PMP working great again tested with mvsata(4), my ahcisata(4) controller unfortunately doesn't support PMP
1.1.2.15	16-Jun-2017	jdolecek	notes from dump testing - TL;DR: broken, fixing
1.1.2.14	13-Jun-2017	jakllsch	Add NCQ support to siisata(4). There are still issues that need to be address before I consider this mergable.
1.1.2.13	12-Jun-2017	jakllsch	Add a few more NCQ branch action items.
1.1.2.12	12-Jun-2017	jdolecek	update notes
1.1.2.11	24-Apr-2017	jdolecek	set NCQ priority field to 'high' for BPRIO_TIMECRITICAL transfers if drive supports it
1.1.2.10	24-Apr-2017	jdolecek	do not use PRD Byte Count for queued commands to determine number of bytes actually transferred; it should not be used and is not guaranteed to be valid according to AHCI spec, underflow is always invalid for queued cmds this fixes my ls problem under QEMU
1.1.2.9	24-Apr-2017	jdolecek	one down
1.1.2.8	23-Apr-2017	jdolecek	note possible problem with ata_exec_xfer()
1.1.2.7	21-Apr-2017	jdolecek	move the wd* at umass? item to bugs, too
1.1.2.6	21-Apr-2017	jdolecek	group items, slightly expand description
1.1.2.5	20-Apr-2017	jdolecek	remove ahci timeouts, fixed by Jonathan
1.1.2.4	20-Apr-2017	jdolecek	use a fake ata_channel struct in umass_isdata.c so that the wd(4) calls to ata_{get,free}_xfer() can work, adjust to ata interface changes compile tested only
1.1.2.3	19-Apr-2017	jdolecek	note the QEMU bug
1.1.2.2	19-Apr-2017	jdolecek	up to 65536 sectors can be transferred for LBA48 commands; adjust wdminphys() and wdstart1() accordingly
1.1.2.1	19-Apr-2017	jdolecek	bunch of random observations
1.3.6.1	10-Jun-2019	christos	Sync with HEAD
1.3.4.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.3.4.2	20-Oct-2018	pgoyette	Sync with head
1.3.4.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.3.2.2	03-Dec-2017	jdolecek	update from HEAD
1.3.2.1	01-Nov-2017	jdolecek	file TODO.ncq was added on branch tls-maxphys on 2017-12-03 11:36:59 +0000
1.4.2.17	21-Oct-2018	jdolecek	testing done (including PMP on siisata(4) and ATAPI cd(4) on ahcisata/siisata)
1.4.2.16	17-Oct-2018	jdolecek	rechecked bad block handling under Parallels and it really doesn't set any error bits in any of the registers (neither IS nor SERR) for out-of-range block read nor write, while it works just fine under VMware, so not our bug
1.4.2.15	15-Oct-2018	jdolecek	change the SATA/NCQ recovery to run in the atabus thread
1.4.2.14	14-Oct-2018	jdolecek	adjust mvsata_bio_intr() so it recognizes the 'tfd' parameter as passed by recovery and hence works; use it also for passing state from mvsata_edma_handle()
1.4.2.13	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.4.2.12	06-Oct-2018	jdolecek	move the entry for error recovery on thread up, it will be done on ncqfixes branch
1.4.2.11	04-Oct-2018	jdolecek	relax the NOERROR + C_CHAOS, just rerun the xfer instead of panic, to handle another variant of bad/missing AHCI error recovery, this time under Parallels
1.4.2.10	03-Oct-2018	jdolecek	change channel reset and drive reset for all ATA controllers to always run via thread, and with channel lock held the whole time; the queue is frozen while reset is pending for this repurpose ata_reset_channel() into new ata_thread_run() also adjust some device printfs to not leak xfer pointer, and avoid aprint_* for non-autoconf messages
1.4.2.9	24-Sep-2018	jdolecek	replace KASSERT() for READ LOG EXT returning without error indication with just log to avoid triggering panic under QEMU - it doesn't implement the command for AHCI emulation
1.4.2.8	22-Sep-2018	jdolecek	one more
1.4.2.7	22-Sep-2018	jdolecek	remove explicit ata_channel_start() calls, it's no longer necessary now that ata_xfer's are allocated via pool and not really limited; replace by just a callout to restart the processing for rare cases where system runs out of memory
1.4.2.6	22-Sep-2018	jdolecek	fix use-after-free in wd(4) dump, detected by switch to the pool change code in wd_dumpblocks() to use it's own non-pool ata_xfer, which skips the deallocation step and thus keeps the contents when the I/O is finished
1.4.2.5	22-Sep-2018	jdolecek	use KM_SLEEP in atabus_alloc_drives() to avoid spurious failures; this is not called from interrupt context
1.4.2.4	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.4.2.3	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.4.2.2	31-Aug-2018	jdolecek	fix comment - dynamically allocated yes, per-drive not
1.4.2.1	31-Aug-2018	jdolecek	list goals for the jdolecek-ncqfixes branch
1.171	17-Feb-2025	jakllsch	Restore ATA DMA mode downgrade support everywhere; it's a necessary part of any system supporting (parallel) ATA DMA. There is hardware out there, including cobalt, macppc, and sparc64 where this functionality is necessary to avoid non-functional disks, either in as-shipped hardware configurations or with add-in cards, or perhaps just with compromised IDE/PATA cables. Should address: PR 58767 PR 59023 PR 59078 If anyone really insists on not having this support they can now turn it off themselves with `options ATA_NO_DOWNGRADE_MODE`
1.170	17-Feb-2025	jakllsch	Fix aprint_error_dev() abuse
1.169	31-May-2022	andvar	branches: 1.169.10; fix various typos in comments, documentation and messages.
1.168	28-May-2022	andvar	fix various typos in comments.
1.167	14-Mar-2022	perseant	Avoid an unaccounted extra channel freeze, if a reset is requested more than once before the thread services the request. Closes PR#56745.
1.166	23-Feb-2022	andvar	fix various typos in comments, mainly immediatly/immediately/, as well shared and recently fixed typos in OpenBSD code by Jonathan Grey.
1.165	12-Nov-2021	skrll	Trailing whitespace
1.164	05-Oct-2021	rin	PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.163	29-Aug-2021	rin	Output missing '\n' for capability list when DMA support is not compiled in.
1.162	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.161	24-Apr-2021	thorpej	branches: 1.161.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.160	03-Oct-2020	riastradh	branches: 1.160.4; autoconf: Blame devices holding up boot with config_pending. Blame message requires `boot -x' (AB_DEBUG). Fix ata so it doesn't mismatch config_pending_incr/decr devices.
1.159	25-May-2020	jdolecek	disable downgrade of ATA mode from DMA, as generally not relevant any more - while it has been instrumental to inadvertedly discover driver bugs in PIO mode under QEMU recently, generally the switch more hurts than helps, so now only warn when DMA errors happen code kept under ATA_DOWNGRADE_MODE ifdef, disabled by default
1.158	25-May-2020	jdolecek	make ata_downgrade_mode() static, it's not used anywhere else
1.157	02-May-2020	thorpej	Back out changes to use a threadpool for now; it's causing trouble for some folks on Thinkpads.
1.156	25-Apr-2020	thorpej	Rather than creating a kthread-per-channel, use a threadpool and a threadpool-job-per-channel for the in-thread-context work that needs to be done (which is rare). On one of my test systems, this results in the total number of LWPs after multi-user boot dropping from 116 to 78.
1.155	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.154	04-Apr-2020	jdolecek	branches: 1.154.2; fix deadlock in wdcwait() when xfer timeout happens while the atabus thread sleeps in wdcwait() - check current lwp rather than relying on global ATACH_TH_RUN channel flag should fix the hang part of the problem reported in http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html thanks to Paul Ripke for providing extensive debugging info
1.153	21-Oct-2019	christos	chuq does not like insomniac allocations so unlock-alloc-lock instead.
1.152	21-Oct-2019	christos	Fix assert_sleepable() panic by allocating with NOSLEEP. The alternative is to unlock and relock the channel, but seems more dangerous to do so.
1.151	15-Oct-2019	chs	convert more KM_NOSLEEP to KM_SLEEP and remove code to handle failures.
1.150	21-Aug-2019	msaitoh	Fix typo (s/controler/controller/).
1.149	25-May-2019	christos	branches: 1.149.2; use PR_ZERO
1.148	06-Apr-2019	uwe	Make sure the aprint_verbose chain is terminated with a newline. Check sep instead of printed, b/c the latter is reset several times.
1.147	11-Dec-2018	jdolecek	in atabus_detach(), shutdown the channel thread only after all the children successfully detach; this is especially important for atapibus, which needs the thread to execute commands fixes hang on poweroff with root on ATAPI cd(4), reported by Christoph Badura
1.146	12-Nov-2018	jdolecek	hold channel lock during whole ata_dmaerr()/ata_downgrade_mode() - according to code inspection this is safe, none of the set_modes hooks execute anything which would be taking the lock adresses PR kern/53714 by Andreas Gustafsson
1.145	24-Oct-2018	jdolecek	ata_kill_active() needs to drop channel drop before calling kill_xfer() hook, otherwise code tries to lock against itself when deactivating the xfer
1.144	24-Oct-2018	jdolecek	add further note about FIS-based switching to atastart()
1.143	24-Oct-2018	jdolecek	detach the controller itself on shutdown; adjust to not detach already detached atabus/channel
1.142	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.141	28-Oct-2017	riastradh	branches: 1.141.2; 1.141.4; 1.141.6; Kill some more extern struct cfdriver declarations. Down with externs in .c!
1.140	20-Oct-2017	jdolecek	move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.139	19-Oct-2017	jdolecek	more detailed debug info; also sync DEBUG_* values in wd.c with ata.c
1.138	18-Oct-2017	jdolecek	add newlines to the debug messages
1.137	15-Oct-2017	jdolecek	move the extra thaw for scheduled ata_reset_channel() to the function itself, so it's done regardless if the actual reset is run from thread context or e.g. call with AT_POLL; fixes a hang after xfer failure and dma downgrade add some debugging printfs, so freeze/thaw issues are easier to track
1.136	10-Oct-2017	jdolecek	split off functions used by 'wd* at umass?' into separate file, unfortunately the previous approach with NATABUS doesn't work for kernels which include MODULAR, but not atabus - such as macppc and evbarm
1.135	08-Oct-2017	jdolecek	as a stopgap measure, include necessary parts of dev/ata/ata.c even if no atabus is configured, so that kernels with just 'wd* at umass?' and no atabus compile; fixes kernel builds for vax, pmax and evbmips after NCQ branch merge XXX wd* at umass? still actually untested after the merge, need to discuss XXX what to do with it
1.134	08-Oct-2017	mlelstv	Don't assert channel lock in polled mode. ahcisata (and maybe others) use ata_delay during initialization where lock isn't taken.
1.133	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.132	10-Sep-2014	matt	branches: 1.132.8; g/u_int[0-9]*_t/ s/u_int/uint/g
1.131	25-Jul-2014	dholland	Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.130	16-Mar-2014	dholland	branches: 1.130.2; Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.129	12-Oct-2013	christos	Pass the device name in, so we can debug what deferred drivers did not work.
1.128	07-Aug-2013	blymn	Make sure a channel has drives attached before resetting it otherwise wdc_reset will panic the kernel due to a null derefrence of ch_drives.
1.127	03-Apr-2013	bouyer	branches: 1.127.4; Fix kernel dump on ahci controller, by making sure we won't sleep while dumping: - introduce ata_delay() which either use delay() or kpause() depending on flags. use it in sata_reset_interface() and some ahci functions - kill ATA_NOSLEEP, it was tested but never set. use ATA_POLL instead. - reduce delay while polling in ahci, to speed up the dump Should fix PR kern/41095
1.126	01-Nov-2012	abs	Allow setting ATADEBUG_MASK to default value of atadebug_mask
1.125	04-Aug-2012	bouyer	branches: 1.125.2; If ch_ndrives is > 0, then ch_drive is not supposed to be NULL.
1.124	31-Jul-2012	bouyer	Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.123	29-Jul-2012	jakllsch	Remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas. Originally part of a commit by bouyer@.
1.122	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.121	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.120	22-Jul-2012	jakllsch	It turns out my previous commit here was off by one.
1.119	15-Jul-2012	jakllsch	Stopgap crash prevention when atadebug_mask includes DEBUG_PROBE bit.
1.118	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.117	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.116	06-Apr-2012	isaki	Fix a typo (that I found 5 years ago). Though the definition of atabusiodetach_args is the same as atabusioscan_args (at the moment).
1.115	30-Apr-2011	jakllsch	branches: 1.115.4; 1.115.8; Add simplistic atabus(4) rescan support.
1.114	18-Apr-2011	rmind	Replace few simple_lock and ltsleep/wakeup uses with mutex(9) and condvar(9). Note to all: please replace old primitives in your code! Thanks.
1.113	28-Mar-2010	snj	branches: 1.113.2; Spell "enough" properly.
1.112	24-Feb-2010	dyoung	branches: 1.112.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.111	21-Jan-2010	dyoung	branches: 1.111.2; Cosmetic: join some lines. Get out early on errors, change switch (...) { ... error = ...; break; } return error; to switch (...) { ... return ...; }
1.110	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.109	12-Nov-2009	dyoung	Remove superfluous activation hook.
1.108	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.107	13-Sep-2009	pooka	Wipe out the last vestiges of POOL_INIT with one swift stroke. In most cases, use a proper constructor. For proplib, give a local equivalent of POOL_INIT for the kernel object implementation. This way the code structure can be preserved, and a local link set is not hazardous anyway (unless proplib is split to several modules, but that'll be the day). tested by booting a kernel in qemu and compile-testing i386/ALL
1.106	12-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.105	07-Apr-2009	dyoung	Cosmetic: flag a potential problem in ata_channel_attach(): it initializes a callout that does not appear to be destroyed anywhere. Every callout_init() should ordinarily be matched by a callout_destroy().
1.104	03-Apr-2009	dyoung	Stop dereferencing a dangling device_t pointer and crashing: skip the drives flagged DRIVE_ATAPI in atabus_activate(,DVACT_DEACTIVATE) just as we skip them in atabus_detach() and in atabus_childdetched(). Make atabus_detach() parallel attachment more closely by calling config_detach() on the child chp->ata_drives[i] instead of on chp->ch_drive[i].drv_softc. Assert that ata_drives[i] and ch_drive[i].drv_softc are equal, and set them both to NULL in atabus_childdetached().
1.103	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.102	16-Nov-2008	bouyer	branches: 1.102.4; cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.101	02-Oct-2008	bouyer	branches: 1.101.2; 1.101.4; Kill ATACH_TH_RUN and use cpu_intr_p() instead.
1.100	01-Oct-2008	bouyer	Use a separate thread to probe/attach atabus's childrens. Fixes a deadlock where the interrupt routine wants to wake up the atabus thread to perform a reset, while the thread is blocked in wd's attach function.
1.99	12-Jun-2008	cegger	branches: 1.99.2; use device_lookup_private to get softc
1.98	18-Mar-2008	cube	branches: 1.98.2; 1.98.4; 1.98.6; 1.98.8; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.97	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.96	10-Jan-2008	dyoung	branches: 1.96.2; 1.96.6; Oops, remove some debug statements that slipped in there.
1.95	10-Jan-2008	dyoung	Use device_t, device_xname(), device_private(). Add atabus_childdetached() to clean up pointers to detached children.
1.94	04-Jan-2008	ad	Start detangling lock.h from intr.h. This is likely to cause short term breakage, but the mess of dependencies has been regularly breaking the build recently anyhow.
1.93	20-Dec-2007	pooka	Don't allocate DEV_BSIZE from the stack, use kmem instead.
1.92	09-Dec-2007	jmcneill	branches: 1.92.2; Merge jmcneill-pm branch.
1.91	19-Oct-2007	ad	branches: 1.91.2; 1.91.4; 1.91.6; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.90	09-Jul-2007	ad	branches: 1.90.6; 1.90.8; 1.90.12; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.89	01-Jul-2007	dsl	Remove 'else' after 'return' for clarity
1.88	17-Mar-2007	dyoung	Let config(1) know that #define ATADEBUG goes in opt_ata.h. In dev/ic/wdc.c and in dev/ata/ata.c, #include "opt_ata.h", and make both the files compile with or without ATADEBUG. Do not compile with ATADEBUG by default.
1.87	12-Mar-2007	ad	branches: 1.87.2; 1.87.4; Pass an ipl argument to pool_init/POOL_INIT to be used when initializing the pool's lock.
1.86	04-Mar-2007	christos	branches: 1.86.2; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.85	09-Feb-2007	ad	branches: 1.85.2; Merge newlock2 to head.
1.84	25-Dec-2006	wiz	Spell "schedule" correctly. From Zafer Aydogan.
1.83	16-Nov-2006	christos	__unused removal on arguments; approved by core.
1.82	25-Oct-2006	bouyer	Add sata registers to struct wdc_regs. Add wdc_sataprobe(), a function probing drives using the standard SATA registers; taken from various PCI sata drivers. Export wdc_drvprobe() too.
1.81	15-Oct-2006	itohy	white space police
1.80	14-Oct-2006	itohy	Add ATAPI Pioneer (info from Linux) and ATA TDK (I have one) as exceptions of string byte order. Add string byte order shuffling for big-endian platforms.
1.79	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.78	30-Sep-2006	itohy	A little effort against kernel bloat.... Exclude ATA DMA support if no ATA DMA capable drivers are compiled in.
1.77	24-Sep-2006	jmcneill	Add "name" parameter to powerhook_establish, to aid debugging. No objections on tech-kern@
1.76	07-Sep-2006	itohy	branches: 1.76.2; 1.76.4; Add PIOBM (busmastering transfer using ATA PIO mode) support. The PIOBM is used by only one driver (will be added later, stay tuned) and intruduce an attribute "ata_piobm" so that it will be conditionally compiled in. The "ata_dma" (busmastering transfer using ATA DMA mode) and "ata_udma" (busmastering transfer using ATA Ultra DMA mode) attributes are also added for consistency, but unused for now.
1.75	27-Aug-2006	christos	Add missing initializer
1.74	29-Mar-2006	thorpej	Use device_cfdata().
1.73	22-Jan-2006	bouyer	branches: 1.73.2; 1.73.4; 1.73.6; 1.73.8; 1.73.10; For an IDENTIFY we have to wait for DRQ after issuing the command.
1.72	11-Dec-2005	christos	branches: 1.72.2; merge ktrace-lwp.
1.71	07-Jul-2005	drochner	tighten the autoconf constraints by passing the atapi/ata/ata_hl attribute to config_found() (after some cleanup we might be able to kill SCSIPI_BUSTYPE_*)
1.70	29-May-2005	christos	branches: 1.70.2; - remove casts [to avoid volatile] - sprinkle const - avoid shadowed variables
1.69	20-May-2005	bouyer	Close a window where (disks) interrupts could be enabled while the ATACH_TH_RUN flag is still set. Should fix panic reported by Jukka Salmi on netbsd-users@.
1.68	16-May-2005	bouyer	Add a ata_queue_idle() function, which freeze a queue and tsleep() until the controller is idle. Change the powerhook function to call ata_queue_idle() on standby/suspend, and ata_reset_channel() on resume (to wake up the disks and start from a clean state). Fix PR kern/30194 by Lubomir Sedlacik.
1.67	11-Apr-2005	matt	Fix mode printing. (don't panic on unmatched devices)
1.66	04-Mar-2005	tacha	branches: 1.66.2; Disestablish powerhook while detaching atabus.
1.65	27-Feb-2005	perry	nuke trailing whitespace
1.64	26-Jan-2005	jmcneill	Add powerhook for atabus.
1.63	14-Nov-2004	soren	branches: 1.63.4; 1.63.6; s/using DMA data transfers/using DMA/ to make it fit in 80 columns. DMA only applies to data transfers anyway..
1.62	30-Oct-2004	bouyer	Allow polled commands to be queued, if ata_exec_xfer() is called in thread context. For this introduce 3 new xfer tags: - C_WAIT, equivalent of AT_WAIT - C_WAITACT, when there is a thread waiting on this xfer to become the active one (that is, to be at the head of the queue) - C_FREE, set by ata_free_xfer() when it can't free the xfer because it's still in use. The holder should then free the xfer ASAP. If ata_exec_xfer() is called with (C_POLL \| C_WAIT), and there is already xfers to be processed, assert C_WAITACT and sleep. atastart() checks for C_WAITACT, and wakeup the thread waiting for this xfer to become active if set. atastart() won't process this xfer, it's the responsability of the thread waked up to handle it. Fix (the right way) kern/27421 by Martin Husemann.
1.61	28-Oct-2004	bouyer	Call xfer->c_start() with the channel from the xfer, not the one from the thread. They may be different if we have a shared queue.
1.60	11-Sep-2004	bouyer	If we're not going to handle a reset request because one if already pending, decrease queue_freeze that we just increased (otherwise the queue stay frozen after the reset). Should fix kern/26873 and port-i386/26910.
1.59	21-Aug-2004	thorpej	Eliminate some gratuitous spl frobbing in atabus_thread().
1.58	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.57	20-Aug-2004	thorpej	- atabus_detach(): Manipulate ch_flags at splbio(). - ata_reset_channel(): Assert that we're at splbio() when called. - Sprinkle some comments about functions that must be called at splbio().
1.56	20-Aug-2004	thorpej	WDC_DIAGNOSTIC -> ATA_DEBUG
1.55	20-Aug-2004	thorpej	- Add an (ata_reset_channel)() member to ata_bustype. - Add an ata_reset_channel() function that performs the common parts of resetting an ATA channel, which uses the (ata_reset_channel)() callback to do the heavy lifting. Adjust callers to use ata_reset_channel() instead of wdc_reset_channel(). This removes the last wdc-specific code from ata.c!
1.54	20-Aug-2004	thorpej	Use ANSI function decls for atabus open/close/ioctl.
1.53	20-Aug-2004	thorpej	wdcreg.h is not needed here any more.
1.52	20-Aug-2004	thorpej	Move common parts of attaching an ata_channel to a new ata_channel_attach() function.
1.51	20-Aug-2004	thorpej	Move atabusconfig() to ata.c.
1.50	20-Aug-2004	thorpej	In ata_get_params(), use the (*ata_exec_command)() available from the ata_bustype rather than using wdc_exec_command() directly.
1.49	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.48	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.47	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.46	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.45	13-Aug-2004	thorpej	Move wdcstart() to ata.c and rename it to atastart().
1.44	13-Aug-2004	thorpej	Move wdc_exec_xfer() to ata.c and rename it ata_exec_xfer().
1.43	12-Aug-2004	thorpej	Move wdc_addref() and wdc_delref() to ata.c and rename them to ata_*.
1.42	12-Aug-2004	thorpej	Move wdc_kill_pending() to ata.c and rename it ata_kill_pending().
1.41	12-Aug-2004	thorpej	- Move wdc_xfer_pool, wdc_get_xfer(), wdc_free_xfer() to ata.c, and rename to ata_*. - Use a static initializer for the ata_xfer_pool.
1.40	12-Aug-2004	thorpej	Move wdc_probe_caps() to ata.c and rename it ata_probe_caps().
1.39	12-Aug-2004	thorpej	Move wdc_downgrade_mode() to ata.c and rename it to ata_downgrade_mode().
1.38	12-Aug-2004	thorpej	Move wdc_print_modes() into ata.c and rename it ata_print_modes().
1.37	12-Aug-2004	thorpej	Rename some constants: WDC_COMPLETE -> ATACMD_COMPLETE WDC_QUEUED -> ATACMD_QUEUED WDC_TRY_AGAIN -> ATACMD_TRY_AGAIN
1.36	12-Aug-2004	thorpej	Rename "struct wdc_command" to "struct ata_command".
1.35	10-Aug-2004	mycroft	Hold a reference between atabus_attach() and atabusconfig() to avoid an extra enable/disable cycle. Now my CF card attaches a bit faster. XXX This code is split strangely between files.
1.34	04-Aug-2004	bouyer	Change wdc_kill_pending() to take a struct ata_drive_datas * as argument, and kill only pending requests for this drive. Implement a DRIVE_WAITDRAIN flag, which will cause the active command to be killed once complete. Other minor fixes. Now it's possible to detach a ATA or ATAPI device from ioctl even when a dd on the raw char partition is running.
1.33	04-Aug-2004	bouyer	Add struct ata_xfer *active_xfer to ata_queue. Now the active xfer isn't the head of the queue any more, this makes a few things easier (this will also help for tagged queuing support). Remove the WDCF_ACTIVE flag, test active_xfer != NULL instead. clean up wdc_free_xfer() and kill_xfer(). Clean up wdc_reset_channel(), and make it issue a ATAPI_SOFT_RESET if the active command is ATAPI. In wdc_atapi_get_params(), use AT_WAIT \| AT_POLL for ATAPI_SOFT_RESET, so that we'll use tsleep() instead of delay(). In wdc_atapi_start(), call wdc_dmawait() at the right place.
1.32	03-Aug-2004	bouyer	First cut at implementing ATABUSIODETACH. Still buggy.
1.31	02-Aug-2004	bouyer	Clean up interraction between wdc_reset_channel and the kernel thread. Move kill_xfer() after the reset, and stop the DMA engine if needed (this will unload the DMA maps).
1.30	01-Aug-2004	bouyer	Implement an atabus control device, and define some ATA bus control IOCTLS. Implement ATABUSIORESET, which will reset the given ATA bus.
1.29	27-May-2004	thorpej	Clean-up of use of "precomp". It is used as "features" everywhere, so rename the fields in wdc_command as appropriate.
1.28	13-Apr-2004	bouyer	Make wdc_channel->ch_flags volatile, and cast it to (void *) when used in tsleep()/wakeup() to appease gcc. Otherwise, the ch_flags value may be cached in a register in atabus_thread(), and when it sets the WDCF_TH_RUN bit after tsleep() it may loose loose the changes made by an interrupt handler or another thread. Problem analysed by Jukka Andberg on tech-kern.
1.27	03-Jan-2004	thorpej	branches: 1.27.2; More structure member namespace cleanup: thread -> ch_thread
1.26	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.25	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.24	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.23	30-Dec-2003	thorpej	Move most of the atabus layer into ata.c.
1.22	30-Dec-2003	thorpej	Tweak function names in debug messages.
1.21	14-Dec-2003	thorpej	De-__P and use ANSI-style function decls.
1.20	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.19	23-Sep-2003	mycroft	Fix more probe delay and/or failure problems: 1) Don't wait for DRQ on an IDENTIFY command -- if it's not set when we see BSY clear, abort the command and ignore the drive. (Do this by testing for DRQ in the read/write cases in __wdccommand_intr().) 2) Don't wait for DRQ to deassert when we finish an IDENTIFY (or any other non-block command that reads data) -- we don't do this for block I/O, and empirically it doesn't clear on my CF cards at all, causing a pointless 1s delay. 3) Add comments to some of the delay()s, and add missing ones in wdcreset() and the WDCC_RECAL in the so-called "pre-ATA" probe. 4) Slightly simplify the reset sequence -- we were doing an extra I/O. 5) Modify the register writability test to make sure that registers are not overlapped -- this can happen in some weird cases with a missing device 1. 6) Check the error register value after the reset -- if it's not 01h or 81h, as appropriate (see ATA spec), punt. Tested with a number of ATA-only, ATAPI-only, mixed ATA-ATAPI, CF, and IDE disk configurations. Also remove the SINGLE_DRIVE nonsense again.
1.18	27-Jan-2003	bouyer	branches: 1.18.2; Bump WDCC_IDENTIFY timeout to 3 seconds, as required by some drives. Reported by Karl Janmar <karlj@mdstud.chalmers.se> on tech-kern.
1.17	09-Apr-2002	bouyer	branches: 1.17.4; In wdc_exec_command(), for data commands, read the status register after the data transfer. This is mandatory for data out commands (although none are used for now), and not forbiddend for data in commands. Also record if we did transfers any data. May solve kern/16159 by making the probe more robust in face of fake identify.
1.16	03-Dec-2001	bouyer	Update my copyrigth.
1.15	03-Dec-2001	bouyer	Allow wd to attach to something else than wdc/pciide (like USB<->ATA bridges): - move some functions from ata.c to ata_wdc.c or wdc.c. - add callbacks to struct ata_bustype so that wd.c doesn't call directly functions from the lower level driver.
1.14	13-Nov-2001	lukem	add RCSID
1.13	15-May-2000	bouyer	branches: 1.13.4; 1.13.6; 1.13.8; - sync my copyrigth - Add usefull message for obsolete error code (part of fix for kern/9856).
1.12	25-Jan-2000	enami	Don't put leading space, otherwise diskerr() prints unnecessary space like this: wd0h: aborted command reading fsbn 650256 of 650256-650271 (wd0 bn 839056; cn 888 tn 13 sn 17), retrying
1.11	17-Jan-2000	bouyer	Improve the downgrade logic: - If UDMA 2 is failing try UDMA 1 first, it helps in some cases - downgrade if we get an error in the first 4000 xfers, or if we get 4 errors in 4000 xfers if the first 4000 went without troubles. While I'm there commit a local change I have since some time to get my CD probed: issue a "blanck" IDENTIFY before the one used to detect slave ghosts, with my drive the first IDENTIFY following a controller reset fails with an aborted command ...
1.10	26-Nov-1999	bouyer	More WDCDEBUG_PRINT.
1.9	15-Apr-1999	bouyer	branches: 1.9.2; 1.9.8; Bump timeout to 10s only for ATAPI devices, 1s is fine for IDE drives once they are ready.
1.8	14-Apr-1999	jonathan	Complete bugfix in rev 1.3: increase read-parameter-block timeout to 10s (as used for the hardware for which rev 1.3 was made.)
1.7	10-Mar-1999	bouyer	branches: 1.7.2; Add a new disk flag: DRIVE_OLD for pre-ATA disks. probe routine will now set DRIVE_OLD, DRIVE_ATA or DRIVE_ATAPI based on register signatures. The attach routine will issue a IDENTIFY command for ATA/ATAPI disk, to detect flase matches by the probe routine. probe/attach should now be fully compliant with ata-4/ata-5. As a side effect, ATAPI drives which improperly use ATA register signatures should now be attached as ATAPI.
1.6	21-Feb-1999	hubertf	Only define WDCDEBUG if not already define (e.g. by kernel option)
1.5	08-Feb-1999	bouyer	There's no ATA draft where it is required for the drive to set DRDY \| DSC when the disk is ready to transfer data, and in ATA-5 the DSC has been obsoleted. So only wait for DRQ to transfer data. This can be made conditional on the ATA version if it's proven to break with some drives (worked with all the drives I have access to). While I'm there correct a few typos.
1.4	18-Jan-1999	bouyer	move wd.c:print_wderror() to ata.c:ata_perror(). In wdc_probe_caps() add code to guess the ATA revision supported (if ATA4 if Ultra-DMA, ATA2 if PIO mode > 2). We can't rely on param.atap_ata_major here, at last one Ultra-DMA drive claims to support only ATA-3. Use the ATA version in ata_perror(), and to try a flush cache command in a shutdown hook for IDE drives.
1.3	16-Nov-1998	bouyer	For ATAPI IDENTiFY, poll only for DRQ, some devices don't assert DRDY at this point.
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file ata.c was initially added on branch bouyer-ide.
1.1.2.4	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.1.2.3	13-Aug-1998	bouyer	- sync with HEAD - better error detection and report for wd drives - better wddump() support - wdcwait() now takes a timout argument, so that we can honnor the timeout argument of scsipi cmds.
1.1.2.2	05-Jun-1998	bouyer	Patches from Leo Weppelman for IDE on atari, with minor fixes by me (needed the addition of a void wdcrestart __P((void*)) function). Quoting his words: These diffs are preliminary because it looks like the free_hw() function does not get called. This is only a problem on the atari falcon. I have not yet been able to track this down. Also, KNF wdc.c.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.7.2.3	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.7.2.2	23-Jan-2000	he	Pull up revision 1.11 (requested by bouyer): Improve the downgrade logic to work better in some cases (especially for Acer Labs M5229 controllers with Ultra-DMA/66 disks).
1.7.2.1	20-Apr-1999	cjs	branches: 1.7.2.1.2; Pull up to HEAD 1.9, per Manuel Bouyer <bouyer@antioche.eu.org>
1.7.2.1.2.1	21-Jun-1999	thorpej	Sync w/ -current.
1.9.8.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.9.8.1	21-Dec-1999	wrstuden	Initial commit of recent changes to make DEV_BSIZE go away. Runs on i386, needs work on other arch's. Main kernel routines should be fine, but a number of the stand programs need help. cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512 byte block devices. vnd, raidframe, and lfs need work. Non 2**n block support is automatic for LKM's and conditional for kernels on "options NON_PO2_BLOCKS".
1.9.2.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.13.8.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.13.8.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.13.6.3	17-Apr-2002	nathanw	Catch up to -current.
1.13.6.2	08-Jan-2002	nathanw	Catch up to -current.
1.13.6.1	14-Nov-2001	nathanw	Catch up to -current.
1.13.4.1	23-Jan-2003	msaitoh	Pull up revision 1.17 (requested by Manuel Bouyer): read status after data commands in wdc_exec_command(). Mandatory for data out commands (nonw are used for now), and doesn't hurt for data in. Also keep track if we did transfer data. Makes the probe more robust and solves kern/16159
1.17.4.1	30-Jun-2003	grant	Pull up revision 1.18 (requested by bouyer in ticket #1353): Bump WDCC_IDENTIFY timeout to 3 seconds, as required by some drives. Reported by Karl Janmar <karlj@mdstud.chalmers.se> on tech-kern.
1.18.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.18.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.18.2.8	04-Feb-2005	skrll	Sync with HEAD.
1.18.2.7	29-Nov-2004	skrll	Sync with HEAD.
1.18.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.18.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.18.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.18.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.18.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.18.2.1	03-Aug-2004	skrll	Sync with HEAD
1.27.2.2	24-May-2005	riz	Pull up revision 1.69 via patch(requested by bouyer in ticket #1543): Close a window where (disks) interrupts could be enabled while the ATACH_TH_RUN flag is still set. Should fix panic reported by Jukka Salmi on netbsd-users@.
1.27.2.1	18-Apr-2004	jmc	branches: 1.27.2.1.2; Pullup rev 1.28 (requested by bouyer in ticket #149) Make wdc_channel->ch_flags volatile, and cast it to (void *) when used in tsleep()/wakeup() to appease gcc. Otherwise, the ch_flags value may be cached in a register in atabus_thread(), and when it sets the WDCF_TH_RUN bit after tsleep() it may loose loose the changes made by an interrupt handler or another thread.
1.27.2.1.2.1	24-May-2005	riz	Pull up revision 1.69 via patch(requested by bouyer in ticket #1543): Close a window where (disks) interrupts could be enabled while the ATACH_TH_RUN flag is still set. Should fix panic reported by Jukka Salmi on netbsd-users@.
1.63.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.63.6.1	12-Feb-2005	yamt	sync with head.
1.63.4.1	29-Apr-2005	kent	sync with -current
1.66.2.2	28-May-2005	tron	Pull up revision 1.69 (requested by bouyer in ticket #340): Close a window where (disks) interrupts could be enabled while the ATACH_TH_RUN flag is still set. Should fix panic reported by Jukka Salmi on netbsd-users@.
1.66.2.1	27-May-2005	riz	Pull up revision 1.68 (requested by bouyer in ticket #328): Add a ata_queue_idle() function, which freeze a queue and tsleep() until the controller is idle. Change the powerhook function to call ata_queue_idle() on standby/suspend, and ata_reset_channel() on resume (to wake up the disks and start from a clean state). Fix PR kern/30194 by Lubomir Sedlacik.
1.70.2.8	24-Mar-2008	yamt	sync with head.
1.70.2.7	17-Mar-2008	yamt	sync with head.
1.70.2.6	21-Jan-2008	yamt	sync with head
1.70.2.5	27-Oct-2007	yamt	sync with head.
1.70.2.4	03-Sep-2007	yamt	sync with head.
1.70.2.3	26-Feb-2007	yamt	sync with head.
1.70.2.2	30-Dec-2006	yamt	sync with head.
1.70.2.1	21-Jun-2006	yamt	sync with head.
1.72.2.1	01-Feb-2006	yamt	sync with head.
1.73.10.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.73.8.1	19-Apr-2006	elad	sync with head.
1.73.6.3	14-Sep-2006	yamt	sync with head.
1.73.6.2	03-Sep-2006	yamt	sync with head.
1.73.6.1	01-Apr-2006	yamt	sync with head.
1.73.4.1	22-Apr-2006	simonb	Sync with head.
1.73.2.1	09-Sep-2006	rpaulo	sync with head
1.76.4.2	10-Dec-2006	yamt	sync with head.
1.76.4.1	22-Oct-2006	yamt	sync with head
1.76.2.3	06-Feb-2007	ad	Quieten noisy boot messages.
1.76.2.2	12-Jan-2007	ad	Sync with head.
1.76.2.1	18-Nov-2006	ad	Sync with head.
1.85.2.2	24-Mar-2007	yamt	sync with head.
1.85.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.86.2.8	23-Oct-2007	ad	Sync with head.
1.86.2.7	15-Jul-2007	ad	Sync with head.
1.86.2.6	01-Jul-2007	ad	Adapt to callout API change.
1.86.2.5	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.86.2.4	10-Apr-2007	ad	Sync with head.
1.86.2.3	10-Apr-2007	ad	Nuke the deferred kthread creation stuff, as it's no longer needed. Pointed out by thorpej@.
1.86.2.2	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.86.2.1	13-Mar-2007	ad	Sync with head.
1.87.4.1	18-Mar-2007	reinoud	First attempt to bring branch in sync with HEAD
1.87.2.1	11-Jul-2007	mjf	Sync with head.
1.90.12.1	25-Oct-2007	bouyer	Sync with HEAD.
1.90.8.3	23-Mar-2008	matt	sync with HEAD
1.90.8.2	09-Jan-2008	matt	sync with HEAD
1.90.8.1	06-Nov-2007	matt	sync with HEAD
1.90.6.4	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.90.6.3	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.90.6.2	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.90.6.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.91.6.1	11-Dec-2007	yamt	sync with head.
1.91.4.1	26-Dec-2007	ad	Sync with head.
1.91.2.1	18-Feb-2008	mjf	Sync with HEAD.
1.92.2.3	10-Jan-2008	bouyer	Sync with HEAD
1.92.2.2	08-Jan-2008	bouyer	Sync with HEAD
1.92.2.1	02-Jan-2008	bouyer	Sync with HEAD
1.96.6.6	17-Jan-2009	mjf	Sync with HEAD.
1.96.6.5	05-Oct-2008	mjf	Sync with HEAD.
1.96.6.4	29-Jun-2008	mjf	Sync with HEAD.
1.96.6.3	06-Apr-2008	mjf	- after some discussion with agc@ i agreed it would be a good idea to move device_unregister_* to device_deregister_* to be more like the pmf(9) functions, especially since a lot of the time the function calls are next to each other. - add device_register_name() support for dk(4).
1.96.6.2	05-Apr-2008	mjf	- add "file-system DEVFS" and "pseudo-device devfsctl" to conf/std seeing as these are always needed. - convert many, many drivers over to the New Devfs World Order. For a list of device drivers yet to be converted see, http://www.netbsd.org/~mjf/devfs-todo.html. - add a new device_unregister_all(device_t) function to remove all device names associated with a device_t, which saves us having to construct device names when the driver is detached. - add a DEV_AUDIO type for devices.
1.96.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.96.2.1	24-Mar-2008	keiichi	sync with head.
1.98.8.1	18-Jun-2008	simonb	Sync with head.
1.98.6.2	10-Oct-2008	skrll	Sync with HEAD.
1.98.6.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.98.4.5	11-Aug-2010	yamt	sync with head.
1.98.4.4	11-Mar-2010	yamt	sync with head
1.98.4.3	16-Sep-2009	yamt	sync with head
1.98.4.2	16-May-2009	yamt	sync with head
1.98.4.1	04-May-2009	yamt	sync with head.
1.98.2.1	17-Jun-2008	yamt	sync with head.
1.99.2.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.99.2.1	19-Oct-2008	haad	Sync with HEAD.
1.101.4.1	20-Nov-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #67): sys/dev/ic/wdc.c: revision 1.256 sys/dev/ata/ata.c: revision 1.102 sys/dev/scsipi/atapi_wdc.c: revision 1.109 sys/dev/ata/atavar.h: revision 1.78 sys/dev/ata/ata_wdc.c: revision 1.91 cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.101.2.2	28-Apr-2009	skrll	Sync with HEAD.
1.101.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.102.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.111.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.112.2.3	31-May-2011	rmind	sync with head
1.112.2.2	21-Apr-2011	rmind	sync with head
1.112.2.1	30-May-2010	rmind	sync with head
1.113.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.115.8.1	29-Apr-2012	mrg	sync to latest -current.
1.115.4.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.115.4.3	16-Jan-2013	yamt	sync with (a bit old) head
1.115.4.2	30-Oct-2012	yamt	sync with head
1.115.4.1	17-Apr-2012	yamt	sync with head
1.125.2.4	03-Dec-2017	jdolecek	update from HEAD
1.125.2.3	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.125.2.2	23-Jun-2013	tls	resync from head
1.125.2.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.127.4.2	18-May-2014	rmind	sync with head
1.127.4.1	28-Aug-2013	rmind	sync with head
1.130.2.1	10-Aug-2014	tls	Rebase.
1.132.8.39	29-Sep-2017	jdolecek	introduce ATA_BSIZE and use it instead of DEV_BSIZE for get params and recovery, where they are by spec 512 bytes and not negotiable
1.132.8.38	27-Sep-2017	jdolecek	restore the atac_claim_hw and atac_free_hw hooks, they are used on atari
1.132.8.37	25-Sep-2017	jdolecek	put recovery command on head of active xfers queue, so that drivers using ata_queue_get_active_xfer() like mvsata(4) will find it and not random other NCQ xfer never return NCQ xfer from ata_queue_get_active_xfer(), if the first xfer is NCQ simply return NULL adjust ata_timo_xfer_check() to not check the expiring flag in C_WAITTIMO branch, if we get there we ought to get the timeout handler aborted
1.132.8.36	23-Sep-2017	jdolecek	note to previous: the KASSERT() insisted actually there is at most one active xfer, which is false during NCQ error recovery
1.132.8.35	23-Sep-2017	jdolecek	remove ata_queue_get_active_xfer() KASSERT() about having at least one active xfer; it can happen we get interrupt while no longer having the cmd active e.g. during recovery, and all callers handle getting NULL as result fixes panic in mvsata(4) during error recovery
1.132.8.34	20-Sep-2017	jdolecek	add forgotten destroy of queue_idle cv in ata_queue_free()
1.132.8.33	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.132.8.32	11-Sep-2017	jdolecek	move debug printf where it belongs
1.132.8.31	10-Sep-2017	jdolecek	refactor code so that xfer c_start() hook is called with channel mutex held, and hence the controller submit code no longer relies on spl tested all the affected drivers - wdc (via piixide), ahci, mvsata, siisata, both disk and atapi I/O
1.132.8.30	10-Sep-2017	jdolecek	fix regression in atabus_thread() when it was converted from spl to mutex - the reset and c_start() routines expect to run on splbio; wrap the calls insite splbio/splx() again for now, since we can't hold the mutex while calling them fixes problem experienced by Jonathan, where drive setup triggered an spurious interrupt and panic due to state < READY
1.132.8.29	15-Aug-2017	jdolecek	explicitly do not try to activate any further commands when running recovery xfer; it was kind of implied since the code would not queue another non-NCQ command when non-NCQ command is active, but this is better for readibility
1.132.8.28	12-Aug-2017	jdolecek	use AT_LBA48 flag for the READ LOG EXT - it's required so that e.g. mvsata() executes the command using wdccommandext(), it fails when executed using wdccommand()
1.132.8.27	12-Aug-2017	jdolecek	take atabus_qlock while inserting into atabus_initq_head to avoid race in attach and rescan; just cleanup, noticed this while doing the thread locking, don't think there is realistic way to trigger this
1.132.8.26	12-Aug-2017	jdolecek	convert the atabus thread to use the channel lock and a condvar, adjust code which sets the relevant channel flags to take the lock while doing so
1.132.8.25	12-Aug-2017	jdolecek	remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status, so that there is less hidden state shared by commands; primary intent is to make the NCQ and non-NCQ paths more similar, and remove possibility of incorrect handling for the NCQ commands tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4), ahcisata(4), mvsata(4) on real hw
1.132.8.24	01-Aug-2017	jdolecek	make atastart() schedule as many commands as possible, instead of always only one; makes it able to pick up pace again after processing non-NCQ command
1.132.8.23	01-Aug-2017	jdolecek	move the drive recovery block to drive struct, it's inherently per-drive
1.132.8.22	29-Jul-2017	jdolecek	reserve the highest slot for error recovery, and also have ata_channel include space for the READ LOG EXT sector, so that it's not necessary to allocate memory on the error handling path; now ata_read_log_ext_ncq() will never fail due to resource shortage
1.132.8.21	22-Jul-2017	jdolecek	provide channel flag when executing NCQ commands, so that e.g. intr handler can use this for handling decisions without checking xfer
1.132.8.20	21-Jul-2017	jdolecek	add checksum verification for data returned by READ LOG EXT; this is mostly just paranoia for eventual driver/hw DMA bugs this doesn't make difference for QEMU, as there the command actually always just fails (it's not implemented)
1.132.8.19	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.132.8.18	27-Jun-2017	jdolecek	attend error paths, more strict asserts and code consistency - atastart() and ata_kill_pending() now KASSERT() that all xfers on queue have same channel - inactive xfers are killed via new reason KILL_GONE_INACTIVE, controller code must not call any resource deactivation in that case - c_intr() must call ata_waitdrain_xfer_check() as first thing, and must not further touch any xfer structures on exit path; any resource cleanup is supposed to be done in c_kill_xfer() - c_kill_xfer() should never call atastart() - ata_waitdrain_check() removed, replaced by ata_waitdrain_xfer_check() - ATA_DRIVE_WAITDRAIN handling converted to use condvar - removed unused ata_c callback
1.132.8.17	24-Jun-2017	jdolecek	change ata_delay() to tsleep for 1 hz rather than indefinitely if provided ms is lower than 1 hz
1.132.8.16	23-Jun-2017	jdolecek	restart I/O processing after freeing xfer, i.e. now even after commands like cache flush or standby; the command handling no longer use on-stack xfer, hence use queue slot and compete with normal I/O for the xfers the restart give change to all drives attached to the same channel in round-robin fashion, for fair usage and to recover from situation when disk is idle due to all xfers being consumed by other drives make special concession for flush cache - ignore any new I/O requests on the particular disk while the flush cache is waiting for xfer, so that I/O queue won't starve the flush cache and the flush cache would be done ASAP tested on piixide(4), ahci(4), siisata(4)
1.132.8.15	21-Jun-2017	jdolecek	hold channel lock for ata_exec_xfer() and most of atastart(), convert C_WAITACT handling to use condvar add comment for the code block with C_FREE and ata_free_xfer() explaining why it's not abstraction layer violation
1.132.8.14	21-Jun-2017	jdolecek	two last forgotten on-stack xfers replaced with using the queue xfer - ata_get_params() and ata_set_mode() fix wdc_atapi_get_params() to free the xfer used for soft reset before calling ata_get_params() - it's now necessary, as ata_get_params() doesn't invent a private xfer own any more
1.132.8.13	21-Jun-2017	jdolecek	change ata_queue_hwslot_to_xfer() and ata_queue_get_active_xfer() to take ata_channel instead of ata_queue as parameter, and lock the channel while traversing the queue
1.132.8.12	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.132.8.11	19-Jun-2017	jdolecek	add ata_channel lock, use it to protect queue manipulation (only that for now); add ata_channel_detach() to destroy the locks change ata_get_xfer() so that it can wait for xfer, convert all on-stack xfer code to use the blocking variant fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(), convert drive probe code also over from slot0 XXX to ata_get_xfer() drive reset and PMP now works on siisata(4) too; changes tested also on piixide(4), ahci(4), mvsata(4)
1.132.8.10	17-Jun-2017	jdolecek	make PMP working great again tested with mvsata(4), my ahcisata(4) controller unfortunately doesn't support PMP
1.132.8.9	16-Jun-2017	jdolecek	adjust reset channel and dump paths - channel reset now always kills active transfer, even on dump path, but now doesn't touch the queued waiting transfers; also kill_xfer hook is always called, so that HBA can free any private xfer resources and thus the dump request has chance to work - kill_xfer routines now always call ata_deactivate_xfer(); added KASSERT()s to ata_free_xfer() to expect deactivated xfer - when called during channel reset before dump, ata_kill_active() drops any queued waiting transfers without processing - do not (re)queue any transfers in wddone() when dumping - kill AT_RST_NOCMD flag This should also hopefully fix the 'polled command has been queued' panic as reported in: PR kern/11811 by John Hawkinson PR kern/47041 by Taylor R Campbell PR kern/51979 by Martin Husemann dump tested working with piixide(4) and ahci(4). mvsata(4) dump times out, but otherwise tested working, will be fixed separately. siisata(4) mechanically changed and not tested.
1.132.8.8	24-Apr-2017	jdolecek	set NCQ priority field to 'high' for BPRIO_TIMECRITICAL transfers if drive supports it
1.132.8.7	23-Apr-2017	jakllsch	Use WDC_QUEUE_DEPTH_MASK to correctly mask the NCQ depth IDENTIFY data.
1.132.8.6	19-Apr-2017	jdolecek	add ATA FUA support
1.132.8.5	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.132.8.4	15-Apr-2017	jdolecek	pass also ata_command via ata_xfer, callers of ata_exec_command() is now responsible for allocation/disposal of the structure change code to allocate ata_xfer for commands on stack same way as previously the ata_command were, using c_slot 0; adjust asserts so that it would allow several xfers with same c_slot, as long as only one such transfer is active at a time
1.132.8.3	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.132.8.2	11-Apr-2017	jdolecek	hide active_xfer from drivers
1.132.8.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.141.6.18	15-Oct-2018	jdolecek	change the SATA/NCQ recovery to run in the atabus thread
1.141.6.17	14-Oct-2018	jdolecek	add comment to kthread_create() call to explain why atabus_thread is actually running under KERNEL_LOCK()
1.141.6.16	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.141.6.15	06-Oct-2018	jdolecek	actually, just make dump use the same queue skip as recovery, and remove the no longer necessary ata_queue_reset() call from wd(4) also for PR kern/47041
1.141.6.14	06-Oct-2018	jdolecek	remove AT_RST_EMERG, do the queue reset explicitly in wd(4) this should explicitly fix PR kern/47041 with sync during heavy disk activity, even thought it was actually already implicitly fixed by calling ata_thread_run() for drive reset in previous commit already, since the function already called ata_queue_reset() drop now unused ch_reset_flags and drive_reset_flags
1.141.6.13	06-Oct-2018	jdolecek	remove extra newline
1.141.6.12	06-Oct-2018	jdolecek	fix ata_thread_run() for drive reset to set and check correct drive_flags
1.141.6.11	03-Oct-2018	jdolecek	change channel reset and drive reset for all ATA controllers to always run via thread, and with channel lock held the whole time; the queue is frozen while reset is pending for this repurpose ata_reset_channel() into new ata_thread_run() also adjust some device printfs to not leak xfer pointer, and avoid aprint_* for non-autoconf messages
1.141.6.10	24-Sep-2018	jdolecek	replace KASSERT() for READ LOG EXT returning without error indication with just log to avoid triggering panic under QEMU - it doesn't implement the command for AHCI emulation
1.141.6.9	22-Sep-2018	jdolecek	fix use-after-free in wd(4) dump, detected by switch to the pool change code in wd_dumpblocks() to use it's own non-pool ata_xfer, which skips the deallocation step and thus keeps the contents when the I/O is finished
1.141.6.8	22-Sep-2018	jdolecek	revert atabus_alloc_drives() to use KM_NOSLEEP again - it's called with ch_lock held hence cannot actually sleep, fixing is out of scope of the branch
1.141.6.7	22-Sep-2018	jdolecek	use KM_SLEEP in atabus_alloc_drives() to avoid spurious failures; this is not called from interrupt context
1.141.6.6	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.141.6.5	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.141.6.4	17-Sep-2018	jdolecek	switch from TAILQ to SIMPLEQ for ata_xfer pending queue to save space, don't need doubly linked queue
1.141.6.3	17-Sep-2018	jdolecek	move low-level protocol handlers hooks from ata_xfer to separate struct, initialized statically primarily to reduce ata_xfer struct size, but also improves readibility, and enforces consistency
1.141.6.2	01-Sep-2018	jdolecek	stop assuming c_slot is immutable, use the xfer pointer rather than c_slot in printfs()
1.141.6.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.141.4.4	21-Apr-2020	martin	Sync with HEAD
1.141.4.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.141.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.141.4.1	10-Jun-2019	christos	Sync with HEAD
1.141.2.2	26-Dec-2018	pgoyette	Sync with HEAD, resolve a few conflicts
1.141.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.149.2.3	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.149.2.2	30-Aug-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1512): sys/dev/ata/ata.c: revision 1.163 Output missing '\n' for capability list when DMA support is not compiled in.
1.149.2.1	15-Mar-2022	martin	Pull up following revision(s) (requested by perseant in ticket #1435): sys/dev/ata/ata.c: revision 1.167 Avoid an unaccounted extra channel freeze, if a reset is requested more than once before the thread services the request. Closes PR#56745.
1.154.2.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.154.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.160.4.1	02-Apr-2021	thorpej	config_found_ia() -> config_found() w/ CFARG_IATTR.
1.161.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.169.10.1	02-Aug-2025	perseant	Sync with HEAD
1.45	07-Aug-2021	thorpej	Merge thorpej-cfargs2.
1.44	24-Apr-2021	thorpej	branches: 1.44.8; Merge thorpej-cfargs branch: Simplify and make extensible the config_search() / config_found() / config_attach() interfaces: rather than having different variants for which arguments you want pass along, just have a single call that takes a variadic list of tag-value arguments. Adjust all call sites: - Simplify wherever possible; don't pass along arguments that aren't actually needed. - Don't be explicit about what interface attribute is attaching if the device only has one. (More simplification.) - Add a config_probe() function to be used in indirect configuiration situations, making is visibly easier to see when indirect config is in play, and allowing for future change in semantics. (As of now, this is just a wrapper around config_match(), but that is an implementation detail.) Remove unnecessary or redundant interface attributes where they're not needed. There are currently 5 "cfargs" defined: - CFARG_SUBMATCH (submatch function for direct config) - CFARG_SEARCH (search function for indirect config) - CFARG_IATTR (interface attribte) - CFARG_LOCATORS (locators array) - CFARG_DEVHANDLE (devhandle_t - wraps OFW, ACPI, etc. handles) ...and a sentinel value CFARG_EOL. Add some extra sanity checking to ensure that interface attributes aren't ambiguous. Use CFARG_DEVHANDLE in MI FDT, OFW, and ACPI code, and macppc and shark ports to associate those device handles with device_t instance. This will trickle trough to more places over time (need back-end for pre-OFW Sun OBP; any others?).
1.43	25-Aug-2020	skrll	branches: 1.43.4; KNF
1.42	26-Apr-2019	pgoyette	Some more empty-string --> NULL conversions for module dependencies
1.41	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.40	22-Jun-2018	pgoyette	branches: 1.40.2; 1.40.4; Call config_cfattach_attach() regardless of whether the module is being built as part of a kernel or as a external module. Addresses kern/53389 - thanks for the report! XXX pullup-8
1.39	27-Sep-2016	pgoyette	branches: 1.39.6; 1.39.8; 1.39.14; Remove some debug printf's that slipped through. Thanks, wiz!
1.38	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.37	14-Jul-2016	msaitoh	branches: 1.37.2; - Use aprint*() more in xxx_attach(). - Add missing aprint_naive("\n"). - Remove extra spaces and tabs. - KNF.
1.36	20-Aug-2015	christos	include "ioconf.h" to get the 'void <driver>attach(int count);' prototype.
1.35	27-Oct-2012	chs	branches: 1.35.14; split device_t/softc for all remaining drivers. replace "struct device *" with "device_t". use device_xname(), device_unit(), etc.
1.34	06-Jul-2010	bsh	branches: 1.34.8; 1.34.18; fix some probles with ataraid(4) and Intel MatrixRaid. * handle volumes largetr than 2TiB. * support multiple arrays for Intel Matrix RAID
1.33	12-May-2009	cegger	branches: 1.33.2; 1.33.4; KNF
1.32	11-Sep-2008	tron	branches: 1.32.8; Add support for Intel MatrixRAID to ataraid(4). It is currently limited to one volume. The driver was provided by Juan RP in PR kern/39511.
1.31	05-Sep-2008	tron	Add JMicron RAID support to ataraid(4). Patch provided by Juan RP in PR kern/39395.
1.30	20-Aug-2008	tacha	Add support for nVidia MediaShield ata raid found on HP ProLiant ML115. Tested on HEAD and netbsd-4 branch with RAID0, RAID1, SPAN with MCP55 SATA controller.
1.29	17-Jun-2008	reinoud	branches: 1.29.2; Mark a buffer `busy` in getnewbuf() when it came from the pool_cache since its not on a free list. Also change buf_init() to not automatically mark buffers `busy' since this only makes sense for bufcache buffers. Mark all buf_init'd buffers 'busy' on the places where they ought to be flagged as such to not confuse the buffer cache. Fixes PR 38923.
1.28	06-May-2008	yamt	branches: 1.28.2; 1.28.4; ata_raid_config_block_rw: getiobuf(vp, NULL) -> getiobuf(vp, false)
1.27	05-Apr-2008	cegger	branches: 1.27.2; 1.27.4; use aprint_*_dev and device_xname
1.26	26-Mar-2008	matt	split softc/device_t. use device_t & cfdata_t
1.25	19-Mar-2008	cube	Don't assume wd_softc embeds a struct device.
1.24	02-Jan-2008	ad	branches: 1.24.2; 1.24.6; Merge vmlocking2 to head.
1.23	09-Jul-2007	ad	branches: 1.23.8; 1.23.14; 1.23.16; 1.23.20; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.22	27-Mar-2007	garbled	Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.21	09-Mar-2007	drochner	branches: 1.21.2; 1.21.4; 1.21.6; after the recent config_attach_pseudo() changes, unit numbers have the same sementics as for real devices, so DVUNIT_ANY=-1 is illegal
1.20	16-Nov-2006	christos	branches: 1.20.2; 1.20.4; __unused removal on arguments; approved by core.
1.19	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.18	04-Jan-2006	yamt	branches: 1.18.18; 1.18.20; - add simple functions to allocate/free a buffer for i/o. - make bufpool static.
1.17	11-Dec-2005	christos	branches: 1.17.2; merge ktrace-lwp.
1.16	25-Aug-2005	drochner	kill a number of autoconf submatch functions which follow the standard scheme: if (<configured> != <wildcard> && <configured> != <real>) then fail else ask device match function This is handled by config_stdsubmatch() now.
1.15	25-Aug-2005	drochner	replace the "locdesc_t" structure carrying the number of locators explicitely by a plain integer array the length in now known to all relevant parties, so this avoids duplication of information, and we can allocate that thing in drivers without hacks
1.14	18-Jul-2005	briggs	Update array of software RAID names in ata_raid.c. Change a boundary check to ensure that we won't accidentally read and use uninitialized memory if ATA_RAID_TYPE_MAX is updated without updating the array. Update comment near ATA_RAID_TYPE_MAX to note that the array in ata_raid.c should be updated if a new ATA_RAID_TYPE_* is added.
1.13	20-Jun-2005	briggs	branches: 1.13.2; Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.12	29-May-2005	christos	- remove casts [to avoid volatile] - sprinkle const - avoid shadowed variables
1.11	28-Oct-2004	yamt	branches: 1.11.10; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.10	15-Oct-2004	thorpej	Change config_attach_pseudo() to take a cfdata * that contains the necessary information to create the pseudo-device instance. Pseudo-device device's will reference this cfdata, just as normal devices reference their corresponding cfdata. Welcome to 2.99.10.
1.9	13-Sep-2004	drochner	a round of autoconf cleanup: -convert submatch() style functions (passed to config_search() or config_found_sm()) to the locator passing variants -pass interface attributes in some cases -make submatch() functions look uniformly as far as possible -avoid macros which just hide cfdata members, and reduce dependencies on "locators.h"
1.8	25-Jan-2004	hannken	Make VOP_STRATEGY(bp) a real VOP as discussed on tech-kern. VOP_STRATEGY(bp) is replaced by one of two new functions: - VOP_STRATEGY(vp, bp) Call the strategy routine of vp for bp. - DEV_STRATEGY(bp) Call the d_strategy routine of bp->b_dev for bp. DEV_STRATEGY(bp) is used only for block-to-block device situations.
1.7	14-Dec-2003	thorpej	Garbage-collect __ATA_DISK_PRIVATE; it is no longer needed.
1.6	29-Sep-2003	grant	typo in comment
1.5	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.4	25-Feb-2003	thorpej	branches: 1.4.2; Add missing splbio() protecting of bufpool access. Add a clarifying comment to <sys/buf.h> reminding everyone of the need for splbio().
1.3	25-Feb-2003	thorpej	Add a new BUF_INIT() macro which initializes b_dep and b_interlock, and use it. This fixes a few places where either b_dep or b_interlock were not properly initialized.
1.2	05-Feb-2003	pk	Make the buffer cache code MP-safe.
1.1	27-Jan-2003	thorpej	Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.4.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.4.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.4.2.4	19-Oct-2004	skrll	Sync with HEAD
1.4.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.4.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.4.2.1	03-Aug-2004	skrll	Sync with HEAD
1.11.10.2	21-Jul-2005	tron	Pull up revision 1.14 (requested by briggs in ticket #604): Update array of software RAID names in ata_raid.c. Change a boundary check to ensure that we won't accidentally read and use uninitialized memory if ATA_RAID_TYPE_MAX is updated without updating the array. Update comment near ATA_RAID_TYPE_MAX to note that the array in ata_raid.c should be updated if a new ATA_RAID_TYPE_* is added.
1.11.10.1	03-Jul-2005	tron	Pull up revision 1.13 (requested by briggs in ticket #521): Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.13.2.5	24-Mar-2008	yamt	sync with head.
1.13.2.4	21-Jan-2008	yamt	sync with head
1.13.2.3	03-Sep-2007	yamt	sync with head.
1.13.2.2	30-Dec-2006	yamt	sync with head.
1.13.2.1	21-Jun-2006	yamt	sync with head.
1.17.2.1	15-Jan-2006	yamt	sync with head.
1.18.20.2	10-Dec-2006	yamt	sync with head.
1.18.20.1	22-Oct-2006	yamt	sync with head
1.18.18.1	18-Nov-2006	ad	Sync with head.
1.20.4.2	15-Apr-2007	yamt	sync with head.
1.20.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.20.2.1	31-Mar-2007	bouyer	Pull up following revision(s) (requested by garbled in ticket #542): sys/dev/ata/ata_raidreg.h: revision 1.4 via patch sys/dev/ata/ata_raidvar.h: revision 1.5via patch sys/dev/ata/files.ata: revision 1.15via patch sys/dev/ata/ata_raid_via.c: revision 1.1 sys/dev/ata/ata_raid.c: revision 1.22via patch Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.21.6.1	29-Mar-2007	reinoud	Pullup to -current
1.21.4.1	11-Jul-2007	mjf	Sync with head.
1.21.2.3	24-Aug-2007	ad	Sync with buffer cache locking changes. See buf.h/vfs_bio.c for details. Some minor portions are incomplete and needs to be verified as a whole.
1.21.2.2	10-Apr-2007	ad	Sync with head.
1.21.2.1	05-Apr-2007	ad	Compile fixes.
1.23.20.1	02-Jan-2008	bouyer	Sync with HEAD
1.23.16.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.23.14.1	18-Feb-2008	mjf	Sync with HEAD.
1.23.8.2	23-Mar-2008	matt	sync with HEAD
1.23.8.1	09-Jan-2008	matt	sync with HEAD
1.24.6.4	28-Sep-2008	mjf	Sync with HEAD.
1.24.6.3	29-Jun-2008	mjf	Sync with HEAD.
1.24.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.24.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.24.2.1	24-Mar-2008	keiichi	sync with head.
1.27.4.4	11-Aug-2010	yamt	sync with head.
1.27.4.3	16-May-2009	yamt	sync with head
1.27.4.2	04-May-2009	yamt	sync with head.
1.27.4.1	16-May-2008	yamt	sync with head.
1.27.2.1	18-May-2008	yamt	sync with head.
1.28.4.1	18-Jun-2008	simonb	Sync with head.
1.28.2.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.28.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.28.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.29.2.1	19-Oct-2008	haad	Sync with HEAD.
1.32.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.33.4.1	05-Mar-2011	rmind	sync with head
1.33.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.34.18.2	03-Dec-2017	jdolecek	update from HEAD
1.34.18.1	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.34.8.1	30-Oct-2012	yamt	sync with head
1.35.14.2	05-Oct-2016	skrll	Sync with HEAD
1.35.14.1	22-Sep-2015	skrll	Sync with HEAD
1.37.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.39.14.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.39.14.1	25-Jun-2018	pgoyette	Sync with HEAD
1.39.8.1	23-Jun-2018	martin	Pull up following revision(s) (requested by pgoyette in ticket #894): sys/dev/ata/ata_raid.c: revision 1.40 Call config_cfattach_attach() regardless of whether the module is being built as part of a kernel or as a external module. Addresses kern/53389 - thanks for the report! XXX pullup-8
1.39.6.3	17-May-2017	pgoyette	It's "device_t x" not "struct device_t x"
1.39.6.2	17-May-2017	pgoyette	Typo
1.39.6.1	17-May-2017	pgoyette	At suggestion of chuq@, modify config_attach_pseudo() to return with a reference held on the device. Adapt callers to expect the reference to exist, and to ensure that the reference is released.
1.40.4.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.40.2.1	10-Jun-2019	christos	Sync with HEAD
1.43.4.3	28-Mar-2021	thorpej	Correct names of arguments passed to rescan function.
1.43.4.2	22-Mar-2021	thorpej	Audit CFARG_IATTR in config_found() calls, and remove it in situations where the interface attribute is not ambiguous.
1.43.4.1	22-Mar-2021	thorpej	Mechanical conversion of config_found_sm_loc() -> config_found(). CFARG_IATTR usage needs to be audited.
1.44.8.1	04-Aug-2021	thorpej	Adapt to CFARGS().
1.13	01-Aug-2023	mrg	don't test arrays against NULL. found by GCC 12.
1.12	19-Mar-2022	hannken	Lock vnode across VOP_OPEN.
1.11	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.10	01-Nov-2017	mlelstv	branches: 1.10.2; 1.10.4; 1.10.6; refactor wd and ataraid drivers to use common disk subroutines.
1.9	15-Sep-2008	tron	branches: 1.9.38; Add support for status reports via bio(4) to ataraid(4). The code was contributed by Juan RP in PR kern/39514.
1.8	18-Mar-2008	cube	branches: 1.8.4; 1.8.6; 1.8.10; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.7	02-Feb-2008	mjf	branches: 1.7.2; 1.7.6; Match up some VOP_OPEN()S with VOP_CLOSE()s. This allows the root file system to be identified properly again.
1.6	26-Nov-2007	pooka	Remove the "struct lwp *" argument from all VFS and VOP interfaces. The general trend is to remove it from all kernel interfaces and this is a start. In case the calling lwp is desired, curlwp should be used. quick consensus on tech-kern
1.5	14-May-2006	elad	branches: 1.5.32; 1.5.34; 1.5.40; integrate kauth.
1.4	28-Mar-2006	thorpej	Use device_unit().
1.3	25-Mar-2006	thorpej	Use device_parent().
1.2	11-Dec-2005	christos	branches: 1.2.4; 1.2.6; 1.2.8; 1.2.10; 1.2.12; merge ktrace-lwp.
1.1	20-Jun-2005	briggs	branches: 1.1.2; 1.1.4; 1.1.10; Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1.10.2	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.1.10.1	20-Jun-2005	skrll	file ata_raid_adaptec.c was added on branch ktrace-lwp on 2005-11-10 14:03:54 +0000
1.1.4.4	24-Mar-2008	yamt	sync with head.
1.1.4.3	04-Feb-2008	yamt	sync with head.
1.1.4.2	07-Dec-2007	yamt	sync with head
1.1.4.1	21-Jun-2006	yamt	sync with head.
1.1.2.2	03-Jul-2005	tron	Pull up revision 1.1 (requested by briggs in ticket #521): Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1.2.1	20-Jun-2005	tron	file ata_raid_adaptec.c was added on branch netbsd-3 on 2005-07-03 21:11:10 +0000
1.2.12.3	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.2.12.2	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.2.12.1	28-Mar-2006	tron	Merge 2006-03-28 NetBSD-current into the "peter-altq" branch.
1.2.10.2	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.2.10.1	19-Apr-2006	elad	sync with head.
1.2.8.2	24-May-2006	yamt	sync with head.
1.2.8.1	01-Apr-2006	yamt	sync with head.
1.2.6.2	01-Jun-2006	kardel	Sync with head.
1.2.6.1	22-Apr-2006	simonb	Sync with head.
1.2.4.1	09-Sep-2006	rpaulo	sync with head
1.5.40.2	18-Feb-2008	mjf	Sync with HEAD.
1.5.40.1	08-Dec-2007	mjf	Sync with HEAD.
1.5.34.2	23-Mar-2008	matt	sync with HEAD
1.5.34.1	09-Jan-2008	matt	sync with HEAD
1.5.32.1	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.7.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.7.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.7.2.1	24-Mar-2008	keiichi	sync with head.
1.8.10.1	19-Oct-2008	haad	Sync with HEAD.
1.8.6.1	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.8.4.1	04-May-2009	yamt	sync with head.
1.9.38.1	03-Dec-2017	jdolecek	update from HEAD
1.10.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.10.4.1	10-Jun-2019	christos	Sync with HEAD
1.10.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.12	01-Aug-2023	mrg	don't test arrays against NULL. found by GCC 12.
1.11	19-Mar-2022	hannken	Lock vnode across VOP_OPEN.
1.10	04-Oct-2019	mrg	use memmove() not memcpy() for overlapping regions.
1.9	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.8	01-Nov-2017	mlelstv	branches: 1.8.2; 1.8.4; 1.8.6; refactor wd and ataraid drivers to use common disk subroutines.
1.7	25-Mar-2014	christos	kill sprintf
1.6	06-Jul-2010	bsh	branches: 1.6.8; 1.6.18; 1.6.22; * use correct interleave factor for RAID-0 of Intel MatrixRaid. * report stripe size in Kibytes to bioctl(8). Unfortunately this change will break compatibility with existing filesystem made by NetBSD on RAID-0 Intel MatrixRAID volumes. Please read: http://mail-index.netbsd.org/current-users/2010/07/02/msg013810.html
1.5	06-Jul-2010	bsh	fix some probles with ataraid(4) and Intel MatrixRaid. * handle volumes largetr than 2TiB. * support multiple arrays for Intel Matrix RAID
1.4	11-May-2009	cegger	branches: 1.4.2; 1.4.4; use device_xname()
1.3	16-Sep-2008	tron	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.14; 1.3.18; Support multiple volumes connected to Intel MatrixRAID controllers. Code contributed by Juan RP in PR kern/39552.
1.2	15-Sep-2008	tron	Add support for status reports via bio(4) to ataraid(4). The code was contributed by Juan RP in PR kern/39514.
1.1	11-Sep-2008	tron	Add support for Intel MatrixRAID to ataraid(4). It is currently limited to one volume. The driver was provided by Juan RP in PR kern/39511.
1.3.18.4	11-Aug-2010	yamt	sync with head.
1.3.18.3	16-May-2009	yamt	sync with head
1.3.18.2	04-May-2009	yamt	sync with head.
1.3.18.1	16-Sep-2008	yamt	file ata_raid_intel.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:34 +0000
1.3.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.3.6.2	19-Oct-2008	haad	Sync with HEAD.
1.3.6.1	16-Sep-2008	haad	file ata_raid_intel.c was added on branch haad-dm on 2008-10-19 22:16:19 +0000
1.3.4.2	28-Sep-2008	mjf	Sync with HEAD.
1.3.4.1	16-Sep-2008	mjf	file ata_raid_intel.c was added on branch mjf-devfs2 on 2008-09-28 10:40:19 +0000
1.3.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.3.2.1	16-Sep-2008	wrstuden	file ata_raid_intel.c was added on branch wrstuden-revivesa on 2008-09-24 16:38:51 +0000
1.4.4.1	05-Mar-2011	rmind	sync with head
1.4.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.6.22.1	18-May-2014	rmind	sync with head
1.6.18.2	03-Dec-2017	jdolecek	update from HEAD
1.6.18.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.8.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.8.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.8.4.2	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.8.4.1	10-Jun-2019	christos	Sync with HEAD
1.8.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.9	01-Aug-2023	mrg	don't test arrays against NULL. found by GCC 12.
1.8	19-Mar-2022	hannken	Lock vnode across VOP_OPEN.
1.7	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.6	01-Nov-2017	mlelstv	branches: 1.6.2; 1.6.4; 1.6.6; refactor wd and ataraid drivers to use common disk subroutines.
1.5	25-Mar-2014	christos	kill sprintf
1.4	11-May-2009	cegger	branches: 1.4.12; 1.4.22; 1.4.26; use device_xname()
1.3	15-Sep-2008	tron	branches: 1.3.2; 1.3.4; 1.3.6; 1.3.14; 1.3.18; Add support for status reports via bio(4) to ataraid(4). The code was contributed by Juan RP in PR kern/39514.
1.2	10-Sep-2008	tron	Don't attach RAID volumes with incomplete config blocks which would lead to panic later. Patch provided by Juan PR in private e-mail.
1.1	05-Sep-2008	tron	Add JMicron RAID support to ataraid(4). Patch provided by Juan RP in PR kern/39395.
1.3.18.3	16-May-2009	yamt	sync with head
1.3.18.2	04-May-2009	yamt	sync with head.
1.3.18.1	15-Sep-2008	yamt	file ata_raid_jmicron.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:34 +0000
1.3.14.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.3.6.2	19-Oct-2008	haad	Sync with HEAD.
1.3.6.1	15-Sep-2008	haad	file ata_raid_jmicron.c was added on branch haad-dm on 2008-10-19 22:16:19 +0000
1.3.4.2	28-Sep-2008	mjf	Sync with HEAD.
1.3.4.1	15-Sep-2008	mjf	file ata_raid_jmicron.c was added on branch mjf-devfs2 on 2008-09-28 10:40:19 +0000
1.3.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.3.2.1	15-Sep-2008	wrstuden	file ata_raid_jmicron.c was added on branch wrstuden-revivesa on 2008-09-18 04:35:02 +0000
1.4.26.1	18-May-2014	rmind	sync with head
1.4.22.2	03-Dec-2017	jdolecek	update from HEAD
1.4.22.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.4.12.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.6.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.6.4.1	10-Jun-2019	christos	Sync with HEAD
1.6.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.5	19-Mar-2022	hannken	Lock vnode across VOP_OPEN.
1.4	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.3	01-Nov-2017	mlelstv	branches: 1.3.2; 1.3.4; 1.3.6; refactor wd and ataraid drivers to use common disk subroutines.
1.2	25-Mar-2014	christos	kill sprintf
1.1	20-Aug-2008	tacha	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.18; 1.1.36; 1.1.46; 1.1.52; Add support for nVidia MediaShield ata raid found on HP ProLiant ML115. Tested on HEAD and netbsd-4 branch with RAID0, RAID1, SPAN with MCP55 SATA controller.
1.1.52.1	18-May-2014	rmind	sync with head
1.1.46.2	03-Dec-2017	jdolecek	update from HEAD
1.1.46.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.1.36.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.1.18.2	04-May-2009	yamt	sync with head.
1.1.18.1	20-Aug-2008	yamt	file ata_raid_nvidia.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:35 +0000
1.1.6.2	19-Oct-2008	haad	Sync with HEAD.
1.1.6.1	20-Aug-2008	haad	file ata_raid_nvidia.c was added on branch haad-dm on 2008-10-19 22:16:19 +0000
1.1.4.2	28-Sep-2008	mjf	Sync with HEAD.
1.1.4.1	20-Aug-2008	mjf	file ata_raid_nvidia.c was added on branch mjf-devfs2 on 2008-09-28 10:40:19 +0000
1.1.2.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.1.2.1	20-Aug-2008	wrstuden	file ata_raid_nvidia.c was added on branch wrstuden-revivesa on 2008-09-18 04:35:02 +0000
1.3.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.3.4.1	10-Jun-2019	christos	Sync with HEAD
1.3.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.14	19-Mar-2022	hannken	Lock vnode across VOP_OPEN.
1.13	22-Oct-2018	jdolecek	convert from malloc(9) to kmem(9)
1.12	01-Nov-2017	mlelstv	branches: 1.12.2; 1.12.4; 1.12.6; refactor wd and ataraid drivers to use common disk subroutines.
1.11	18-Mar-2008	cube	branches: 1.11.48; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.10	02-Feb-2008	mjf	branches: 1.10.2; 1.10.6; Match up some VOP_OPEN()S with VOP_CLOSE()s. This allows the root file system to be identified properly again.
1.9	26-Nov-2007	pooka	Remove the "struct lwp *" argument from all VFS and VOP interfaces. The general trend is to remove it from all kernel interfaces and this is a start. In case the calling lwp is desired, curlwp should be used. quick consensus on tech-kern
1.8	14-May-2006	elad	branches: 1.8.32; 1.8.34; 1.8.40; integrate kauth.
1.7	28-Mar-2006	thorpej	Use device_unit().
1.6	11-Dec-2005	christos	branches: 1.6.4; 1.6.6; 1.6.8; 1.6.10; 1.6.12; merge ktrace-lwp.
1.5	27-Feb-2005	perry	branches: 1.5.4; nuke trailing whitespace
1.4	28-Oct-2004	yamt	branches: 1.4.4; 1.4.6; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.3	14-Dec-2003	thorpej	Garbage-collect __ATA_DISK_PRIVATE; it is no longer needed.
1.2	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.1	27-Jan-2003	thorpej	branches: 1.1.2; Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.1.2.5	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.1.2.4	02-Nov-2004	skrll	Sync with HEAD.
1.1.2.3	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.2	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.1	03-Aug-2004	skrll	Sync with HEAD
1.4.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.4.4.1	29-Apr-2005	kent	sync with -current
1.5.4.4	24-Mar-2008	yamt	sync with head.
1.5.4.3	04-Feb-2008	yamt	sync with head.
1.5.4.2	07-Dec-2007	yamt	sync with head
1.5.4.1	21-Jun-2006	yamt	sync with head.
1.6.12.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.6.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.6.10.2	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.6.10.1	19-Apr-2006	elad	sync with head.
1.6.8.2	24-May-2006	yamt	sync with head.
1.6.8.1	01-Apr-2006	yamt	sync with head.
1.6.6.2	01-Jun-2006	kardel	Sync with head.
1.6.6.1	22-Apr-2006	simonb	Sync with head.
1.6.4.1	09-Sep-2006	rpaulo	sync with head
1.8.40.2	18-Feb-2008	mjf	Sync with HEAD.
1.8.40.1	08-Dec-2007	mjf	Sync with HEAD.
1.8.34.2	23-Mar-2008	matt	sync with HEAD
1.8.34.1	09-Jan-2008	matt	sync with HEAD
1.8.32.1	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.10.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.10.2.1	24-Mar-2008	keiichi	sync with head.
1.11.48.1	03-Dec-2017	jdolecek	update from HEAD
1.12.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.12.4.1	10-Jun-2019	christos	Sync with HEAD
1.12.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.5	16-Mar-2025	andvar	ataraid(4), ld(4): ensure RAID disk status is set before resolving vnode. Given a RAID array disk is missing (due to removal, failure, or a code bug), its configuration is skipped. This causes ld(4) attachment at ataraid(4) to fail because adi->adi_dev is NULL. The fix will treat disks without known status (ADI_S_ONLINE, ADI_S_SPARE, ADI_S_ASSIGNED) as if ata_raid_disk_vnode_find() returned NULL, more gracefully solving above situation and avoiding the crash. Additionally, replace vnode block device opening logic with vn_bdev_open() (vn_subr(9)). This function provides identical functionality while also incrementing vp->v_writecount, preventing a 'vp->v_writecount > 0` assertion failure on vn_close(). More details are available in the tech-kern mailing list thread: the https://mail-index.netbsd.org/tech-kern/2025/03/10/msg030179.html thread and PR thread. Should fix PR kern/43986 Also observed in PR kern/59130 (due to a different bug). pullups are needed to netbsd-9, netbsd-10.
1.4	19-Mar-2022	hannken	branches: 1.4.4; 1.4.10; Lock vnode across VOP_OPEN.
1.3	25-Jun-2017	maxv	dumb instruction
1.2	24-Jun-2010	hannken	branches: 1.2.18; 1.2.36; Clean up vnode lock operations pass 2: VOP_UNLOCK(vp, flags) -> VOP_UNLOCK(vp): Remove the unneeded flags argument. Welcome to 5.99.32. Discussed on tech-kern.
1.1	16-Sep-2008	tron	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.18; 1.1.22; 1.1.24; Support multiple volumes connected to Intel MatrixRAID controllers. Code contributed by Juan RP in PR kern/39552.
1.1.24.1	03-Jul-2010	rmind	sync with head
1.1.22.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.1.18.3	11-Aug-2010	yamt	sync with head.
1.1.18.2	04-May-2009	yamt	sync with head.
1.1.18.1	16-Sep-2008	yamt	file ata_raid_subr.c was added on branch yamt-nfs-mp on 2009-05-04 08:12:35 +0000
1.1.6.2	19-Oct-2008	haad	Sync with HEAD.
1.1.6.1	16-Sep-2008	haad	file ata_raid_subr.c was added on branch haad-dm on 2008-10-19 22:16:19 +0000
1.1.4.2	28-Sep-2008	mjf	Sync with HEAD.
1.1.4.1	16-Sep-2008	mjf	file ata_raid_subr.c was added on branch mjf-devfs2 on 2008-09-28 10:40:19 +0000
1.1.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.1.2.1	16-Sep-2008	wrstuden	file ata_raid_subr.c was added on branch wrstuden-revivesa on 2008-09-24 16:38:51 +0000
1.2.36.1	28-Aug-2017	skrll	Sync with HEAD
1.2.18.1	03-Dec-2017	jdolecek	update from HEAD
1.4.10.1	02-Aug-2025	perseant	Sync with HEAD
1.4.4.1	07-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1100): sys/dev/ata/ld_ataraid.c: revision 1.51 sys/dev/ata/ata_raid_subr.c: revision 1.5 ataraid(4), ld(4): ensure RAID disk status is set before resolving vnode. Given a RAID array disk is missing (due to removal, failure, or a code bug), its configuration is skipped. This causes ld(4) attachment at ataraid(4) to fail because adi->adi_dev is NULL. The fix will treat disks without known status (ADI_S_ONLINE, ADI_S_SPARE, ADI_S_ASSIGNED) as if ata_raid_disk_vnode_find() returned NULL, more gracefully solving above situation and avoiding the crash. Additionally, replace vnode block device opening logic with vn_bdev_open() (vn_subr(9)). This function provides identical functionality while also incrementing vp->v_writecount, preventing a 'vp->v_writecount > 0` assertion failure on vn_close(). More details are available in the tech-kern mailing list thread: the https://mail-index.netbsd.org/tech-kern/2025/03/10/msg030179.html thread and PR thread. Should fix PR kern/43986 Also observed in PR kern/59130 (due to a different bug).
1.11	11-Mar-2025	andvar	ataraid(4): obtain VIA V-RAID drive index from configuration blocks. The original logic resolved disk index based on the drive channel it was attached to. However, VX800 and some other VIA controllers have two SATA ports per channel, which causes unintended overrides when two disks share the same channel. VIA V-RAID configuration blocks store the disk index as a value incrementing by 0x04 (0x00, 0x04, 0x08, 0x0C). Therefore, shift the value left by 2 obtains the disk number. Additionally, limit disk number to 4, which is the maximum supported by any VIA RAID configuration. Fixes PR kern/59130. Tested on VX800/VT6421A/VT8251 RAID controllers. pull-ups needed for netbsd-9, netbsd-10.
1.10	19-Mar-2022	hannken	branches: 1.10.4; 1.10.10; Lock vnode across VOP_OPEN.
1.9	22-Oct-2018	jdolecek	branches: 1.9.4; convert from malloc(9) to kmem(9)
1.8	01-Nov-2017	mlelstv	branches: 1.8.2; 1.8.4; 1.8.6; refactor wd and ataraid drivers to use common disk subroutines.
1.7	25-Mar-2014	christos	kill sprintf
1.6	19-Sep-2008	christos	branches: 1.6.28; 1.6.38; 1.6.44; PR/46593: Patrick Welche: Compute both the original and the ~VIA_T_BOOTABLE hacked checksums and compare against both.
1.5	18-Mar-2008	cube	branches: 1.5.4; 1.5.6; 1.5.10; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.4	07-Feb-2008	xtraeme	branches: 1.4.2; 1.4.6; Move a debug printf into the ATA_RAID_DEBUG block.
1.3	02-Feb-2008	mjf	Match up some VOP_OPEN()S with VOP_CLOSE()s. This allows the root file system to be identified properly again.
1.2	26-Nov-2007	pooka	Remove the "struct lwp *" argument from all VFS and VOP interfaces. The general trend is to remove it from all kernel interfaces and this is a start. In case the calling lwp is desired, curlwp should be used. quick consensus on tech-kern
1.1	27-Mar-2007	garbled	branches: 1.1.2; 1.1.4; 1.1.8; 1.1.10; 1.1.16; 1.1.18; 1.1.20; 1.1.26; Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.1.26.2	18-Feb-2008	mjf	Sync with HEAD.
1.1.26.1	08-Dec-2007	mjf	Sync with HEAD.
1.1.20.6	24-Mar-2008	yamt	sync with head.
1.1.20.5	11-Feb-2008	yamt	sync with head.
1.1.20.4	04-Feb-2008	yamt	sync with head.
1.1.20.3	07-Dec-2007	yamt	sync with head
1.1.20.2	03-Sep-2007	yamt	sync with head.
1.1.20.1	27-Mar-2007	yamt	file ata_raid_via.c was added on branch yamt-lazymbuf on 2007-09-03 14:33:26 +0000
1.1.18.2	23-Mar-2008	matt	sync with HEAD
1.1.18.1	09-Jan-2008	matt	sync with HEAD
1.1.16.1	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.1.10.2	11-Jul-2007	mjf	Sync with head.
1.1.10.1	27-Mar-2007	mjf	file ata_raid_via.c was added on branch mjf-ufs-trans on 2007-07-11 20:05:17 +0000
1.1.8.2	15-Apr-2007	yamt	sync with head.
1.1.8.1	27-Mar-2007	yamt	file ata_raid_via.c was added on branch yamt-idlelwp on 2007-04-15 16:03:18 +0000
1.1.4.2	10-Apr-2007	ad	Sync with head.
1.1.4.1	27-Mar-2007	ad	file ata_raid_via.c was added on branch vmlocking on 2007-04-10 13:24:22 +0000
1.1.2.2	31-Mar-2007	bouyer	Pull up following revision(s) (requested by garbled in ticket #542): sys/dev/ata/ata_raidreg.h: revision 1.4 via patch sys/dev/ata/ata_raidvar.h: revision 1.5via patch sys/dev/ata/files.ata: revision 1.15via patch sys/dev/ata/ata_raid_via.c: revision 1.1 sys/dev/ata/ata_raid.c: revision 1.22via patch Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.1.2.1	27-Mar-2007	bouyer	file ata_raid_via.c was added on branch netbsd-4 on 2007-03-31 16:38:04 +0000
1.4.6.2	28-Sep-2008	mjf	Sync with HEAD.
1.4.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.4.2.1	24-Mar-2008	keiichi	sync with head.
1.5.10.1	19-Oct-2008	haad	Sync with HEAD.
1.5.6.1	10-Oct-2008	skrll	Sync with HEAD.
1.5.4.1	04-May-2009	yamt	sync with head.
1.6.44.1	18-May-2014	rmind	sync with head
1.6.38.2	03-Dec-2017	jdolecek	update from HEAD
1.6.38.1	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.6.28.1	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.8.6.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.8.4.1	10-Jun-2019	christos	Sync with HEAD
1.8.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.9.4.1	07-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1942): sys/dev/ata/ata_raid_via.c: revision 1.11 ataraid(4): obtain VIA V-RAID drive index from configuration blocks. The original logic resolved disk index based on the drive channel it was attached to. However, VX800 and some other VIA controllers have two SATA ports per channel, which causes unintended overrides when two disks share the same channel. VIA V-RAID configuration blocks store the disk index as a value incrementing by 0x04 (0x00, 0x04, 0x08, 0x0C). Therefore, shift the value left by 2 obtains the disk number. Additionally, limit disk number to 4, which is the maximum supported by any VIA RAID configuration. Fixes PR kern/59130. Tested on VX800/VT6421A/VT8251 RAID controllers.
1.10.10.1	02-Aug-2025	perseant	Sync with HEAD
1.10.4.1	07-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1099): sys/dev/ata/ata_raid_via.c: revision 1.11 ataraid(4): obtain VIA V-RAID drive index from configuration blocks. The original logic resolved disk index based on the drive channel it was attached to. However, VX800 and some other VIA controllers have two SATA ports per channel, which causes unintended overrides when two disks share the same channel. VIA V-RAID configuration blocks store the disk index as a value incrementing by 0x04 (0x00, 0x04, 0x08, 0x0C). Therefore, shift the value left by 2 obtains the disk number. Additionally, limit disk number to 4, which is the maximum supported by any VIA RAID configuration. Fixes PR kern/59130. Tested on VX800/VT6421A/VT8251 RAID controllers.
1.9	10-Sep-2014	matt	g/u_int[0-9]*_t/ s/u_int/uint/g
1.8	11-Sep-2008	tron	branches: 1.8.38; Add support for Intel MatrixRAID to ataraid(4). It is currently limited to one volume. The driver was provided by Juan RP in PR kern/39511.
1.7	05-Sep-2008	tron	Add JMicron RAID support to ataraid(4). Patch provided by Juan RP in PR kern/39395.
1.6	20-Aug-2008	tacha	Add support for nVidia MediaShield ata raid found on HP ProLiant ML115. Tested on HEAD and netbsd-4 branch with RAID0, RAID1, SPAN with MCP55 SATA controller.
1.5	25-Dec-2007	perry	branches: 1.5.6; 1.5.10; 1.5.12; 1.5.16; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.4	27-Mar-2007	garbled	branches: 1.4.10; 1.4.16; 1.4.18; 1.4.22; Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.3	11-Dec-2005	christos	branches: 1.3.24; 1.3.26; 1.3.30; 1.3.32; 1.3.34; merge ktrace-lwp.
1.2	20-Jun-2005	briggs	branches: 1.2.2; Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1	27-Jan-2003	thorpej	branches: 1.1.2; 1.1.16; Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.1.16.1	03-Jul-2005	tron	Pull up revision 1.2 (requested by briggs in ticket #521): Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1.2.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.2	21-Jan-2008	yamt	sync with head
1.2.2.1	03-Sep-2007	yamt	sync with head.
1.3.34.1	29-Mar-2007	reinoud	Pullup to -current
1.3.32.1	11-Jul-2007	mjf	Sync with head.
1.3.30.1	10-Apr-2007	ad	Sync with head.
1.3.26.1	15-Apr-2007	yamt	sync with head.
1.3.24.1	31-Mar-2007	bouyer	Pull up following revision(s) (requested by garbled in ticket #542): sys/dev/ata/ata_raidreg.h: revision 1.4 via patch sys/dev/ata/ata_raidvar.h: revision 1.5via patch sys/dev/ata/files.ata: revision 1.15via patch sys/dev/ata/ata_raid_via.c: revision 1.1 sys/dev/ata/ata_raid.c: revision 1.22via patch Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.4.22.1	02-Jan-2008	bouyer	Sync with HEAD
1.4.18.1	26-Dec-2007	ad	Sync with head.
1.4.16.1	18-Feb-2008	mjf	Sync with HEAD.
1.4.10.1	09-Jan-2008	matt	sync with HEAD
1.5.16.1	19-Oct-2008	haad	Sync with HEAD.
1.5.12.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.5.12.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.5.10.1	04-May-2009	yamt	sync with head.
1.5.6.1	28-Sep-2008	mjf	Sync with HEAD.
1.8.38.1	03-Dec-2017	jdolecek	update from HEAD
1.12	06-Jul-2010	bsh	fix some probles with ataraid(4) and Intel MatrixRaid. * handle volumes largetr than 2TiB. * support multiple arrays for Intel Matrix RAID
1.11	12-May-2009	cegger	branches: 1.11.2; 1.11.4; struct device * -> device_t, no functional changes intended.
1.10	16-Sep-2008	tron	branches: 1.10.8; Support multiple volumes connected to Intel MatrixRAID controllers. Code contributed by Juan RP in PR kern/39552.
1.9	15-Sep-2008	tron	Add support for status reports via bio(4) to ataraid(4). The code was contributed by Juan RP in PR kern/39514.
1.8	11-Sep-2008	tron	Add support for Intel MatrixRAID to ataraid(4). It is currently limited to one volume. The driver was provided by Juan RP in PR kern/39511.
1.7	05-Sep-2008	tron	Add JMicron RAID support to ataraid(4). Patch provided by Juan RP in PR kern/39395.
1.6	20-Aug-2008	tacha	Add support for nVidia MediaShield ata raid found on HP ProLiant ML115. Tested on HEAD and netbsd-4 branch with RAID0, RAID1, SPAN with MCP55 SATA controller.
1.5	27-Mar-2007	garbled	branches: 1.5.30; 1.5.34; 1.5.36; 1.5.40; Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.4	11-Dec-2005	christos	branches: 1.4.24; 1.4.26; 1.4.30; 1.4.32; 1.4.34; merge ktrace-lwp.
1.3	18-Jul-2005	briggs	Update array of software RAID names in ata_raid.c. Change a boundary check to ensure that we won't accidentally read and use uninitialized memory if ATA_RAID_TYPE_MAX is updated without updating the array. Update comment near ATA_RAID_TYPE_MAX to note that the array in ata_raid.c should be updated if a new ATA_RAID_TYPE_* is added.
1.2	20-Jun-2005	briggs	branches: 1.2.2; Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1	27-Jan-2003	thorpej	branches: 1.1.2; 1.1.16; Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.1.16.2	21-Jul-2005	tron	Pull up revision 1.3 (requested by briggs in ticket #604): Update array of software RAID names in ata_raid.c. Change a boundary check to ensure that we won't accidentally read and use uninitialized memory if ATA_RAID_TYPE_MAX is updated without updating the array. Update comment near ATA_RAID_TYPE_MAX to note that the array in ata_raid.c should be updated if a new ATA_RAID_TYPE_* is added.
1.1.16.1	03-Jul-2005	tron	Pull up revision 1.2 (requested by briggs in ticket #521): Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.1.2.1	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.2.2.2	03-Sep-2007	yamt	sync with head.
1.2.2.1	21-Jun-2006	yamt	sync with head.
1.4.34.1	29-Mar-2007	reinoud	Pullup to -current
1.4.32.1	11-Jul-2007	mjf	Sync with head.
1.4.30.1	10-Apr-2007	ad	Sync with head.
1.4.26.1	15-Apr-2007	yamt	sync with head.
1.4.24.1	31-Mar-2007	bouyer	Pull up following revision(s) (requested by garbled in ticket #542): sys/dev/ata/ata_raidreg.h: revision 1.4 via patch sys/dev/ata/ata_raidvar.h: revision 1.5via patch sys/dev/ata/files.ata: revision 1.15via patch sys/dev/ata/ata_raid_via.c: revision 1.1 sys/dev/ata/ata_raid.c: revision 1.22via patch Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.5.40.1	19-Oct-2008	haad	Sync with HEAD.
1.5.36.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.5.36.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.5.34.3	11-Aug-2010	yamt	sync with head.
1.5.34.2	16-May-2009	yamt	sync with head
1.5.34.1	04-May-2009	yamt	sync with head.
1.5.30.1	28-Sep-2008	mjf	Sync with HEAD.
1.10.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.11.4.1	05-Mar-2011	rmind	sync with head
1.11.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.4	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.3	04-Apr-2020	jdolecek	branches: 1.3.2; stop xfer timeouts during recovery, all xfers will be requeued anyway this avoids race with the timeout routine when processing the xfers for requeueing should fix PR kern/54790 by Izumi Tsutsui
1.2	22-Oct-2018	jdolecek	branches: 1.2.2; 1.2.6; 1.2.8; Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.1	11-Oct-2018	jdolecek	branches: 1.1.2; file ata_recovery.c was initially added on branch jdolecek-ncqfixes.
1.1.2.2	15-Oct-2018	jdolecek	change the SATA/NCQ recovery to run in the atabus thread
1.1.2.1	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.2.8.2	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.2.8.1	08-Feb-2022	martin	Pull up following revision(s) (requested by jdolecek in ticket #1426): sys/dev/ata/ata_recovery.c: revision 1.3 sys/dev/ata/ata_subr.c: revision 1.9 stop xfer timeouts during recovery, all xfers will be requeued anyway this avoids race with the timeout routine when processing the xfers for requeueing should fix PR kern/54790 by Izumi Tsutsui
1.2.6.4	21-Apr-2020	martin	Sync with HEAD
1.2.6.3	08-Apr-2020	martin	Merge changes from current as of 20200406
1.2.6.2	10-Jun-2019	christos	Sync with HEAD
1.2.6.1	22-Oct-2018	christos	file ata_recovery.c was added on branch phil-wifi on 2019-06-10 22:07:05 +0000
1.2.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.2.2.1	22-Oct-2018	pgoyette	file ata_recovery.c was added on branch pgoyette-compat on 2018-11-26 01:52:30 +0000
1.3.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.13	23-Dec-2020	skrll	Add missing '\n' in debug
1.12	19-Dec-2020	jmcneill	ata_timeout: restore spl in ATACH_RECOVERING path
1.11	02-May-2020	thorpej	branches: 1.11.2; Back out changes to use a threadpool for now; it's causing trouble for some folks on Thinkpads.
1.10	25-Apr-2020	thorpej	Rather than creating a kthread-per-channel, use a threadpool and a threadpool-job-per-channel for the in-thread-context work that needs to be done (which is rare). On one of my test systems, this results in the total number of LWPs after multi-user boot dropping from 116 to 78.
1.9	04-Apr-2020	jdolecek	branches: 1.9.2; stop xfer timeouts during recovery, all xfers will be requeued anyway this avoids race with the timeout routine when processing the xfers for requeueing should fix PR kern/54790 by Izumi Tsutsui
1.8	07-Nov-2018	jdolecek	branches: 1.8.4; don't touch the queue structure in ata_queue_active() if atabus is already detached from the channel fixes panic on null pointer derefence during shutdown reported privately by Launey Thomas
1.7	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.6	10-Aug-2018	jdolecek	branches: 1.6.2; fix race in wd_lastclose() on systems with two ide disks on same channel, which happened when one disk had pending I/O while the other disk executed the final disk flush - need to restart bufq processing once xfer is freed in this case it could happen e.g. on boot when system executes fsck on different partitions on the two drives in parallell and hence open and closes the disk devices repeatedly add KASSERT() for empty bufq on wd_lastclose(), and fix similar issue also on suspend/standby path this was introduced by the NCQ merge and not dksubr - before the merge each drive had their own xfer, so they could not block each other fixes PR kern/52783 by Onno van der Linden; many thanks for extensive help with tracking this down
1.5	06-Aug-2018	jdolecek	add wddebug() which dumps some status for attached disks; indended for debugging of PR kern/52783
1.4	20-Oct-2017	jdolecek	branches: 1.4.2; 1.4.4; 1.4.6; move ata_queue_alloc(1) and ata_queue_free() calls to ata_channel_init() and ata_channel_destroy() respectively, to make attachment code simpler, and to make it easier to spot special queue manipulation like cmdide(4) on topic of PR kern/52606
1.3	19-Oct-2017	jdolecek	more detailed debug info; also sync DEBUG_* values in wd.c with ata.c
1.2	17-Oct-2017	jdolecek	reintroduce ATACH_IRQ_WAIT flag for attachments using wdcintr(), only process the interrupt when the flag is set - this fixes spurious interrupt during post-reset drive setup in wdc_ata_bio_start(), and wdc_atapi_start() while those functions set WDCTL_IDS, this seems to be ignored by certain (maybe all) PCI-IDE controllers; usually the implicit KERNEL_LOCK() would prevent the interrupt anyway, but not when the start routine is started from the atabus thread, which doesn't take it fixes 'panic: wdc_ata_bio_intr: bad state' reported on current-users by Chavdar Ivanov
1.1	10-Oct-2017	jdolecek	split off functions used by 'wd* at umass?' into separate file, unfortunately the previous approach with NATABUS doesn't work for kernels which include MODULAR, but not atabus - such as macppc and evbarm
1.4.6.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.4.6.1	10-Jun-2019	christos	Sync with HEAD
1.4.4.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.4.4.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.4.2.2	03-Dec-2017	jdolecek	update from HEAD
1.4.2.1	20-Oct-2017	jdolecek	file ata_subr.c was added on branch tls-maxphys on 2017-12-03 11:36:59 +0000
1.6.2.8	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.6.2.7	06-Oct-2018	jdolecek	actually, just make dump use the same queue skip as recovery, and remove the no longer necessary ata_queue_reset() call from wd(4) also for PR kern/47041
1.6.2.6	22-Sep-2018	jdolecek	remove explicit ata_channel_start() calls, it's no longer necessary now that ata_xfer's are allocated via pool and not really limited; replace by just a callout to restart the processing for rare cases where system runs out of memory
1.6.2.5	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.6.2.4	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.6.2.3	17-Sep-2018	jdolecek	switch from TAILQ to SIMPLEQ for ata_xfer pending queue to save space, don't need doubly linked queue
1.6.2.2	17-Sep-2018	jdolecek	move low-level protocol handlers hooks from ata_xfer to separate struct, initialized statically primarily to reduce ata_xfer struct size, but also improves readibility, and enforces consistency
1.6.2.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.8.4.1	08-Feb-2022	martin	Pull up following revision(s) (requested by jdolecek in ticket #1426): sys/dev/ata/ata_recovery.c: revision 1.3 sys/dev/ata/ata_subr.c: revision 1.9 stop xfer timeouts during recovery, all xfers will be requeued anyway this avoids race with the timeout routine when processing the xfers for requeueing should fix PR kern/54790 by Izumi Tsutsui
1.9.2.1	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.11.2.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.120	05-Oct-2021	rin	PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.119	25-Dec-2020	skrll	Use designated initializers for struct ata_bustype
1.118	21-May-2020	jdolecek	branches: 1.118.2; stop timeout handler while scheduling another part of partial I/O, to avoid race between the timeout and I/O submission; the I/O submission can sleep with xfer while waiting for the controller to be ready once it gets to thread context, and timeout might cause the xfer to be freed, leading to crashes due to use-after-free this fixes another type of crashes with slow devices under QEMU reported by Paul Ripke - thanks a lot with extensive debugging help
1.117	19-May-2020	jdolecek	only start the timeout machinery once the I/O is completely setup and successful, particularly after PIO write is finished fixes crashes in case the setup is so slow that timeout is triggered e.g. while still waiting in wdc_wait_for_unbusy() or shortly after, without drive actually having chance to complete the I/O, as seen in some configuration under QEMU by Paul Ripke
1.116	15-May-2020	jdolecek	whitespace (bad indent)
1.115	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.114	04-Apr-2020	jdolecek	branches: 1.114.2; fix deadlock in wdcwait() when xfer timeout happens while the atabus thread sleeps in wdcwait() - check current lwp rather than relying on global ATACH_TH_RUN channel flag should fix the hang part of the problem reported in http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html thanks to Paul Ripke for providing extensive debugging info
1.113	12-Nov-2018	jdolecek	branches: 1.113.4; hold channel lock during whole ata_dmaerr()/ata_downgrade_mode() - according to code inspection this is safe, none of the set_modes hooks execute anything which would be taking the lock adresses PR kern/53714 by Andreas Gustafsson
1.112	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.111	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.110	01-Jun-2018	macallan	branches: 1.110.2; 1.110.4; add a flag to start DMA before issuing commands - needed to work around a bug in some SATA chips which get confused if the disk responds too fast Mostly for K2 SATA / svwsata found in G5 Macs adapted from OpenBSD
1.109	17-Oct-2017	jdolecek	branches: 1.109.2; reintroduce ATACH_IRQ_WAIT flag for attachments using wdcintr(), only process the interrupt when the flag is set - this fixes spurious interrupt during post-reset drive setup in wdc_ata_bio_start(), and wdc_atapi_start() while those functions set WDCTL_IDS, this seems to be ignored by certain (maybe all) PCI-IDE controllers; usually the implicit KERNEL_LOCK() would prevent the interrupt anyway, but not when the start routine is started from the atabus thread, which doesn't take it fixes 'panic: wdc_ata_bio_intr: bad state' reported on current-users by Chavdar Ivanov
1.108	15-Oct-2017	jdolecek	revert the logic for wdc_wait_for_unbusy() in wdc_ata_bio_intr() to pre-NCQ, wdcintr() and wdctimeout() need that; follows same change in atapi_wdc.c rev. 1.128
1.107	08-Oct-2017	christos	Re-use the err label more so it is not unused for drivers that don't support dma.
1.106	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.105	02-Jan-2015	christos	branches: 1.105.6; We have three sets of DTYPE_ constants in the kernel: altq Drop Type disklabel Disk Type file Descriptor Type (not to mention constants that contain the string DTYPE). Let's make them two, by changing the disklabel one to be DisK TYPE since the other disklabel constants seem to do that. Not many userland programs use these constants (and the ones that they do are mostly in ifdefs). They will be fixed shortly.
1.104	10-Sep-2014	matt	branches: 1.104.2; g/u_int[0-9]*_t/ s/u_int/uint/g
1.103	03-Feb-2013	jakllsch	Add argument to wdccommandext() to allow the entire contents of the device/head register to be specified. Needed for upcoming port multipler support in mvsata(4).
1.102	19-Nov-2012	rkujawa	Introduce WDC_CAPABILITY_NO_AUXCTL flag. For lame controllers that don't have aux control registers (driver coming soon).
1.101	31-Jul-2012	bouyer	branches: 1.101.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.100	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.99	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.98	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.97	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.96	09-Jan-2012	jakllsch	Instead of assume that 'features' is 0 in wdccommandext(), pass it as an argument, as is done for wdccommand().
1.95	28-Aug-2011	christos	branches: 1.95.2; 1.95.6; make this compile.
1.94	27-Aug-2011	bouyer	The loongon2f+cs5526+jmicron PATA->SATA bridge cause an interresting issue: 1) because the CS5536 is not associated with a x86 CPU, interrupts are not ack'ed as it expects so interrupts cannot configured as edge-triggered (as is expected for a PCIIDE in compat mode) 2) the PATA->SATA bridge ignores the WDC_IDS (interrupt disable bit) so the PATA IRQ line gets asserted when resetting or running some polled commands. It also wrongly asserts IRQ when the (nonexistent) slave device is selected 2) wouldn't be an issue with edge-triggered interrupt because we would get a spurious interrupt and continue operation, a new interrupt only shows up when the PATA IRQ line goes low and high again. But because of 1), we get an unclearable interrupt instead, and the system loops on the interrupt handler. To workaround this, introduce a WDC_NO_IDS compile option which runs all polled commands (including reset) at splbio() and without sleeps, so that the controller's interrupt is effectively disabled and won't be reenabled before the interrupt can be cleared. The conditions triggering this problem are speficic enough to handle this via a compile-time option; no need for a run-time (e.g. a config(9), device property or callback to disable interrupts) solution.
1.93	28-Mar-2010	snj	Spell "enough" properly.
1.92	19-Oct-2009	bouyer	branches: 1.92.2; 1.92.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.91	16-Nov-2008	bouyer	cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.90	02-Oct-2008	bouyer	branches: 1.90.2; 1.90.4; Kill ATACH_TH_RUN and use cpu_intr_p() instead.
1.89	28-Apr-2008	martin	branches: 1.89.2; 1.89.6; Remove clause 3 and 4 from TNF licenses
1.88	18-Mar-2008	cube	branches: 1.88.2; 1.88.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.87	19-Oct-2007	ad	branches: 1.87.12; 1.87.16; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.86	08-Apr-2007	dyoung	branches: 1.86.8; 1.86.10; 1.86.14; Include opt_ata.h for ATADEBUG definition instead of #defining it unconditionally. Make this compile when ATADEBUG is not #defined.
1.85	30-Sep-2006	itohy	branches: 1.85.4; 1.85.8; 1.85.10; A little effort against kernel bloat.... Exclude ATA DMA support if no ATA DMA capable drivers are compiled in.
1.84	07-Sep-2006	itohy	branches: 1.84.2; 1.84.4; Add PIOBM (busmastering transfer using ATA PIO mode) support. The PIOBM is used by only one driver (will be added later, stay tuned) and intruduce an attribute "ata_piobm" so that it will be conditionally compiled in. The "ata_dma" (busmastering transfer using ATA DMA mode) and "ata_udma" (busmastering transfer using ATA Ultra DMA mode) attributes are also added for consistency, but unused for now.
1.83	11-Dec-2005	christos	branches: 1.83.4; 1.83.8; merge ktrace-lwp.
1.82	06-Jul-2005	thorpej	branches: 1.82.2; - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.81	07-Jun-2005	peter	Fix typos.
1.80	29-May-2005	christos	- remove casts [to avoid volatile] - sprinkle const - avoid shadowed variables
1.79	27-Feb-2005	perry	branches: 1.79.2; nuke trailing whitespace
1.78	16-Jan-2005	bouyer	branches: 1.78.2; Also call ata_dmaerr() on DMA errors (as reported by the underlying controller driver). We want to downgrade to PIO mode if DMA isn't functionnal.
1.77	08-Dec-2004	bouyer	branches: 1.77.2; Move ata_dmaerr() so that it's called only when there really has been an error (ata_bio->r_error doesn't containt anything meaningfull if no error occured). Fix kern/28255 by Izumi Tsutsui, but with a different patch.
1.76	28-Oct-2004	yamt	move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.75	21-Aug-2004	thorpej	- Move to48() to atareg.h and rename it atacmd_to48(). - Add an atacmd_tosataq() that converts a 32-bit command to a Native SATA Queued command.
1.74	20-Aug-2004	thorpej	- Add an (ata_reset_channel)() member to ata_bustype. - Add an ata_reset_channel() function that performs the common parts of resetting an ATA channel, which uses the (ata_reset_channel)() callback to do the heavy lifting. Adjust callers to use ata_reset_channel() instead of wdc_reset_channel(). This removes the last wdc-specific code from ata.c!
1.73	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.72	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.71	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.70	13-Aug-2004	thorpej	Don't bother with bits that tell of the presence of optional callbacks; just check the function pointers for NULL.
1.69	13-Aug-2004	thorpej	Move wdcstart() to ata.c and rename it to atastart().
1.68	13-Aug-2004	thorpej	Move wdc_exec_xfer() to ata.c and rename it ata_exec_xfer().
1.67	12-Aug-2004	thorpej	Move wdc_addref() and wdc_delref() to ata.c and rename them to ata_*.
1.66	12-Aug-2004	thorpej	Move wdc_kill_pending() to ata.c and rename it ata_kill_pending().
1.65	12-Aug-2004	thorpej	- Move wdc_xfer_pool, wdc_get_xfer(), wdc_free_xfer() to ata.c, and rename to ata_*. - Use a static initializer for the ata_xfer_pool.
1.64	12-Aug-2004	thorpej	Rename some constants: WDC_COMPLETE -> ATACMD_COMPLETE WDC_QUEUED -> ATACMD_QUEUED WDC_TRY_AGAIN -> ATACMD_TRY_AGAIN
1.63	11-Aug-2004	mycroft	Make datain_pio and dataout_pio function pointers in wdc_softc, which can be overridden by the backend if desired. Add experimental code to wdc_pcmcia to use this in memory-mapped mode, disabled by default.
1.62	11-Aug-2004	mycroft	Add two helper functions -- wdc_datain_pio() and wdc_dataout_pio() -- which encapsulate the logic for the various methods of transferring data. Use these throughout.
1.61	04-Aug-2004	bouyer	Change wdc_kill_pending() to take a struct ata_drive_datas * as argument, and kill only pending requests for this drive. Implement a DRIVE_WAITDRAIN flag, which will cause the active command to be killed once complete. Other minor fixes. Now it's possible to detach a ATA or ATAPI device from ioctl even when a dd on the raw char partition is running.
1.60	04-Aug-2004	bouyer	Add struct ata_xfer *active_xfer to ata_queue. Now the active xfer isn't the head of the queue any more, this makes a few things easier (this will also help for tagged queuing support). Remove the WDCF_ACTIVE flag, test active_xfer != NULL instead. clean up wdc_free_xfer() and kill_xfer(). Clean up wdc_reset_channel(), and make it issue a ATAPI_SOFT_RESET if the active command is ATAPI. In wdc_atapi_get_params(), use AT_WAIT \| AT_POLL for ATAPI_SOFT_RESET, so that we'll use tsleep() instead of delay(). In wdc_atapi_start(), call wdc_dmawait() at the right place.
1.59	01-Aug-2004	bouyer	Implement an atabus control device, and define some ATA bus control IOCTLS. Implement ATABUSIORESET, which will reset the given ATA bus.
1.58	31-Jul-2004	bouyer	Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes (problem found and analysed on sparc64 by Martin Husemann). This should even make core dumps marginally more reliable on ATA drives.
1.57	22-Jun-2004	mycroft	Update copyright.
1.56	22-Jun-2004	mycroft	Do not force ATA_SINGLE on devices that don't support READ/WRITE MULTI. We just have to take an interrupt for each sector. Tested with one laptop disk (which normally runs in DMA mode and was forced to single-sector transfers) and 3 CF cards. Increases the performance of the CF cards substantially (760KB/s->1240KB/s in one case, 410KB/s->750KB/s in the other two cases).
1.55	01-Jun-2004	mycroft	Only downgrade modes due to an actual CRC error. Downgrading on other errors is anti-social -- especially given that there's no way to upgrade again short of rebooting. Also, downgrade UDMA modes more slowly. It's entirely possible that they're using an 80-wire cable, but it's just too long for the higher modes, or there is minor crosstalk.
1.54	08-May-2004	bouyer	Add a delay(10) after re-enabling interrupts in the control register. Some controllers/drives (e.g. SataLink 3114 with WD Raptor) require it. Should fix kern/23808 by Chris Gilbert, patch suplied by Chris Gilbert on tech-kern, extended to all places enabling interrupts by me.
1.53	02-Mar-2004	fvdl	branches: 1.53.2; During a control operation (like a reset), wait 10us after selecting the drive. This fixes (or rather, works around) a timing problem with WD Raptor drives attached to a Sil3114 SATA controller. Should fix PR 23808, it fixes the same problem for me.
1.52	03-Jan-2004	thorpej	More structure member namespace cleanup: thread -> ch_thread
1.51	03-Jan-2004	thorpej	More wdc_channel structure member namespace cleanup: - channel -> ch_channel - wdc -> ch_wdc
1.50	03-Jan-2004	thorpej	Rename "struct channel_softc" to "struct wdc_channel".
1.49	01-Jan-2004	thorpej	Prepend "wdc_" to wait_for_drq, wait_for_unbusy, and wait_for_ready.
1.48	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.47	14-Dec-2003	thorpej	No need to include wdvar.h anymore.
1.46	14-Dec-2003	thorpej	Instead of calling wddone() directly, set a (*drv_done)() callback in the ata_drive_datas structure and invoke that.
1.45	14-Dec-2003	thorpej	Liberally apply static.
1.44	14-Dec-2003	thorpej	De-__P and use ANSI-style function decls.
1.43	27-Nov-2003	fvdl	There are some cards that map the ATA control and IDE DMA registers in a different fashion. Individually, they have the same functionality, but their layout is different. An example of such a chipset is the Promise 203xx. To be able to deal with this, transform the cmd and dma bus_space handles into an array of handles, each seperately created with bus_space_subregion. The code generated by using the extra indirection shouldn't change much, since the extra indirection is negated by having the offset calculation already done in bus_space_subregion. E.g. bus_space_write_4(tag, handle, offset, value) becomes bus_space_write_4(tag, handles[offset], 0, value) Reviewed by Manuel Bouyer. Tested on wdc_isa, wdc_pcmcia, viaide, piixide (i386) and on cmdide (sparc64).
1.42	29-Oct-2003	bouyer	freese->freeze, as pointed out by Frederick Bruck.
1.41	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.40	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.39	28-Apr-2003	nakayama	branches: 1.39.2; Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.38	27-Sep-2002	provos	remove trailing \n in panic(). approved perry.
1.37	26-Jul-2002	wiz	enouth -> enough.
1.36	23-Apr-2002	bouyer	branches: 1.36.2; 1.36.4; More copyright fixes, pointed out by Thomas. Thanks !
1.35	14-Jan-2002	christos	add the dma command ext translations. pointed out by manuel.
1.34	13-Jan-2002	christos	Add LBA48 support based on patches from Keisuke YOSHIDA. KNF a bit while I am there.
1.33	03-Dec-2001	bouyer	Update my copyrigth.
1.32	03-Dec-2001	bouyer	Allow wd to attach to something else than wdc/pciide (like USB<->ATA bridges): - move some functions from ata.c to ata_wdc.c or wdc.c. - add callbacks to struct ata_bustype so that wd.c doesn't call directly functions from the lower level driver.
1.31	13-Nov-2001	lukem	add RCSID
1.30	13-Jun-2001	bjh21	branches: 1.30.2; Add explicit support for IDE and SCSI adaptors which don't support interrupts. On such adaptors, all transfers are done in polling mode. OK'ed by Manuel on tech-kern.
1.29	18-Apr-2001	jdolecek	fix typo in comment - s/splio()/splbio()/
1.28	12-Jun-2000	bouyer	branches: 1.28.4; Add a callback (*irqack), for controllers that need special action to ack the interrupt once it has been ack'd on the drive.
1.27	15-May-2000	bouyer	branches: 1.27.2; Sync my copyrigth notice
1.26	01-Apr-2000	bouyer	- DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op if an IRQ was not detected, unless the force flag was given. Use this to detect if the IRQ was for us (closer to shared IRQ for controllers which don't have their own IRQ handler in pciide.c) and to poll for DMA xfer. Also makes the timeout recovery code simpler. - ATAPI cleanup: don't call controller-specific functions from atapiconf.c (wdc_), so that it's possible to attach an atapibus to something else than a wdc/pciide (Hi Lennart :). Overload struct scsi_adapter with struct atapi_adapter, defined as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points to an scsi_adapter, atapi code casts it to atapi_adapter if needed. Move atapi_softc to atapiconf.h so that it can be used by the underlying controller code (e.g. atapi_wdc.c). Add an atapi-specific callback atapi_probedev(), which probe a drive in a controller-specific way, allocate the sc_link and fills in the ataparams if needed. It then calls atapi_probedev() (from atapiconf.c) to do the generic initialisations and attach the device. - While I'm there merge and centralise the state definitions in atavar.h. It should now be possible to use a common ata/atapi routine to set the drive's modes (will do later).
1.25	25-Mar-2000	bouyer	Remove unneeded assignement to nblks
1.24	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.23	17-Jan-2000	bouyer	Improve the downgrade logic: - If UDMA 2 is failing try UDMA 1 first, it helps in some cases - downgrade if we get an error in the first 4000 xfers, or if we get 4 errors in 4000 xfers if the first 4000 went without troubles. While I'm there commit a local change I have since some time to get my CD probed: issue a "blanck" IDENTIFY before the one used to detect slave ghosts, with my drive the first IDENTIFY following a controller reset fails with an aborted command ...
1.22	20-Oct-1999	enami	Cancel active transfers on aic/wdc detach. Also makes LS-120 drive works for me again.
1.21	09-Aug-1999	bouyer	branches: 1.21.2; 1.21.4; 1.21.6; Don't use C_NEEDDONE, it's deprecated.
1.20	06-Aug-1999	bouyer	- Add some debug printf (WDCPROBE) in _wdcreset_wait(), I've needed these 2 times in the past - Set up timeout per xfer instead of per interrupt. This helps with PIO transfer (we would call timeout()/untimout() several times for a transfer). - If we missed an interrupt for a PIO transfer, reset and restart it immedialy, don't try to recover and continue. If we missed an interrupt we may have lost a read/write cycle on the IDE bus. If this happens 1) we corrupt data and 2) we enter an interrupt loop at the end of the xfer, as the drive has some more data to read/write, but the host thinks the xfer is done. This last change fix the (or at last some of the) 'lookup after lost interrupt' some peoples have been experiencing.
1.19	01-Apr-1999	bouyer	branches: 1.19.2; - change the interrupt routines to take a 3rd arguments, set to 1 if we are called from the interrupt or timeout handler, 0 otherwise. - use this to know if we can busy-wait for wait_for_unbusy or wait_for_ready This fixes a bug where CDs withot the DRQ_INTR capability would not busy-wait for the CMDOUT phase. While I'm there change 2 delay() to DELAY() for consistency, and garbage-collect some old code from wdcintr() which has been ifdef'd out for some time now.
1.18	25-Mar-1999	bouyer	Avoid busy-waiting when possible if we are not polling (real IRQ or timeout callback). Shared PCI IRQ should now work (but still untested).
1.17	17-Mar-1999	bouyer	Revert to a 10s timeout, 1s is too low for drives in sleep mode.
1.16	07-Mar-1999	bouyer	Keep track of CRC errors in Ultra-DMA mode. If we noticed a CRC error and we need to downgrade, downgrade to PIO, as it has been shown if we got CRC errors in Ultra-DMA mode, we will have silent data corruption in multiword DMA mode (isn't IDE wonderfull ? :). Set timeout to 1s for "normal" ata I/O, to minimise the effects of missed interrupts.
1.15	21-Feb-1999	hubertf	Only define WDCDEBUG if not already define (e.g. by kernel option)
1.14	08-Feb-1999	bouyer	There's no ATA draft where it is required for the drive to set DRDY \| DSC when the disk is ready to transfer data, and in ATA-5 the DSC has been obsoleted. So only wait for DRQ to transfer data. This can be made conditional on the ATA version if it's proven to break with some drives (worked with all the drives I have access to). While I'm there correct a few typos.
1.13	29-Jan-1999	bouyer	Cosmetic changes, from Soren S. Jorvan.
1.12	09-Jan-1999	thorpej	Make this compile again after void * arithmetic changes.
1.11	08-Jan-1999	augustss	Avoid arithmetic on `void *' since that's not ANSI C.
1.10	16-Dec-1998	bouyer	Keep track of DMA errors, and downgrade the transert mode (UDMA ->DMA, DMA->PIO) in case of 2 consecutive errors. Don't downgrade if the PIO/DMA/UDMA modes were forced by a config flag.
1.9	21-Nov-1998	drochner	in wdc_softc: access the per-channel data via a pointer array instead of an array of fixed-sized channel_softc elements. This way IDE controllers which more than 1 channel (pciide) can extend the channel data easily for private needs. To avoid the double dereference at runtime, change the argument of wdcstart() to the channel data pointer instead of the array index.
1.8	20-Nov-1998	thorpej	Implement controller reference counting on ATA devices.
1.7	11-Nov-1998	bouyer	- clearify the boot messages (features supported vs features used). Thanks to Havard Eidnes for his complains about this :) - fix some typo in comments - hoppefully better detection of drives reporting bogus PIO modes.
1.6	13-Oct-1998	bouyer	Add a new drive flag, DRIVE_MODE, set when the drive properly reported its PIO/DMA modes. Don't try a SET_FEATURE if it didn't report its mode.
1.5	13-Oct-1998	bouyer	In debug message, always print controller:channel:drive
1.4	13-Oct-1998	bouyer	Properly separate DEBUF_XFERS and DEBUG_FUNCS: DEBUF_XFERS is for data transfers (may be called often), where DEBUG_FUNCS is for setup functions (not used for normal contitions).
1.3	13-Oct-1998	thorpej	Default to "no debugging messages" (still patchable).
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file ata_wdc.c was initially added on branch bouyer-ide.
1.1.2.15	05-Oct-1998	bouyer	Rename WDC_CAPABILITY_PIO to WDC_CAPABILITY_MODE, as it's used to tell wether the controller's driver can tell which timing mode it uses. Check this before setting DMA modes too. This allow the generic DMA code to work again.
1.1.2.14	04-Oct-1998	bouyer	Remove an unneeded delay().
1.1.2.13	04-Oct-1998	bouyer	New flag for wdc_xfer: C_DMA, set when a cmd will use DMA. This reduce some if() in ata_wdc.c and atapi_wdc.c from 3 tests to one.
1.1.2.12	02-Oct-1998	bouyer	Cleanup a few error message, remove some dead code. Re-add a reset at end of attach, unless we have WDC_NO_EXTRA_RESETS. Add support for data32iot/data32ioh from -current: each controller pass WDC_CAPABILITY_DATA32 if they can do 32bit, and WDC_CAPABILITY_DATA16 if they can do 16bit. For controller that support both, the usual autodetect mechanism is still used.
1.1.2.11	20-Sep-1998	bouyer	Init up xfer->drive in wdc_ata_bio(), or bad things will happen for drive 1. No need to init xfer->channel, it's done in wdc_exec_xfer().
1.1.2.10	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.1.2.9	21-Aug-1998	bouyer	Sync with -current
1.1.2.8	13-Aug-1998	bouyer	- sync with HEAD - better error detection and report for wd drives - better wddump() support - wdcwait() now takes a timout argument, so that we can honnor the timeout argument of scsipi cmds.
1.1.2.7	23-Jun-1998	leo	Add necessary cruft to handle swapped byte-lanes on the data-register. It basically boils down to the fact that there are 2 new options: WDC_CAPABILITY_{ATA,ATAPI}_NOSTREAM with which you are able to control the usage bus_space_read_multi_X() vs. bus_space_read_multi_stream_X().
1.1.2.6	09-Jun-1998	bouyer	If an error occured while setting up the drive, reset the drive's state to 0.
1.1.2.5	09-Jun-1998	bouyer	- sync with trunck - don't try to set the drive's PIO mode if we don't know how to set it on the controller. This seems to confuse some 'intelligent' ISA controllers.
1.1.2.4	05-Jun-1998	bouyer	_DMA -> _UDMA, + cosmetic changes. Ultra DMA still doesn't work.
1.1.2.3	05-Jun-1998	bouyer	- call wdc_ata_bio_done() only once. - better error recovery: if an error occured, retry the transfer in single sector PIO mode.
1.1.2.2	05-Jun-1998	bouyer	A few bux fixes (untested): - If the PIO mode of the controller is unknown, put the drive in the hightest mode it supports. A drive in a faster mode than the controller should work. - Create DMA maps even for unknown controllers in pciide.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.19.2.3	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.19.2.2	23-Jan-2000	he	Pull up revision 1.23 (requested by bouyer): Improve the downgrade logic to work better in some cases (especially for Acer Labs M5229 controllers with Ultra-DMA/66 disks).
1.19.2.1	25-Aug-1999	he	Pull up revision 1.19->1.20: Add some debug printfs, change timeout handling to be more efficient and less error-prone, and fix the "lockup after lost interrupt" problem some users have been reporting. (bouyer)
1.21.6.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.21.4.1	15-Nov-1999	fvdl	Sync with -current
1.21.2.3	21-Apr-2001	bouyer	Sync with HEAD
1.21.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.21.2.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.27.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.28.4.7	18-Oct-2002	nathanw	Catch up to -current.
1.28.4.6	01-Aug-2002	nathanw	Catch up to -current.
1.28.4.5	20-Jun-2002	nathanw	Catch up to -current.
1.28.4.4	28-Feb-2002	nathanw	Catch up to -current.
1.28.4.3	08-Jan-2002	nathanw	Catch up to -current.
1.28.4.2	14-Nov-2001	nathanw	Catch up to -current.
1.28.4.1	21-Jun-2001	nathanw	Catch up to -current.
1.30.2.5	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.30.2.4	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.30.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.30.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.30.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.36.4.1	28-Apr-2003	tron	Pull up revision 1.39 (requested by nakayama in ticket #1275): Make DMA mode works on Promise Ultra66/100 with 48-bit LBA drives. Ok'ed by bouyer in tech-kern@netbsd.org.
1.36.2.1	29-Aug-2002	gehenna	catch up with -current.
1.39.2.10	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.39.2.9	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.39.2.8	17-Jan-2005	skrll	Sync with HEAD.
1.39.2.7	18-Dec-2004	skrll	Sync with HEAD.
1.39.2.6	02-Nov-2004	skrll	Sync with HEAD.
1.39.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.39.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.39.2.3	25-Aug-2004	skrll	Sync with HEAD.
1.39.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.39.2.1	03-Aug-2004	skrll	Sync with HEAD
1.53.2.3	11-Aug-2004	jmc	branches: 1.53.2.3.2; Pullup rev 1.58 (requested by bouyer in ticket #733) Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes
1.53.2.2	02-Jul-2004	he	Pull up revisions 1.56-1.57 (requested by mycroft in ticket #572): Several fixes mostly related to USB: o Don't force ATA_SINGLE on devices which don't support READ/WRITE MULTI. Increases performance with several CF cards substantially. o Update copyright.
1.53.2.1	29-May-2004	tron	Pull up revision 1.54 (requested by bouyer in ticket #396): Add a delay(10) after re-enabling interrupts in the control register. Some controllers/drives (e.g. SataLink 3114 with WD Raptor) require it. Should fix kern/23808 by Chris Gilbert, patch suplied by Chris Gilbert on tech-kern, extended to all places enabling interrupts by me.
1.53.2.3.2.4	18-Jul-2005	riz	Pull up revision 1.82 (requested by thorpej in ticket #2072): - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.53.2.3.2.3	16-Mar-2005	tron	Pull up revision 1.78 (requested by bouyer in ticket #1325): Also call ata_dmaerr() on DMA errors (as reported by the underlying controller driver). We want to downgrade to PIO mode if DMA isn't functionnal.
1.53.2.3.2.2	16-Mar-2005	tron	Pull up revision 1.77 (requested by bouyer in ticket #1324): Move ata_dmaerr() so that it's called only when there really has been an error (ata_bio->r_error doesn't containt anything meaningfull if no error occured). Fix kern/28255 by Izumi Tsutsui, but with a different patch.
1.53.2.3.2.1	16-Mar-2005	jmc	Pullup rev 1.55 (requested by mycroft in ticket #1018) Only downgrade modes due to an actual CRC error. Downgrading on other errors is anti-social -- especially given that there's no way to upgrade again short of rebooting. Also, downgrade UDMA modes more slowly. It's entirely possible that they're using an 80-wire cable, but it's just too long for the higher modes, or there is minor crosstalk.
1.77.2.1	29-Apr-2005	kent	sync with -current
1.78.2.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.79.2.1	06-Jul-2005	tron	Pull up revision 1.82 (requested by thorpej in ticket #553): - When starting an ATA or ATAPI transfer, handle the case where (*dma_init)() returns EINVAL, indicating that DMA cannot be done for this transfer. Fall back to PIO in this case. - Add a geodeide_dma_init() routine that checks to make sure that transfers start on a 16 byte boundary, returning EINVAL if not. Works around a chip bug that causes a hard system hang. Problem reported and patch tested by Erik Fair.
1.82.2.4	24-Mar-2008	yamt	sync with head.
1.82.2.3	27-Oct-2007	yamt	sync with head.
1.82.2.2	03-Sep-2007	yamt	sync with head.
1.82.2.1	30-Dec-2006	yamt	sync with head.
1.83.8.1	14-Sep-2006	yamt	sync with head.
1.83.4.1	09-Sep-2006	rpaulo	sync with head
1.84.4.1	22-Oct-2006	yamt	sync with head
1.84.2.1	18-Nov-2006	ad	Sync with head.
1.85.10.1	11-Jul-2007	mjf	Sync with head.
1.85.8.2	23-Oct-2007	ad	Sync with head.
1.85.8.1	10-Apr-2007	ad	Sync with head.
1.85.4.1	15-Apr-2007	yamt	sync with head.
1.86.14.1	25-Oct-2007	bouyer	Sync with HEAD.
1.86.10.2	23-Mar-2008	matt	sync with HEAD
1.86.10.1	06-Nov-2007	matt	sync with HEAD
1.86.8.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.87.16.4	17-Jan-2009	mjf	Sync with HEAD.
1.87.16.3	05-Oct-2008	mjf	Sync with HEAD.
1.87.16.2	02-Jun-2008	mjf	Sync with HEAD.
1.87.16.1	03-Apr-2008	mjf	Sync with HEAD.
1.87.12.1	24-Mar-2008	keiichi	sync with head.
1.88.4.3	11-Aug-2010	yamt	sync with head.
1.88.4.2	11-Mar-2010	yamt	sync with head
1.88.4.1	16-May-2008	yamt	sync with head.
1.88.2.1	18-May-2008	yamt	sync with head.
1.89.6.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.89.6.1	19-Oct-2008	haad	Sync with HEAD.
1.89.2.1	10-Oct-2008	skrll	Sync with HEAD.
1.90.4.1	20-Nov-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #67): sys/dev/ic/wdc.c: revision 1.256 sys/dev/ata/ata.c: revision 1.102 sys/dev/scsipi/atapi_wdc.c: revision 1.109 sys/dev/ata/atavar.h: revision 1.78 sys/dev/ata/ata_wdc.c: revision 1.91 cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.90.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.92.4.1	30-May-2010	rmind	sync with head
1.92.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.95.6.1	18-Feb-2012	mrg	merge to -current.
1.95.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.95.2.3	16-Jan-2013	yamt	sync with (a bit old) head
1.95.2.2	30-Oct-2012	yamt	sync with head
1.95.2.1	17-Apr-2012	yamt	sync with head
1.101.2.2	03-Dec-2017	jdolecek	update from HEAD
1.101.2.1	25-Feb-2013	tls	resync with head
1.104.2.1	06-Apr-2015	skrll	Sync with HEAD
1.105.6.12	30-Sep-2017	jdolecek	must drop channel lock before calling ata_dmaerr() to avoid 'locking against myself' in case of consecutive errors from ata_reset_channel() called via ata_downgrade_mode()
1.105.6.11	26-Sep-2017	jdolecek	make compile without NATA_DMA
1.105.6.10	21-Sep-2017	jdolecek	add missing ata_channel_unlock() in bailout path of wdc_ata_bio_intr()
1.105.6.9	10-Sep-2017	jdolecek	refactor code so that xfer c_start() hook is called with channel mutex held, and hence the controller submit code no longer relies on spl tested all the affected drivers - wdc (via piixide), ahci, mvsata, siisata, both disk and atapi I/O
1.105.6.8	12-Aug-2017	jdolecek	convert the atabus thread to use the channel lock and a condvar, adjust code which sets the relevant channel flags to take the lock while doing so
1.105.6.7	12-Aug-2017	jdolecek	remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status, so that there is less hidden state shared by commands; primary intent is to make the NCQ and non-NCQ paths more similar, and remove possibility of incorrect handling for the NCQ commands tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4), ahcisata(4), mvsata(4) on real hw
1.105.6.6	27-Jun-2017	jdolecek	attend error paths, more strict asserts and code consistency - atastart() and ata_kill_pending() now KASSERT() that all xfers on queue have same channel - inactive xfers are killed via new reason KILL_GONE_INACTIVE, controller code must not call any resource deactivation in that case - c_intr() must call ata_waitdrain_xfer_check() as first thing, and must not further touch any xfer structures on exit path; any resource cleanup is supposed to be done in c_kill_xfer() - c_kill_xfer() should never call atastart() - ata_waitdrain_check() removed, replaced by ata_waitdrain_xfer_check() - ATA_DRIVE_WAITDRAIN handling converted to use condvar - removed unused ata_c callback
1.105.6.5	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.105.6.4	16-Jun-2017	jdolecek	adjust reset channel and dump paths - channel reset now always kills active transfer, even on dump path, but now doesn't touch the queued waiting transfers; also kill_xfer hook is always called, so that HBA can free any private xfer resources and thus the dump request has chance to work - kill_xfer routines now always call ata_deactivate_xfer(); added KASSERT()s to ata_free_xfer() to expect deactivated xfer - when called during channel reset before dump, ata_kill_active() drops any queued waiting transfers without processing - do not (re)queue any transfers in wddone() when dumping - kill AT_RST_NOCMD flag This should also hopefully fix the 'polled command has been queued' panic as reported in: PR kern/11811 by John Hawkinson PR kern/47041 by Taylor R Campbell PR kern/51979 by Martin Husemann dump tested working with piixide(4) and ahci(4). mvsata(4) dump times out, but otherwise tested working, will be fixed separately. siisata(4) mechanically changed and not tested.
1.105.6.3	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.105.6.2	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.105.6.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.109.2.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.109.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.109.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.110.4.4	15-Oct-2018	jdolecek	change the SATA/NCQ recovery to run in the atabus thread
1.110.4.3	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.110.4.2	17-Sep-2018	jdolecek	move low-level protocol handlers hooks from ata_xfer to separate struct, initialized statically primarily to reduce ata_xfer struct size, but also improves readibility, and enforces consistency
1.110.4.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.110.2.3	21-Apr-2020	martin	Sync with HEAD
1.110.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.110.2.1	10-Jun-2019	christos	Sync with HEAD
1.113.4.1	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.114.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.118.2.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.5	10-Oct-2017	jdolecek	split off functions used by 'wd* at umass?' into separate file, unfortunately the previous approach with NATABUS doesn't work for kernels which include MODULAR, but not atabus - such as macppc and evbarm
1.4	08-Oct-2017	jdolecek	as a stopgap measure, include necessary parts of dev/ata/ata.c even if no atabus is configured, so that kernels with just 'wd* at umass?' and no atabus compile; fixes kernel builds for vax, pmax and evbmips after NCQ branch merge XXX wd* at umass? still actually untested after the merge, need to discuss XXX what to do with it
1.3	13-Feb-2009	apb	Use "defopt MODULAR" in sys/conf/files, and #include "opt_modular.h" in all kernel sources that use the MODULAR option. Proposed in tech-kern on 18 Jan 2009.
1.2	12-Nov-2008	ad	branches: 1.2.4; Remove LKMs and switch to the module framework, pass 1. Proposed on tech-kern@.
1.1	07-Sep-2006	itohy	branches: 1.1.4; 1.1.6; 1.1.12; 1.1.58; 1.1.62; 1.1.68; 1.1.72; Add PIOBM (busmastering transfer using ATA PIO mode) support. The PIOBM is used by only one driver (will be added later, stay tuned) and intruduce an attribute "ata_piobm" so that it will be conditionally compiled in. The "ata_dma" (busmastering transfer using ATA DMA mode) and "ata_udma" (busmastering transfer using ATA Ultra DMA mode) attributes are also added for consistency, but unused for now.
1.1.72.2	03-Mar-2009	skrll	Sync with HEAD.
1.1.72.1	19-Jan-2009	skrll	Sync with HEAD.
1.1.68.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.1.62.1	04-May-2009	yamt	sync with head.
1.1.58.1	17-Jan-2009	mjf	Sync with HEAD.
1.1.12.2	30-Dec-2006	yamt	sync with head.
1.1.12.1	07-Sep-2006	yamt	file ataconf.h was added on branch yamt-lazymbuf on 2006-12-30 20:47:54 +0000
1.1.6.2	14-Sep-2006	yamt	sync with head.
1.1.6.1	07-Sep-2006	yamt	file ataconf.h was added on branch yamt-pdpolicy on 2006-09-14 12:31:27 +0000
1.1.4.2	09-Sep-2006	rpaulo	sync with head
1.1.4.1	07-Sep-2006	rpaulo	file ataconf.h was added on branch rpaulo-netinet-merge-pcb on 2006-09-09 02:49:44 +0000
1.2.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.46	05-Jul-2022	andvar	s/retrys/retries/ in comments.
1.45	15-May-2020	jdolecek	remove unused atacmd_tostatq()
1.44	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.43	30-Oct-2013	drochner	branches: 1.43.18; -recognize CF cards by the magic value in inquiry data -kill CFG_ATAPI_MASK, didn't see anything in the specs supporting that it exists
1.42	09-Jan-2013	riastradh	branches: 1.42.2; Name the ATA SECURITY commands.
1.41	19-Oct-2012	drochner	Implement experimental support to pass notifications that a file was deleted from the filesystem to the disk driver, commonly known as "discard" or "trim". fs/driver support is in ffs and ata wd for now. This is what was posted here: http://mail-index.netbsd.org/tech-kern/2012/02/28/msg012813.html with minor cleanup, and the global switch replaced by a mount option.
1.40	24-Oct-2011	jakllsch	branches: 1.40.2; 1.40.12; SATA_SIGNAL_GEN3 definition for Serial ATA capabilities word in IDENTIFY DEVICE structure.
1.39	24-Oct-2011	jakllsch	Add defines for SET FEATURES for the SATA, PUIS and AAM feature sets.
1.38	25-Jan-2010	jakllsch	Add some more elements to struct ataparams. These mostly concern the new Long Physical Sectors and Long Logical Sectors feature sets. Information from a draft of ACS-2, with some additional in-between members and bits being borrowed from OpenBSD. There's lots more in ACS-2 we could add, but for the most part it would be superfluous. XXX: Some elements marked as reserved no longer are, and some elements we still reference are now obsolete.
1.37	22-Jan-2010	jakllsch	Correct an error in previous commit.
1.36	21-Jan-2010	jakllsch	Convert to C99 types. Make whitespace more consistent.
1.35	10-Jan-2010	tsutsui	Provide atacmd_to48() in case of #if defined(_STANDALONE) too because standalone wdc driver in cobalt bootloader requires it for LBA48 support.
1.34	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.33	08-Jun-2009	jakllsch	Add atap_max_lba member to ataparams for LBA48 Maximum Address. From OpenBSD. While here, renumber some __reserved member names to be sequential. Also, add a member for the World Wide Name. Use atap_max_lba in wd.c instead of offsets into a __reserved member.
1.32	25-Dec-2007	perry	branches: 1.32.10; 1.32.20; 1.32.24; 1.32.28; Convert many of the uses of __attribute__ to equivalent __packed, __unused and __dead macros from cdefs.h
1.31	18-Nov-2007	christos	branches: 1.31.2; 1.31.6; From Marco Trillo (marcotrillo at gmail dot com): Add Advanced power management in atactl(8)
1.30	21-Aug-2007	bouyer	branches: 1.30.2; 1.30.6; 1.30.8; Improved SATA support, from Jonathan A. Kollasch in PR bin/36772: * use full 5-bit command queue depth that SATA supports * decode SATA bits
1.29	16-Nov-2006	christos	branches: 1.29.8; 1.29.18; 1.29.22; __unused removal on arguments; approved by core.
1.28	30-Sep-2006	itohy	A little effort against kernel bloat.... Exclude ATA DMA support if no ATA DMA capable drivers are compiled in.
1.27	24-Sep-2006	xtraeme	Typo: WCDD_SECURITY_FREEZE -> WDCC_SECURITY_FREEZE Noticed by Sergey Svishchev.
1.26	16-Feb-2006	perry	branches: 1.26.14; 1.26.16; Change "inline" back to "__inline" in .h files -- C99 is still too new, and some apps compile things in C89 mode. C89 keywords stay. As per core@.
1.25	24-Dec-2005	perry	branches: 1.25.2; 1.25.4; 1.25.6; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.24	11-Dec-2005	christos	merge ktrace-lwp.
1.23	12-Apr-2005	drochner	branches: 1.23.2; add definition for "security freeze"
1.22	27-Feb-2005	perry	branches: 1.22.2; nuke trailing whitespace
1.21	23-Aug-2004	he	branches: 1.21.4; 1.21.6; Put an #ifdef _KERNEL around inline functions using panic(). Fixes compile problem for the atactl program for the vax target.
1.20	21-Aug-2004	thorpej	- Move to48() to atareg.h and rename it atacmd_to48(). - Add an atacmd_tosataq() that converts a 32-bit command to a Native SATA Queued command.
1.19	12-Aug-2004	thorpej	Add SATA queued read/write commands.
1.18	03-May-2004	bouyer	When using LBA48, use Flush cache extended command if supported. From Christos Zoulas in PR 25366, with minor adjustements from me.
1.17	14-Mar-2004	wiz	branches: 1.17.2; It's extension, not extention. From Miod Vallat (miod at online fr).
1.16	30-Dec-2003	thorpej	Move the register definitions for ATA Task File registers to atareg.h; they are also used by controllers that do not map the registers into the system address space, and are thus distinct from the "controller" registers.
1.15	30-Dec-2003	thorpej	SMART info is logically part of the register set; move it from atavar.h to atareg.h
1.14	30-Nov-2003	yamt	add definitions of new feature bits.
1.13	30-Nov-2003	yamt	- rename WDC_CAP_LBA48 to ATA_CMD2_LBA48. - remove WDC_CAP_QUEUE. (there's already ATA_CMD2_RWQ.) as discussed on tech-kern@.
1.12	26-Nov-2003	yamt	remove ATA_CMD2_SMART, which seems wrong and isn't used.
1.11	02-Nov-2003	wiz	Fix two typos. From Tom Cosgrove via jmc@openbsd.
1.10	23-Sep-2003	skd	Additional definitions as per SATA specifications. Approved by Manuel Bouyer.
1.9	05-Aug-2002	soren	branches: 1.9.6; Add ATA SMART reporting support from Ben Collver in kern/12787.
1.8	13-Jan-2002	christos	branches: 1.8.8; 1.8.10; Add LBA48 support based on patches from Keisuke YOSHIDA. KNF a bit while I am there.
1.7	05-Jul-2001	toshii	branches: 1.7.2; Fix typo. s/extention/extension/
1.6	29-Mar-2000	augustss	branches: 1.6.6; Fix a couple of misspelled comments.
1.5	18-Jan-1999	bouyer	branches: 1.5.2; 1.5.8; move wd.c:print_wderror() to ata.c:ata_perror(). In wdc_probe_caps() add code to guess the ATA revision supported (if ATA4 if Ultra-DMA, ATA2 if PIO mode > 2). We can't rely on param.atap_ata_major here, at last one Ultra-DMA drive claims to support only ATA-3. Use the ATA version in ata_perror(), and to try a flush cache command in a shutdown hook for IDE drives.
1.4	01-Dec-1998	tsubai	struct ataparam has endian dependence, so add big-endian case.
1.3	18-Nov-1998	kenh	Don't protect this with #ifdef _KERNEL
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file atareg.h was initially added on branch bouyer-ide.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.5.8.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.5.2.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.6.6.3	13-Aug-2002	nathanw	Catch up to -current.
1.6.6.2	28-Feb-2002	nathanw	Catch up to -current.
1.6.6.1	24-Aug-2001	nathanw	Catch up with -current.
1.7.2.2	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.7.2.1	11-Feb-2002	jdolecek	Sync w/ -current.
1.8.10.1	14-May-2004	jdc	Pull up revision 1.14 via patch (requested by bouyer in ticket #1690) add definitions of new feature bits.
1.8.8.1	29-Aug-2002	gehenna	catch up with -current.
1.9.6.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.9.6.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.9.6.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.9.6.4	18-Sep-2004	skrll	Sync with HEAD.
1.9.6.3	25-Aug-2004	skrll	Sync with HEAD.
1.9.6.2	12-Aug-2004	skrll	Sync with HEAD.
1.9.6.1	03-Aug-2004	skrll	Sync with HEAD
1.17.2.1	09-May-2004	jdc	Pull up revision 1.18 (requested by bouyer in ticket #262) When using LBA48, use Flush cache extended command if supported. From Christos Zoulas in PR 25366, with minor adjustements from me.
1.21.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.21.4.1	29-Apr-2005	kent	sync with -current
1.22.2.1	15-Apr-2005	tron	Pull up revision 1.23 (requested by drochner in ticket #156): add definition for "security freeze"
1.23.2.4	21-Jan-2008	yamt	sync with head
1.23.2.3	07-Dec-2007	yamt	sync with head
1.23.2.2	03-Sep-2007	yamt	sync with head.
1.23.2.1	30-Dec-2006	yamt	sync with head.
1.25.6.1	22-Apr-2006	simonb	Sync with head.
1.25.4.1	09-Sep-2006	rpaulo	sync with head
1.25.2.1	18-Feb-2006	yamt	sync with head.
1.26.16.2	10-Dec-2006	yamt	sync with head.
1.26.16.1	22-Oct-2006	yamt	sync with head
1.26.14.1	18-Nov-2006	ad	Sync with head.
1.29.22.2	21-Nov-2007	joerg	Sync with HEAD.
1.29.22.1	03-Sep-2007	jmcneill	Sync with HEAD.
1.29.18.1	03-Sep-2007	skrll	Sync with HEAD.
1.29.8.1	09-Oct-2007	ad	Sync with head.
1.30.8.2	18-Feb-2008	mjf	Sync with HEAD.
1.30.8.1	19-Nov-2007	mjf	Sync with HEAD.
1.30.6.1	18-Nov-2007	bouyer	Sync with HEAD
1.30.2.1	09-Jan-2008	matt	sync with HEAD
1.31.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.31.2.1	26-Dec-2007	ad	Sync with head.
1.32.28.1	21-Apr-2010	matt	sync to netbsd-5
1.32.24.1	23-Jul-2009	jym	Sync with HEAD.
1.32.20.1	10-Jan-2010	snj	Pull up following revision(s) (requested by dsl in ticket #1231): sys/dev/ata/atareg.h: revision 1.33 sys/dev/ata/wd.c: revision 1.378 Add atap_max_lba member to ataparams for LBA48 Maximum Address. From OpenBSD. While here, renumber some __reserved member names to be sequential. Also, add a member for the World Wide Name. Use atap_max_lba in wd.c instead of offsets into a __reserved member.
1.32.10.2	11-Mar-2010	yamt	sync with head
1.32.10.1	20-Jun-2009	yamt	sync with head
1.40.12.5	03-Dec-2017	jdolecek	update from HEAD
1.40.12.4	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.40.12.3	25-Feb-2013	tls	resync with head
1.40.12.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.40.12.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.40.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.40.2.2	23-Jan-2013	yamt	sync with head
1.40.2.1	30-Oct-2012	yamt	sync with head
1.42.2.1	18-May-2014	rmind	sync with head
1.43.18.3	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.43.18.2	24-Apr-2017	jdolecek	set NCQ priority field to 'high' for BPRIO_TIMECRITICAL transfers if drive supports it
1.43.18.1	19-Apr-2017	jdolecek	add ATA FUA support
1.109	05-Oct-2021	rin	PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.108	25-May-2020	jdolecek	make ata_downgrade_mode() static, it's not used anywhere else
1.107	02-May-2020	thorpej	Back out changes to use a threadpool for now; it's causing trouble for some folks on Thinkpads.
1.106	25-Apr-2020	thorpej	Rather than creating a kthread-per-channel, use a threadpool and a threadpool-job-per-channel for the in-thread-context work that needs to be done (which is rare). On one of my test systems, this results in the total number of LWPs after multi-user boot dropping from 116 to 78.
1.105	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.104	04-Apr-2020	jdolecek	branches: 1.104.2; fix deadlock in wdcwait() when xfer timeout happens while the atabus thread sleeps in wdcwait() - check current lwp rather than relying on global ATACH_TH_RUN channel flag should fix the hang part of the problem reported in http://mail-index.netbsd.org/netbsd-users/2020/03/12/msg024249.html thanks to Paul Ripke for providing extensive debugging info
1.103	05-Apr-2019	bouyer	branches: 1.103.4; drop AT_RST_NOCMD, it's a cut'n'paste side effect
1.102	05-Apr-2019	christos	unbreak the build, someone forgot to commit more stuff :-)
1.101	24-Oct-2018	jdolecek	detach the controller itself on shutdown; adjust to not detach already detached atabus/channel
1.100	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.99	10-Aug-2018	jdolecek	branches: 1.99.2; fix race in wd_lastclose() on systems with two ide disks on same channel, which happened when one disk had pending I/O while the other disk executed the final disk flush - need to restart bufq processing once xfer is freed in this case it could happen e.g. on boot when system executes fsck on different partitions on the two drives in parallell and hence open and closes the disk devices repeatedly add KASSERT() for empty bufq on wd_lastclose(), and fix similar issue also on suspend/standby path this was introduced by the NCQ merge and not dksubr - before the merge each drive had their own xfer, so they could not block each other fixes PR kern/52783 by Onno van der Linden; many thanks for extensive help with tracking this down
1.98	06-Aug-2018	jdolecek	add wddebug() which dumps some status for attached disks; indended for debugging of PR kern/52783
1.97	01-Jun-2018	macallan	branches: 1.97.2; add a flag to start DMA before issuing commands - needed to work around a bug in some SATA chips which get confused if the disk responds too fast Mostly for K2 SATA / svwsata found in G5 Macs adapted from OpenBSD
1.96	16-Apr-2018	jdolecek	remove superfluous semicolon
1.95	17-Oct-2017	jdolecek	branches: 1.95.2; reintroduce ATACH_IRQ_WAIT flag for attachments using wdcintr(), only process the interrupt when the flag is set - this fixes spurious interrupt during post-reset drive setup in wdc_ata_bio_start(), and wdc_atapi_start() while those functions set WDCTL_IDS, this seems to be ignored by certain (maybe all) PCI-IDE controllers; usually the implicit KERNEL_LOCK() would prevent the interrupt anyway, but not when the start routine is started from the atabus thread, which doesn't take it fixes 'panic: wdc_ata_bio_intr: bad state' reported on current-users by Chavdar Ivanov
1.94	10-Oct-2017	jdolecek	split off functions used by 'wd* at umass?' into separate file, unfortunately the previous approach with NATABUS doesn't work for kernels which include MODULAR, but not atabus - such as macppc and evbarm
1.93	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.92	10-Sep-2014	matt	branches: 1.92.8; g/u_int[0-9]*_t/ s/u_int/uint/g
1.91	03-Apr-2013	bouyer	Fix kernel dump on ahci controller, by making sure we won't sleep while dumping: - introduce ata_delay() which either use delay() or kpause() depending on flags. use it in sata_reset_interface() and some ahci functions - kill ATA_NOSLEEP, it was tested but never set. use ATA_POLL instead. - reduce delay while polling in ahci, to speed up the dump Should fix PR kern/41095
1.90	31-Jul-2012	bouyer	branches: 1.90.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.89	29-Jul-2012	jakllsch	Remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas. Originally part of a commit by bouyer@.
1.88	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.87	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.86	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.85	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.84	24-Jan-2012	jakllsch	Rework struct ata_command to support LBA28 and LBA48-protocol commands. Implement 28- and 48-bit command issuance and parameter read-back in the various ATA host drivers. Add LBA28-protocol support to ATAIOCCOMMAND ioctl.
1.83	29-Oct-2011	jakllsch	branches: 1.83.2; 1.83.6; Some comment improvements; spelling and grammar.
1.82	30-Apr-2011	jakllsch	Add simplistic atabus(4) rescan support.
1.81	18-Apr-2011	rmind	Replace few simple_lock and ltsleep/wakeup uses with mutex(9) and condvar(9). Note to all: please replace old primitives in your code! Thanks.
1.80	19-Oct-2009	bouyer	branches: 1.80.4; 1.80.6; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.79	12-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.78	16-Nov-2008	bouyer	branches: 1.78.4; cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.77	02-Oct-2008	bouyer	branches: 1.77.2; 1.77.4; Kill ATACH_TH_RUN and use cpu_intr_p() instead.
1.76	18-Mar-2008	cube	branches: 1.76.4; 1.76.6; 1.76.10; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.75	09-Dec-2007	jmcneill	branches: 1.75.6; 1.75.10; Merge jmcneill-pm branch.
1.74	09-Jul-2007	ad	branches: 1.74.6; 1.74.8; 1.74.16; 1.74.18; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.73	30-Sep-2006	itohy	branches: 1.73.8; 1.73.10; A little effort against kernel bloat.... Exclude ATA DMA support if no ATA DMA capable drivers are compiled in.
1.72	07-Sep-2006	itohy	branches: 1.72.2; 1.72.4; Add PIOBM (busmastering transfer using ATA PIO mode) support. The PIOBM is used by only one driver (will be added later, stay tuned) and intruduce an attribute "ata_piobm" so that it will be conditionally compiled in. The "ata_dma" (busmastering transfer using ATA DMA mode) and "ata_udma" (busmastering transfer using ATA Ultra DMA mode) attributes are also added for consistency, but unused for now.
1.71	24-Dec-2005	perry	branches: 1.71.4; 1.71.8; Remove leading __ from __(const\|inline\|signed\|volatile) -- it is obsolete.
1.70	11-Dec-2005	christos	merge ktrace-lwp.
1.69	07-Jun-2005	peter	branches: 1.69.2; Fix typos.
1.68	16-May-2005	bouyer	Add a ata_queue_idle() function, which freeze a queue and tsleep() until the controller is idle. Change the powerhook function to call ata_queue_idle() on standby/suspend, and ata_reset_channel() on resume (to wake up the disks and start from a clean state). Fix PR kern/30194 by Lubomir Sedlacik.
1.67	27-Feb-2005	perry	branches: 1.67.2; nuke trailing whitespace
1.66	26-Jan-2005	jmcneill	Add powerhook for atabus.
1.65	30-Oct-2004	bouyer	branches: 1.65.4; 1.65.6; Allow polled commands to be queued, if ata_exec_xfer() is called in thread context. For this introduce 3 new xfer tags: - C_WAIT, equivalent of AT_WAIT - C_WAITACT, when there is a thread waiting on this xfer to become the active one (that is, to be at the head of the queue) - C_FREE, set by ata_free_xfer() when it can't free the xfer because it's still in use. The holder should then free the xfer ASAP. If ata_exec_xfer() is called with (C_POLL \| C_WAIT), and there is already xfers to be processed, assert C_WAITACT and sleep. atastart() checks for C_WAITACT, and wakeup the thread waiting for this xfer to become active if set. atastart() won't process this xfer, it's the responsability of the thread waked up to handle it. Fix (the right way) kern/27421 by Martin Husemann.
1.64	20-Aug-2004	thorpej	- Add an (ata_reset_channel)() member to ata_bustype. - Add an ata_reset_channel() function that performs the common parts of resetting an ATA channel, which uses the (ata_reset_channel)() callback to do the heavy lifting. Adjust callers to use ata_reset_channel() instead of wdc_reset_channel(). This removes the last wdc-specific code from ata.c!
1.63	20-Aug-2004	thorpej	Move common parts of attaching an ata_channel to a new ata_channel_attach() function.
1.62	20-Aug-2004	thorpej	- Add an atac_bustype_ata member to atac_softc to hold a pointer to the ata_bustype for the controller. Fill it in and use it instead of referencing wdc_ata_bustype directly. - Add an atac_atapibus_attach member to atac_softc to hold a pointer to the function that attaches the ATAPI bus for the controller. Fill it in and use it instead of referencing wdc_atapibus_attach directly.
1.61	20-Aug-2004	thorpej	Rename the ata_bustype "ata_reset_channel" member to "ata_reset_drive" to reflect the new reality after Manuel added the bus reset ioctls.
1.60	20-Aug-2004	thorpej	Move most of wdc_softc into a new atac_softc structure that contains info common to all types of ATA controllers.
1.59	14-Aug-2004	thorpej	- Split the register handles out of struct wdc_channel into a separate wdc_regs structure, and array of which (indexed per channel) is pointed to by struct wdc_softc. - Move the resulting wdc_channel structure to atavar.h and rename it to ata_channel. Rename the corresponding flags. - Add a "ch_ndrive" member to struct ata_channel, which indicates the maximum number of drives that can be present on the channel. For now, this is always 2. Add an ATA_MAXDRIVES constant that places an upper limit on this value, also currently 2.
1.58	13-Aug-2004	thorpej	Move wdcstart() to ata.c and rename it to atastart().
1.57	13-Aug-2004	thorpej	Move wdc_exec_xfer() to ata.c and rename it ata_exec_xfer().
1.56	12-Aug-2004	thorpej	Move wdc_addref() and wdc_delref() to ata.c and rename them to ata_*.
1.55	12-Aug-2004	thorpej	Move wdc_kill_pending() to ata.c and rename it ata_kill_pending().
1.54	12-Aug-2004	thorpej	- Move wdc_xfer_pool, wdc_get_xfer(), wdc_free_xfer() to ata.c, and rename to ata_*. - Use a static initializer for the ata_xfer_pool.
1.53	12-Aug-2004	thorpej	Move wdc_probe_caps() to ata.c and rename it ata_probe_caps().
1.52	12-Aug-2004	thorpej	Move wdc_downgrade_mode() to ata.c and rename it to ata_downgrade_mode().
1.51	12-Aug-2004	thorpej	Move wdc_print_modes() into ata.c and rename it ata_print_modes().
1.50	12-Aug-2004	thorpej	Rename some constants: WDC_COMPLETE -> ATACMD_COMPLETE WDC_QUEUED -> ATACMD_QUEUED WDC_TRY_AGAIN -> ATACMD_TRY_AGAIN
1.49	12-Aug-2004	thorpej	Rename "struct wdc_command" to "struct ata_command".
1.48	11-Aug-2004	mycroft	Add two helper functions -- wdc_datain_pio() and wdc_dataout_pio() -- which encapsulate the logic for the various methods of transferring data. Use these throughout.
1.47	04-Aug-2004	bouyer	Change wdc_kill_pending() to take a struct ata_drive_datas * as argument, and kill only pending requests for this drive. Implement a DRIVE_WAITDRAIN flag, which will cause the active command to be killed once complete. Other minor fixes. Now it's possible to detach a ATA or ATAPI device from ioctl even when a dd on the raw char partition is running.
1.46	04-Aug-2004	bouyer	Add struct ata_xfer *active_xfer to ata_queue. Now the active xfer isn't the head of the queue any more, this makes a few things easier (this will also help for tagged queuing support). Remove the WDCF_ACTIVE flag, test active_xfer != NULL instead. clean up wdc_free_xfer() and kill_xfer(). Clean up wdc_reset_channel(), and make it issue a ATAPI_SOFT_RESET if the active command is ATAPI. In wdc_atapi_get_params(), use AT_WAIT \| AT_POLL for ATAPI_SOFT_RESET, so that we'll use tsleep() instead of delay(). In wdc_atapi_start(), call wdc_dmawait() at the right place.
1.45	01-Aug-2004	bouyer	Implement an atabus control device, and define some ATA bus control IOCTLS. Implement ATABUSIORESET, which will reset the given ATA bus.
1.44	31-Jul-2004	bouyer	Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes (problem found and analysed on sparc64 by Martin Husemann). This should even make core dumps marginally more reliable on ATA drives.
1.43	01-Jun-2004	mycroft	Only downgrade modes due to an actual CRC error. Downgrading on other errors is anti-social -- especially given that there's no way to upgrade again short of rebooting. Also, downgrade UDMA modes more slowly. It's entirely possible that they're using an 80-wire cable, but it's just too long for the higher modes, or there is minor crosstalk.
1.42	27-May-2004	thorpej	Clean-up of use of "precomp". It is used as "features" everywhere, so rename the fields in wdc_command as appropriate.
1.41	14-Apr-2004	minoura	Change the value of DRIVE_ATAPIST which was same as that of DRIVE_DMAERR. Ok'd by bouyer.
1.40	03-Jan-2004	thorpej	branches: 1.40.2; Rename "struct channel_softc" to "struct wdc_channel".
1.39	01-Jan-2004	thorpej	Rename: - wdc_xfer to ata_xfer - channel_queue to ata_queue and move them to <dev/ata/atavar.h> so they can be used by non-wdc ATA controllers. Clean up the member names of these structures while at it.
1.38	30-Dec-2003	thorpej	SMART info is logically part of the register set; move it from atavar.h to atareg.h
1.37	30-Dec-2003	thorpej	Move most of the atabus layer into ata.c.
1.36	20-Dec-2003	lha	add more fields to ata_smart_attr, add structures for selftest log
1.35	14-Dec-2003	thorpej	Instead of calling wddone() directly, set a (*drv_done)() callback in the ata_drive_datas structure and invoke that.
1.34	14-Dec-2003	thorpej	Move struct ata_device from wdvar.h to atavar.h.
1.33	14-Dec-2003	thorpej	Move struct ata_bustype from wdvar.h to atavar.h.
1.32	14-Dec-2003	thorpej	Move struct ata_bio from wdvar.h to atavar.h.
1.31	14-Dec-2003	thorpej	Tidy up the whitespace.
1.30	14-Dec-2003	thorpej	De-__P and use ANSI-style function decls.
1.29	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.28	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.27	08-Jul-2003	itojun	function prototype must not have variable name
1.26	05-Aug-2002	soren	branches: 1.26.6; Add ATA SMART reporting support from Ben Collver in kern/12787.
1.25	23-Apr-2002	bouyer	branches: 1.25.2; 1.25.4; More copyright fixes, pointed out by Thomas. Thanks !
1.24	09-Apr-2002	bouyer	In wdc_exec_command(), for data commands, read the status register after the data transfer. This is mandatory for data out commands (although none are used for now), and not forbiddend for data in commands. Also record if we did transfers any data. May solve kern/16159 by making the probe more robust in face of fake identify.
1.23	03-Dec-2001	bouyer	Update my copyrigth.
1.22	03-Dec-2001	bouyer	Allow wd to attach to something else than wdc/pciide (like USB<->ATA bridges): - move some functions from ata.c to ata_wdc.c or wdc.c. - add callbacks to struct ata_bustype so that wd.c doesn't call directly functions from the lower level driver.
1.21	02-Dec-2001	bouyer	Clean up attach of wd/atapibus: kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead of ata_atapi_attach as attach arch. Create a ata_device, compatible with scsipi_channel, to attach wd.
1.20	01-Dec-2001	bouyer	For ATAPI tape drives, poll for DSC (using a callout), to get the real status of the command, and make sure the drive is ready for the next one.
1.19	25-Apr-2001	bouyer	branches: 1.19.2; Pull up the thorpej_scsipi branch to main branch. This is a completely rewritten scsipi_xfer execution engine, and the associated changes to HBA drivers. Overview of changes & features: - All xfers are queued in the mid-layer, rather than doing so in an ad-hoc fashion in individual adapter drivers. - Adapter/channel resource management in the mid-layer, avoids even trying to start running an xfer if the adapter/channel doesn't have the resources. - Better communication between the mid-layer and the adapters. - Asynchronous event notification mechanism from adapter to mid-layer and peripherals. - Better peripheral queue management: freeze/thaw, sorted requeueing during recovery, etc. - Clean separation of peripherals, adapters, and adapter channels (no more scsipi_link). - Kernel thread for each scsipi_channel makes error recovery much easier (no more dealing with interrupt context when recovering from an error). - Mid-layer support for tagged queueing: commands can have the tag type set explicitly, tag IDs are allocated in the mid-layer (thus eliminating the need to use buggy tag ID allocation schemes in many adapter drivers). - support for QUEUE FULL and CHECK CONDITION status in mid-layer; the command will be requeued, or a REQUEST SENSE will be sent as appropriate. Just before the merge syssrc has been tagged with thorpej_scsipi_beforemerge
1.18	15-May-2000	bouyer	branches: 1.18.4; 1.18.6; Sync my copyrigth notice
1.17	02-Apr-2000	augustss	Change the initial field in struct ata_atapi_attach and struct scsipi_link slightly to allow scsibus and atapibus to attach to the same device. Furthermore, only attach a scsibus when the bus type is BUS_SCSI.
1.16	01-Apr-2000	bouyer	- DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op if an IRQ was not detected, unless the force flag was given. Use this to detect if the IRQ was for us (closer to shared IRQ for controllers which don't have their own IRQ handler in pciide.c) and to poll for DMA xfer. Also makes the timeout recovery code simpler. - ATAPI cleanup: don't call controller-specific functions from atapiconf.c (wdc_), so that it's possible to attach an atapibus to something else than a wdc/pciide (Hi Lennart :). Overload struct scsi_adapter with struct atapi_adapter, defined as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points to an scsi_adapter, atapi code casts it to atapi_adapter if needed. Move atapi_softc to atapiconf.h so that it can be used by the underlying controller code (e.g. atapi_wdc.c). Add an atapi-specific callback atapi_probedev(), which probe a drive in a controller-specific way, allocate the sc_link and fills in the ataparams if needed. It then calls atapi_probedev() (from atapiconf.c) to do the generic initialisations and attach the device. - While I'm there merge and centralise the state definitions in atavar.h. It should now be possible to use a common ata/atapi routine to set the drive's modes (will do later).
1.15	17-Jan-2000	bouyer	Improve the downgrade logic: - If UDMA 2 is failing try UDMA 1 first, it helps in some cases - downgrade if we get an error in the first 4000 xfers, or if we get 4 errors in 4000 xfers if the first 4000 went without troubles. While I'm there commit a local change I have since some time to get my CD probed: issue a "blanck" IDENTIFY before the one used to detect slave ghosts, with my drive the first IDENTIFY following a controller reset fails with an aborted command ...
1.14	20-Oct-1999	enami	Cancel active transfers on aic/wdc detach. Also makes LS-120 drive works for me again.
1.13	10-Mar-1999	bouyer	branches: 1.13.2; 1.13.8; 1.13.10; 1.13.12; Add a new disk flag: DRIVE_OLD for pre-ATA disks. probe routine will now set DRIVE_OLD, DRIVE_ATA or DRIVE_ATAPI based on register signatures. The attach routine will issue a IDENTIFY command for ATA/ATAPI disk, to detect flase matches by the probe routine. probe/attach should now be fully compliant with ata-4/ata-5. As a side effect, ATAPI drives which improperly use ATA register signatures should now be attached as ATAPI.
1.12	07-Mar-1999	bouyer	Keep track of CRC errors in Ultra-DMA mode. If we noticed a CRC error and we need to downgrade, downgrade to PIO, as it has been shown if we got CRC errors in Ultra-DMA mode, we will have silent data corruption in multiword DMA mode (isn't IDE wonderfull ? :). Set timeout to 1s for "normal" ata I/O, to minimise the effects of missed interrupts.
1.11	29-Jan-1999	bouyer	Cosmetic changes, from Soren S. Jorvan.
1.10	18-Jan-1999	bouyer	move wd.c:print_wderror() to ata.c:ata_perror(). In wdc_probe_caps() add code to guess the ATA revision supported (if ATA4 if Ultra-DMA, ATA2 if PIO mode > 2). We can't rely on param.atap_ata_major here, at last one Ultra-DMA drive claims to support only ATA-3. Use the ATA version in ata_perror(), and to try a flush cache command in a shutdown hook for IDE drives.
1.9	16-Dec-1998	bouyer	Keep track of DMA errors, and downgrade the transert mode (UDMA ->DMA, DMA->PIO) in case of 2 consecutive errors. Don't downgrade if the PIO/DMA/UDMA modes were forced by a config flag.
1.8	04-Dec-1998	bouyer	AT_ERROR should be different from AT_TIMEOUT. pointed out by Matthias Drochner.
1.7	02-Dec-1998	bouyer	- change drive_flags from u_int8_t to u_int16_t - keep the modes supported by the drive in struct ata_drive_datas (will be later used for downgrading the DMA/PIO mode on error) - use config flags to force/disable PIO/DMA/UDMA modes - For the CMD PCI0643/6 setup DMA mode to DMA Read multiple.
1.6	23-Nov-1998	kenh	Add a new flag to wdc_command: ATA_READREG. This will indicate that all device registers should be read back into the wdc_command structure after successfull command completion. Use this this in wdioctl() for ATAIOCCOMMAND.
1.5	20-Nov-1998	thorpej	Implement controller reference counting on ATA devices.
1.4	20-Oct-1998	bouyer	Donc reset the disk state to 0 in wd.c, set a flag which will cause it to be reset to 0 at wdcstart(). This fixes a race condition between normal I/O and wdioclt().
1.3	13-Oct-1998	bouyer	Add a new drive flag, DRIVE_MODE, set when the drive properly reported its PIO/DMA modes. Don't try a SET_FEATURE if it didn't report its mode.
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file atavar.h was initially added on branch bouyer-ide.
1.1.2.5	04-Oct-1998	bouyer	atavar.h: drv_softc is a struct device * instead of void*, as it's mostly used for dv_xname wd.c: convert for drv_softc type change, printf cleanup wdc.c: always call ata_get_params() (params was used initialised with non-32bit controllers, leaving to bogus PIO/DMA mode report). Cleaup of the PIO/DMA mode message. pciide_piix_reg.h: Fix definition of PIIX_IDETIM_CLEAR (unused before) pciide.c: add a method do disable a channel on know device. If a channel doesn't have any drive, we disable it and free its resources if disable was successfull. This should help with laptops where the second channel of the PIIX4 is unused but not disabled by BIOS. On such laptops, irq15 can be used for PCMCIA but it was claimed by pciide. Misc printf cleanup. wdc_isa.c: printf cleanup.
1.1.2.4	20-Sep-1998	bouyer	Kill the 'old drive' flag. The probe for IDE controllers now looks like this : - if status = 0xff for both master and slave, no drive at all -> exit. - issue a reset (timeout = 31s). - test signatures: if ATAPI signature, flag a (possible) ATAPI devices here else test register writability, then wait for the drive to become ready, issue a diagnostic command and wait for the drive to become ready if ok, flag a real ATA device. This looks more like the old probe. For ATA devices, if IDENTIFY fails, then it's an old drive and use a fake disklabel/geometry.
1.1.2.3	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.1.2.2	13-Aug-1998	bouyer	- sync with HEAD - better error detection and report for wd drives - better wddump() support - wdcwait() now takes a timout argument, so that we can honnor the timeout argument of scsipi cmds.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.13.12.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.13.12.1	21-Dec-1999	wrstuden	Initial commit of recent changes to make DEV_BSIZE go away. Runs on i386, needs work on other arch's. Main kernel routines should be fine, but a number of the stand programs need help. cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512 byte block devices. vnd, raidframe, and lfs need work. Non 2**n block support is automatic for LKM's and conditional for kernels on "options NON_PO2_BLOCKS".
1.13.10.1	15-Nov-1999	fvdl	Sync with -current
1.13.8.3	27-Mar-2001	bouyer	Kill scsipi_link from comment.
1.13.8.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.13.8.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.13.2.2	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.13.2.1	23-Jan-2000	he	Pull up revision 1.15 (requested by bouyer): Improve the downgrade logic to work better in some cases (especially for Acer Labs M5229 controllers with Ultra-DMA/66 disks).
1.18.6.5	13-Aug-2002	nathanw	Catch up to -current.
1.18.6.4	20-Jun-2002	nathanw	Catch up to -current.
1.18.6.3	17-Apr-2002	nathanw	Catch up to -current.
1.18.6.2	08-Jan-2002	nathanw	Catch up to -current.
1.18.6.1	21-Jun-2001	nathanw	Catch up to -current.
1.18.4.1	23-Jan-2003	msaitoh	Pull up revision 1.24 (requested by Manuel Bouyer): read status after data commands in wdc_exec_command(). Mandatory for data out commands (nonw are used for now), and doesn't hurt for data in. Also keep track if we did transfer data. Makes the probe more robust and solves kern/16159
1.19.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.19.2.2	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.19.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.25.4.1	12-May-2004	tron	Pull up revision 1.41 via patch (requested by bouyer in ticket #1689): Change the value of DRIVE_ATAPIST which was same as that of DRIVE_DMAERR. Ok'd by bouyer.
1.25.2.1	29-Aug-2002	gehenna	catch up with -current.
1.26.6.9	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.26.6.8	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.26.6.7	04-Feb-2005	skrll	Sync with HEAD.
1.26.6.6	02-Nov-2004	skrll	Sync with HEAD.
1.26.6.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.26.6.4	18-Sep-2004	skrll	Sync with HEAD.
1.26.6.3	25-Aug-2004	skrll	Sync with HEAD.
1.26.6.2	12-Aug-2004	skrll	Sync with HEAD.
1.26.6.1	03-Aug-2004	skrll	Sync with HEAD
1.40.2.2	11-Aug-2004	jmc	branches: 1.40.2.2.2; Pullup rev 1.44 (requested by bouyer in ticket #733) Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes
1.40.2.1	16-Apr-2004	tron	Pull up revision 1.41 (requested by minoura in ticket #137): Change the value of DRIVE_ATAPIST which was same as that of DRIVE_DMAERR. Ok'd by bouyer.
1.40.2.2.2.2	16-Apr-2005	tron	Pull up revision 1.65 via patch (requested by bouyer in ticket #1105): Allow polled commands to be queued, if ata_exec_xfer() is called in thread context. For this introduce 3 new xfer tags: - C_WAIT, equivalent of AT_WAIT - C_WAITACT, when there is a thread waiting on this xfer to become the active one (that is, to be at the head of the queue) - C_FREE, set by ata_free_xfer() when it can't free the xfer because it's still in use. The holder should then free the xfer ASAP. If ata_exec_xfer() is called with (C_POLL \| C_WAIT), and there is already xfers to be processed, assert C_WAITACT and sleep. atastart() checks for C_WAITACT, and wakeup the thread waiting for this xfer to become active if set. atastart() won't process this xfer, it's the responsability of the thread waked up to handle it. Fix (the right way) kern/27421 by Martin Husemann.
1.40.2.2.2.1	16-Mar-2005	jmc	Pullup rev 1.43 (requested by mycroft in ticket #1018) Only downgrade modes due to an actual CRC error. Downgrading on other errors is anti-social -- especially given that there's no way to upgrade again short of rebooting. Also, downgrade UDMA modes more slowly. It's entirely possible that they're using an 80-wire cable, but it's just too long for the higher modes, or there is minor crosstalk.
1.65.6.2	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.65.6.1	12-Feb-2005	yamt	sync with head.
1.65.4.1	29-Apr-2005	kent	sync with -current
1.67.2.1	27-May-2005	riz	Pull up revision 1.68 (requested by bouyer in ticket #328): Add a ata_queue_idle() function, which freeze a queue and tsleep() until the controller is idle. Change the powerhook function to call ata_queue_idle() on standby/suspend, and ata_reset_channel() on resume (to wake up the disks and start from a clean state). Fix PR kern/30194 by Lubomir Sedlacik.
1.69.2.5	24-Mar-2008	yamt	sync with head.
1.69.2.4	21-Jan-2008	yamt	sync with head
1.69.2.3	03-Sep-2007	yamt	sync with head.
1.69.2.2	30-Dec-2006	yamt	sync with head.
1.69.2.1	21-Jun-2006	yamt	sync with head.
1.71.8.1	14-Sep-2006	yamt	sync with head.
1.71.4.1	09-Sep-2006	rpaulo	sync with head
1.72.4.1	22-Oct-2006	yamt	sync with head
1.72.2.1	18-Nov-2006	ad	Sync with head.
1.73.10.1	11-Jul-2007	mjf	Sync with head.
1.73.8.1	09-Apr-2007	ad	- Add two new arguments to kthread_create1: pri_t pri, bool mpsafe. - Fork kthreads off proc0 as new LWPs, not new processes.
1.74.18.1	11-Dec-2007	yamt	sync with head.
1.74.16.1	26-Dec-2007	ad	Sync with head.
1.74.8.2	23-Mar-2008	matt	sync with HEAD
1.74.8.1	09-Jan-2008	matt	sync with HEAD
1.74.6.2	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.74.6.1	03-Aug-2007	jmcneill	Pull in power management changes from private branch.
1.75.10.3	17-Jan-2009	mjf	Sync with HEAD.
1.75.10.2	05-Oct-2008	mjf	Sync with HEAD.
1.75.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.75.6.1	24-Mar-2008	keiichi	sync with head.
1.76.10.2	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.76.10.1	19-Oct-2008	haad	Sync with HEAD.
1.76.6.1	10-Oct-2008	skrll	Sync with HEAD.
1.76.4.2	11-Mar-2010	yamt	sync with head
1.76.4.1	16-May-2009	yamt	sync with head
1.77.4.1	20-Nov-2008	snj	Pull up following revision(s) (requested by bouyer in ticket #67): sys/dev/ic/wdc.c: revision 1.256 sys/dev/ata/ata.c: revision 1.102 sys/dev/scsipi/atapi_wdc.c: revision 1.109 sys/dev/ata/atavar.h: revision 1.78 sys/dev/ata/ata_wdc.c: revision 1.91 cpu_intr_p() doesn't account for software interrupts (e.g. callouts) so we can't use it here. Rssurect ATACH_TH_RUN, backing out src/sys/dev/ata/ata.c 1.101 src/sys/dev/ata/ata_wdc.c 1.90 src/sys/dev/ata/atavar.h 1.77 src/sys/dev/ic/wdc.c 1.255 src/sys/dev/scsipi/atapi_wdc.c 1.108 Should fix kern/39927 and kern/39725.
1.77.2.1	19-Jan-2009	skrll	Sync with HEAD.
1.78.4.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.80.6.1	06-Jun-2011	jruoho	Sync with HEAD.
1.80.4.2	31-May-2011	rmind	sync with head
1.80.4.1	21-Apr-2011	rmind	sync with head
1.83.6.1	18-Feb-2012	mrg	merge to -current.
1.83.2.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.83.2.2	30-Oct-2012	yamt	sync with head
1.83.2.1	17-Apr-2012	yamt	sync with head
1.90.2.2	03-Dec-2017	jdolecek	update from HEAD
1.90.2.1	23-Jun-2013	tls	resync from head
1.92.8.29	29-Sep-2017	jdolecek	introduce ATA_BSIZE and use it instead of DEV_BSIZE for get params and recovery, where they are by spec 512 bytes and not negotiable
1.92.8.28	27-Sep-2017	jdolecek	restore the atac_claim_hw and atac_free_hw hooks, they are used on atari
1.92.8.27	26-Sep-2017	jdolecek	make compile without NATA_DMA
1.92.8.26	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.92.8.25	10-Sep-2017	jdolecek	refactor code so that xfer c_start() hook is called with channel mutex held, and hence the controller submit code no longer relies on spl tested all the affected drivers - wdc (via piixide), ahci, mvsata, siisata, both disk and atapi I/O
1.92.8.24	12-Aug-2017	jdolecek	do not reset drive after successful NCQ error recovery
1.92.8.23	12-Aug-2017	jdolecek	convert the atabus thread to use the channel lock and a condvar, adjust code which sets the relevant channel flags to take the lock while doing so
1.92.8.22	12-Aug-2017	jdolecek	remove all logic around ATACH_IRQ_WAIT and channel-global ch_error/ch_status, so that there is less hidden state shared by commands; primary intent is to make the NCQ and non-NCQ paths more similar, and remove possibility of incorrect handling for the NCQ commands tested both disk and ATAPI - piixide(4) on QEMU, and siisata(4), ahcisata(4), mvsata(4) on real hw
1.92.8.21	01-Aug-2017	jdolecek	move the drive recovery block to drive struct, it's inherently per-drive
1.92.8.20	29-Jul-2017	jdolecek	reserve the highest slot for error recovery, and also have ata_channel include space for the READ LOG EXT sector, so that it's not necessary to allocate memory on the error handling path; now ata_read_log_ext_ncq() will never fail due to resource shortage
1.92.8.19	22-Jul-2017	jdolecek	provide channel flag when executing NCQ commands, so that e.g. intr handler can use this for handling decisions without checking xfer
1.92.8.18	21-Jul-2017	jdolecek	KASSERT() that chaosed xfer actually ends up with error; might end up being returned as successful due to bugs in error recovery code
1.92.8.17	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.92.8.16	27-Jun-2017	jdolecek	attend error paths, more strict asserts and code consistency - atastart() and ata_kill_pending() now KASSERT() that all xfers on queue have same channel - inactive xfers are killed via new reason KILL_GONE_INACTIVE, controller code must not call any resource deactivation in that case - c_intr() must call ata_waitdrain_xfer_check() as first thing, and must not further touch any xfer structures on exit path; any resource cleanup is supposed to be done in c_kill_xfer() - c_kill_xfer() should never call atastart() - ata_waitdrain_check() removed, replaced by ata_waitdrain_xfer_check() - ATA_DRIVE_WAITDRAIN handling converted to use condvar - removed unused ata_c callback
1.92.8.15	23-Jun-2017	jdolecek	restart I/O processing after freeing xfer, i.e. now even after commands like cache flush or standby; the command handling no longer use on-stack xfer, hence use queue slot and compete with normal I/O for the xfers the restart give change to all drives attached to the same channel in round-robin fashion, for fair usage and to recover from situation when disk is idle due to all xfers being consumed by other drives make special concession for flush cache - ignore any new I/O requests on the particular disk while the flush cache is waiting for xfer, so that I/O queue won't starve the flush cache and the flush cache would be done ASAP tested on piixide(4), ahci(4), siisata(4)
1.92.8.14	21-Jun-2017	jdolecek	hold channel lock for ata_exec_xfer() and most of atastart(), convert C_WAITACT handling to use condvar add comment for the code block with C_FREE and ata_free_xfer() explaining why it's not abstraction layer violation
1.92.8.13	21-Jun-2017	jdolecek	change ata_queue_hwslot_to_xfer() and ata_queue_get_active_xfer() to take ata_channel instead of ata_queue as parameter, and lock the channel while traversing the queue
1.92.8.12	21-Jun-2017	jdolecek	actually remove the KASSERT() for queue_active in wdc.c as it's superfluous, and make struct ata_queue completly opaque to attachment code; fix two other places which were not converted to ata_queue_alloc() which were surfaced by this change
1.92.8.11	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.92.8.10	19-Jun-2017	jdolecek	add ata_channel lock, use it to protect queue manipulation (only that for now); add ata_channel_detach() to destroy the locks change ata_get_xfer() so that it can wait for xfer, convert all on-stack xfer code to use the blocking variant fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(), convert drive probe code also over from slot0 XXX to ata_get_xfer() drive reset and PMP now works on siisata(4) too; changes tested also on piixide(4), ahci(4), mvsata(4)
1.92.8.9	16-Jun-2017	jdolecek	adjust reset channel and dump paths - channel reset now always kills active transfer, even on dump path, but now doesn't touch the queued waiting transfers; also kill_xfer hook is always called, so that HBA can free any private xfer resources and thus the dump request has chance to work - kill_xfer routines now always call ata_deactivate_xfer(); added KASSERT()s to ata_free_xfer() to expect deactivated xfer - when called during channel reset before dump, ata_kill_active() drops any queued waiting transfers without processing - do not (re)queue any transfers in wddone() when dumping - kill AT_RST_NOCMD flag This should also hopefully fix the 'polled command has been queued' panic as reported in: PR kern/11811 by John Hawkinson PR kern/47041 by Taylor R Campbell PR kern/51979 by Martin Husemann dump tested working with piixide(4) and ahci(4). mvsata(4) dump times out, but otherwise tested working, will be fixed separately. siisata(4) mechanically changed and not tested.
1.92.8.8	24-Apr-2017	jdolecek	set NCQ priority field to 'high' for BPRIO_TIMECRITICAL transfers if drive supports it
1.92.8.7	20-Apr-2017	jdolecek	use a fake ata_channel struct in umass_isdata.c so that the wd(4) calls to ata_{get,free}_xfer() can work, adjust to ata interface changes compile tested only
1.92.8.6	19-Apr-2017	jdolecek	add ATA FUA support
1.92.8.5	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.92.8.4	15-Apr-2017	jdolecek	pass also ata_command via ata_xfer, callers of ata_exec_command() is now responsible for allocation/disposal of the structure change code to allocate ata_xfer for commands on stack same way as previously the ata_command were, using c_slot 0; adjust asserts so that it would allow several xfers with same c_slot, as long as only one such transfer is active at a time
1.92.8.3	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.92.8.2	11-Apr-2017	jdolecek	hide active_xfer from drivers
1.92.8.1	10-Apr-2017	jdolecek	ATA infrastructure improvements to eventually support more outstanding commands patch by Matt Thomas
1.95.2.4	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.95.2.3	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.95.2.2	25-Jun-2018	pgoyette	Sync with HEAD
1.95.2.1	22-Apr-2018	pgoyette	Sync with HEAD
1.97.2.3	21-Apr-2020	martin	Sync with HEAD
1.97.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.97.2.1	10-Jun-2019	christos	Sync with HEAD
1.99.2.12	15-Oct-2018	jdolecek	change the SATA/NCQ recovery to run in the atabus thread
1.99.2.11	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.99.2.10	06-Oct-2018	jdolecek	actually, just make dump use the same queue skip as recovery, and remove the no longer necessary ata_queue_reset() call from wd(4) also for PR kern/47041
1.99.2.9	06-Oct-2018	jdolecek	remove AT_RST_EMERG, do the queue reset explicitly in wd(4) this should explicitly fix PR kern/47041 with sync during heavy disk activity, even thought it was actually already implicitly fixed by calling ata_thread_run() for drive reset in previous commit already, since the function already called ata_queue_reset() drop now unused ch_reset_flags and drive_reset_flags
1.99.2.8	03-Oct-2018	jdolecek	change channel reset and drive reset for all ATA controllers to always run via thread, and with channel lock held the whole time; the queue is frozen while reset is pending for this repurpose ata_reset_channel() into new ata_thread_run() also adjust some device printfs to not leak xfer pointer, and avoid aprint_* for non-autoconf messages
1.99.2.7	22-Sep-2018	jdolecek	remove explicit ata_channel_start() calls, it's no longer necessary now that ata_xfer's are allocated via pool and not really limited; replace by just a callout to restart the processing for rare cases where system runs out of memory
1.99.2.6	22-Sep-2018	jdolecek	fix use-after-free in wd(4) dump, detected by switch to the pool change code in wd_dumpblocks() to use it's own non-pool ata_xfer, which skips the deallocation step and thus keeps the contents when the I/O is finished
1.99.2.5	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.99.2.4	17-Sep-2018	jdolecek	move ATAPI-only members of ata_xfer to an union struct to further save space
1.99.2.3	17-Sep-2018	jdolecek	switch from TAILQ to SIMPLEQ for ata_xfer pending queue to save space, don't need doubly linked queue
1.99.2.2	17-Sep-2018	jdolecek	move low-level protocol handlers hooks from ata_xfer to separate struct, initialized statically primarily to reduce ata_xfer struct size, but also improves readibility, and enforces consistency
1.99.2.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.103.4.1	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.104.2.2	25-Apr-2020	bouyer	Sync with bouyer-xenpvh-base2 (HEAD)
1.104.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.33	17-Feb-2025	jakllsch	Restore ATA DMA mode downgrade support everywhere; it's a necessary part of any system supporting (parallel) ATA DMA. There is hardware out there, including cobalt, macppc, and sparc64 where this functionality is necessary to avoid non-functional disks, either in as-shipped hardware configurations or with add-in cards, or perhaps just with compromised IDE/PATA cables. Should address: PR 58767 PR 59023 PR 59078 If anyone really insists on not having this support they can now turn it off themselves with `options ATA_NO_DOWNGRADE_MODE`
1.32	22-Sep-2024	uwe	files.ata: defparam ATADEBUG_WD_MASK
1.31	17-Jul-2023	riastradh	branches: 1.31.6; ata(4): Add ATA_DOWNGRADE_MODE to opt_ata.h. This way adding it to kernel config will trigger recompilation.
1.30	24-Oct-2018	jdolecek	only include umass_isdata.c if atabus present, and likewise ata_subr.c
1.29	24-Oct-2018	jdolecek	only include ata_recovery.c with atabus (same as ata.c which provides ata_thread_run() used here)
1.28	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.27	10-Oct-2017	jdolecek	branches: 1.27.2; 1.27.4; 1.27.6; split off functions used by 'wd* at umass?' into separate file, unfortunately the previous approach with NATABUS doesn't work for kernels which include MODULAR, but not atabus - such as macppc and evbarm
1.26	08-Oct-2017	jdolecek	as a stopgap measure, include necessary parts of dev/ata/ata.c even if no atabus is configured, so that kernels with just 'wd* at umass?' and no atabus compile; fixes kernel builds for vax, pmax and evbmips after NCQ branch merge XXX wd* at umass? still actually untested after the merge, need to discuss XXX what to do with it
1.25	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.24	31-Jul-2012	bouyer	branches: 1.24.2; 1.24.28; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.23	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.22	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.21	17-Jun-2009	jakllsch	branches: 1.21.12; ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.20	16-Sep-2008	tron	branches: 1.20.4; 1.20.8; 1.20.12; Support multiple volumes connected to Intel MatrixRAID controllers. Code contributed by Juan RP in PR kern/39552.
1.19	11-Sep-2008	tron	Add support for Intel MatrixRAID to ataraid(4). It is currently limited to one volume. The driver was provided by Juan RP in PR kern/39511.
1.18	05-Sep-2008	tron	Add JMicron RAID support to ataraid(4). Patch provided by Juan RP in PR kern/39395.
1.17	20-Aug-2008	tacha	Add support for nVidia MediaShield ata raid found on HP ProLiant ML115. Tested on HEAD and netbsd-4 branch with RAID0, RAID1, SPAN with MCP55 SATA controller.
1.16	10-Jun-2008	drochner	branches: 1.16.2; use new "defpseudodev" where appropriate, and remove dummy interface attribute definitions which were only to trick config(8)
1.15	27-Mar-2007	garbled	branches: 1.15.30; 1.15.32; 1.15.34; 1.15.36; 1.15.38; Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.14	25-Oct-2006	bouyer	branches: 1.14.2; 1.14.4; 1.14.8; 1.14.10; 1.14.12; Exclude wdc_sataprobe() when no SATA controllers are present. Fix build of *_TINY kernels.
1.13	11-Dec-2005	christos	branches: 1.13.20; 1.13.22; merge ktrace-lwp.
1.12	20-Jun-2005	briggs	branches: 1.12.2; Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.11	30-Aug-2004	drochner	branches: 1.11.10; make the software managed bad-sector list optional (kernel option WD_SOFTBADSECT)
1.10	12-Aug-2004	thorpej	Don't assume wdc-specific ATA / ATAPI code should be brought in if atabus or atapibus are configured; use a separate wdc_common attribute to indicate that the shared wdc code is also present.
1.9	28-May-2004	thorpej	Add common routines for SATA controllers. Right now, we have sata_speed(), which reports the speed encoded in the SStatus register.
1.8	08-Oct-2003	bouyer	Following Matt Thomas's request, rename ata attribute to ata_hl, and wdc_base to ata. We can now have atabus* at ata? in kernel config files.
1.7	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.6	27-Jan-2003	thorpej	branches: 1.6.2; Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.5	24-Feb-2001	lukem	whitespace police
1.4	10-Apr-2000	bouyer	Add dependancy to wdc for ata.c, so that 'atapibus at umass' will compile without wdc or pciide. Closes kern/9842 from Gary Duzan.
1.3	12-Oct-1998	bouyer	branches: 1.3.12; Merge bouyer-ide
1.2	16-Feb-1998	thorpej	branches: 1.2.2; Remove "class" declarations, and add "devclass" declarations where appropriate. Fix several inconsistencies between device class and attributes. Mostly from Chris Demetriou.
1.1	14-Jan-1998	cgd	Shuffle wd/wdc driver files around (via repository copies). wdc now lives in dev/ic, wd now lives in dev/ata. there's now a 'ata' interface attribute defined in conf/files, but wdc can't go there yet because some ports still use private versions based on the old ISA version.
1.2.2.2	15-Jun-1998	bouyer	include ata_wdc.c only if wd is defined.
1.2.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.3.12.2	12-Mar-2001	bouyer	Sync with HEAD.
1.3.12.1	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.6.2.6	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.6.2.5	21-Sep-2004	skrll	Fix the sync with head I botched.
1.6.2.4	18-Sep-2004	skrll	Sync with HEAD.
1.6.2.3	03-Sep-2004	skrll	Sync with HEAD
1.6.2.2	12-Aug-2004	skrll	Sync with HEAD.
1.6.2.1	03-Aug-2004	skrll	Sync with HEAD
1.11.10.1	03-Jul-2005	tron	Pull up revision 1.12 (requested by briggs in ticket #521): Basic support for RAID0 and RAID1 for the Adaptec HostRAID format as found on the Intel 6300ESB on-board RAID. Adapted from FreeBSD.
1.12.2.2	03-Sep-2007	yamt	sync with head.
1.12.2.1	30-Dec-2006	yamt	sync with head.
1.13.22.1	10-Dec-2006	yamt	sync with head.
1.13.20.1	18-Nov-2006	ad	Sync with head.
1.14.12.1	29-Mar-2007	reinoud	Pullup to -current
1.14.10.1	11-Jul-2007	mjf	Sync with head.
1.14.8.1	10-Apr-2007	ad	Sync with head.
1.14.4.1	15-Apr-2007	yamt	sync with head.
1.14.2.1	31-Mar-2007	bouyer	Pull up following revision(s) (requested by garbled in ticket #542): sys/dev/ata/ata_raidreg.h: revision 1.4 via patch sys/dev/ata/ata_raidvar.h: revision 1.5via patch sys/dev/ata/files.ata: revision 1.15via patch sys/dev/ata/ata_raid_via.c: revision 1.1 sys/dev/ata/ata_raid.c: revision 1.22via patch Add support for VIA V-TECH ata raid. Tested on 4.0 with RAID0, RAID1 and SPAN over a pair of sata drives on a VT8237A SATA Controller.
1.15.38.1	18-Jun-2008	simonb	Sync with head.
1.15.36.3	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.15.36.2	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.15.36.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.15.34.2	20-Jun-2009	yamt	sync with head
1.15.34.1	04-May-2009	yamt	sync with head.
1.15.32.1	17-Jun-2008	yamt	sync with head.
1.15.30.2	28-Sep-2008	mjf	Sync with HEAD.
1.15.30.1	29-Jun-2008	mjf	Sync with HEAD.
1.16.2.1	19-Oct-2008	haad	Sync with HEAD.
1.20.12.1	21-Apr-2010	matt	sync to netbsd-5
1.20.8.1	23-Jul-2009	jym	Sync with HEAD.
1.20.4.1	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/conf/files: revision 1.947 sys/dev/ata/files.ata: revision 1.21 sys/dev/ata/satafis_subr.c: revision 1.1 sys/dev/ata/satafisreg.h: revision 1.1 sys/dev/ata/satafisvar.h: revision 1.1 ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.21.12.1	30-Oct-2012	yamt	sync with head
1.24.28.1	19-Jul-2017	jdolecek	defflag WD_CHAOS_MONKEY into opt_wd.h together with WD_SOFTBADSECT to set/unset this more easily
1.24.2.1	03-Dec-2017	jdolecek	update from HEAD
1.27.6.1	11-Oct-2018	jdolecek	refactor shared parts of the SATA error recovery into new function ata_recovery_resume() and use for ahcisata/siisata/mvsata, also replace per-controller hold/unhold with generic version move the shared recovery code into separate file ata_recovery.c
1.27.4.1	10-Jun-2019	christos	Sync with HEAD
1.27.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.31.6.1	02-Aug-2025	perseant	Sync with HEAD
1.52	13-Apr-2025	rin	ld(4): Convert blkno argument for sc_dump() to daddr_t PR kern/59153 (1) For backends that accept 64-bit block address, i.e., nvme(4), virtio(4), aac(4), iop(4), and mainbus(usermode/4), this should enable to dump beyond 2Gi blocks. (2) sdmmc(4) backend allows to dump up to the last block. (3) For other backends, block address is handled as `int`. Some of them may support blocks up to 4Gi, but I do not have enough time to examine datasheets. So, continue to reject >2Gi blocks as before. XXX This is KABI change, and cannot be pulled up into netbsd-{10,9}. XXX Compile-test only (for amd64/ALL) due to lack of large SSDs ;) Thanks mlelstv@ for discussion and careful review!!
1.51	16-Mar-2025	andvar	ataraid(4), ld(4): ensure RAID disk status is set before resolving vnode. Given a RAID array disk is missing (due to removal, failure, or a code bug), its configuration is skipped. This causes ld(4) attachment at ataraid(4) to fail because adi->adi_dev is NULL. The fix will treat disks without known status (ADI_S_ONLINE, ADI_S_SPARE, ADI_S_ASSIGNED) as if ata_raid_disk_vnode_find() returned NULL, more gracefully solving above situation and avoiding the crash. Additionally, replace vnode block device opening logic with vn_bdev_open() (vn_subr(9)). This function provides identical functionality while also incrementing vp->v_writecount, preventing a 'vp->v_writecount > 0` assertion failure on vn_close(). More details are available in the tech-kern mailing list thread: the https://mail-index.netbsd.org/tech-kern/2025/03/10/msg030179.html thread and PR thread. Should fix PR kern/43986 Also observed in PR kern/59130 (due to a different bug). pullups are needed to netbsd-9, netbsd-10.
1.50	17-Jan-2020	ad	branches: 1.50.24; 1.50.30; Acquire kernel_lock in the bp->b_iodone callback.
1.49	27-Jan-2019	pgoyette	branches: 1.49.6; Merge the [pgoyette-compat] branch
1.48	23-Oct-2018	jdolecek	add support for DIOCCACHESYNC (!), and DIOCGCACHE; code adapted from ccd(4) tested with Intel ATA Raid
1.47	22-Oct-2018	jdolecek	no need for <sys/malloc.h> here
1.46	03-Sep-2018	riastradh	Rename min/max -> uimin/uimax for better honesty. These functions are defined on unsigned int. The generic name min/max should not silently truncate to 32 bits on 64-bit systems. This is purely a name change -- no functional change intended. HOWEVER! Some subsystems have #define min(a, b) ((a) < (b) ? (a) : (b)) #define max(a, b) ((a) > (b) ? (a) : (b)) even though our standard name for that is MIN/MAX. Although these may invite multiple evaluation bugs, these do _not_ cause integer truncation. To avoid `fixing' these cases, I first changed the name in libkern, and then compile-tested every file where min/max occurred in order to confirm that it failed -- and thus confirm that nothing shadowed min/max -- before changing it. I have left a handful of bootloaders that are too annoying to compile-test, and some dead code: cobalt ews4800mips hp300 hppa ia64 luna68k vax acorn32/if_ie.c (not included in any kernels) macppc/if_gm.c (superseded by gem(4)) It should be easy to fix the fallout once identified -- this way of doing things fails safe, and the goal here, after all, is to _avoid_ silent integer truncations, not introduce them. Maybe one day we can reintroduce min/max as type-generic things that never silently truncate. But we should avoid doing that for a while, so that existing code has a chance to be detected by the compiler for conversion to uimin/uimax without changing the semantics until we can properly audit it all. (Who knows, maybe in some cases integer truncation is actually intended!)
1.45	03-Jun-2018	martin	branches: 1.45.2; 1.45.4; Whitespace
1.44	27-Sep-2016	pgoyette	branches: 1.44.14; Remove some debug printf's that slipped through. Thanks, wiz!
1.43	27-Sep-2016	pgoyette	Modularize the ld driver and all of its attachments. Ensure that all parents are capable of rescan (or otherwise provide a means of attaching children post-initialization).
1.42	16-Sep-2016	jdolecek	modify ldattach() to have default strategy as a parameter
1.41	02-May-2016	christos	branches: 1.41.2; move scsipi_strvis -> libkern:strnvisx() change the prototype to match userland fix sizes of strings passed to it
1.40	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.39	02-Feb-2012	tls	branches: 1.39.6; 1.39.24; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.38	12-Jun-2011	rmind	branches: 1.38.2; 1.38.6; Welcome to 5.99.53! Merge rmind-uvmplock branch: - Reorganize locking in UVM and provide extra serialisation for pmap(9). New lock order: [vmpage-owner-lock] -> pmap-lock. - Simplify locking in some pmap(9) modules by removing P->V locking. - Use lock object on vmobjlock (and thus vnode_t::v_interlock) to share the locks amongst UVM objects where necessary (tmpfs, layerfs, unionfs). - Rewrite and optimise x86 TLB shootdown code, make it simpler and cleaner. Add TLBSTATS option for x86 to collect statistics about TLB shootdowns. - Unify /dev/mem et al in MI code and provide required locking (removes kernel-lock on some ports). Also, avoid cache-aliasing issues. Thanks to Andrew Doran and Joerg Sonnenberger, as their initial patches formed the core changes of this branch.
1.37	06-Jul-2010	bsh	branches: 1.37.6; * use correct interleave factor for RAID-0 of Intel MatrixRaid. * report stripe size in Kibytes to bioctl(8). Unfortunately this change will break compatibility with existing filesystem made by NetBSD on RAID-0 Intel MatrixRAID volumes. Please read: http://mail-index.netbsd.org/current-users/2010/07/02/msg013810.html
1.36	06-Jul-2010	bsh	fix LOCKDEBUG panic at ataraid(4). the patch was provided by Juan RP in response to PR kern 38273. (http://mail-index.netbsd.org/tech-kern/2008/09/17/msg002734.html)
1.35	12-May-2009	cegger	branches: 1.35.2; 1.35.4; struct device * -> device_t, no functional changes intended.
1.34	12-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.33	15-Oct-2008	wrstuden	branches: 1.33.8; Merge wrstuden-revivesa into HEAD.
1.32	16-Sep-2008	tron	Support multiple volumes connected to Intel MatrixRAID controllers. Code contributed by Juan RP in PR kern/39552.
1.31	15-Sep-2008	tron	Only report volume names if they are non empty. Patch provided by Juan RP in PR kern/39514.
1.30	15-Sep-2008	tron	Add support for status reports via bio(4) to ataraid(4). The code was contributed by Juan RP in PR kern/39514.
1.29	09-Sep-2008	tron	Complete device_t/softc split for the ld(4) attachments. This should prevent crashes while attaching a drive. Patch supplied by Juan RP in PR kern/39468.
1.28	24-Aug-2008	hannken	Destroy cb_buf before pool_put(). Resolves PR #38636
1.27	04-May-2008	xtraeme	branches: 1.27.2; 1.27.6; device_t/softc split.
1.26	05-Apr-2008	cegger	branches: 1.26.2; 1.26.4; use aprint_*_dev and device_xname
1.25	21-Mar-2008	ad	Catch up with descriptor handling changes. See kern_descrip.c revision 1.173 for details.
1.24	02-Jan-2008	ad	branches: 1.24.6; Another driver that should use getiobuf().
1.23	02-Jan-2008	ad	Merge vmlocking2 to head.
1.22	26-Nov-2007	pooka	branches: 1.22.2; 1.22.6; Remove the "struct lwp *" argument from all VFS and VOP interfaces. The general trend is to remove it from all kernel interfaces and this is a start. In case the calling lwp is desired, curlwp should be used. quick consensus on tech-kern
1.21	29-Jul-2007	ad	branches: 1.21.4; 1.21.6; 1.21.12; 1.21.14; It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.20	12-Mar-2007	ad	branches: 1.20.8; Pass an ipl argument to pool_init/POOL_INIT to be used when initializing the pool's lock.
1.19	04-Mar-2007	christos	branches: 1.19.2; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.18	16-Nov-2006	christos	branches: 1.18.4; __unused removal on arguments; approved by core.
1.17	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.16	14-May-2006	elad	branches: 1.16.8; 1.16.10; integrate kauth.
1.15	28-Mar-2006	thorpej	Use device_unit().
1.14	11-Dec-2005	christos	branches: 1.14.4; 1.14.6; 1.14.8; 1.14.10; 1.14.12; merge ktrace-lwp.
1.13	28-Oct-2004	yamt	branches: 1.13.12; move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.12	17-Sep-2004	enami	Minimal support for mirror component. The relation between original and mirror component learned from FreeBSD driver.
1.11	22-Apr-2004	itojun	sprintf -> snprintf
1.10	25-Jan-2004	hannken	Make VOP_STRATEGY(bp) a real VOP as discussed on tech-kern. VOP_STRATEGY(bp) is replaced by one of two new functions: - VOP_STRATEGY(vp, bp) Call the strategy routine of vp for bp. - DEV_STRATEGY(bp) Call the d_strategy routine of bp->b_dev for bp. DEV_STRATEGY(bp) is used only for block-to-block device situations.
1.9	14-Dec-2003	thorpej	Cosmetic changes in some autoconfiguration messages.
1.8	14-Jul-2003	lukem	add missing __KERNEL_RCSID()
1.7	29-Jun-2003	fvdl	branches: 1.7.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.6	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.5	07-Jun-2003	thorpej	Always queue transfers in the ld driver to prevent transient resource shortages in the back-ends from aborting transfers. Reviewed by ad@.
1.4	25-Feb-2003	thorpej	Add a new BUF_INIT() macro which initializes b_dep and b_interlock, and use it. This fixes a few places where either b_dep or b_interlock were not properly initialized.
1.3	29-Jan-2003	thorpej	Don't set LDF_ENABLED unless the device is really okay. Noted by Andrew Doran.
1.2	27-Jan-2003	thorpej	G/C something left over from an earlier version of this code.
1.1	27-Jan-2003	thorpej	Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.7.2.5	02-Nov-2004	skrll	Sync with HEAD.
1.7.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.7.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.7.2.2	03-Aug-2004	skrll	Sync with HEAD
1.7.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.13.12.6	24-Mar-2008	yamt	sync with head.
1.13.12.5	21-Jan-2008	yamt	sync with head
1.13.12.4	07-Dec-2007	yamt	sync with head
1.13.12.3	03-Sep-2007	yamt	sync with head.
1.13.12.2	30-Dec-2006	yamt	sync with head.
1.13.12.1	21-Jun-2006	yamt	sync with head.
1.14.12.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.14.12.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.14.10.2	06-May-2006	christos	- Move kauth_cred_t declaration to <sys/types.h> - Cleanup struct ucred; forward declarations that are unused. - Don't include <sys/kauth.h> in any header, but include it in the c files that need it. Approved by core.
1.14.10.1	19-Apr-2006	elad	sync with head.
1.14.8.2	24-May-2006	yamt	sync with head.
1.14.8.1	01-Apr-2006	yamt	sync with head.
1.14.6.2	01-Jun-2006	kardel	Sync with head.
1.14.6.1	22-Apr-2006	simonb	Sync with head.
1.14.4.1	09-Sep-2006	rpaulo	sync with head
1.16.10.2	10-Dec-2006	yamt	sync with head.
1.16.10.1	22-Oct-2006	yamt	sync with head
1.16.8.1	18-Nov-2006	ad	Sync with head.
1.18.4.2	24-Mar-2007	yamt	sync with head.
1.18.4.1	12-Mar-2007	rmind	Sync with HEAD.
1.19.2.6	09-Oct-2007	ad	Sync with head.
1.19.2.5	24-Aug-2007	ad	Sync with buffer cache locking changes. See buf.h/vfs_bio.c for details. Some minor portions are incomplete and needs to be verified as a whole.
1.19.2.4	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.19.2.3	17-Jun-2007	ad	- Increase the number of thread priorities from 128 to 256. How the space is set up is to be revisited. - Implement soft interrupts as kernel threads. A generic implementation is provided, with hooks for fast-path MD code that can run the interrupt threads over the top of other threads executing in the kernel. - Split vnode::v_flag into three fields, depending on how the flag is locked (by the interlock, by the vnode lock, by the file system). - Miscellaneous locking fixes and improvements.
1.19.2.2	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.19.2.1	13-Mar-2007	ad	Sync with head.
1.20.8.1	15-Aug-2007	skrll	Sync with HEAD.
1.21.14.2	29-Jul-2007	ad	It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.21.14.1	29-Jul-2007	ad	file ld_ataraid.c was added on branch matt-mips64 on 2007-07-29 12:50:20 +0000
1.21.12.2	18-Feb-2008	mjf	Sync with HEAD.
1.21.12.1	08-Dec-2007	mjf	Sync with HEAD.
1.21.6.1	09-Jan-2008	matt	sync with HEAD
1.21.4.1	27-Nov-2007	joerg	Sync with HEAD. amd64 Xen support needs testing.
1.22.6.2	08-Jan-2008	bouyer	Sync with HEAD
1.22.6.1	02-Jan-2008	bouyer	Sync with HEAD
1.22.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.24.6.4	17-Jan-2009	mjf	Sync with HEAD.
1.24.6.3	28-Sep-2008	mjf	Sync with HEAD.
1.24.6.2	02-Jun-2008	mjf	Sync with HEAD.
1.24.6.1	03-Apr-2008	mjf	Sync with HEAD.
1.26.4.4	11-Aug-2010	yamt	sync with head.
1.26.4.3	16-May-2009	yamt	sync with head
1.26.4.2	04-May-2009	yamt	sync with head.
1.26.4.1	16-May-2008	yamt	sync with head.
1.26.2.1	18-May-2008	yamt	sync with head.
1.27.6.1	19-Oct-2008	haad	Sync with HEAD.
1.27.2.2	24-Sep-2008	wrstuden	Merge in changes between wrstuden-revivesa-base-2 and wrstuden-revivesa-base-3.
1.27.2.1	18-Sep-2008	wrstuden	Sync with wrstuden-revivesa-base-2.
1.33.8.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.35.4.2	05-Mar-2011	rmind	sync with head
1.35.4.1	16-Mar-2010	rmind	Change struct uvm_object::vmobjlock to be dynamically allocated with mutex_obj_alloc(). It allows us to share the locks among UVM objects.
1.35.2.1	17-Aug-2010	uebayasi	Sync with HEAD.
1.37.6.1	23-Jun-2011	cherry	Catchup with rmind-uvmplock merge.
1.38.6.1	18-Feb-2012	mrg	merge to -current.
1.38.2.1	17-Apr-2012	yamt	sync with head
1.39.24.3	05-Oct-2016	skrll	Sync with HEAD
1.39.24.2	29-May-2016	skrll	Sync with HEAD
1.39.24.1	06-Jun-2015	skrll	Sync with HEAD
1.39.6.1	03-Dec-2017	jdolecek	update from HEAD
1.41.2.1	04-Nov-2016	pgoyette	Sync with HEAD
1.44.14.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.44.14.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.44.14.1	25-Jun-2018	pgoyette	Sync with HEAD
1.45.4.1	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.45.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.45.2.1	10-Jun-2019	christos	Sync with HEAD
1.49.6.1	17-Jan-2020	ad	Sync with head.
1.50.30.1	02-Aug-2025	perseant	Sync with HEAD
1.50.24.1	07-May-2025	martin	Pull up following revision(s) (requested by andvar in ticket #1100): sys/dev/ata/ld_ataraid.c: revision 1.51 sys/dev/ata/ata_raid_subr.c: revision 1.5 ataraid(4), ld(4): ensure RAID disk status is set before resolving vnode. Given a RAID array disk is missing (due to removal, failure, or a code bug), its configuration is skipped. This causes ld(4) attachment at ataraid(4) to fail because adi->adi_dev is NULL. The fix will treat disks without known status (ADI_S_ONLINE, ADI_S_SPARE, ADI_S_ASSIGNED) as if ata_raid_disk_vnode_find() returned NULL, more gracefully solving above situation and avoiding the crash. Additionally, replace vnode block device opening logic with vn_bdev_open() (vn_subr(9)). This function provides identical functionality while also incrementing vp->v_writecount, preventing a 'vp->v_writecount > 0` assertion failure on vn_close(). More details are available in the tech-kern mailing list thread: the https://mail-index.netbsd.org/tech-kern/2025/03/10/msg030179.html thread and PR thread. Should fix PR kern/43986 Also observed in PR kern/59130 (due to a different bug).
1.24	21-Jun-2018	jdolecek	split the port status reporting to new function sata_interpret_det() so it can be called separately from sata_reset_interface() do not treat PHY offline as an error, it's pretty normal when there is no device actually connected debugging aid for PR kern/52372
1.23	07-Oct-2017	jdolecek	branches: 1.23.2; Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.22	10-May-2017	msaitoh	branches: 1.22.2; KNF. No functional change.
1.21	03-Apr-2013	bouyer	branches: 1.21.12; 1.21.24; 1.21.26; Fix kernel dump on ahci controller, by making sure we won't sleep while dumping: - introduce ata_delay() which either use delay() or kpause() depending on flags. use it in sata_reset_interface() and some ahci functions - kill ATA_NOSLEEP, it was tested but never set. use ATA_POLL instead. - reduce delay while polling in ahci, to speed up the dump Should fix PR kern/41095
1.20	31-Jul-2012	bouyer	branches: 1.20.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.19	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.18	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.17	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.16	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.15	15-May-2012	bouyer	If the PHY reports link up but no communication, wait 5 more seconds for communications to be established. This seems to be needed for some configurations.
1.14	12-Dec-2010	jakllsch	branches: 1.14.8; 1.14.12; Actually, 31244 does not refer to the Silicon Image SiI3124, but the now-obscure Intel GD31244.
1.13	11-Dec-2010	matt	Fix typo (31224 -> 3124) and constify an array.
1.12	18-Nov-2008	cegger	branches: 1.12.8; SATA GEN3 speed is 6.0Gb/s
1.11	28-Apr-2008	martin	branches: 1.11.6; 1.11.8; 1.11.10; Remove clause 3 and 4 from TNF licenses
1.10	18-Mar-2008	cube	branches: 1.10.2; 1.10.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.9	11-Dec-2007	lukem	branches: 1.9.4; 1.9.8; use __KERNEL_RCSID()
1.8	30-Nov-2006	bouyer	branches: 1.8.24; 1.8.34; 1.8.36; 1.8.38; It's 3.0Gb/s, not GB/s. Pointed out by Thomas E. Spanjaard.
1.7	27-Nov-2006	dogcow	revert last change, as havard's fix is more general.
1.6	22-Nov-2006	dogcow	#include <sys/device.h> to fix build failure on sparc64. (why only sparc64? no idea.)
1.5	22-Nov-2006	bouyer	Cosmetic: remove one ":" from the port status string, it looks better this way.
1.4	22-Nov-2006	bouyer	In sata_reset_interface() return only DET bits of SStatus, as the return value is compared against SStatus_DET_*
1.3	20-Nov-2006	bouyer	Move part of wdc_sataprobe() to sys/dev/ata/sata_subr.c so that it can be shared with non-wdc SATA controllers.
1.2	11-Dec-2005	christos	branches: 1.2.20; 1.2.22; merge ktrace-lwp.
1.1	28-May-2004	thorpej	branches: 1.1.2; 1.1.14; Add common routines for SATA controllers. Right now, we have sata_speed(), which reports the speed encoded in the SStatus register.
1.1.14.3	24-Mar-2008	yamt	sync with head.
1.1.14.2	21-Jan-2008	yamt	sync with head
1.1.14.1	30-Dec-2006	yamt	sync with head.
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	03-Aug-2004	skrll	Sync with HEAD
1.1.2.1	28-May-2004	skrll	file sata_subr.c was added on branch ktrace-lwp on 2004-08-03 10:45:46 +0000
1.2.22.1	10-Dec-2006	yamt	sync with head.
1.2.20.1	12-Jan-2007	ad	Sync with head.
1.8.38.1	13-Dec-2007	bouyer	Sync with HEAD
1.8.36.1	11-Dec-2007	yamt	sync with head.
1.8.34.1	26-Dec-2007	ad	Sync with head.
1.8.24.2	23-Mar-2008	matt	sync with HEAD
1.8.24.1	09-Jan-2008	matt	sync with HEAD
1.9.8.3	17-Jan-2009	mjf	Sync with HEAD.
1.9.8.2	02-Jun-2008	mjf	Sync with HEAD.
1.9.8.1	03-Apr-2008	mjf	Sync with HEAD.
1.9.4.1	24-Mar-2008	keiichi	sync with head.
1.10.4.2	04-May-2009	yamt	sync with head.
1.10.4.1	16-May-2008	yamt	sync with head.
1.10.2.1	18-May-2008	yamt	sync with head.
1.11.10.1	22-Nov-2008	snj	branches: 1.11.10.1.8; Pull up following revision(s) (requested by cegger in ticket #99): sys/dev/ata/sata_subr.c: revision 1.12 SATA GEN3 speed is 6.0Gb/s
1.11.10.1.8.1	07-Jan-2011	matt	small fixes (constification + comment type).
1.11.8.1	19-Jan-2009	skrll	Sync with HEAD.
1.11.6.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.12.8.1	05-Mar-2011	rmind	sync with head
1.14.12.1	02-Jun-2012	mrg	sync to latest -current.
1.14.8.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.14.8.2	30-Oct-2012	yamt	sync with head
1.14.8.1	23-May-2012	yamt	sync with head.
1.20.2.2	03-Dec-2017	jdolecek	update from HEAD
1.20.2.1	23-Jun-2013	tls	resync from head
1.21.26.1	11-May-2017	pgoyette	Sync with HEAD
1.21.24.1	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.21.12.1	28-Aug-2017	skrll	Sync with HEAD
1.22.2.1	24-Feb-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1805): sys/dev/ata/satavar.h: revision 1.10 sys/dev/ata/sata_subr.c: revision 1.24 split the port status reporting to new function sata_interpret_det() so it can be called separately from sata_reset_interface() do not treat PHY offline as an error, it's pretty normal when there is no device actually connected debugging aid for PR kern/52372
1.23.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.9	15-Aug-2019	gson	Set the ATAPI "BYTE COUNT LIMIT" field in the SATA case like we already do in the ATA case, to make NetBSD work with the virtual SATA CD-ROM of "qemu-system-i386 -machine q35". Fixes PR kern/54389. OK mlelstv.
1.8	07-Oct-2017	jdolecek	branches: 1.8.4; 1.8.10; Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.7	22-Jul-2012	jakllsch	branches: 1.7.2; 1.7.28; SATA uses the head portion of the device register in the command protocol for port multiplier register access. Thus, in the LBA48 case, pass the complete r(hd\|dh)_dh byte back and forth to the upper layers. (This is irrelevant in the LBA28/CHS code path for what should be obvious reasons.)
1.6	24-Jan-2012	jakllsch	Rework struct ata_command to support LBA28 and LBA48-protocol commands. Implement 28- and 48-bit command issuance and parameter read-back in the various ATA host drivers. Add LBA28-protocol support to ATAIOCCOMMAND ioctl.
1.5	07-Apr-2010	jakllsch	branches: 1.5.2; 1.5.10; 1.5.14; satafis: - Add function to parse RDH FIS for use in implementing AT_READREG. - Correct and clean up some structure definitions. - Sprinkle a bit of const. - Remove dependency on <dev/ic/wdcreg.h>, WDCTL_4BIT doesn't seem to be specified by any recent ATA standard, and it seems to make no difference in practice. - Stop using WDSD_IBM, these bits have been obsolete since before SATA. siisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>. - Whitespace and slight debug code cleanup. - Some possibly-uncessary code reordering. ahcisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>.
1.4	19-Oct-2009	bouyer	branches: 1.4.2; 1.4.4; Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.3	04-Jul-2009	jakllsch	branches: 1.3.2; Further corrections for siisata. Reset ch_status and ch_error at the beginning of any operation. The chip only writes the RDH FIS to the SRAM when the error bit is set in a RDH FIS (per the datasheet). Thus, satafis_sdb_parse() is replaced with satafis_rhd_parse(). A valid FIS type field seems to not be written to the SRAM, so don't bother checking it for the correct magic. Additionally, append 2009 to copyright year set, and remove obsolete CVS Id lines from when this was in my repository.
1.2	17-Jun-2009	jakllsch	branches: 1.2.2; Use abstracted FIS code for ahcisata(4). Have the FIS code pre-zero the FIS.
1.1	17-Jun-2009	jakllsch	ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.2.2.5	11-Aug-2010	yamt	sync with head.
1.2.2.4	11-Mar-2010	yamt	sync with head
1.2.2.3	18-Jul-2009	yamt	sync with head.
1.2.2.2	20-Jun-2009	yamt	sync with head
1.2.2.1	17-Jun-2009	yamt	file satafis_subr.c was added on branch yamt-nfs-mp on 2009-06-20 07:20:20 +0000
1.3.2.4	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ata/satafis_subr.c: revision 1.3 sys/dev/ata/satafisvar.h: revision 1.2 sys/dev/ic/siisata.c: revision 1.7 sys/dev/ic/siisatareg.h: revision 1.4 Further corrections for siisata. Reset ch_status and ch_error at the beginning of any operation. The chip only writes the RDH FIS to the SRAM when the error bit is set in a RDH FIS (per the datasheet). Thus, satafis_sdb_parse() is replaced with satafis_rhd_parse(). A valid FIS type field seems to not be written to the SRAM, so don't bother checking it for the correct magic. Additionally, append 2009 to copyright year set, and remove obsolete CVS Id lines from when this was in my repository.
1.3.2.3	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ata/satafis_subr.c: revision 1.2 Have the FIS code pre-zero the FIS.
1.3.2.2	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/conf/files: revision 1.947 sys/dev/ata/files.ata: revision 1.21 sys/dev/ata/satafis_subr.c: revision 1.1 sys/dev/ata/satafisreg.h: revision 1.1 sys/dev/ata/satafisvar.h: revision 1.1 ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.3.2.1	04-Jul-2009	snj	file satafis_subr.c was added on branch netbsd-5 on 2009-09-28 00:13:02 +0000
1.4.4.1	30-May-2010	rmind	sync with head
1.4.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.5.14.1	18-Feb-2012	mrg	merge to -current.
1.5.10.2	30-Oct-2012	yamt	sync with head
1.5.10.1	17-Apr-2012	yamt	sync with head
1.5.2.3	24-Mar-2014	matt	Merge satafis changes from HEAD for ahcisata
1.5.2.2	21-Apr-2010	matt	sync to netbsd-5
1.5.2.1	07-Apr-2010	matt	file satafis_subr.c was added on branch matt-nb5-mips64 on 2010-04-21 00:27:34 +0000
1.7.28.3	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.7.28.2	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.7.28.1	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.7.2.1	03-Dec-2017	jdolecek	update from HEAD
1.8.10.1	05-Sep-2019	martin	Pull up following revision(s) (requested by gson in ticket #179): sys/dev/ata/satafis_subr.c: revision 1.9 Set the ATAPI "BYTE COUNT LIMIT" field in the SATA case like we already do in the ATA case, to make NetBSD work with the virtual SATA CD-ROM of "qemu-system-i386 -machine q35". Fixes PR kern/54389. OK mlelstv.
1.8.4.1	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.4	24-Jan-2012	jakllsch	Rework struct ata_command to support LBA28 and LBA48-protocol commands. Implement 28- and 48-bit command issuance and parameter read-back in the various ATA host drivers. Add LBA28-protocol support to ATAIOCCOMMAND ioctl.
1.3	02-Nov-2011	jakllsch	branches: 1.3.4; Drop trailing whitespace.
1.2	07-Apr-2010	jakllsch	branches: 1.2.2; 1.2.10; satafis: - Add function to parse RDH FIS for use in implementing AT_READREG. - Correct and clean up some structure definitions. - Sprinkle a bit of const. - Remove dependency on <dev/ic/wdcreg.h>, WDCTL_4BIT doesn't seem to be specified by any recent ATA standard, and it seems to make no difference in practice. - Stop using WDSD_IBM, these bits have been obsolete since before SATA. siisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>. - Whitespace and slight debug code cleanup. - Some possibly-uncessary code reordering. ahcisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>.
1.1	17-Jun-2009	jakllsch	branches: 1.1.2; 1.1.4; 1.1.6; 1.1.8; ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.1.8.1	30-May-2010	rmind	sync with head
1.1.6.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.1.4.2	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/conf/files: revision 1.947 sys/dev/ata/files.ata: revision 1.21 sys/dev/ata/satafis_subr.c: revision 1.1 sys/dev/ata/satafisreg.h: revision 1.1 sys/dev/ata/satafisvar.h: revision 1.1 ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.1.4.1	17-Jun-2009	snj	file satafisreg.h was added on branch netbsd-5 on 2009-09-28 00:13:02 +0000
1.1.2.3	11-Aug-2010	yamt	sync with head.
1.1.2.2	20-Jun-2009	yamt	sync with head
1.1.2.1	17-Jun-2009	yamt	file satafisreg.h was added on branch yamt-nfs-mp on 2009-06-20 07:20:20 +0000
1.2.10.2	17-Apr-2012	yamt	sync with head
1.2.10.1	10-Nov-2011	yamt	sync with head
1.2.2.3	24-Mar-2014	matt	Merge satafis changes from HEAD for ahcisata
1.2.2.2	21-Apr-2010	matt	sync to netbsd-5
1.2.2.1	07-Apr-2010	matt	file satafisreg.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:34 +0000
1.3.4.1	18-Feb-2012	mrg	merge to -current.
1.4	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.3	07-Apr-2010	jakllsch	branches: 1.3.2; 1.3.20; 1.3.50; satafis: - Add function to parse RDH FIS for use in implementing AT_READREG. - Correct and clean up some structure definitions. - Sprinkle a bit of const. - Remove dependency on <dev/ic/wdcreg.h>, WDCTL_4BIT doesn't seem to be specified by any recent ATA standard, and it seems to make no difference in practice. - Stop using WDSD_IBM, these bits have been obsolete since before SATA. siisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>. - Whitespace and slight debug code cleanup. - Some possibly-uncessary code reordering. ahcisata: - Implement AT_READREG. - Reap now-unneeded <dev/ic/wdcreg.h>.
1.2	04-Jul-2009	jakllsch	branches: 1.2.2; 1.2.4; 1.2.6; Further corrections for siisata. Reset ch_status and ch_error at the beginning of any operation. The chip only writes the RDH FIS to the SRAM when the error bit is set in a RDH FIS (per the datasheet). Thus, satafis_sdb_parse() is replaced with satafis_rhd_parse(). A valid FIS type field seems to not be written to the SRAM, so don't bother checking it for the correct magic. Additionally, append 2009 to copyright year set, and remove obsolete CVS Id lines from when this was in my repository.
1.1	17-Jun-2009	jakllsch	branches: 1.1.2; ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.1.2.4	11-Aug-2010	yamt	sync with head.
1.1.2.3	18-Jul-2009	yamt	sync with head.
1.1.2.2	20-Jun-2009	yamt	sync with head
1.1.2.1	17-Jun-2009	yamt	file satafisvar.h was added on branch yamt-nfs-mp on 2009-06-20 07:20:20 +0000
1.2.6.1	30-May-2010	rmind	sync with head
1.2.4.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.2.2.3	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/dev/ata/satafis_subr.c: revision 1.3 sys/dev/ata/satafisvar.h: revision 1.2 sys/dev/ic/siisata.c: revision 1.7 sys/dev/ic/siisatareg.h: revision 1.4 Further corrections for siisata. Reset ch_status and ch_error at the beginning of any operation. The chip only writes the RDH FIS to the SRAM when the error bit is set in a RDH FIS (per the datasheet). Thus, satafis_sdb_parse() is replaced with satafis_rhd_parse(). A valid FIS type field seems to not be written to the SRAM, so don't bother checking it for the correct magic. Additionally, append 2009 to copyright year set, and remove obsolete CVS Id lines from when this was in my repository.
1.2.2.2	28-Sep-2009	snj	Pull up following revision(s) (requested by jakllsch in ticket #1023): sys/conf/files: revision 1.947 sys/dev/ata/files.ata: revision 1.21 sys/dev/ata/satafis_subr.c: revision 1.1 sys/dev/ata/satafisreg.h: revision 1.1 sys/dev/ata/satafisvar.h: revision 1.1 ahcisata(4) and siisata(4) use similar SATA FIS functions, share them.
1.2.2.1	04-Jul-2009	snj	file satafisvar.h was added on branch netbsd-5 on 2009-09-28 00:13:02 +0000
1.3.50.1	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.3.20.1	03-Dec-2017	jdolecek	update from HEAD
1.3.2.3	24-Mar-2014	matt	Merge satafis changes from HEAD for ahcisata
1.3.2.2	21-Apr-2010	matt	sync to netbsd-5
1.3.2.1	07-Apr-2010	matt	file satafisvar.h was added on branch matt-nb5-mips64 on 2010-04-21 00:27:34 +0000
1.16	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.15	22-Oct-2018	jdolecek	branches: 1.15.4; 1.15.10; Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.14	29-Jul-2018	jdolecek	branches: 1.14.2; mark satapmp_rescan() static, it's not to be used outside satapmp_subr.c
1.13	07-Oct-2017	jdolecek	branches: 1.13.2; 1.13.4; Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.12	03-May-2013	jakllsch	branches: 1.12.24; Wait for DRDY after the PMP commands and don't wait before. Should fix the worst part of PR kern/47793.
1.11	17-Apr-2013	jakllsch	Also print port multiplier port number when register accesses fail.
1.10	01-Aug-2012	bouyer	branches: 1.10.2; 1.10.4; Make this compile again after DRIVET rename
1.9	01-Aug-2012	bouyer	Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.8	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.7	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.6	22-Jul-2012	jakllsch	Port multiplier registers are 64-bits wide (although the ones we need only implement 32-bits). Implement the access routines using 64-bit values and add 32-bit wrappers thereto.
1.5	22-Jul-2012	jakllsch	The vendor and product IDs are 16-bit, print out as such.
1.4	22-Jul-2012	jakllsch	Correctly issue port multiplier register access commands as LBA48 now that the FIS code doesn't get in the way of doing that.
1.3	22-Jul-2012	jakllsch	KASSERT if we attempt to access an invalid port. Also, small whitespace change for internal consistency.
1.2	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.1	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.10.4.3	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.10.4.2	30-Oct-2012	yamt	sync with head
1.10.4.1	01-Aug-2012	yamt	file satapmp_subr.c was added on branch yamt-pagecache on 2012-10-30 17:20:53 +0000
1.10.2.2	03-Dec-2017	jdolecek	update from HEAD
1.10.2.1	23-Jun-2013	tls	resync from head
1.12.24.6	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.12.24.5	26-Jul-2017	jdolecek	write the device and channel for port multiplier attach (not just the atabus), so it's easier to check
1.12.24.4	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.12.24.3	19-Jun-2017	jdolecek	add ata_channel lock, use it to protect queue manipulation (only that for now); add ata_channel_detach() to destroy the locks change ata_get_xfer() so that it can wait for xfer, convert all on-stack xfer code to use the blocking variant fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(), convert drive probe code also over from slot0 XXX to ata_get_xfer() drive reset and PMP now works on siisata(4) too; changes tested also on piixide(4), ahci(4), mvsata(4)
1.12.24.2	17-Jun-2017	jdolecek	make PMP working great again tested with mvsata(4), my ahcisata(4) controller unfortunately doesn't support PMP
1.12.24.1	15-Apr-2017	jdolecek	pass also ata_command via ata_xfer, callers of ata_exec_command() is now responsible for allocation/disposal of the structure change code to allocate ata_xfer for commands on stack same way as previously the ata_command were, using c_slot 0; adjust asserts so that it would allow several xfers with same c_slot, as long as only one such transfer is active at a time
1.13.4.2	21-Apr-2020	martin	Sync with HEAD
1.13.4.1	10-Jun-2019	christos	Sync with HEAD
1.13.2.2	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.13.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.14.2.2	20-Oct-2018	jdolecek	hold lock during drive reset (new locking protocol) erase xfer before each reuse for read/write port commands
1.14.2.1	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.15.10.1	20-Apr-2020	bouyer	Sync with HEAD
1.15.4.1	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.6	31-Jul-2012	bouyer	Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.5	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.4	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.3	28-Apr-2008	martin	branches: 1.3.34; Remove clause 3 and 4 from TNF licenses
1.2	11-Dec-2005	christos	branches: 1.2.70; 1.2.72; 1.2.74; merge ktrace-lwp.
1.1	21-Aug-2004	thorpej	branches: 1.1.2; Register description for SATA port multipliers.
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	25-Aug-2004	skrll	Sync with HEAD.
1.1.2.1	21-Aug-2004	skrll	file satapmpreg.h was added on branch ktrace-lwp on 2004-08-25 06:57:34 +0000
1.2.74.1	16-May-2008	yamt	sync with head.
1.2.72.1	18-May-2008	yamt	sync with head.
1.2.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.3.34.1	30-Oct-2012	yamt	sync with head
1.5	29-Jul-2018	jdolecek	mark satapmp_rescan() static, it's not to be used outside satapmp_subr.c
1.4	07-Oct-2017	jdolecek	branches: 1.4.2; 1.4.4; Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.3	01-Aug-2012	bouyer	branches: 1.3.2; 1.3.4; 1.3.30; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.2	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.1	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.3.30.1	19-Sep-2017	jdolecek	replace all remaining tsleep()/wakeup() calls with condition variables, or calls to ata_delay(), as appropriate; change ata_delay() to require the channel lock on entry, and pass the lock to kpause() for unlocking while sleeping
1.3.4.2	30-Oct-2012	yamt	sync with head
1.3.4.1	01-Aug-2012	yamt	file satapmpvar.h was added on branch yamt-pagecache on 2012-10-30 17:20:53 +0000
1.3.2.1	03-Dec-2017	jdolecek	update from HEAD
1.4.4.1	10-Jun-2019	christos	Sync with HEAD
1.4.2.1	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.7	16-Feb-2022	andvar	fix various typos, mainly in comments.
1.6	27-Dec-2020	jmcneill	Add G3 and DevSleep definitions. This changes the mask used by SControl_IPM_NONE from 0x3 to 0x7.
1.5	28-Apr-2008	martin	branches: 1.5.94; 1.5.102; Remove clause 3 and 4 from TNF licenses
1.4	11-Dec-2005	christos	branches: 1.4.70; 1.4.72; 1.4.74; merge ktrace-lwp.
1.3	23-May-2004	wiz	branches: 1.3.2; Fix typo reported by Alexander Yurchenko grange at rt mipt ru in private mail.
1.2	17-Dec-2003	thorpej	Some additional SATA-II registers.
1.1	14-Dec-2003	thorpej	Serial ATA register definitions.
1.3.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.3.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.3.2.2	03-Aug-2004	skrll	Sync with HEAD
1.3.2.1	23-May-2004	skrll	file satareg.h was added on branch ktrace-lwp on 2004-08-03 10:45:46 +0000
1.4.74.1	16-May-2008	yamt	sync with head.
1.4.72.1	18-May-2008	yamt	sync with head.
1.4.70.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.102.1	03-Jan-2021	thorpej	Sync w/ HEAD.
1.5.94.1	30-Dec-2020	martin	Pull up following revision(s) (requested by jmcneill in ticket #1167): sys/dev/ic/ahcisata_core.c: revision 1.84 sys/dev/ic/ahcisata_core.c: revision 1.85 sys/dev/ic/ahcisata_core.c: revision 1.88 sys/dev/ic/ahcisata_core.c: revision 1.89 sys/arch/arm/nvidia/tegra_ahcisata.c: revision 1.13 sys/dev/ic/ahcisatavar.h: revision 1.26 sys/dev/ic/ahcisata_core.c: revision 1.90 sys/dev/ic/ahcisata_core.c: revision 1.91 sys/dev/ic/ahcisata_core.c: revision 1.92 sys/dev/ata/satareg.h: revision 1.6 ahci_exec_fis: wait for the correct amount of time when AT_WAIT is set Retry clearing WDCTL_RST a few times before giving up. Makes SATA work in Solidrun Honeycomb LX2K. AHCI 1.3.1 specification says that it is good practice for system software to 'zero-out' the memory allocated and referenced by PxCLB and PxFB. ahci_intr: use ffs in the port bitmask instead of looping over all 32 bits AHCI 1.3.1 section 5.5.3 "Processing Completed Commands" says that we should clear PxIS before IS.IPS. Add G3 and DevSleep definitions. This changes the mask used by SControl_IPM_NONE from 0x3 to 0x7. Make sure to ack IS after PxIS when polling and when using multiple MSI-X messages. Remove the AHCI_QUIRK_SKIP_RESET quirk now that the underlying issue is fixed.
1.10	21-Jun-2018	jdolecek	split the port status reporting to new function sata_interpret_det() so it can be called separately from sata_reset_interface() do not treat PHY offline as an error, it's pretty normal when there is no device actually connected debugging aid for PR kern/52372
1.9	03-Apr-2013	bouyer	branches: 1.9.28; 1.9.34; Fix kernel dump on ahci controller, by making sure we won't sleep while dumping: - introduce ata_delay() which either use delay() or kpause() depending on flags. use it in sata_reset_interface() and some ahci functions - kill ATA_NOSLEEP, it was tested but never set. use ATA_POLL instead. - reduce delay while polling in ahci, to speed up the dump Should fix PR kern/41095
1.8	31-Jul-2012	bouyer	branches: 1.8.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.7	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.6	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.5	28-Apr-2008	martin	branches: 1.5.34; Remove clause 3 and 4 from TNF licenses
1.4	19-Oct-2007	ad	branches: 1.4.16; 1.4.18; 1.4.20; machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.3	20-Nov-2006	bouyer	branches: 1.3.8; 1.3.22; 1.3.24; 1.3.28; Move part of wdc_sataprobe() to sys/dev/ata/sata_subr.c so that it can be shared with non-wdc SATA controllers.
1.2	11-Dec-2005	christos	branches: 1.2.20; 1.2.22; merge ktrace-lwp.
1.1	28-May-2004	thorpej	branches: 1.1.2; 1.1.14; Add common routines for SATA controllers. Right now, we have sata_speed(), which reports the speed encoded in the SStatus register.
1.1.14.2	27-Oct-2007	yamt	sync with head.
1.1.14.1	30-Dec-2006	yamt	sync with head.
1.1.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.1.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.1.2.2	03-Aug-2004	skrll	Sync with HEAD
1.1.2.1	28-May-2004	skrll	file satavar.h was added on branch ktrace-lwp on 2004-08-03 10:45:46 +0000
1.2.22.1	10-Dec-2006	yamt	sync with head.
1.2.20.1	12-Jan-2007	ad	Sync with head.
1.3.28.1	25-Oct-2007	bouyer	Sync with HEAD.
1.3.24.1	06-Nov-2007	matt	sync with HEAD
1.3.22.1	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.3.8.1	23-Oct-2007	ad	Sync with head.
1.4.20.1	16-May-2008	yamt	sync with head.
1.4.18.1	18-May-2008	yamt	sync with head.
1.4.16.1	02-Jun-2008	mjf	Sync with HEAD.
1.5.34.2	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.5.34.1	30-Oct-2012	yamt	sync with head
1.8.2.1	23-Jun-2013	tls	resync from head
1.9.34.1	25-Jun-2018	pgoyette	Sync with HEAD
1.9.28.1	24-Feb-2023	martin	Pull up following revision(s) (requested by msaitoh in ticket #1805): sys/dev/ata/satavar.h: revision 1.10 sys/dev/ata/sata_subr.c: revision 1.24 split the port status reporting to new function sata_interpret_det() so it can be called separately from sata_reset_interface() do not treat PHY offline as an error, it's pretty normal when there is no device actually connected debugging aid for PR kern/52372
1.474	13-Apr-2025	jakllsch	Add physical sector and alignment info to struct disk_geom and the geometry plist, and handle in partutil. Bump version for disk_geom addition. Collect DIOCGSECTORALIGN handling into one place.
1.473	27-Feb-2025	jakllsch	Ensure dsa_firstaligned returned from DIOCGSECTORALIGN is actually the first
1.472	23-Feb-2025	jakllsch	atap_logical_align is already what we want, we don't have to subtract it from the logical blocks per physical block number.
1.471	17-Feb-2025	jakllsch	Fix aprint_error_dev() abuse
1.470	22-Sep-2024	uwe	wd.c: consistent whitespace within one #define group
1.469	05-Feb-2024	andvar	branches: 1.469.2; fix various typos in comments.
1.468	24-Jan-2023	mlelstv	Sanitize capacity values.
1.467	28-Mar-2022	riastradh	wd(4): Use d_cfdriver/devtounit to avoid open/detach races.
1.466	28-Dec-2021	riastradh	wd(4): Fix bugs in softbadsect handling. - Don't copyout kernel virtual addresses (of SLIST entries) that userland won't use anyway. => The structure still has space for this pointer; it's just always null when userland gets it now. - Don't copyout under a lock. - Stop and return error if copyout fails (unless we've already copied some out). - Don't kmem_free under a lock. XXX Unclear whether anyone actually uses WD_SOFTBADSECT or why -- it's always been disabled by default. Maybe we should just remove it?
1.465	28-Sep-2020	jakllsch	fix typo that prevented bytes/physsect reporting from working
1.464	27-Sep-2020	christos	de-quadruplicate, remove unused argument
1.463	24-May-2020	jdolecek	account for already transferred data (partially done I/O) when retrying an xfer, to avoid reading/writing data from/to wrong offset, and eventually beyond the end of data buffer fixes data corruption under QEMU observed by Paul Ripke for emulated IDE drives
1.462	13-Apr-2020	jdolecek	fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson
1.461	13-Apr-2020	maxv	constify
1.460	07-Apr-2020	jdolecek	branches: 1.460.2; drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait()
1.459	02-Mar-2020	riastradh	New ioctl DIOCGSECTORALIGN returns sector alignment parameters. struct disk_sectoralign { /* First aligned sector number. / uint32_t dsa_firstaligned; / Number of sectors per aligned unit. */ uint32_t dsa_alignment; }; - Teach wd(4) to get it from ATA. - Teach cgd(4) to pass it through from the underlying disk. - Teach dk(4) to pass it through with adjustments. - Teach zpool (zfs) to take advantage of it. => XXX zpool doesn't seem to understand when the vdev's starting sector is misaligned. Missing: - ccd(4) and raidframe(4) support -- these should support _using_ DIOCGSECTORALIGN to decide where to start putting ccd or raid stripes on disk, and these should perhaps _implement_ DIOCGSECTORALIGN by reporting the stripe/interleave factor. - sd(4) support -- I don't know any obvious way to get it from SCSI, but if any SCSI wizards know better than I, please feel free to teach sd(4) about it! - any ld(4) attachments -- might be worth teaching the ld drivers for nvme and various raid controllers to get the aligned sector size There's some duplicate logic here for now. I'm doing it this way, rather than gathering the logic into a new disklabel_sectoralign function or something, so that this change is limited to adding a new ioctl, without any new kernel symbols, in order to make it easy to pull up to netbsd-9 without worrying about the module ABI.
1.458	01-Mar-2020	riastradh	Add a flag to dk_dump for virtual disk devices. If a disk is backed by a physical medium other than itself, such as cgd(4), then it passes DK_DUMP_RECURSIVE to disable the recursion detection for dk_dump. If, however, a device represents a physical medium on its own, such as wd(4), then it passes 0 instead. With this, I can now dump to dk on cgd on dk on wd.
1.457	18-Jan-2020	simonb	Tidy quirk table and remove outdated quick from the quirk format string.
1.456	18-Jan-2020	simonb	Revert kern/54790 and kern/54855 NCQ fix that penalised all Samsung EVO 860 drives. ok jdolecek@
1.455	14-Jan-2020	jdolecek	enable the BAD_NCQ quirk for all 860 EVO drives XXX work-in-progress, it's not clear whether this is driver or controller XXX problem
1.454	13-Jan-2020	jdolecek	disable NCQ by default for "Samsung SSD 860 EVO 1TB" and "Samsung SSD 860 EVO 500GB" - these drives have known broken NCQ support particularly when used with AMD SB710/750 chipsets, problem occur also under Linux and Windows https://eu.community.samsung.com/t5/Cameras-IT-Everything-Else/860-EVO-250GB-causing-freezes-on-AMD-system/td-p/575813 https://bugzilla.kernel.org/show_bug.cgi?id=201693 It seems there is no Samsung firmware update to fix this even. Disable NCQ regardless of the controller, it's likely same problem exists with other controllers too. This should fix PR kern/54790 and PR kern/54855
1.453	27-Dec-2019	msaitoh	branches: 1.453.2; s/transfered/transferred/
1.452	06-Jun-2019	mlelstv	branches: 1.452.2; Take channel lock for calling reset_drive. Should fix PR 54217.
1.451	06-Jun-2019	mlelstv	Count only the initial start of a transfer, not the retries. Should fix kern/54166. Thanks to macallan@ for spotting the issue.
1.450	24-May-2019	mlelstv	Also schedule timeouts when all openings are in use.
1.449	07-Apr-2019	bouyer	Really implement WDF_DIRTY. patch(1) did something silly here ...
1.448	05-Apr-2019	bouyer	drop AT_RST_NOCMD, it's a cut'n'paste side effect
1.447	05-Apr-2019	bouyer	Implement a DIRTY flag (copied from sd(4)) so avoid flushing the cache if there has been no write. This avoids a (long) timeout on the flush cache command triggered by atactl sleep, when the device is open only by the atactl command itself. If a drive has no partition open and goes to sleep, the WDF_LOADED flag is clear, and the next open will issue wd_get_params() command. But to wake up the drive a reset is required, and wd_get_params() doens't issue a reset on timeout, so there's no way to wake up the disk. Add a retry after reset to wd_get_params(). Tested by Hauke Fath; fixes PR kern/49457
1.446	19-Mar-2019	mlelstv	The NCQ support added a private request queue to the wd driver. This makes the regular buffer queue ineffective, it also allowed to queue an unlimited number of requests. Fix this by limiting the number of requests queued to the driver to the possible number of concurrent NCQ transactions.
1.445	19-Mar-2019	mlelstv	Set disk model name as disk type. The information can be queried through drvctl(4).
1.444	19-Mar-2019	mlelstv	Move standby on detach after wedges deletion in case wedges trigger I/O on the parent disk. Add debug messages.
1.443	24-Oct-2018	jdolecek	move the comment (and expand) about NCQ TRIM to wd_trim()
1.442	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.441	10-Aug-2018	jdolecek	branches: 1.441.2; fix race in wd_lastclose() on systems with two ide disks on same channel, which happened when one disk had pending I/O while the other disk executed the final disk flush - need to restart bufq processing once xfer is freed in this case it could happen e.g. on boot when system executes fsck on different partitions on the two drives in parallell and hence open and closes the disk devices repeatedly add KASSERT() for empty bufq on wd_lastclose(), and fix similar issue also on suspend/standby path this was introduced by the NCQ merge and not dksubr - before the merge each drive had their own xfer, so they could not block each other fixes PR kern/52783 by Onno van der Linden; many thanks for extensive help with tracking this down
1.440	06-Aug-2018	jdolecek	add wddebug() which dumps some status for attached disks; indended for debugging of PR kern/52783
1.439	03-Jun-2018	jdolecek	branches: 1.439.2; take mutex around check for pending flush, as the code before dksubr conversion had, to avoid possible race on my system doesn't really change behaviour, besides the test runs being slightly faster (3x parallell pkgsrc archive extraction, up to 5% difference), thought that can just be noise done as part of investigation for PR kern/53183 by Sevan Janiyan
1.438	07-Jan-2018	mlelstv	branches: 1.438.2; Fix block address calculation for bad sectors.
1.437	13-Dec-2017	pgoyette	Fix build for WD_SOFTBADSECT option. PR kern/52814 XXX No clue if this option actually works. This fix just makes it XXX compile without error.
1.436	07-Nov-2017	mlelstv	Make wddone poll all drives of a channel again.
1.435	03-Nov-2017	mlelstv	Add WDF_OPEN flag to really disallow opening of a disk that has been invalidated. Restore wdbiorestart function to actually retry the failed I/O request instead of just restarting the queue. Fix compilation without ATADEBUG.
1.434	01-Nov-2017	mlelstv	refactor wd and ataraid drivers to use common disk subroutines.
1.433	19-Oct-2017	jdolecek	more detailed debug info; also sync DEBUG_* values in wd.c with ata.c
1.432	14-Oct-2017	jdolecek	do not use the NCQ priority by default; seems it negatively affects performance at least with some drives, so this needs better understood first
1.431	14-Oct-2017	jdolecek	only call drive reset with AT_POLL when the command itself was polled, so that the logic for AT_POLL matches how e.g. ata_dmaerr() is called; this was the original intent of the change in 1.428.2.25, to make the error handling safe wrt. polled xfers this is stopgap fix for ATA channel wedge after DMA error, as reported by Martin Husemann in PR kern/52606, and PR kern/52605 problem happened due to ata_reset_channel() being called once in ata_dmaerr() with flags == 0, which freezed channel and set flag to reset via thread, then ata_reset_channel() was called via wdc_drive_reset() with AT_POLL, which just executed the reset and cleared the flag, without clearing the extra freeze; that logic will be refactored in separate commit
1.430	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.429	24-Apr-2017	jdolecek	branches: 1.429.2; remove the workaround for Seagate 'mod15write' bug, now driver only prints error on boog; unfortunately the code actually doesn't work, and there is little point trying to fix
1.428	05-Mar-2017	mlelstv	branches: 1.428.2; Enhance disk metrics by calculating a weighted sum that is incremented by the number of concurrent I/O requests. Also introduce a new disk_wait() function to measure requests waiting in a bufq. iostat -y now reports data about waiting and active requests. So far only drivers using dksubr and dk, ccd, wd and xbd collect data about waiting requests.
1.427	20-Nov-2016	pgoyette	branches: 1.427.2; Avoid calling bufq_free() from critical section.
1.426	05-Aug-2016	christos	CID 1364758: Integer handling issues, avoid sign extension to 64 bits.
1.425	22-Jul-2016	jakllsch	Space before tab and trailing whitespace fixes.
1.424	22-Jul-2016	jakllsch	Add ATA8-ACS Long Logical Sector Feature Set support to wd(4).
1.423	21-Jul-2016	jakllsch	Call wd_params_to_properties() after softc is sufficently initialized.
1.422	21-Jul-2016	jakllsch	Remove unused 'params' argument of local function wd_params_to_properties()
1.421	19-Jun-2016	bouyer	branches: 1.421.2; Add a new config_detach() flag, DETACH_POWEROFF, which is set when detaching devices at shutdown time with RB_POWERDOWN. When detaching wd(4), put the drive in standby before detach for DETACH_POWEROFF. Fix PR kern/51252
1.420	26-Apr-2015	mlelstv	Use C99-style initializers for struct dkdriver.
1.419	13-Apr-2015	riastradh	Convert sys/dev to use <sys/rndsource.h>.
1.418	02-Jan-2015	christos	We have three sets of DTYPE_ constants in the kernel: altq Drop Type disklabel Disk Type file Descriptor Type (not to mention constants that contain the string DTYPE). Let's make them two, by changing the disklabel one to be DisK TYPE since the other disklabel constants seem to do that. Not many userland programs use these constants (and the ones that they do are mostly in ifdefs). They will be fixed shortly.
1.417	31-Dec-2014	christos	make more drivers use disk_ioctl, and add a dev parameter to it so that we can merge the "easy" disklabel ioctls to it. Ultimately all this will go do dk_ioctl once all the drivers have been converted.
1.416	31-Dec-2014	christos	Centralize wedge ioctls in disk_ioctl.
1.415	04-Nov-2014	mlelstv	branches: 1.415.2; support DIOCMWEDGES ioctl.
1.414	11-Oct-2014	mlelstv	The partition size is always computed in native blocks. The code also assumes that native blocks are always DEVB_SIZE (a few lines earlier) which makes the current calculation a no-op.
1.413	10-Sep-2014	matt	g/u_int[0-9]*_t/ s/u_int/uint/g
1.412	10-Aug-2014	tls	branches: 1.412.2; Merge tls-earlyentropy branch into HEAD.
1.411	25-Jul-2014	dholland	Drop the old discard/trim ioctls from wd.c.
1.410	25-Jul-2014	dholland	Implement d_discard for wd.
1.409	25-Jul-2014	dholland	Add d_discard to all struct cdevsw instances I could find. All have been set to "nodiscard"; some should get a real implementation.
1.408	25-Jul-2014	dholland	Add d_discard to all struct bdevsw instances I could find. I've set them all to nodiscard. Some of them (wd, dk, vnd, ld, raidframe, maybe cgd) should be implemented for real.
1.407	16-Mar-2014	dholland	branches: 1.407.2; Change (mostly mechanically) every cdevsw/bdevsw I can find to use designated initializers. I have not built every extant kernel so I have probably broken at least one build; however I've also found and fixed some wrong cdevsw/bdevsw entries so even if so I think we come out ahead.
1.406	30-Oct-2013	drochner	play the addref/delref game on suspend, prevents crash if the disk/CF Card is eg. in a PCMCIA adapter and not mounted
1.405	19-Aug-2013	skrll	Don't return EPASSTHROUGH when succeeding in ODIOCGDINFO or ODIOCGDEFLABEL
1.404	06-Aug-2013	soren	TRIM showed up sometime between ATA-7 and ATA-8.
1.403	29-May-2013	christos	branches: 1.403.2; phase 1 of disk geometry cleanup: - centralize the geometry -> plist code so that we don't have n useless copies of it.
1.402	09-Jan-2013	riastradh	Initialize b_resid before biodone in wdioctlstrategy error branch. Otherwise we may panic when physio's biodone callback kasserts something about b_resid, since nothing will have initialized it.
1.401	19-Oct-2012	drochner	Implement experimental support to pass notifications that a file was deleted from the filesystem to the disk driver, commonly known as "discard" or "trim". fs/driver support is in ffs and ata wd for now. This is what was posted here: http://mail-index.netbsd.org/tech-kern/2012/02/28/msg012813.html with minor cleanup, and the global switch replaced by a mount option.
1.400	31-Jul-2012	bouyer	branches: 1.400.2; Apply back changes that were reverted on Jul 24 and Jul 26 (general ata/wdc cleanup and SATA PMP support), now that I'm back to fix the fallouts.
1.399	26-Jul-2012	jakllsch	Revert, with intention of restoring in a less invasive way, the SATA Port Multiplier code. ok christos@
1.398	24-Jul-2012	jakllsch	Revert dsl@'s changes of Sun, 15 Jul 2012 10:55:35 +0000 and Sun, 15 Jul 2012 10:56:50 +0000, excepting the kernel version bump. First step in reverting regressions to ata(4) subsystem during the addition of port multiplier support.
1.397	22-Jul-2012	jakllsch	FLUSH CACHE EXT is a 48-bit command, issue as such. Both FLUSH CACHE commands will return the LBA of first failure in the return task file, specifiy that we want it read, even if we don't look at the value.
1.396	22-Jul-2012	jakllsch	ATA-6 and ACS-2 drafts both say the flush cache commands may take longer than 30 seconds to complete. For now assume 5 minutes will always be enough.
1.395	15-Jul-2012	dsl	Some namespace protection (and add greppablity). Prefix the DRIVE_ and DRIVET_ constants from atavar.h with ATA_. Don't use an enum for drive_type - you don't know how big it will be. Move driver_type to avoid implicit structure padding (esp on arm). This change is purely lexical and mechanical. Update to 6.99.9 - this wasn't done when the SATA PMP changes were made - I'm sure they warranted a bump.
1.394	02-Jul-2012	bouyer	Add sata Port MultiPlier (PMP) support to the ata bus layer, as described in http://mail-index.netbsd.org/tech-kern/2012/06/23/msg013442.html PMP support in integrated to the atabus layer. struct ata_channel's ch_drive[] is not dynamically allocated, and ch_ndrive (renamed to ch_ndrives) closely reflects the size of the ch_drive[] array. Add helper functions atabus_alloc_drives() and atabus_free_drives() to manage ch_drive[]/ch_ndrives. Add wdc_maxdrives to struct wdc_softc so that bus front-end can specify how much drive they really support (master/slave or single). ata_reset_drive() callback gains a uint32_t *sigp argument which, when not NULL, will contain the signature of the device being reset. While there, some cosmetic changes: - added a drive_type enum to ata_drive_datas, and stop encoding the probed drive type in drive_flags (we were out of drive flags anyway). - rename DRIVE_ATAPIST to DRIVE_ATAPIDSCW to better reflect what this really is - remove ata_channel->ata_drives, it's redundant with the pointer in ata_drive_datas - factor out the interpretation of SATA signatures in sata_interpet_sig() propagate these changes to the ATA HBA drivers, and add support for PMP to ahcisata(4) and siisata(4). Thanks to: - Protocase (http://www.protocase.com/) which provided a system with lots of controllers, SATA PMP and drive slots - Conservation Genomics Laboratory, Department of Biology, New Mexico State University for hosting the above system - Brook Milligan, who set up remote access and has been very responsive when SATA cable move was needed
1.393	26-Jun-2012	bouyer	In some case, when an error is reported by the disk, the ahci controller still reports a number of bytes transfered equal to bcount. This then triggers a KASSERT in physio_biodone: if (done == todo) KASSERT(bp->b_error == 0); Detect this case in wd(4) (so that the workaround works for other controllers too if they have the same issue, or if the issue is with the drive) and claim we didn't read/write anything.
1.392	02-Feb-2012	tls	branches: 1.392.2; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.391	24-Jan-2012	jakllsch	Rework struct ata_command to support LBA28 and LBA48-protocol commands. Implement 28- and 48-bit command issuance and parameter read-back in the various ATA host drivers. Add LBA28-protocol support to ATAIOCCOMMAND ioctl.
1.390	25-Nov-2011	joerg	branches: 1.390.2; Replace broken memset with M_ZERO
1.389	27-Oct-2011	jakllsch	branches: 1.389.2; Attempt to complete the WD_QUIRK_SPLIT_MOD15_WRITE list. Per Seagate Publication number 100221381, Rev. B, the ST360015AS is the only other drive in the Seagate Barracuda Serial ATA V family that was not already listed. Also, correct spelling of "globing".
1.388	05-Oct-2011	jakllsch	Limit wd(4) transfers to 128 (512-byte) logical sectors, as the traditional MAXPHYS value has for at least the past decade. There are issues in wd(4) and possibly in host adapter drivers that need work before this should be raised.
1.387	30-Jul-2011	jakllsch	Add some additional consistency checking paranoia with KASSERT.
1.386	10-Feb-2011	enami	Start new line before issuing IDENTIFY command since messages from other driver may interfere during waiting for command completion.
1.385	05-Nov-2010	dyoung	branches: 1.385.2; 1.385.4; Mark a buffer with EIO in wdstrategy() if the device is not "enabled" (more like "present"), not if it is merely asleep. If the device is not awake, get out of wdstart() right away. Brett Lymn and KAMADA Ken'ichi report that this stops the kernel from crashing after a suspend/resume cycle.
1.384	24-Feb-2010	dyoung	branches: 1.384.2; A pointer typedef entails trading too much flexibility to declare const and non-const types, and the kernel uses both const and non-const PMF qualifiers and device suspensors, so change the pmf_qual_t and device_suspensor_t typedefs from "pointers to const" to non-pointer, non-const types.
1.383	23-Jan-2010	bouyer	branches: 1.383.2; struct buf::b_iodone is not called at splbio() any more. Make sure non-MPsafe iodone callbacks raise the SPL as appropriate.
1.382	19-Jan-2010	pooka	Get rid of last "easy" kernel symbols starting with __: __assert -> kern_assert __sigtimedwait1 -> sigtimedwait1 __wdstart -> wdstart1 The rest are MD and/or shared with userspace, so they will require a little more involvement than what is available for this quick "ride the 5.99.24 bump" action.
1.381	08-Jan-2010	dyoung	Expand PMF_FN_* macros.
1.380	17-Dec-2009	bouyer	ATA sepcs changed the LBA48 boundary from 0xfffffff to 0xffffffe between ATA6r1 and ATA6r3, which caused drives conform to ATA6r3 or later to reject LBA28 commands at sector 0xfffffff. Get the right idea from the LBA48 boundary from IDENTIFY words 60-61. Remove the WD_QUIRK_FORCE_LBA48 quirk, associated tables entries and autodetect code, it's not needed any more. Based on patch sent to teck-kern by Christoph Badura, use of words 60-61 instead of a constant for the LBA48 boundary by me.
1.379	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.378	08-Jun-2009	jakllsch	Add atap_max_lba member to ataparams for LBA48 Maximum Address. From OpenBSD. While here, renumber some __reserved member names to be sequential. Also, add a member for the World Wide Name. Use atap_max_lba in wd.c instead of offsets into a __reserved member.
1.377	20-May-2009	dyoung	On second thought, let's call disk_predetach() disk_begindetach(). Verbs are good.
1.376	19-May-2009	dyoung	Use disk_predetach().
1.375	19-May-2009	dyoung	Delete wdactivate() and sdactivate(). They were no-ops, but as a side-effect of registering them, config_detach(9) cleared DVF_ACTIVE before it called wddetach() or sddetach(). Even though sd(4)'s detachment may have subsequently failed with EBUSY, we could not begin new disk I/O on sd(4) because the device had been deactivated. By analogy to sdstrategy(), test device_is_active() in wdstrategy() and if it is false set b_errno to EIO instead of initiating new disk I/O. XXX We should decline more politely to start new I/O, since XXX !device_is_active() may mean simply that the device is suspended. XXX I suppose that EIO is safe as long as system suspension is XXX all-or-nothing.
1.374	15-May-2009	dyoung	Do not detach a wd(4) instance if it is open, but return EBUSY, instead. If the detach is forced, do what wdclose() does on the last close: flush the cache, clear the flag WDF_LOADED, and delete our reference to the ATA bus device.
1.373	12-May-2009	cegger	struct device * -> device_t, no functional changes intended.
1.372	12-May-2009	cegger	struct cfdata * -> cfdata_t, no functional changes intended.
1.371	02-Apr-2009	dyoung	During shutdown, detach devices in an orderly fashion. Call the detach routine for every device in the device tree, starting with the leaves and moving toward the root, expecting that each (pseudo-)device driver will use the opportunity to gracefully commit outstandings transactions to the underlying (pseudo-)device and to relinquish control of the hardware to the system BIOS. Detaching devices is not suitable for every shutdown: in an emergency, or if the system state is inconsistent, we should resort to a fast, simple shutdown that uses only the pmf(9) shutdown hooks and the (deprecated) shutdownhooks. For now, if the flag RB_NOSYNC is set in boothowto, opt for the fast, simple shutdown. Add a device flag, DVF_DETACH_SHUTDOWN, that indicates by its presence that it is safe to detach a device during shutdown. Introduce macros CFATTACH_DECL3() and CFATTACH_DECL3_NEW() for creating autoconf attachments with default device flags. Add DVF_DETACH_SHUTDOWN to configuration attachments for atabus(4), atw(4) at cardbus(4), cardbus(4), cardslot(4), com(4) at isa(4), elanpar(4), elanpex(4), elansc(4), gpio(4), npx(4) at isa(4), nsphyter(4), pci(4), pcib(4), pcmcia(4), ppb(4), sip(4), wd(4), and wdc(4) at isa(4). Add a device-detachment "reason" flag, DETACH_SHUTDOWN, that tells the autoconf code and a device driver that the reason for detachment is system shutdown. Add a sysctl, kern.detachall, that tells the system to try to detach every device at shutdown, regardless of any device's DVF_DETACH_SHUTDOWN flag. The default for kern.detachall is 0. SET IT TO 1, PLEASE, TO HELP TEST AND DEBUG DEVICE DETACHMENT AT SHUTDOWN. This is a work in progress. In future work, I aim to treat pseudo-devices more thoroughly, and to gracefully tear down a stack of (pseudo-)disk drivers and filesystems, including cgd(4), vnd(4), and raid(4) instances at shutdown. Also commit some changes that are not easily untangled from the rest: (1) begin to simplify device_t locking: rename struct pmf_private to device_lock, and incorporate device_lock into struct device. (2) #include <sys/device.h> in sys/pmf.h in order to get some definitions that it needs. Stop unnecessarily #including <sys/device.h> in sys/arch/x86/include/pic.h to keep the amd64, xen, and i386 releases building.
1.370	10-Feb-2009	tron	Backout LBA 48 quick entries which were added to fix one aspect of PR kern/40569 because of objections by Manual Bouyer.
1.369	09-Feb-2009	tron	Add two more entries to the quirk table for hard disks which need the LBA 48 work around. The first entry will watch the Seagate ST3160815AS (and similar models), the second one HP's OEM version of the same drive. This avoids the RAID rebuild problems described in PR kern/40569.
1.368	06-Feb-2009	drochner	branches: 1.368.2; -in the wdc@pcmcia driver, don't try to powerup/down while cold or shutting down -- this is only supposed to happen from the pcmcia kernel thread while the system is running -in wd_shutdown, call ata_addref first, so that the adapter is either active or we know that it isn't This fixes a hang on shutdown-p reported by Dennis den Brok per PR kern/40531. Verified by by Dennis and me. (reported against 5.0-Beta, so this is a pullup candidate)
1.367	13-Jan-2009	yamt	g/c BUFQ_FOO() macros and use bufq_foo() directly.
1.366	16-Dec-2008	christos	replace bitmask_snprintf(9) with snprintb(3)
1.365	13-Dec-2008	christos	more debugging in the dump path.
1.364	05-Dec-2008	dyoung	In wddetach(), avoid a crash by destroying the callout sc_restart_ch.
1.363	08-Jun-2008	tsutsui	branches: 1.363.4; 1.363.6; 1.363.8; Use device_lookup_private() rathter than device_private(device_lookup()).
1.362	02-May-2008	plunky	branches: 1.362.2; disk_destroy() is needed when the disk structure is no longer required, this fixes PR kern/38131
1.361	28-Apr-2008	martin	Remove clause 3 and 4 from TNF licenses
1.360	18-Mar-2008	cube	branches: 1.360.2; 1.360.4; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.359	29-Feb-2008	bouyer	The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.358	29-Feb-2008	dyoung	Use PMF_FN_ARGS, PMF_FN_PROTO.
1.357	28-Feb-2008	drochner	use pmf again to register the shutdown handler, instead of shutdownhook_establish()
1.356	21-Feb-2008	drochner	put back the old shutdown handlers to flush the disk cache, the pmf API can't deal with all the different suspend/resume/reboot cases well yet, so better keep suspend/resume and reboot/halt/poweroff clearly seperated
1.355	02-Jan-2008	ad	branches: 1.355.2; 1.355.6; Merge vmlocking2 to head.
1.354	18-Dec-2007	joerg	Restore suspend handler. Put the disc into standby on suspend and for powerdown, but not for normal halt or reboot or panic. Fixes PR 37508.
1.353	11-Dec-2007	jmcneill	Make this compile again.
1.352	11-Dec-2007	jmcneill	Disable the wd_suspend power handler on the proper branch, workaround for kern/37508
1.351	09-Dec-2007	jmcneill	branches: 1.351.2; Merge jmcneill-pm branch.
1.350	07-Nov-2007	itohy	branches: 1.350.2; 1.350.4; more KNF changes
1.349	07-Nov-2007	itohy	KNF or make it compile by gcc 2
1.348	31-Oct-2007	jnemeth	branches: 1.348.2; PR/37251 - Brian Buhrow -- FORCE_LBA48 on seagate drives larger then 1TB
1.347	19-Oct-2007	ad	machine/{bus,cpu,intr}.h -> sys/{bus,cpu,intr}.h
1.346	08-Oct-2007	ad	branches: 1.346.2; Merge disk init changes from the vmlocking branch. These seperate init / destroy of 'struct disk' from attach / detach.
1.345	16-Sep-2007	bouyer	branches: 1.345.2; Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.344	05-Sep-2007	jnemeth	yet more caddr_t fallout
1.343	30-Jul-2007	taca	branches: 1.343.4; 1.343.6; 1.343.8; Make it compile with WD_SOFTBADSECT option.
1.342	29-Jul-2007	ad	It's not a good idea for device drivers to modify b_flags, as they don't need to understand the locking around that field. Instead of setting B_ERROR, set b_error instead. b_error is 'owned' by whoever completes the I/O request.
1.341	21-Jul-2007	ad	Replace some uses of lockmgr().
1.340	09-Jul-2007	ad	branches: 1.340.2; Merge some of the less invasive changes from the vmlocking branch: - kthread, callout, devsw API changes - select()/poll() improvements - miscellaneous MT safety improvements
1.339	01-Jul-2007	dsl	Allow callers of the ATAIOCCOMMAND ioctl to request the WDCC_IDENTIFY to specify a buffer that is less than 512 bytes.
1.338	08-Apr-2007	dyoung	Include opt_ata.h for ATADEBUG definition instead of #defining it unconditionally.
1.337	04-Mar-2007	christos	branches: 1.337.2; 1.337.4; Kill caddr_t; there will be some MI fallout, but it will be fixed shortly.
1.336	09-Feb-2007	ad	branches: 1.336.2; Merge newlock2 to head.
1.335	16-Nov-2006	christos	branches: 1.335.2; 1.335.4; __unused removal on arguments; approved by core.
1.334	09-Nov-2006	bouyer	Yet another broken seagate drive.
1.333	26-Oct-2006	thorpej	Use the new prop_dictionary_util functions.
1.332	25-Oct-2006	thorpej	- Add a new disk ioctl (DIOCGDISKINFO) to get the disk-info dictionary for the disk. - Add a new function, disk_ioctl(), that does generic disk ioctl handling. DIOCGDISKINFO is handled here now, and others will be added in the future. - In the wd driver, fill in the dk_info member of struct disk and use the new disk_ioctl() function.
1.331	12-Oct-2006	christos	- sprinkle __unused on function decls. - fix a couple of unused bugs - no more -Wno-unused for i386
1.330	22-Sep-2006	thorpej	- Define disk information, disk geometry, and disk partition dictionary schemas. Disk information and disk geometry are designed to replace information currently conveyed to user space using struct disklabel. - Add a dk_info member to struct disk; a reference to a disk information dictionary. This dictionary is to be allocated and the reference stored in struct disk by individual drivers. - disk_detach0() will release dk_info if non-NULL. - Convert the wd(4) driver to stash geometry and other disk properties as the "disk-info" property in its properties dictionary. This needs some cleanup, but will serve as an example of what to do with other disk drivers.
1.329	27-Aug-2006	christos	branches: 1.329.2; 1.329.4; Oops, this is not ready yet.
1.328	27-Aug-2006	christos	Add missing initializer
1.327	01-Aug-2006	lukem	If there's any wd quirks in use, display them.
1.326	24-Jun-2006	gendalia	Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.325	05-Jun-2006	bouyer	branches: 1.325.2; Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.324	18-May-2006	bouyer	branches: 1.324.2; Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.323	05-Apr-2006	cherry	Backing out previous commit for more review.
1.322	04-Apr-2006	cherry	Sleeps hard disk just before system shutdown. Some hardware (eg: Apple iBooks) don't do this automatically and the harddisk behaves as if the power plug were pulled. Closes PR: kern/32766
1.321	28-Mar-2006	thorpej	Use device_unit().
1.320	01-Mar-2006	yamt	branches: 1.320.2; 1.320.4; 1.320.6; merge yamt-uio_vmspace branch. - use vmspace rather than proc or lwp where appropriate. the latter is more natural to specify an address space. (and less likely to be abused for random purposes.) - fix a swdmover race.
1.319	20-Feb-2006	thorpej	Use device_is_active() rather than testing dv_flags for DVF_ACTIVE directly.
1.318	15-Jan-2006	abs	branches: 1.318.2; 1.318.4; cast 16bit atap_capacity to u_int64_t before << 16. (assigned to a u_int64_t variable). Fixes 1 to 2 TB disks in qemu. Tested with 1200GB and to a lesser extent 2047GB sparse files. For reference an empty FFS1 filesystem on a 1200GB sparse file uses 19GB of disk.
1.317	07-Jan-2006	yamt	remove B_EINTR as it isn't used anymore.
1.316	04-Jan-2006	yamt	- add simple functions to allocate/free a buffer for i/o. - make bufpool static.
1.315	26-Dec-2005	yamt	branches: 1.315.2; - add ioctls to set/get disk bufq strategy. - implement them for some drivers.
1.314	11-Dec-2005	christos	merge ktrace-lwp.
1.313	01-Nov-2005	martin	Rework the funny "user_strat: No ioctl" error messages - the functions have not been called user_strat for quite some time and the message is confusing if you do not know the code.
1.312	15-Oct-2005	yamt	branches: 1.312.2; - change the way to specify a bufq strategy. (by string rather than by number) - rather than embedding bufq_state in driver softc, have a pointer to the former. - move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c. - rename method to strategy for consistency. - move some definitions which don't need to be exposed to the rest of kernel from sys/bufq.h to sys/bufq_impl.h. (is it better to move it to kern/ or somewhere?) - fix some obvious breakage in dev/qbus/ts.c. (not tested)
1.311	05-Sep-2005	riz	Fix typo in comment from previous revision.
1.310	05-Sep-2005	bouyer	Use a pmatch(9) expression which should catch all present and future seagate drives larger than 200GB for the WD_QUIRK_FORCE_LBA48 quirks. Based on infos from http://www.seagate.com/support/kb/disc/howto/interpret_model.html
1.309	05-Sep-2005	bouyer	Add quirk for ST3200826A. From rudolf (netbsd at eq.cz) on tech-kern.
1.308	29-Aug-2005	bouyer	wd->atabus->ata_bio() should handle the multi-sector PIO transfers just fine, and in a better way than what is done in wddump() (which also does it for DMA transfers, limiting the dump speed significantly). So remove code splitting transfers in wd->sc_multi chunks from wddump(), and call wd->atabus->ata_bio() with the whole transfers. Problem reported by Frank Kardel, and patch tested by him.
1.307	26-Aug-2005	briggs	ST3300831A also needs WD_QUIRK_FORCE_LBA48.
1.306	25-Aug-2005	bouyer	Add a comment explaining why we have WD_QUIRK_FORCE_LBA48, instead of always using LBA48 for sector 0xffffff.
1.305	02-Jul-2005	dsainty	branches: 1.305.2; Force LBA48 access to ST3160827AS drives. Without this change the entire system hangs whenever a read or write request to sector 0xfffffff occurs. The hang appears to happen when: status = bus_space_read_1(wdr->cmd_iot, wdr->cmd_iohs[wd_status], 0); is executed in __wdcwait(), and it is not possible to enter the debugger or make the system respond at all after this point. Whilst not necessarily the most prudent solution to handling the large number of drives that exhibit this problem, it is currently the popular workaround. [An alternative approach is to just always drop to LBA48 if the request happens to pass by sector 0xfffffff and the drive reported as larger than 0xfffffff sectors. My understanding is that the 32 bit addressing below 0x10000000 is purely there as a performance booster, not to resolve a compatibility issue.]
1.304	19-Jun-2005	bouyer	The ST3160023A* also need the LBA48 quirk. From Timo Schoeler.
1.303	16-Jun-2005	bouyer	Use pmatch() metacharacters to match both SATA and PATA versions of seagate drives in the quirk table.
1.302	16-Jun-2005	martti	Added LBA48 quirk for Seagate ST3200822AS. Tested by Teemu Rinta-aho.
1.301	29-May-2005	christos	- remove casts [to avoid volatile] - sprinkle const - avoid shadowed variables
1.300	16-May-2005	bouyer	Add another broken seagate device to the quirk list. Problem reported by rudolf <netbsd at eq.cz> on netbsd-help@
1.299	31-Mar-2005	yamt	introduce a function to drain bufq and use it where appropriate.
1.298	27-Feb-2005	perry	branches: 1.298.2; nuke trailing whitespace
1.297	18-Nov-2004	bouyer	branches: 1.297.4; 1.297.6; Some drives reports ATA version < 4, and still support (and need) the WDCC_FLUSHCACHE command. So for drive < ATA4, test WDC_CMD2_FC in atap_cmd_set2 before returning ENODEV. Problem reported and fix tested by Michael Gerhards on port-sparc64.
1.296	28-Oct-2004	yamt	move buffer queue related stuffs from buf.h to their own header, bufq.h.
1.295	28-Sep-2004	bouyer	Add a WD_QUIRK_FORCE_LBA48 quirck, which forces all access to use LBA48. Use it for ST3200822A, which has a broken firmware and return an error when accessing sector 0xfffffff with a LBA command. Patch tested by David Brownlee.
1.294	25-Sep-2004	thorpej	Add support for wedges to the ATA disk driver.
1.293	16-Sep-2004	bouyer	In case of error, reset the channel with AT_RST_NOCMD. We know the reset won't abort any command at this point.
1.292	30-Aug-2004	drochner	make the software managed bad-sector list optional (kernel option WD_SOFTBADSECT)
1.291	21-Aug-2004	thorpej	atastart() (called only at splbio(), and from interrupts) can change drive_flags, to make sure all drive_flags manipulations are done at splbio().
1.290	20-Aug-2004	thorpej	Rename the ata_bustype "ata_reset_channel" member to "ata_reset_drive" to reflect the new reality after Manuel added the bus reset ioctls.
1.289	13-Aug-2004	thorpej	WDCDEBUG -> ATADEBUG.
1.288	12-Aug-2004	thorpej	Rename some constants: WDC_COMPLETE -> ATACMD_COMPLETE WDC_QUEUED -> ATACMD_QUEUED WDC_TRY_AGAIN -> ATACMD_TRY_AGAIN
1.287	12-Aug-2004	thorpej	Rename "struct wdc_command" to "struct ata_command".
1.286	10-Aug-2004	mycroft	A few things: * Kill off device nodes before we clear out the queue, to help prevent any new I/O being queued. * More useful error reporting in wd_setcache() and wd_flushcache(). * Add a wd_standby() (tested but not used yet). And the most important: * In wddetach(), if the device was open, call adapt_delref() so that we propagate the disable up to our PCMCIA socket.
1.285	04-Aug-2004	bouyer	Change wdc_kill_pending() to take a struct ata_drive_datas * as argument, and kill only pending requests for this drive. Implement a DRIVE_WAITDRAIN flag, which will cause the active command to be killed once complete. Other minor fixes. Now it's possible to detach a ATA or ATAPI device from ioctl even when a dd on the raw char partition is running.
1.284	03-Aug-2004	bouyer	First cut at implementing ATABUSIODETACH. Still buggy.
1.283	03-Aug-2004	bouyer	wddetach(): move lockmgr(LK_DRAIN) near the end; otherwise we'll panic in wddone() (vdevgone() may end up calling it) trying to get the lock. wdopen(): check for DVF_ACTIVE.
1.282	03-Aug-2004	bouyer	move cleaning of the bad block list in wddetach() after the call to disk_detach(), to avoid a race condition with the interrupt code.
1.281	31-Jul-2004	bouyer	Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes (problem found and analysed on sparc64 by Martin Husemann). This should even make core dumps marginally more reliable on ATA drives.
1.280	22-Jun-2004	mycroft	Update copyright.
1.279	22-Jun-2004	mycroft	Do not force ATA_SINGLE on devices that don't support READ/WRITE MULTI. We just have to take an interrupt for each sector. Tested with one laptop disk (which normally runs in DMA mode and was forced to single-sector transfers) and 3 CF cards. Increases the performance of the CF cards substantially (760KB/s->1240KB/s in one case, 410KB/s->750KB/s in the other two cases).
1.278	01-Jun-2004	mycroft	Fix an extremely obvious bug in the handling of the bad block list: the "max" block was being set 512x further out than it should be, causing rather severe escalation of the error. XXX WTF is the point of this shit, anyway? In most cases, the way you're supposed to fix a bad block on an ATA disk is to rewrite it -- which will either just transparently fix it, or spare it. This code actively prevents that.
1.277	27-May-2004	thorpej	Clean-up of use of "precomp". It is used as "features" everywhere, so rename the fields in wdc_command as appropriate.
1.276	21-May-2004	bouyer	Add support for the DIOCCACHESYNC ioctl.
1.275	03-May-2004	bouyer	When using LBA48, use Flush cache extended command if supported. From Christos Zoulas in PR 25366, with minor adjustements from me.
1.274	28-Feb-2004	yamt	branches: 1.274.2; change the way to handle NEW_BUFQ_STRATEGY option. instead of putting #ifdefs into each drivers, use a global variable to indicate default strategy. XXX should have a way to specify other strategies.
1.273	10-Jan-2004	yamt	store a i/o priority hint in struct buf for buffer queue discipline.
1.272	14-Dec-2003	thorpej	Move the wddone() prototype to wd.c
1.271	14-Dec-2003	thorpej	Garbage-collect __ATA_DISK_PRIVATE; it is no longer needed.
1.270	14-Dec-2003	thorpej	Instead of calling wddone() directly, set a (*drv_done)() callback in the ata_drive_datas structure and invoke that.
1.269	04-Dec-2003	keihan	netbsd.org -> NetBSD.org This was the last commit of this kind to src/sys, which is now totally "NetBSD.org clean". Thanks for the patiance, and sorry for all the commits.
1.268	30-Nov-2003	yamt	- rename WDC_CAP_LBA48 to ATA_CMD2_LBA48. - remove WDC_CAP_QUEUE. (there's already ATA_CMD2_RWQ.) as discussed on tech-kern@.
1.267	07-Nov-2003	mycroft	Replace home-grown locks with lockmgr().
1.266	29-Oct-2003	mycroft	Back out the bogus initializer -- the compiler bug is fixed.
1.265	28-Oct-2003	cl	note 'm68k {u,}int64_t used uninitialized' bug. add reference to gcc bug report. mark all (known) occurrences.
1.264	27-Oct-2003	fredb	Appease gcc3 "-Wno-uninitialized".
1.263	08-Oct-2003	bouyer	Following Matt Thomas's request, rename ata attribute to ata_hl, and wdc_base to ata. We can now have atabus* at ata? in kernel config files.
1.262	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.261	19-Sep-2003	mycroft	1) Use config_interrupts() to attach IDE and ATAPI drives. This eliminates most polling. 2) Clean up some goofiness in pciide -- get rid of the whole "candisable" path (it's gratuitous) and simplify the code by calling pciide_map_compat_intr(), *_set_modes() and wdc_print_modes() from central locations. 3) Add a register writability and register ghost test to eliminate phantom drives more quickly.
1.260	03-Aug-2003	bouyer	For LBA drives, compute the number of cylinders from the drive capacity in the default disklabel and the boot message, instead of using the value reported by the drive (which is 16383 if the drive is larger than 8G). Should fix PR 9864
1.259	11-Jul-2003	dsl	KNF and whitespace (and indent a contuation line)
1.258	10-Jul-2003	matt	Add LBA48_THRESHOLD #define and set it to the correct value. Use it in place of "0xffffff" (which is one too few 'f').
1.257	29-Jun-2003	fvdl	branches: 1.257.2; Back out the lwp/ktrace changes. They contained a lot of colateral damage, and need to be examined and discussed more.
1.256	29-Jun-2003	bouyer	Don't use LBA48 if we don't have to (that is, if the request is inside the first 128GB). Tested by Andreas Johansson.
1.255	29-Jun-2003	bouyer	Remove unused WAITTIME definition. Pointed out by Andreas Johansson.
1.254	28-Jun-2003	darrenr	Pass lwp pointers throughtout the kernel, as required, so that the lwpid can be inserted into ktrace records. The general change has been to replace "struct proc " with "struct lwp " in various function prototypes, pass the lwp through and use l_proc to get the process pointer when needed. Bump the kernel rev up to 1.6V
1.253	17-May-2003	thorpej	* Use aprint(). Add Ultra/133 to wdc_print_modes().
1.252	10-May-2003	thorpej	Change bounds_check_with_label() to take a pointer to the disk structure, rather than the label itself. This paves the way for some future changes.
1.251	02-May-2003	dsl	Change return type of readdisklabel() to const char * I hope I've found all the correct places!
1.250	27-Apr-2003	bouyer	Fix typo.
1.249	27-Apr-2003	bouyer	Acceptable cache operations are r and rw, not none and w Properly initialise the timeout value.
1.248	26-Apr-2003	bouyer	Add support for the getcache/setcache ioctls.
1.247	18-Apr-2003	explorer	add my 80 GB SATA drive to the quirk table.
1.246	18-Apr-2003	mason	From current-users: On Thu, Apr 17, 2003 at 11:26:26AM -0500, Dave Huang wrote: > Typo police here :) "whose," not "who's."
1.245	16-Apr-2003	darrenr	add missing setting of b_error (thanks Manuel)
1.244	15-Apr-2003	nathanw	Add a missing paren.
1.243	15-Apr-2003	darrenr	Implement (and document!) "dkctl <dev> badsector .." to `manage' the bad sector information in the kernel. Doing this uncovered some shortcomings that should have been pretty obvious with the code committed prior, addressing the major kludge with a new struct - disk_bacsecinfo to be passed into DIOCBSLIST.
1.242	15-Apr-2003	dogcow	fix typos
1.241	15-Apr-2003	darrenr	Automatically build up a list of bad sectors from failed read operations and prempt read operations on matching regions with a failure rather than waiting for the device to return a failure. The I/O operation must have already failed MAXRETRIES times before being added to the list - this can generally take up to 12 seconds. List is made accessible to userspace via DIOCBSLIST and DIOCBSFLUSH.
1.240	03-Apr-2003	fvdl	Check RAW_PART against the media size instead of the disklabel. Add the media size in 512-byte sectors to the softc, to avoid some 64 bit computations. Bump the capacity stored in softcs for disks to 64 bits.
1.239	21-Mar-2003	thorpej	Work around a buggy PHY in some Seagate S-ATA drives which get confused by the way some S-ATA controllers packetize data when the sector count of the transfer mod 15 is 1.
1.238	23-Feb-2003	yamt	initialize b_interlock properly. (for ioctl)
1.237	27-Jan-2003	thorpej	Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.236	23-Jan-2003	bad	Being able to make image backups of your whole disk, not to mention not causing certain ata disks to lock up by reading beyond the last block, beats catering to broken devices. bounds_check_with_label() RAW_PART too.
1.235	20-Jan-2003	simonb	The Double-Semi-Colon Police.
1.234	07-Jan-2003	fvdl	Fix error in ODIOCGDINFO handling caused by previous change to it.
1.233	15-Dec-2002	fvdl	Avoid using a lot of kernel stack.
1.232	01-Nov-2002	mrg	implement separate read/write disk statistics: - disk_unbusy() gets a new parameter to tell the IO direction. - struct disk_sysctl gets 4 new members for read/write bytes/transfers. when processing hw.diskstats, add the read&write bytes/transfers for the old combined stats to attempt to keep backwards compatibility. unfortunately, due to multiple bugs, this will cause new kernels and old vmstat/iostat/systat programs to fail. however, the next time this is change it will not fail again. this is just the kernel portion.
1.231	23-Oct-2002	jdolecek	merge kqueue branch into -current kqueue provides a stateful and efficient event notification framework currently supported events include socket, file, directory, fifo, pipe, tty and device changes, and monitoring of processes and signals kqueue is supported by all writable filesystems in NetBSD tree (with exception of Coda) and all device drivers supporting poll(2) based on work done by Jonathan Lemon for FreeBSD initial NetBSD port done by Luke Mewburn and Jason Thorpe
1.230	18-Oct-2002	junyoung	Add option NEW_BUFQ_STRATEGY, a new buffer queue strategy for better system responsiness under high disk I/O load. OK'ed by Juergen Hannken-Illjes.
1.229	02-Oct-2002	thorpej	Add trailing ; to CFATTACH_DECL.
1.228	30-Sep-2002	thorpej	Use CFATTACH_DECL().
1.227	27-Sep-2002	thorpej	Declare all cfattach structures const.
1.226	27-Sep-2002	drochner	Remove a check whether the drive is initialized before crashdump. It prevents dumping on an otherwise unused disk and seems unnecessary anyway - the driver will deal with initialization. fixes my PR kern/18135
1.225	06-Sep-2002	gehenna	Merge the gehenna-devsw branch into the trunk. This merge changes the device switch tables from static array to dynamically generated by config(8). - All device switches is defined as a constant structure in device drivers. - The new grammer ``device-major'' is introduced to ``files''. device-major <prefix> char <num> [block <num>] [<rules>] - All device major numbers must be listed up in port dependent majors.<arch> by using this grammer. - Added the new naming convention. The name of the device switch must be <prefix>_[bc]devsw for auto-generation of device switch tables. - The backward compatibility of loading block/character device switch by LKM framework is broken. This is necessary to convert from block/character device major to device name in runtime and vice versa. - The restriction to assign device major by LKM is completely removed. We don't need to reserve LKM entries for dynamic loading of device switch. - In compile time, device major numbers list is packed into the kernel and the LKM framework will refer it to assign device major number dynamically.
1.224	21-Jul-2002	hannken	Rename bufq_init() to bufq_alloc(). Add bufq_free() to remove a buffer queue. Avoid MALLOC while holding a spinlock. From Chuck Silvers.
1.223	16-Jul-2002	hannken	Implement a new device buffer queue interface. One basic struct, a function to setup a queue with a specific strategy and three macros to put buf's into the queue, get and remove the next buf or get the next buf without removal. The BUFQ_XXX interface will be removed in the future. The B_ORDERED flag is not longer supported. Approved by: Jason R. Thorpe <thorpej@wasabisystems.com>
1.222	28-Jun-2002	yamt	do previous differently for wddump case.
1.221	28-Jun-2002	yamt	- remove big buffers on stack. discussed on tech-kern. - constify.
1.220	13-Jan-2002	christos	branches: 1.220.8; 1.220.10; Add LBA48 support based on patches from Keisuke YOSHIDA. KNF a bit while I am there.
1.219	12-Jan-2002	tsutsui	Call malloc(9) with M_ZERO flag instead of memset() after malloc().
1.218	03-Dec-2001	bouyer	Update my copyrigth.
1.217	03-Dec-2001	bouyer	Allow wd to attach to something else than wdc/pciide (like USB<->ATA bridges): - move some functions from ata.c to ata_wdc.c or wdc.c. - add callbacks to struct ata_bustype so that wd.c doesn't call directly functions from the lower level driver.
1.216	02-Dec-2001	bouyer	Clean up attach of wd/atapibus: kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead of ata_atapi_attach as attach arch. Create a ata_device, compatible with scsipi_channel, to attach wd.
1.215	13-Nov-2001	lukem	add RCSID
1.214	13-Jun-2001	bjh21	branches: 1.214.2; 1.214.4; Add explicit support for IDE and SCSI adaptors which don't support interrupts. On such adaptors, all transfers are done in polling mode. OK'ed by Manuel on tech-kern.
1.213	06-May-2001	drochner	Since __HAVE_OLD_DISKLABEL, DISKMINOR(u, p) isn't continuous over "p" anymore. So we can't pass a range to vdevgone() but have to loop over the partitions explicitely. (otherwise the next device unit will be killed)
1.212	08-Jan-2001	fvdl	branches: 1.212.2; Return error in the case of using ODIOCGDINFO or ODIOCGDEFLABEL when the number of partitions is > OLDMAXPARTITIONS. This is better than silently truncating the label (don't want to silently throw away partitions when using an old disklabel binary on a label with > 8 partitions). From Enami Tsugutomo.
1.211	07-Jan-2001	fvdl	Adapt all disk devices in MI directories to handle ODIOC* calls for ports that have bumped MAXPARTITIONS (and thus define __HAVE_OLD_DISKLABEL).
1.210	14-Dec-2000	lukem	s/pio/PIO/ and s/lba/LBA/ in messages output to users
1.209	06-Jul-2000	thorpej	Use device_lookup() more.
1.208	06-Jul-2000	thorpej	Use device_lookup().
1.207	05-Jul-2000	thorpej	Garbage-collect __BDEVSW_DUMP_OLD_TYPE.
1.206	28-Jun-2000	mrg	remove include of <vm/vm.h>
1.205	27-May-2000	bouyer	branches: 1.205.2; Ops, state should be RESET, not RECAL here.
1.204	15-May-2000	bouyer	branches: 1.204.2; Sync my copyrigth notice
1.203	23-Mar-2000	thorpej	New callout mechanism with two major improvements over the old timeout()/untimeout() API: - Clients supply callout handle storage, thus eliminating problems of resource allocation. - Insertion and removal of callouts is constant time, important as this facility is used quite a lot in the kernel. The old timeout()/untimeout() API has been removed from the kernel.
1.202	07-Feb-2000	thorpej	Fix a bug in disksort_*() which caused non-optimal ordering when multiple active partitions were on a single spindle. Add a b_rawblkno member to struct buf which contains the non-partition-relative block number to sort by.
1.201	24-Jan-2000	enami	Use format_bytes to convert drive capacity into human familier string (like sd.c does).
1.200	21-Jan-2000	thorpej	Update for sys/buf.h/disksort_*() changes.
1.199	23-Dec-1999	leo	Implement the DIOCKLABEL ioctl. Through this ioctl it it possible keep the disklabel on the last close or to drop it.
1.198	10-Nov-1999	leo	Give lp->d_type a value other than 'unknown'. When the model is "ST506", the d_type field becomes DTYPE_ST506, otherwise it will be set to DTYPE_ESDI.
1.197	20-Oct-1999	enami	Cancel active transfers on aic/wdc detach. Also makes LS-120 drive works for me again.
1.196	23-Sep-1999	enami	branches: 1.196.2; 1.196.4; 1.196.6; Allow to detach wdc, atapibus, wd and cd.
1.195	22-Sep-1999	enami	Fix indentation of close curly brace.
1.194	09-Aug-1999	bouyer	Don't call wddone() from _wdstart(), the ata_wdc system will do it now. call wd_flushcache() with the proper flags (especially don't set AT_WAIT \| AT_POLL).
1.193	28-Feb-1999	explorer	branches: 1.193.2; Update to slightly altered rnd_attach_source() api
1.192	21-Feb-1999	hubertf	Only define WDCDEBUG if not already defined (e.g. by some kernel option)
1.191	18-Feb-1999	bouyer	Add a missing ':'. pointed out by Dave Huang.
1.190	18-Jan-1999	bouyer	move wd.c:print_wderror() to ata.c:ata_perror(). In wdc_probe_caps() add code to guess the ATA revision supported (if ATA4 if Ultra-DMA, ATA2 if PIO mode > 2). We can't rely on param.atap_ata_major here, at last one Ultra-DMA drive claims to support only ATA-3. Use the ATA version in ata_perror(), and to try a flush cache command in a shutdown hook for IDE drives.
1.189	16-Dec-1998	bouyer	Switch to single-sector transfert mode only for the last retry, to let a chance to the underlying driver to downgrade the mode first.
1.188	23-Nov-1998	kenh	Add a new flag to wdc_command: ATA_READREG. This will indicate that all device registers should be read back into the wdc_command structure after successfull command completion. Use this this in wdioctl() for ATAIOCCOMMAND.
1.187	20-Nov-1998	thorpej	Implement controller reference counting on ATA devices.
1.186	19-Nov-1998	kenh	Rename WDCIOCCOMMAND (and related structures) to ATAIOCCOMMAND
1.185	19-Nov-1998	kenh	Add support for the WDCIOCCOMMAND ioctl. This allows the sending of arbitrary ATA commands from user-space. Concepts mostly taken from the implementation of the SCIOCCOMMAND ioctl.
1.184	11-Nov-1998	bouyer	- clearify the boot messages (features supported vs features used). Thanks to Havard Eidnes for his complains about this :) - fix some typo in comments - hoppefully better detection of drives reporting bogus PIO modes.
1.183	20-Oct-1998	bouyer	Donc reset the disk state to 0 in wd.c, set a flag which will cause it to be reset to 0 at wdcstart(). This fixes a race condition between normal I/O and wdioclt(). Also use interrupts for wd_get_params in wdopen().
1.182	13-Oct-1998	bouyer	Properly separate DEBUF_XFERS and DEBUG_FUNCS: DEBUF_XFERS is for data transfers (may be called often), where DEBUG_FUNCS is for setup functions (not used for normal contitions).
1.181	12-Oct-1998	bouyer	Merge bouyer-ide
1.180	07-Sep-1998	drochner	make the default disklabel consistent to the SCSI/ATAPI case
1.179	15-Aug-1998	rvb	Use scsi style geometry printout
1.178	15-Aug-1998	mycroft	Assign my copyrights to TNF.
1.177	13-Aug-1998	rvb	Report "true" disk geometry; like sd disks do
1.176	09-Jun-1998	thorpej	Nuke __BROKEN_INDIRECT_CONFIG.
1.175	04-Jun-1998	enami	branches: 1.175.2; Print newline like others do when tracing wdunlock.
1.174	09-May-1998	mycroft	Disable the change to check wdp_ataversion, since it's unreliable with old devices. (Yuck.)
1.173	26-Apr-1998	mycroft	Update copyright.
1.172	26-Apr-1998	mycroft	In LBA mode, always use the `total capacity' count, rather than calculating the capacity based on the c/h/s numbers. In fact, don't use the c/h/s numbers for much of anything. For ATA-4 drives or later, always use LBA mode, since it's now required. Collectively, this allows >8GB disks (like the 12GB Bigfoot) to work.
1.171	23-Apr-1998	bouyer	From Robert V. Baron: Compute the disk block addr at command queing time rather than exec time. This fix a bug which could lead to data corruption on disk: when a command was reexecuted after an error condition (from wdcunwedge), the partition offset was re-added to the block addr, leading to a transfert at the wrong disk block. This should fix the problem reported by some laptop's users, where the first disk read after a suspend/resume returned garbage.
1.170	25-Mar-1998	leo	Conditionalize badblock handling on 'HAS_BAD144_HANDLING'. Some ports (like atari) do not support it.
1.169	15-Jan-1998	cgd	kill a couple of unnecessary #ifs
1.168	14-Jan-1998	cgd	Various cleanups and bug fixes to the wdc/wd code: * Fix bug in wdc that would overflow ATAPI transfer length. * Improve wdc probe code so that 'wdc' is probed in if present even if there are no drives attached, and so that it works properly even if the only device is an ATAPI slave. * bus_space-ify. * split the ISA attachment from the wdc driver, and remove ISA dependencies from non-ISA files. * claim that wd and wdc are now machine-independent (probably not completely true, but mostly so; they at least work on arm32 and i386). * Various other minor fixups and cleanups, some of which were pointed out by Kazuki Sakamoto.
1.167	12-Jan-1998	thorpej	Update for config changes.
1.166	13-Oct-1997	explorer	o Make usage of /dev/random dependant on pseudo-device rnd # /dev/random and in-kernel generator in config files. o Add declaration to all architectures. o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include that this code is derived in part from Ted Tyso's linux code.
1.165	10-Oct-1997	explorer	Add hooks to insert timing info into the random system
1.164	08-Oct-1997	thorpej	Implement DIOCGDEFLABEL.
1.163	27-Aug-1997	bouyer	Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.162	18-Jul-1997	fvdl	branches: 1.162.2; Work around possible race condition with 2 drives on one controller in wd_get_parms. PR 3773, from Onno van der Linden (onno@simplex.nl)
1.161	17-Jul-1997	jtk	use locator defines in "locators.h" to index cf_loc[] still to be done and/or decided not to be done: replace all *UNK uses in isa code with the native names from locators.h
1.160	24-Jun-1997	thorpej	branches: 1.160.2; foosize()'s return value is in DEV_BSIZE units; adjust the size obtained from the disklabel accordingly.
1.159	18-Jun-1997	pk	Remove spurious line left over in previous commit.
1.158	18-Jun-1997	pk	wdsize(): only call wdopen()/wdclose() if not already open, per the current `openmask'.
1.157	06-Jun-1997	thorpej	Pull thorpej-bus-dma branch into mainline.
1.156	17-Jan-1997	perry	branches: 1.156.6; Fixes to wd driver from cgd -- corrects several nasty bugs typically associated with coming out of suspends/sleeps on laptops. Approved by fvdl.
1.155	27-Nov-1996	thorpej	In wdattach(), when printing out info about the drive, make the output look a little more like a SCSI disk's. This has the pleasant side-effect of preventing the output from annoyingly extending past 80 columns.
1.154	07-Nov-1996	mikel	Fix WDDEBUG code; PR kern/1480.
1.153	13-Oct-1996	christos	backout kprintf changes
1.152	10-Oct-1996	christos	printf -> kprintf, sprintf -> ksprintf
1.151	27-Aug-1996	cgd	change cfprint_t type definition to take a const char , rather than a char , because that's what was really intended, and because if the print function modifies the string, various things could become unhappy (so the string should _not_ be modified).
1.150	12-May-1996	mycroft	Use intr.h.
1.149	29-Apr-1996	christos	- gcc -Wall fixes
1.148	11-Apr-1996	cgd	update for addition of a machine-dependent cookie as the first argument to isa_intr_{,dis}establish().
1.147	17-Mar-1996	thorpej	New device attachment scheme: - split softc size and match/attach out from cfdriver into a new struct cfattach. - new "attach" directive for files.*. May specify the name of the cfattach structure, so that devices may be easily attached to parents with different autoconfiguration semantics.
1.146	01-Mar-1996	mycroft	Add support to the ISA DMA framework for auto-initialize mode. Add experimental SB16 code, disabled for now.
1.145	08-Jan-1996	mycroft	Make sure we disk_busy() for every transfer we start.
1.144	07-Jan-1996	thorpej	New generic disk framework. Highlights: - New metrics handling. Metrics are now kept in the new `struct disk'. Busy time is now stored as a timeval, and transfer count in bytes. - Storage for disklabels is now dynamically allocated, so that the size of the disk structure is not machine-dependent. - Several new functions for attaching and detaching disks, and handling metrics calculation. Old-style instrumentation is still supported in drivers that did it before. However, old-style instrumentation is being deprecated, and will go away once the userland utilities are updated for the new framework. For usage and architectural details, see the forthcoming disk(9) manual page.
1.143	24-Dec-1995	mycroft	The IST_* and IPL_* constants are not bus-specific; don't treat them as such. Change splimp -> splnet in Ethernet, ARCnet, and FDDI drivers.
1.142	05-Aug-1995	mycroft	Use an intermediate variable to shorten label initialization code. Make wddump() work on more machines.
1.141	04-Jul-1995	mycroft	Make each disk and tape driver define its own read and write functions. Deprecate rawread() and rawwrite() completely. Remove d_strategy from cdevsw to force the abstraction barrier.
1.140	26-Jun-1995	cgd	implement new-style block device dump routine if __BDEVSW_DUMP_OLD_TYPE not defined, otherwise use an "not implemented" stub.
1.139	17-Apr-1995	cgd	clean up several ISA device interfaces: autoconfiguration, header inclusion, and interrupt configuration. more work still needs to be done, but it's getting better...
1.138	15-Apr-1995	mycroft	Don't boundary check I/O to the `raw' partition.
1.137	10-Apr-1995	mycroft	kernel_pmap --> pmap_kernel()
1.136	01-Apr-1995	mycroft	Add a missing unlock.
1.135	29-Mar-1995	mycroft	Revamp the locking mechanism slightly. Update some outdated comments, and add a few new ones.
1.134	25-Mar-1995	mycroft	Remove the write protect check altogether, and rely on the drive to do it.
1.133	23-Mar-1995	mycroft	Check for read-only media in open(), not write().
1.132	23-Mar-1995	mycroft	Allow DIOCWDINFO even if label is not `writable'.
1.131	27-Feb-1995	cgd	use wdc_softc in wdccd, rather than wd_softc; from chuck cranor. (pr 828)
1.130	29-Jan-1995	cgd	lip service to making ISA support machine-independent. This is the first round: moving the drivers into a machine-independent directory. Some drivers (e.g. fd.c) not moved because they use other pc features (e.g. CMOS settings), and none of the non-driver files moved, because they're still pretty much PC specific. eventually (when other ports with ISA busses really start using this code), more 'high-level' ISA support will live here.
1.129	13-Jan-1995	mycroft	(Almost) allow sector sizes other than 512 bytes. Reduce some expensive operations.
1.128	13-Jan-1995	mycroft	Always boundary check I/O.
1.127	13-Jan-1995	mycroft	Fix another bug in the DMA code.
1.126	13-Jan-1995	mycroft	Minor rearrangement.
1.125	13-Jan-1995	mycroft	Change the way sc_skip is used. Fix a bug in the (incomplete) DMA code. Add a new sc_blkno (which will be useful later).
1.124	07-Jan-1995	mycroft	Fix a race condition between normal I/O and wd_get_parms().
1.123	03-Jan-1995	mycroft	Add interrupt sharing types.
1.122	14-Dec-1994	mycroft	Remove dkbad.h.
1.121	01-Dec-1994	mycroft	Ignore ERR during diagnostics.
1.120	30-Nov-1994	mycroft	Clear the controller active flag after a timeout.
1.119	23-Nov-1994	mycroft	Officially credit Onno.
1.118	23-Nov-1994	mycroft	There's no point in bothering to unlock what we didn't lock.
1.117	23-Nov-1994	mycroft	Fix an outdated comment.
1.116	23-Nov-1994	mycroft	Use logical block addressing if available. Be a little more careful about maintaining timeouts. Clean up wddump() a bit.
1.115	22-Nov-1994	mycroft	Implement 32-bit reads and writes, but for now assume 16-bit mode.
1.114	22-Nov-1994	mycroft	Add a missing newline in some error messages.
1.113	22-Nov-1994	mycroft	Implement EIDE multi-sector PIO and DMA modes, based on code from Onno van der Linden. Also, various structural cleanup, and fix two bogons that crept in, and remove all of the old clustering nonsense.
1.112	22-Nov-1994	mycroft	Minor fix to debugging code.
1.111	22-Nov-1994	mycroft	Fudge a couple more fields for ancient disks.
1.110	22-Nov-1994	mycroft	Change some constant names, etc., to better match the ATA spec.
1.109	22-Nov-1994	mycroft	Add still more locking.
1.108	20-Nov-1994	mycroft	Add some missing locking, and some general cleanup.
1.107	18-Nov-1994	mycroft	Convert port, IRQ, and DRQ numbers to ints.
1.106	04-Nov-1994	mycroft	Fix stupid bug in last change.
1.105	04-Nov-1994	mycroft	Add a simplified version of wdcommand() that can be used in some cases. Avoid needing a wd_softc inside wdprobe().
1.104	04-Nov-1994	mycroft	No longer needs icu.h.
1.103	03-Nov-1994	mycroft	intr() arg does not need to be void.
1.102	03-Nov-1994	mycroft	Use indirect config for subdevices, and add dkdrivers where missing.
1.101	30-Oct-1994	cgd	be more careful with types, also pull in headers where necessary.
1.100	27-Oct-1994	cgd	new RCS ID format.
1.99	20-Oct-1994	mycroft	Be careful not to stack multiple timeouts.
1.98	20-Oct-1994	mycroft	Remove the local timeout mechanism, and just use explicit timeout()/untimeout() calls.
1.97	20-Oct-1994	mycroft	Always set d_secsize to 512 for now.
1.96	20-Oct-1994	mycroft	Don't switch to single-sector mode just because we got a corrected error.
1.95	20-Oct-1994	mycroft	Rearrange wdcstart() in a more logical fashion, and eliminate many redundant calculations. Also, speed up bad block searching a little.
1.94	20-Oct-1994	mycroft	openpart -> openmask
1.93	20-Oct-1994	mycroft	Various changes: * Add a dkdevice; move sc_openpart, sc_label, and sc_cpulabel into it. Turn sc_wlabel into a flag, and don't set it automatically in DIOCWLABEL. * If there is no label, disallow all I/O except to the raw partition. * Don't allow I/O to `unused' partitions. * Beginnings of support for block sizes other than 512 bytes. Other minor changes.
1.92	14-Oct-1994	cgd	move arch dependence to archs. clean up a bit. deal with weird MAXPARTITONS.
1.91	09-Oct-1994	mycroft	Remove unneeded #includes.
1.90	07-Oct-1994	mycroft	Don't poll the BUSY bit until the SDH register is set.
1.89	12-Sep-1994	mycroft	Turn off the delay debugging messages again.
1.88	15-Aug-1994	mycroft	Eliminate struct pte and struct pde.
1.87	15-Aug-1994	mycroft	Don't use altsts. Sigh.
1.86	26-Jul-1994	mycroft	Normalize label handling.
1.85	21-Jul-1994	mycroft	Return EINVAL if the requested transfer size is weird.
1.84	16-Jun-1994	mycroft	branches: 1.84.2; b_un.b_addr -> b_data
1.83	09-Jun-1994	pk	Set `uio_procp' field in case we ever need it.
1.82	05-May-1994	mycroft	Remove now-bogus casts.
1.81	05-May-1994	cgd	lots of changes: prototype migration, move lots of variables, definitions, and structure elements around. kill some unnecessary type and macro definitions. standardize clock handling. More changes than you'd want.
1.80	25-Apr-1994	mycroft	Fix uninitialized pointer, pointed out by Michael VanLoon.
1.79	24-Apr-1994	mycroft	Don't need isa.h.
1.78	20-Apr-1994	mycroft	Use TAILQ_*() to maintain the drive activation queues.
1.77	09-Apr-1994	mycroft	Don't print a duplicate message about stray interrupts.
1.76	07-Apr-1994	mycroft	Implement dynamic IRQ configuration and IRQ sharing. Inline spl*() calls. Reorganize and clean up the relevant code.
1.75	30-Mar-1994	mycroft	Turn off the timeout debugging info by default.
1.74	29-Mar-1994	mycroft	Updates for new autoconfig.
1.73	12-Mar-1994	mycroft	Prefix the busy-wait messages with `ignore this:' so people stop reporting it as a bug.
1.72	12-Mar-1994	mycroft	Fix problem with reading the disk label if it's not at the beginning of the disk.
1.71	12-Mar-1994	mycroft	Remove backward compatibility hacks for config(8), as they are no longer useful.
1.70	12-Mar-1994	mycroft	Update to match config(8).
1.69	11-Mar-1994	mycroft	Various: Put WDF_ERROR in the wdc_softc, to avoid gratuitously forcing non-active drives into single-sector mode on the next transfer. Arrange to wait for an interrupt after wdsetctlr(); this avoids long busy-wait delays, and gets rid of the `extra interrupt' messages (except for one immediately after autoconfig on some machines). Replace some uses of wdsetctlr() with simply lowering wd->sc_state; no point in forcing this immediately. Allow control operations to time out. Enable the warnings for long busy-wait delays by default. Some other minor things not worth mentioning.
1.68	10-Mar-1994	mycroft	Add a heuristic for detecting old config(8)s. This is very temporary.
1.67	10-Mar-1994	mycroft	Now that the disk probe routines are actually called, they need to return something.
1.66	10-Mar-1994	mycroft	More updates for new config.
1.65	10-Mar-1994	mycroft	Update for new config.
1.64	10-Mar-1994	mycroft	Cleanup to fit standard coding conventions, and many bugs fixed.
1.63	07-Mar-1994	mycroft	Some changes inspired by Bruce Evans's driver: Consolidate error reporting in one function. Actually use the dk_status and dk_error fields, and pass around only the error bit to avoid lots of unneeded assignments and tests. No functional differences.
1.62	07-Mar-1994	mycroft	Remove DKFL_QUIET.
1.61	06-Mar-1994	mycroft	DELAY() --> delay(). This is not a macro.
1.60	05-Mar-1994	mycroft	Move the drive select and various other things into wdcommand(), to avoid duplicating code.
1.59	04-Mar-1994	mycroft	Make sure we clear the interrupt when necessary. Why this isn't needed on my hardware is a mystery.
1.58	04-Mar-1994	mycroft	Add another missing splx().
1.57	04-Mar-1994	mycroft	Remember to splx() on some error conditions.
1.56	04-Mar-1994	mycroft	Start the timeout loop after we've decided there is a disk there.
1.55	04-Mar-1994	mycroft	De-ANSIfy. Move the timeout counter into struct disk. Remove unneeded args to wd[gs]etctlr(). Other minor changes.
1.54	03-Mar-1994	mycroft	Watch `seek complete' bit per IBM documentation. Clean up wdioctl(). Other minor cleanup.
1.53	02-Mar-1994	mycroft	Be more careful about clearing the controller and drive active flags.
1.52	02-Mar-1994	mycroft	Don't busy-wait for read or write completion.
1.51	02-Mar-1994	mycroft	Make wdc_wait() do an implicit wait_for_unbusy(). Pass structure pointers rahter than unit numbers where appropriate. Fix conflict if two drives have I/O pending at the same time. Add some more sanity checks. Some other minor cleanup.
1.50	26-Feb-1994	mycroft	Get the residual count right.
1.49	26-Feb-1994	mycroft	Clean up busy-wait code a little, and we need a special disksort() to skip over bufs in a combined transfer.
1.48	26-Feb-1994	mycroft	Fix bug from 1.43 correctly.
1.47	26-Feb-1994	mycroft	Fix typo.
1.46	26-Feb-1994	mycroft	Various cleanup, and fix last night's bug.
1.45	26-Feb-1994	mycroft	Fix typo.
1.44	26-Feb-1994	mycroft	Some more stylistic cleanup.
1.43	25-Feb-1994	mycroft	Fix an obvious bug that could stall the queue.
1.42	25-Feb-1994	mycroft	Remove some redundant code.
1.41	25-Feb-1994	mycroft	Increase polling delay.
1.40	25-Feb-1994	mycroft	Generalize busy-wait loops.
1.39	25-Feb-1994	mycroft	Don't bother #ifdefing things that are needed and always enabled.
1.38	25-Feb-1994	mycroft	Don't reimplement disksort().
1.37	25-Feb-1994	mycroft	`We are the KNF. You will be assimilated.'
1.36	10-Feb-1994	mycroft	Back pointers are not useful for disksort().
1.35	09-Feb-1994	mycroft	All ioctl routines take a struct proc * now.
1.34	06-Feb-1994	mycroft	Remove another use of b_actl.
1.33	06-Feb-1994	mycroft	Use b_actf, not av_forw.
1.32	03-Jan-1994	mycroft	Inline all the [lr]cr[023]() `functions' in cpufunc.h.
1.31	03-Jan-1994	mycroft	Inline all the [lr]cr[023]() `functions' in cpufunc.h.
1.30	24-Dec-1993	ws	Correct dumping to WD (was destroyed by MACHINE_NONCONTIG code)
1.29	20-Dec-1993	mycroft	Canonicalize all #includes, and add pio.h where appropriate.
1.28	16-Dec-1993	mycroft	Eliminate reference to Maxmem.
1.27	13-Dec-1993	cgd	from Soeren Schmidt <sos@login.dkuug.dk>: Basically it does a timeout on lost interrupts, starting the operation again and logging and error message on the console. additionally fixes some of the (newer ending) while loops (that made it work with two IDE disks !)
1.26	01-Aug-1993	mycroft	branches: 1.26.2; Add RCS identifiers (this time on the correct side of the branch), and incorporate recent changes in netbsd-0-9 branch.
1.25	28-Jul-1993	cgd	incorporate changes from 0-9-base to 0-9-ALPHA
1.24	08-Jul-1993	deraadt	branches: 1.24.2; documentation error
1.23	07-Jul-1993	deraadt	added a timeout debugging mode to wd.c so that people can see how slow their controllers really are. Look for the WDCNDELAY_DEBUG definition.
1.22	07-Jul-1993	deraadt	ok, 5 second command timeouts..
1.21	06-Jul-1993	deraadt	number of changes: 1. controller reset code moved into one function, called from more places now. 2. more places now do timeouts. TIPCAT code turned on. 3. blew away the unused WDOPENLOCK test code. 4. delay is now done using WDCNDELAY repetitions of DELAY(25). Up to 2.5 second pause accepted from controller (some controllers have some commands that are actually this slow) 5. some of the old timeout code was really whacked.
1.20	05-Jul-1993	deraadt	added a missing {/} pair that barsoom seems to have taken out, but which the standard 386bsd 0.1 driver did have. This is for a piece of code in the bad block handling. found/fixed by hpeyerl.
1.19	05-Jul-1993	deraadt	all timeout code now calls DELAY() so the timeout delays are the same length on all processors. Where possible, the controller is reset if it times out.
1.18	05-Jul-1993	deraadt	increased the timeout inside wdcommand(). Some controllers impliment some commands very slowly.
1.17	18-Jun-1993	cgd	fix the divide-by-zero problem w/old disks. this patch is ws's "fault."
1.16	28-May-1993	deraadt	1. It is now possible to build a kernel that does not have a pc0 device driver. 2. "press any key to reboot" reads the key from the console. 3. wddump() still needs a non-blocking getc() routine (or flush)
1.15	22-May-1993	cgd	add rcsids to everything and clean up headers
1.14	20-May-1993	deraadt	First cut at cpu independent disklabels. There will be niggly little details no doubt..
1.13	10-May-1993	deraadt	This patch allows the wd device driver to tell how big the disks are on WD1007-derived controllers. In this example, wdc0 is a WD1007-clone, and wdc1 is a WD1003-clone. WD1007 controllers are generally ESDI and IDE controllers. wdc0 at 0x1f0-0x1f7 irq 14 on isa wd0 at wdc0 targ 0: 322MB 1224 cyl, 15 head, 36 sec <disktype> wdc1 at 0x170-0x17f irq 7 on isa wd2 at wdc1 targ 0: (unknown size) <disktype>
1.12	09-May-1993	deraadt	i thought i deleted this line before!
1.11	04-May-1993	deraadt	support for making dev->id_alive be set, this is for iostat to find disk devices. wee bit of a kludge. sub-device attach() routines must now return 1 for successful attach(), 0 otherwise. Other bsd's do this too..
1.10	18-Apr-1993	deraadt	slight code quality improvement. not related to the bug with very old controllers or disks! :-)
1.9	15-Apr-1993	deraadt	Once more: some changes to the ioconf.c structure. It will now handle devices hanging off controllers, any kind of controller. A device on isa0 is called at probe(self), then attach(self) A controller on isa0 is called at probe(self), then all it's children with defined unit numbers are initialized by calling attach(subdev); next all subdevices with unit ? are initialized by calling attach(sundev). Almost all device entry points is now like the vax/sun model (intr being the weird one)
1.8	15-Apr-1993	deraadt	wd driver would fail on calling wdgetctlr() for older drives and controllers. New behaviour is that if the standard method fails, try to recalibrate the drive. Success means the drive exists. patch by hpeyerl@novatel.cuc.ab.ca (who owns several of these historical artifacts)
1.7	12-Apr-1993	deraadt	new scsi subsystem. changes also in config/mkioconf.c & sys/scsi/*
1.6	09-Apr-1993	cgd	have probe return size of io space on successful return, rather than simply 1.
1.5	08-Apr-1993	deraadt	dmesg output at boottime now tries to print out information as soon as it is available. The output looks much more like Sunos.
1.4	06-Apr-1993	deraadt	ripped out a message
1.3	06-Apr-1993	deraadt	Mod-barsoom wd additions. config program + kernel config files changed too. fd support had to change slightly too.
1.2	21-Mar-1993	cgd	after 0.2.2 "stable" patches applied
1.1	21-Mar-1993	cgd	branches: 1.1.1; Initial revision
1.1.1.1	21-Mar-1993	cgd	initial import of 386bsd-0.1 sources
1.24.2.3	28-Jul-1993	deraadt	wd.c: fixes 1 drive systems fd.c: improves reliability changes from wolfgang
1.24.2.2	22-Jul-1993	cgd	fix comment for previous commit...
1.24.2.1	20-Jul-1993	cgd	change WDCNDELAY to 400000 -- delay -> 10s, at theo's request
1.26.2.5	06-Feb-1994	mycroft	Use b_actf, not av_forw.
1.26.2.4	13-Dec-1993	cgd	partial (hopefully correct) update from trunk
1.26.2.3	27-Oct-1993	mycroft	Minor cleanup.
1.26.2.2	24-Sep-1993	mycroft	clock.c: Finish conversion. Make rtcput() work if VRT bit not set. Speed up time conversion in delay(). fd.c: Minor changes to autoconfig. Use dv_parent rather than storing another pointer to the fdc_softc in each fd_softc. Change disk label stuff for our kernel. icu.s: Make splx() and splnone() return previous cpl for now. isa.c: Minor autoconfig changes. isa.h: rtcin() is now nvram() and is declare in nvram.h. Make isaphysmem a caddr_t. pccons.c: XXXX: Maybe we should insist they open /dev/io now. timerreg.h: #define TIMER_NPORTS. wd.c: Delete unused wdnoreloc() and wddospart(), and use remaining two bits in unit number. struct disk --> struct wd_softc. struct board --> struct wdc_softc. Still needs much work.
1.26.2.1	14-Sep-1993	mycroft	New i386 code.
1.84.2.7	11-Oct-1994	mycroft	Update from trunk.
1.84.2.6	07-Oct-1994	mycroft	Update from trunk.
1.84.2.5	12-Sep-1994	cgd	from trunk.
1.84.2.4	15-Aug-1994	mycroft	update from trunk
1.84.2.3	15-Aug-1994	mycroft	update from trunk
1.84.2.2	26-Jul-1994	cgd	update from trunk.
1.84.2.1	22-Jul-1994	cgd	from trunk.
1.156.6.1	13-May-1997	thorpej	Adapt to new software interface to the on-board ISA DMA controller.
1.160.2.3	17-Jul-1997	bouyer	Use "locator.h".
1.160.2.2	17-Jul-1997	bouyer	Update from trunck.
1.160.2.1	01-Jul-1997	bouyer	Add the ide atapi driver from my original atapi driver, ported to scsipi.
1.162.2.2	14-Oct-1997	thorpej	Update marc-pcmcia branch from trunk.
1.162.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.175.2.13	04-Oct-1998	bouyer	atavar.h: drv_softc is a struct device * instead of void*, as it's mostly used for dv_xname wd.c: convert for drv_softc type change, printf cleanup wdc.c: always call ata_get_params() (params was used initialised with non-32bit controllers, leaving to bogus PIO/DMA mode report). Cleaup of the PIO/DMA mode message. pciide_piix_reg.h: Fix definition of PIIX_IDETIM_CLEAR (unused before) pciide.c: add a method do disable a channel on know device. If a channel doesn't have any drive, we disable it and free its resources if disable was successfull. This should help with laptops where the second channel of the PIIX4 is unused but not disabled by BIOS. On such laptops, irq15 can be used for PCMCIA but it was claimed by pciide. Misc printf cleanup. wdc_isa.c: printf cleanup.
1.175.2.12	20-Sep-1998	bouyer	Kill the 'old drive' flag. The probe for IDE controllers now looks like this : - if status = 0xff for both master and slave, no drive at all -> exit. - issue a reset (timeout = 31s). - test signatures: if ATAPI signature, flag a (possible) ATAPI devices here else test register writability, then wait for the drive to become ready, issue a diagnostic command and wait for the drive to become ready if ok, flag a real ATA device. This looks more like the old probe. For ATA devices, if IDENTIFY fails, then it's an old drive and use a fake disklabel/geometry.
1.175.2.11	20-Sep-1998	bouyer	Init up xfer->drive in wdc_ata_bio(), or bad things will happen for drive 1. No need to init xfer->channel, it's done in wdc_exec_xfer().
1.175.2.10	20-Sep-1998	bouyer	- always call config_found() from wdcattach(), instead of printing our own the "not configured" message ourselve. When no atapibus is configured, use a fake ata_atapi_attach, else call wdc_atapibus_attach(). This way we don't have to include the whole atapi support in an ata-only config. - the dma_* function pointers take a 'int flags' as last argument, instead of an 'int read'. The 2 valid flags are: WDC_DMA_READ (to select read/write) and WDC_DMA_POLL, to signal interrupt-less mode (for core dumps). - Reworked wdcprobe() so look more like the old one. A status of 0xff is interpreted as "no drive" (freebsd does this); this this speed up the probe for non-IDE machines (a reset timeout has been bumped to 31s, to match the specs). The probe set ups the drive flags to either ATA or ATAPI (depend on register signature) or OLD (if register signature don't match, but a working drive appears to be there). Later the ATA code will look at the OLD flag to decide wether there is no drive, or an old (ST506) one if WDC_IDENTIFY fails. - For known PCI chips, don't use the wdcprobe() heuristic to find wether a channel is enabled/disabled; use the chip's specific registers for this. - Clear the command queue when we have a polled command. This allows dumps to work even if a command was being handled by the controller at panic time. - Increase the number of available DMA segments by one, as user requests may not be aligned on a page boundary.
1.175.2.9	11-Sep-1998	bouyer	Sync with HEAD.
1.175.2.8	21-Aug-1998	bouyer	Sync with -current
1.175.2.7	13-Aug-1998	bouyer	- sync with HEAD - better error detection and report for wd drives - better wddump() support - wdcwait() now takes a timout argument, so that we can honnor the timeout argument of scsipi cmds.
1.175.2.6	13-Jun-1998	bouyer	Sync with HEAD.
1.175.2.5	10-Jun-1998	bouyer	Print the minimum and recomended DMA timings if DEBUG_PROBE is defined.
1.175.2.4	09-Jun-1998	bouyer	- sync with trunck - don't try to set the drive's PIO mode if we don't know how to set it on the controller. This seems to confuse some 'intelligent' ISA controllers.
1.175.2.3	05-Jun-1998	bouyer	- call wdc_ata_bio_done() only once. - better error recovery: if an error occured, retry the transfer in single sector PIO mode.
1.175.2.2	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.175.2.1	04-Jun-1998	bouyer	file wd.c was added on branch bouyer-ide on 1998-06-04 16:52:35 +0000
1.193.2.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.196.6.2	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.196.6.1	21-Dec-1999	wrstuden	Initial commit of recent changes to make DEV_BSIZE go away. Runs on i386, needs work on other arch's. Main kernel routines should be fine, but a number of the stand programs need help. cd, fd, ccd, wd, and sd have been updated. sd has been tested with non-512 byte block devices. vnd, raidframe, and lfs need work. Non 2**n block support is automatic for LKM's and conditional for kernels on "options NON_PO2_BLOCKS".
1.196.4.1	15-Nov-1999	fvdl	Sync with -current
1.196.2.4	18-Jan-2001	bouyer	Sync with head (for UBC+NFS fixes, mostly).
1.196.2.3	05-Jan-2001	bouyer	Sync with HEAD
1.196.2.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.196.2.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.204.2.1	22-Jun-2000	minoura	Sync w/ netbsd-1-5-base.
1.205.2.2	06-May-2001	he	Pull up revision 1.213 (requested by fvdl): Compensate for the fact that after __HAVE_OLD_DISKLABEL, DISKMINOR(u,p) is no longer contiguous over ``p'', so ranges cannot be passed to vdevgone().
1.205.2.1	01-May-2001	he	Pull up revisions 1.211-1.212 (requested by fvdl): Increase the number of BSD disklabel partitions on i386 to 16.
1.212.2.10	07-Jan-2003	thorpej	Sync with HEAD.
1.212.2.9	19-Dec-2002	thorpej	Sync with HEAD.
1.212.2.8	11-Nov-2002	nathanw	Catch up to -current
1.212.2.7	18-Oct-2002	nathanw	Catch up to -current.
1.212.2.6	17-Sep-2002	nathanw	Catch up to -current.
1.212.2.5	01-Aug-2002	nathanw	Catch up to -current.
1.212.2.4	28-Feb-2002	nathanw	Catch up to -current.
1.212.2.3	08-Jan-2002	nathanw	Catch up to -current.
1.212.2.2	14-Nov-2001	nathanw	Catch up to -current.
1.212.2.1	21-Jun-2001	nathanw	Catch up to -current.
1.214.4.3	26-Sep-2001	fvdl	* add a VCLONED vnode flag that indicates a vnode representing a cloned device. * rename REVOKEALL to REVOKEALIAS, and add a REVOKECLONE flag, to pass to VOP_REVOKE * the revoke system call will revoke all aliases, as before, but not the clones * vdevgone is called when detaching a device, so make it use REVOKECLONE to get rid of all clones as well * clean up all uses of VOP_OPEN wrt. locking. * add a few VOPS to spec_vnops that need to do something when it's a clone vnode (access and getattr) * add a copy of the vnode vattr structure of the original 'master' vnode to the specinfo of a cloned vnode. could possibly redirect getattr to the 'master' vnode, but this has issues with revoke * add a vdev_reassignvp function that disassociates a vnode from its original device, and reassociates it with the specified dev_t. to be used by cloning devices only, in case a new minor is allocated. * change all direct references in drivers to v_devcookie and v_rdev to vdev_privdata(vp) and vdev_rdev(vp). for diagnostic purposes when debugging race conditions that still exist wrt. locking and revoking vnodes. * make the locking state of a vnode consistent when passed to d_open and d_close (unlocked). locked would be better, but has some deadlock issues
1.214.4.2	18-Sep-2001	fvdl	Various changes to make cloning devices possible: * Add an extra argument (struct vnode *) to VOP_OPEN. If it is not NULL, specfs will create a cloned (aliased) vnode during the call, and return it there. The caller should release and unlock the original vnode if a new vnode was returned. The new vnode is returned locked. Add a flag field to the cdevsw and bdevsw structures. DF_CLONING indicates that it wants a new vnode for each open (XXX is there a better way? devprop?) * If a device is cloning, always call the close entry point for a VOP_CLOSE. Also, rewrite cons.c to do the right thing with vnodes. Use VOPs rather then direct device entry calls. Suggested by mycroft@ Light to moderate testing done an i386 system (arch doesn't matter though, these are MI changes).
1.214.4.1	07-Sep-2001	thorpej	Commit my "devvp" changes to the thorpej-devvp branch. This replaces the use of dev_t in most places with a struct vnode *. This will form the basic infrastructure for real cloning device support (besides being architecurally cleaner -- it'll be good to get away from using numbers to represent objects).
1.214.2.4	10-Oct-2002	jdolecek	sync kqueue with -current; this includes merge of gehenna-devsw branch, merge of i386 MP branch, and part of autoconf rototil work
1.214.2.3	06-Sep-2002	jdolecek	sync kqueue branch with HEAD
1.214.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.214.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.220.10.7	14-May-2004	jdc	Pull up revision 1.275 via patch (requested by bouyer in ticket #1691) When using LBA48, use Flush cache extended command if supported. From Christos Zoulas in PR 25366, with minor adjustements from me.
1.220.10.6	28-Mar-2004	jmc	Pullup rev 1.260 via patch (requested by bouyer in ticket #1629) For LBA drives, compute the number of cylinders from the drive capacity in the default disklabel and the boot message, instead of using the value reported by the drive (which is 16383 if the drive is larger than 8G). Fixes PR#9864
1.220.10.5	28-Mar-2004	jmc	Pullup rev 1.258 via patch (requested by bouyer in ticket #1628) Add LBA48_THRESHOLD #define and set it to the correct value. Use it in place of 0xffffff (which is one too few 'f').
1.220.10.4	15-Aug-2003	tron	Pull up revision 1.256 (requested by bouyer in ticket #1370): Don't use LBA48 if we don't have to (that is, if the request is inside the first 128GB). Tested by Andreas Johansson.
1.220.10.3	30-Jun-2003	grant	Apply patch (requested by bouyer in ticket #1354): Add support for the getcache/setcache ioctls. Acceptable cache operations are r and rw, not none and w Properly initialise the timeout value. Fix typo.
1.220.10.2	22-Jul-2002	lukem	Pull up revision 1.222 (requested by yamt in ticket #536): do previous differently for wddump case.
1.220.10.1	22-Jul-2002	lukem	Pull up revision 1.221 (requested by yamt in ticket #536): - remove big buffers on stack. discussed on tech-kern. - constify.
1.220.8.4	29-Aug-2002	gehenna	catch up with -current.
1.220.8.3	20-Jul-2002	gehenna	catch up with -current.
1.220.8.2	15-Jul-2002	gehenna	catch up with -current.
1.220.8.1	16-May-2002	gehenna	Add the block/character device switches. Replace the access to devsw table with calling devsw API.
1.257.2.14	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.257.2.13	01-Apr-2005	skrll	Sync with HEAD.
1.257.2.12	09-Mar-2005	skrll	Adapt commented out code to branch.
1.257.2.11	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.257.2.10	29-Nov-2004	skrll	Sync with HEAD.
1.257.2.9	02-Nov-2004	skrll	Sync with HEAD.
1.257.2.8	19-Oct-2004	skrll	Sync with HEAD
1.257.2.7	21-Sep-2004	skrll	Fix the sync with head I botched.
1.257.2.6	18-Sep-2004	skrll	Sync with HEAD.
1.257.2.5	03-Sep-2004	skrll	Sync with HEAD
1.257.2.4	25-Aug-2004	skrll	Sync with HEAD.
1.257.2.3	12-Aug-2004	skrll	Sync with HEAD.
1.257.2.2	03-Aug-2004	skrll	Sync with HEAD
1.257.2.1	02-Jul-2003	darrenr	Apply the aborted ktrace-lwp changes to a specific branch. This is just for others to review, I'm concerned that patch fuziness may have resulted in some errant code being generated but I'll look at that later by comparing the diff from the base to the branch with the file I attempt to apply to it. This will, at the very least, put the changes in a better context for others to review them and attempt to tinker with removing passing of 'struct lwp' through the kernel.
1.274.2.14	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #11369): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.274.2.13	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10747): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.274.2.12	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10746): sys/dev/ata/wd.c: revision 1.326 Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.274.2.11	06-Aug-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10644): sys/dev/ata/wd.c: revision 1.325 via patch Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.274.2.10	24-May-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10625): sys/dev/ata/wd.c: revision 1.324 via patch Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.274.2.9	28-Mar-2006	riz	Pull up patch (requested by bouyer in ticket #10397): sync the list of quirky seagate drives with -current
1.274.2.8	01-Oct-2004	jmc	branches: 1.274.2.8.2; Pullup rev 1.295 (requested by bouyer in ticket #882) Add a WD_QUIRK_FORCE_LBA48 quirck, which forces all access to use LBA48. Use it for ST3200822A, which has a broken firmware and return an error when accessing sector 0xfffffff with a LBA command.
1.274.2.7	17-Sep-2004	jmc	Pullup patch (requested by bouyer in ticket #853) In case of error, reset the channel with AT_RST_NOCMD. We know the reset wont abort any command at this point.
1.274.2.6	11-Aug-2004	jmc	Pullup rev 1.282-1.283 (requested by bouyer in ticket #735) Move cleaning of the bad block list in wddetach() after the call to disk_detach(), to avoid a race condition with the interrupt code. wddetach(): move lockmgr(LK_DRAIN) near the end; otherwise we'll panic in wddone() (vdevgone() may end up calling it) trying to get the lock. wdopen(): check for DVF_ACTIVE.
1.274.2.5	11-Aug-2004	jmc	Pullup rev 1.281 (requested by bouyer in ticket #733) Implement asynchronous channel reset. Use this to reset the channel before doing a dump, instead of the hack in wdc_exec_xfer() based on C_POLL. This hack was causing problems on controllers with a shared queue, because we now can have C_POLL set during concurent channels probes
1.274.2.4	23-Jul-2004	tron	Pull up revision 1.278 (requested by mycroft in ticket #704): Fix an extremely obvious bug in the handling of the bad block list: the "max" block was being set 512x further out than it should be, causing rather severe escalation of the error.
1.274.2.3	02-Jul-2004	he	Pull up revisions 1.279-1.280 (requested by mycroft in ticket #572): Several fixes mostly related to USB: o Don't force ATA_SINGLE on devices which don't support READ/WRITE MULTI. Increases performance with several CF cards substantially. o Update copyright.
1.274.2.2	29-May-2004	tron	Pull up revision 1.276 (requested by bouyer in ticket #398): Add support for the DIOCCACHESYNC ioctl.
1.274.2.1	09-May-2004	jdc	Pull up revision 1.275 (requested by bouyer in ticket #262) When using LBA48, use Flush cache extended command if supported. From Christos Zoulas in PR 25366, with minor adjustements from me.
1.274.2.8.2.12	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #11369): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.274.2.8.2.11	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10747): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.274.2.8.2.10	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10746): sys/dev/ata/wd.c: revision 1.326 Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.274.2.8.2.9	06-Aug-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10644): sys/dev/ata/wd.c: revision 1.325 via patch Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.274.2.8.2.8	24-May-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10625): sys/dev/ata/wd.c: revision 1.324 Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.274.2.8.2.7	28-Mar-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10396): sys/dev/ata/wd.c: revision 1.310 Use a pmatch(9) expression which should catch all present and future seagate drives larger than 200GB for the WD_QUIRK_FORCE_LBA48 quirks. Based on infos from http://www.seagate.com/support/kb/disc/howto/interpret_model.html
1.274.2.8.2.6	06-Sep-2005	riz	branches: 1.274.2.8.2.6.2; Pull up following revision(s) (requested by bouyer in ticket #5822): sys/dev/ata/wd.c: revision 1.309 Add quirk for ST3200826A. From rudolf (netbsd at eq.cz) on tech-kern.
1.274.2.8.2.5	06-Sep-2005	riz	Pull up following revision(s) (requested by bouyer in ticket #5821): sys/dev/ata/wd.c: revision 1.307 ST3300831A also needs WD_QUIRK_FORCE_LBA48.
1.274.2.8.2.4	05-Jul-2005	riz	Pull up revision 1.304 (requested by bouyer in ticket #2028): The ST3160023A* also need the LBA48 quirk. From Timo Schoeler.
1.274.2.8.2.3	05-Jul-2005	riz	Pull up revision 1.303 via patch (requested by bouyer in ticket #2027): Use pmatch() metacharacters to match both SATA and PATA versions of seagate drives in the quirk table.
1.274.2.8.2.2	16-May-2005	riz	Pull up revision 1.300 (requested by bouyer in ticket #1522): Add another broken seagate device to the quirk list. Problem reported by rudolf <netbsd at eq.cz> on netbsd-help@
1.274.2.8.2.1	06-Feb-2005	jmc	Pullup rev 1.297 (requested by bouyer in ticket #986) Some drives reports ATA version < 4, and still support (and need) the WDCC_FLUSHCACHE command. So for drive < ATA4, test WDC_CMD2_FC in atap_cmd_set2 before returning ENODEV.
1.274.2.8.2.6.2.6	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #11369): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.274.2.8.2.6.2.5	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10747): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.274.2.8.2.6.2.4	20-Nov-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10746): sys/dev/ata/wd.c: revision 1.326 Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.274.2.8.2.6.2.3	06-Aug-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #10644): sys/dev/ata/wd.c: revision 1.325 via patch Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.274.2.8.2.6.2.2	24-May-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10625): sys/dev/ata/wd.c: revision 1.324 Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.274.2.8.2.6.2.1	28-Mar-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #10396): sys/dev/ata/wd.c: revision 1.310 Use a pmatch(9) expression which should catch all present and future seagate drives larger than 200GB for the WD_QUIRK_FORCE_LBA48 quirks. Based on infos from http://www.seagate.com/support/kb/disc/howto/interpret_model.html
1.297.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.297.4.1	29-Apr-2005	kent	sync with -current
1.298.2.15	08-Apr-2008	jdc	Pull up revision 1.359 (requested by bouyer in ticket #1910). The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.298.2.14	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1848): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.298.2.13	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1583): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.298.2.12	30-Jun-2006	ghen	branches: 1.298.2.12.2; Pull up following revision(s) (requested by gendalia in ticket #1396): sys/dev/ata/wd.c: revision 1.326 Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.298.2.11	12-Jun-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #1375): sys/dev/ata/wd.c: revision 1.325 via patch Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.298.2.10	24-May-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1337): sys/dev/ata/wd.c: revision 1.324 Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.298.2.9	28-Mar-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1227): sys/dev/ata/wd.c: revision 1.310 Use a pmatch(9) expression which should catch all present and future seagate drives larger than 200GB for the WD_QUIRK_FORCE_LBA48 quirks. Based on infos from http://www.seagate.com/support/kb/disc/howto/interpret_model.html
1.298.2.8	27-Jan-2006	tron	Pull up following revision(s) (requested by abs in ticket #1132): sys/dev/ata/wd.c: revision 1.318 cast 16bit atap_capacity to u_int64_t before << 16. (assigned to a u_int64_t variable). Fixes 1 to 2 TB disks in qemu. Tested with 1=0000GB and to a lesser extent 2047GB sparse files. For reference an empty FFS1 filesystem on a 1200GB sparse file uses 19GB of disk.
1.298.2.7	08-Sep-2005	tron	branches: 1.298.2.7.2; Pull up following revision(s) (requested by bouyer in ticket #750): sys/dev/ata/wd.c: revision 1.309 Add quirk for ST3200826A. From rudolf (netbsd at eq.cz) on tech-kern.
1.298.2.6	08-Sep-2005	tron	Pull up following revision(s) (requested by bouyer in ticket #749): sys/dev/ata/wd.c: revision 1.307 ST3300831A also needs WD_QUIRK_FORCE_LBA48.
1.298.2.5	02-Jul-2005	tron	Pull up revision 1.304 (requested by bouyer in ticket #497): The ST3160023A* also need the LBA48 quirk. From Timo Schoeler.
1.298.2.4	02-Jul-2005	tron	Pull up revision 1.303 (requested by bouyer in ticket #496): Use pmatch() metacharacters to match both SATA and PATA versions of seagate drives in the quirk table.
1.298.2.3	16-Jun-2005	tron	Pull up revision 1.302 (requested by martti in ticket #459): Added LBA48 quirk for Seagate ST3200822AS. Tested by Teemu Rinta-aho.
1.298.2.2	19-May-2005	tron	Pull up revision 1.300 (requested by bouyer in ticket #310): Add another broken seagate device to the quirk list. Problem reported by rudolf <netbsd at eq.cz> on netbsd-help@
1.298.2.1	06-Apr-2005	tron	Pull up revision 1.299 (requested by yamt in ticket #112): introduce a function to drain bufq and use it where appropriate.
1.298.2.12.2.3	08-Apr-2008	jdc	Pull up revision 1.359 (requested by bouyer in ticket #1910). The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.298.2.12.2.2	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1848): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.298.2.12.2.1	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1583): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.298.2.7.2.8	08-Apr-2008	jdc	Pull up revision 1.359 (requested by bouyer in ticket #1910). The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.298.2.7.2.7	15-Oct-2007	riz	Pull up following revision(s) (requested by bouyer in ticket #1848): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.298.2.7.2.6	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1583): sys/dev/ata/wd.c: revision 1.334 Yet another broken seagate drive.
1.298.2.7.2.5	20-Nov-2006	ghen	Pull up following revision(s) (requested by bouyer in ticket #1582): sys/dev/ata/wd.c: revision 1.326 Add yet another broken 160GB seagate drive. Mine. ST3160021A.
1.298.2.7.2.4	12-Jun-2006	tron	Pull up following revision(s) (requested by bouyer in ticket #1375): sys/dev/ata/wd.c: revision 1.325 via patch Adjust quirk entry for seagate drives. Should fix kern/33637 by Shigeya Suzuki.
1.298.2.7.2.3	24-May-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1337): sys/dev/ata/wd.c: revision 1.324 Add yet another broken 160Gb seagate drive. I though we had found all of them ... reported by Damian Lubosch on current-users.
1.298.2.7.2.2	28-Mar-2006	riz	Pull up following revision(s) (requested by bouyer in ticket #1227): sys/dev/ata/wd.c: revision 1.310 Use a pmatch(9) expression which should catch all present and future seagate drives larger than 200GB for the WD_QUIRK_FORCE_LBA48 quirks. Based on infos from http://www.seagate.com/support/kb/disc/howto/interpret_model.html
1.298.2.7.2.1	27-Jan-2006	tron	Pull up following revision(s) (requested by abs in ticket #1132): sys/dev/ata/wd.c: revision 1.318 cast 16bit atap_capacity to u_int64_t before << 16. (assigned to a u_int64_t variable). Fixes 1 to 2 TB disks in qemu. Tested with 1=0000GB and to a lesser extent 2047GB sparse files. For reference an empty FFS1 filesystem on a 1200GB sparse file uses 19GB of disk.
1.305.2.10	24-Mar-2008	yamt	sync with head.
1.305.2.9	17-Mar-2008	yamt	sync with head.
1.305.2.8	27-Feb-2008	yamt	sync with head.
1.305.2.7	21-Jan-2008	yamt	sync with head
1.305.2.6	15-Nov-2007	yamt	sync with head.
1.305.2.5	27-Oct-2007	yamt	sync with head.
1.305.2.4	03-Sep-2007	yamt	sync with head.
1.305.2.3	26-Feb-2007	yamt	sync with head.
1.305.2.2	30-Dec-2006	yamt	sync with head.
1.305.2.1	21-Jun-2006	yamt	sync with head.
1.312.2.1	02-Nov-2005	yamt	sync with head.
1.315.2.4	01-Mar-2006	yamt	sync with head.
1.315.2.3	05-Feb-2006	yamt	adapt wd.
1.315.2.2	01-Feb-2006	yamt	sync with head.
1.315.2.1	15-Jan-2006	yamt	sync with head.
1.318.4.3	07-Jun-2006	kardel	Sync with head.
1.318.4.2	01-Jun-2006	kardel	Sync with head.
1.318.4.1	22-Apr-2006	simonb	Sync with head.
1.318.2.1	09-Sep-2006	rpaulo	sync with head
1.320.6.2	24-May-2006	tron	Merge 2006-05-24 NetBSD-current into the "peter-altq" branch.
1.320.6.1	31-Mar-2006	tron	Merge 2006-03-31 NetBSD-current into the "peter-altq" branch.
1.320.4.1	19-Apr-2006	elad	sync with head.
1.320.2.4	11-Aug-2006	yamt	sync with head
1.320.2.3	26-Jun-2006	yamt	sync with head.
1.320.2.2	24-May-2006	yamt	sync with head.
1.320.2.1	01-Apr-2006	yamt	sync with head.
1.324.2.1	19-Jun-2006	chap	Sync with head.
1.325.2.1	13-Jul-2006	gdamore	Merge from HEAD.
1.329.4.2	10-Dec-2006	yamt	sync with head.
1.329.4.1	22-Oct-2006	yamt	sync with head
1.329.2.2	06-Feb-2007	ad	Quieten noisy boot messages.
1.329.2.1	18-Nov-2006	ad	Sync with head.
1.335.4.2	03-Jun-2008	skrll	Sync with netbsd-4.
1.335.4.1	30-Sep-2007	wrstuden	Catch up on netbsd-4 as of a few days ago.
1.335.2.2	30-Mar-2008	jdc	Pull up revision 1.359 (requested by bouyer in ticket #1094). The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.335.2.1	25-Sep-2007	xtraeme	branches: 1.335.2.1.4; Pull up following revision(s) (requested by bouyer in ticket #899): sys/dev/ata/wd.c: revision 1.345 Add a workaround for drives with the LBA48 bug: if we get a "ID not found" error for a transfer crossing LBA48_THRESHOLD, and the drive is larger than 128GB, automatically add WD_QUIRK_FORCE_LBA48 to the drive's quirks and retry the transfers. Hopefully this will obsolete the WD_QUIRK_FORCE_LBA48 quirk list ...
1.335.2.1.4.1	30-Mar-2008	jdc	Pull up revision 1.359 (requested by bouyer in ticket #1094). The Hitachi HDP725025GLA380 returns "aborted command" instead of "id not found" when hitting the LBA48 bug, so also install the LBA48 bug workaround on "aborted command" errors.
1.336.2.2	15-Apr-2007	yamt	sync with head.
1.336.2.1	12-Mar-2007	rmind	Sync with HEAD.
1.337.4.1	11-Jul-2007	mjf	Sync with head.
1.337.2.11	23-Oct-2007	ad	Sync with head.
1.337.2.10	09-Oct-2007	ad	Sync with head.
1.337.2.9	24-Aug-2007	ad	Sync with buffer cache locking changes. See buf.h/vfs_bio.c for details. Some minor portions are incomplete and needs to be verified as a whole.
1.337.2.8	20-Aug-2007	ad	- Alter disk attach/detach to fix a panic when closing a vnd device. - Sync with HEAD.
1.337.2.7	19-Aug-2007	ad	- Back out the biodone() changes. - Eliminate B_ERROR (from HEAD).
1.337.2.6	15-Jul-2007	ad	Sync with head.
1.337.2.5	01-Jul-2007	ad	Adapt to callout API change.
1.337.2.4	17-Jun-2007	ad	- Increase the number of thread priorities from 128 to 256. How the space is set up is to be revisited. - Implement soft interrupts as kernel threads. A generic implementation is provided, with hooks for fast-path MD code that can run the interrupt threads over the top of other threads executing in the kernel. - Split vnode::v_flag into three fields, depending on how the flag is locked (by the interlock, by the vnode lock, by the file system). - Miscellaneous locking fixes and improvements.
1.337.2.3	13-May-2007	ad	- Pass the error number and residual count to biodone(), and let it handle setting error indicators. Prepare to eliminate B_ERROR. - Add a flag argument to brelse() to be set into the buf's flags, instead of doing it directly. Typically used to set B_INVAL. - Add a "struct cpu_info *" argument to kthread_create(), to be used to create bound threads. Change "bool mpsafe" to "int flags". - Allow exit of LWPs in the IDL state when (l != curlwp). - More locking fixes & conversion to the new API.
1.337.2.2	10-Apr-2007	ad	Sync with head.
1.337.2.1	13-Mar-2007	ad	Pull in the initial set of changes for the vmlocking branch.
1.340.2.2	10-Sep-2007	skrll	Sync with HEAD.
1.340.2.1	15-Aug-2007	skrll	Sync with HEAD.
1.343.8.2	30-Jul-2007	taca	Make it compile with WD_SOFTBADSECT option.
1.343.8.1	30-Jul-2007	taca	file wd.c was added on branch matt-mips64 on 2007-07-30 06:59:14 +0000
1.343.6.4	23-Mar-2008	matt	sync with HEAD
1.343.6.3	09-Jan-2008	matt	sync with HEAD
1.343.6.2	08-Nov-2007	matt	sync with -HEAD
1.343.6.1	06-Nov-2007	matt	sync with HEAD
1.343.4.9	11-Dec-2007	jmcneill	Wrong branch, backout previous.
1.343.4.8	11-Dec-2007	jmcneill	Disable the wd_suspend power handler.
1.343.4.7	08-Dec-2007	jmcneill	Rename pnp(9) -> pmf(9), as requested by many.
1.343.4.6	11-Nov-2007	joerg	Sync with HEAD.
1.343.4.5	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.343.4.4	31-Oct-2007	joerg	Sync with HEAD.
1.343.4.3	26-Oct-2007	joerg	Sync with HEAD. Follow the merge of pmap.c on i386 and amd64 and move pmap_init_tmp_pgtbl into arch/x86/x86/pmap.c. Modify the ACPI wakeup code to restore CR4 before jumping back into kernel space as the large page option might cover that.
1.343.4.2	16-Oct-2007	joerg	Replace shutdown hook to flush cache with PNP style power handler to flush the disk cache also on suspend.
1.343.4.1	02-Oct-2007	joerg	Sync with HEAD.
1.345.2.1	14-Oct-2007	yamt	sync with head.
1.346.2.2	13-Nov-2007	bouyer	Sync with HEAD
1.346.2.1	25-Oct-2007	bouyer	Sync with HEAD.
1.348.2.3	18-Feb-2008	mjf	Sync with HEAD.
1.348.2.2	27-Dec-2007	mjf	Sync with HEAD.
1.348.2.1	19-Nov-2007	mjf	Sync with HEAD.
1.350.4.1	11-Dec-2007	yamt	sync with head.
1.350.2.2	26-Dec-2007	ad	Sync with head.
1.350.2.1	04-Dec-2007	ad	Pull the vmlocking changes into a new branch.
1.351.2.2	02-Jan-2008	bouyer	Sync with HEAD
1.351.2.1	13-Dec-2007	bouyer	Sync with HEAD
1.355.6.7	17-Jan-2009	mjf	Sync with HEAD.
1.355.6.6	29-Jun-2008	mjf	Sync with HEAD.
1.355.6.5	02-Jun-2008	mjf	Sync with HEAD.
1.355.6.4	06-Apr-2008	mjf	- after some discussion with agc@ i agreed it would be a good idea to move device_unregister_* to device_deregister_* to be more like the pmf(9) functions, especially since a lot of the time the function calls are next to each other. - add device_register_name() support for dk(4).
1.355.6.3	04-Apr-2008	mjf	* Catch up with selnotify() changes. * Catch up with some softc changes in HEAD. * More s/dctl/devfsctl/ changes. * Remove debugging printfs from init(8) that slipped in by mistake. * Cosmetic changes in devfsd, allow multiple attributes to be set in one rule. Switch from err() to syslog() messages. * Fix etc/devfsd.conf, I'd made some changes previously that had syntax errors.
1.355.6.2	03-Apr-2008	mjf	Sync with HEAD.
1.355.6.1	29-Mar-2008	mjf	- etc/devfsd.conf: Add some rules to give nodes like /dev/tty and /dev/null better default modes, i.e. 0666. - sbin/init: Run devfsd -s before going to multiuser. - sys/arch: Provide arm32, i386, sparc with a mem_init() function to request device nodes for /dev/null, /dev/zero, etc. - sys/dev: Convert rnd, wd, agp, raid, cd, sd, wsdisplay, wskbd, wsmouse, wsmux, tty, bpf, swap to devfs New World Order. - sys/fs/devfs: Make the visibility attribute of device nodes configurable. Also provide a function to mount a devfs on boot. - sys/kern: Add a new boot flag, -n. This disables devfs support. Unless the -n flag is specified the kernel will mount a devfs file system on boot.
1.355.2.1	24-Mar-2008	keiichi	sync with head.
1.360.4.5	11-Mar-2010	yamt	sync with head
1.360.4.4	20-Jun-2009	yamt	sync with head
1.360.4.3	16-May-2009	yamt	sync with head
1.360.4.2	04-May-2009	yamt	sync with head.
1.360.4.1	16-May-2008	yamt	sync with head.
1.360.2.2	17-Jun-2008	yamt	sync with head.
1.360.2.1	18-May-2008	yamt	sync with head.
1.362.2.1	23-Jun-2008	wrstuden	Sync w/ -current. 34 merge conflicts to follow.
1.363.8.4	30-Jan-2010	snj	Pull up following revision(s) (requested by bouyer in ticket #1269): sys/dev/ata/wd.c: revision 1.383 via patch struct buf::b_iodone is not called at splbio() any more. Make sure non-MPsafe iodone callbacks raise the SPL as appropriate.
1.363.8.3	10-Jan-2010	snj	Pull up following revision(s) (requested by dsl in ticket #1220): sys/dev/ata/wd.c: revision 1.380 sys/dev/ata/wdvar.h: revision 1.38 ATA sepcs changed the LBA48 boundary from 0xfffffff to 0xffffffe between ATA6r1 and ATA6r3, which caused drives conform to ATA6r3 or later to reject LBA28 commands at sector 0xfffffff. Get the right idea from the LBA48 boundary from IDENTIFY words 60-61. Remove the WD_QUIRK_FORCE_LBA48 quirk, associated tables entries and autodetect code, it's not needed any more. Based on patch sent to teck-kern by Christoph Badura, use of words 60-61 instead of a constant for the LBA48 boundary by me.
1.363.8.2	10-Jan-2010	snj	Pull up following revision(s) (requested by dsl in ticket #1231): sys/dev/ata/atareg.h: revision 1.33 sys/dev/ata/wd.c: revision 1.378 Add atap_max_lba member to ataparams for LBA48 Maximum Address. From OpenBSD. While here, renumber some __reserved member names to be sequential. Also, add a member for the World Wide Name. Use atap_max_lba in wd.c instead of offsets into a __reserved member.
1.363.8.1	18-Feb-2009	snj	branches: 1.363.8.1.2; 1.363.8.1.4; Pull up following revision(s) (requested by drochner in ticket #441): sys/dev/ata/wd.c: revision 1.368 sys/dev/pcmcia/wdc_pcmcia.c: revision 1.113 -in the wdc@pcmcia driver, don't try to powerup/down while cold or shutting down -- this is only supposed to happen from the pcmcia kernel thread while the system is running -in wd_shutdown, call ata_addref first, so that the adapter is either active or we know that it isn't This fixes a hang on shutdown-p reported by Dennis den Brok per PR kern/40531. Verified by by Dennis and me. (reported against 5.0-Beta, so this is a pullup candidate)
1.363.8.1.4.1	21-Apr-2010	matt	sync to netbsd-5
1.363.8.1.2.1	30-Jan-2010	snj	Pull up following revision(s) (requested by bouyer in ticket #1269): sys/dev/ata/wd.c: revision 1.383 via patch struct buf::b_iodone is not called at splbio() any more. Make sure non-MPsafe iodone callbacks raise the SPL as appropriate.
1.363.6.3	28-Apr-2009	skrll	Sync with HEAD.
1.363.6.2	03-Mar-2009	skrll	Sync with HEAD.
1.363.6.1	19-Jan-2009	skrll	Sync with HEAD.
1.363.4.1	13-Dec-2008	haad	Update haad-dm branch to haad-dm-base2.
1.368.2.2	23-Jul-2009	jym	Sync with HEAD.
1.368.2.1	13-May-2009	jym	Sync with HEAD. Commit is split, to avoid a "too many arguments" protocol error.
1.383.2.2	06-Nov-2010	uebayasi	Sync with HEAD.
1.383.2.1	30-Apr-2010	uebayasi	Sync with HEAD.
1.384.2.1	05-Mar-2011	rmind	sync with head
1.385.4.1	17-Feb-2011	bouyer	Sync with HEAD
1.385.2.1	06-Jun-2011	jruoho	Sync with HEAD.
1.389.2.4	22-May-2014	yamt	sync with head. for a reference, the tree before this commit was tagged as yamt-pagecache-tag8. this commit was splitted into small chunks to avoid a limitation of cvs. ("Protocol error: too many arguments")
1.389.2.3	23-Jan-2013	yamt	sync with head
1.389.2.2	30-Oct-2012	yamt	sync with head
1.389.2.1	17-Apr-2012	yamt	sync with head
1.390.2.1	18-Feb-2012	mrg	merge to -current.
1.392.2.1	03-Jul-2012	jdc	Pull up revision 1.393 (requested by bouyer in ticket #381). In some case, when an error is reported by the disk, the ahci controller still reports a number of bytes transfered equal to bcount. This then triggers a KASSERT in physio_biodone: if (done == todo) KASSERT(bp->b_error == 0); Detect this case in wd(4) (so that the workaround works for other controllers too if they have the same issue, or if the issue is with the drive) and claim we didn't read/write anything.
1.400.2.6	03-Dec-2017	jdolecek	update from HEAD
1.400.2.5	20-Aug-2014	tls	Rebase to HEAD as of a few days ago.
1.400.2.4	23-Jun-2013	tls	resync from head
1.400.2.3	25-Feb-2013	tls	resync with head
1.400.2.2	20-Nov-2012	tls	Resync to 2012-11-19 00:00:00 UTC
1.400.2.1	09-Oct-2012	bouyer	Support transfers of up to MACHINE_MAXPHYS in all pciide variants, and ahci. wd(4) limits its maxphys depending on the drives's capability (64k sectors for LBA48, 256 sectors for LBA and 128 sectors for older devices). I assumed all pciide controllers could do MACHINE_MAXPHYS transfers, but this may not be true. The capabilities of each controller variants should be looked at more closely.
1.403.2.2	18-May-2014	rmind	sync with head
1.403.2.1	28-Aug-2013	rmind	sync with head
1.407.2.2	10-Aug-2014	tls	Rebase.
1.407.2.1	07-Apr-2014	tls	Be a little more clear and consistent about harvesting entropy from devices: 1) deprecate RND_FLAG_NO_ESTIMATE 2) define RND_FLAG_COLLECT_TIME, RND_FLAG_COLLECT_VALUE 3) define RND_FLAG_ESTIMATE_TIME, RND_FLAG_ESTIMATE_VALUE 4) define RND_FLAG_DEFAULT: RND_FLAG_COLLECT_TIME\| RND_FLAG_COLLECT_VALUE\|RND_FLAG_ESTIMATE_TIME 5) Make entropy harvesting from environmental sensors a little more generic and remove it from individual sensor drivers. 6) Remove individual open-coded delta-estimators for values from a few places in the tree (uvm, environmental drivers). 7) 0 -> RND_FLAG_DEFAULT, actually gather entropy from various drivers that had stubbed out code, other minor cleanups.
1.412.2.2	05-Jul-2016	snj	Pull up following revision(s) (requested by bouyer in ticket #1186): sys/dev/ata/wd.c: revision 1.421 sys/kern/subr_autoconf.c: revision 1.242 via patch sys/sys/device.h: revision 1.149 Add a new config_detach() flag, DETACH_POWEROFF, which is set when detaching devices at shutdown time with RB_POWERDOWN. When detaching wd(4), put the drive in standby before detach for DETACH_POWEROFF. Fix PR kern/51252
1.412.2.1	11-Nov-2014	martin	Pull up following revision(s) (requested by mlelstv in ticket #201): sbin/dkctl/dkctl.8: revision 1.24 sbin/dkctl/dkctl.8: revision 1.25 sys/dev/scsipi/sd.c: revision 1.310 sys/dev/ata/wd.c: revision 1.415 sbin/dkctl/dkctl.c: revision 1.21 sys/dev/raidframe/rf_netbsdkintf.c: revision 1.315 sys/dev/ld.c: revision 1.78 sys/dev/vnd.c: revision 1.234 sys/dev/dksubr.c: revision 1.54 sys/sys/dkio.h: revision 1.20 sys/dev/dkwedge/dk.c: revision 1.74 Add ioctl to autodiscover wedges. Implement DIOCMWEDGES ioctl that triggers wedge autodiscovery. Also fix a reference counting bug and clean up some code. support DIOCMWEDGES ioctl. Add 'makewedges' option to autodiscover wedges from a changed label. New sentence, new line. Bump date for previous.
1.415.2.6	28-Aug-2017	skrll	Sync with HEAD
1.415.2.5	05-Dec-2016	skrll	Sync with HEAD
1.415.2.4	05-Oct-2016	skrll	Sync with HEAD
1.415.2.3	09-Jul-2016	skrll	Sync with HEAD
1.415.2.2	06-Jun-2015	skrll	Sync with HEAD
1.415.2.1	06-Apr-2015	skrll	Sync with HEAD
1.421.2.5	26-Apr-2017	pgoyette	Sync with HEAD
1.421.2.4	20-Mar-2017	pgoyette	Sync with HEAD
1.421.2.3	07-Jan-2017	pgoyette	Sync with HEAD. (Note that most of these changes are simply $NetBSD$ tag issues.)
1.421.2.2	06-Aug-2016	pgoyette	Sync with HEAD
1.421.2.1	26-Jul-2016	pgoyette	Sync with HEAD
1.427.2.1	21-Apr-2017	bouyer	Sync with HEAD
1.428.2.36	07-Oct-2017	jdolecek	make usage of NCQ 'high' priority for BPRIO_TIMECRITICAL xfers settable via sysctl, too
1.428.2.35	28-Sep-2017	jdolecek	add sysctls to control if NCQ is being used, and how many max tags; I have a drive which is significantly slower with NCQ than non-NCQ, and it's generally useful to have this easily overridable while here, also move the frequency settings for WD_CHAOS_MONKEY to a sysctl and make it per-drive
1.428.2.34	13-Aug-2017	jdolecek	by default make the chaos monkey do nothing, so it's easier to have this compiled in all the time for testing; the vars can be set via DDB when needed
1.428.2.33	13-Aug-2017	jdolecek	restore the fallback to non-NCQ on retries, do this after WDIORETRIES_SINGLE retries, but only for non-FUA I/O; also only do the ATA_SINGLE fallback when non-FUA this makes sure that bio with ATA_SINGLE is not attempted as NCQ - the ATA_SINGLE I/O is usually done using PIO by drivers which actually support it, and thus are not compatible with DMA-only NCQ
1.428.2.32	12-Aug-2017	jdolecek	do not reset drive after successful NCQ error recovery
1.428.2.31	30-Jul-2017	jdolecek	actually change the retry to also use NCQ - with one drive I see very frequent fatal errors on siisata when switching often between NCQ and non-NCQ I/O commands with chaos monkey (basically, always the next NCQ command on a slot which had non-NCQ I/O throws fatal error); they completely vanish when just using NCQ all the time
1.428.2.30	29-Jul-2017	jdolecek	actually count the REQUEUE as retry also, so that it will be retried as non-NCQ, will not be subject to chaos monkey, and reported as fixed once finished, too
1.428.2.29	23-Jul-2017	jdolecek	for wd, only call disk_busy() on the first try, do not call it on retries, as unbusy is called just once when the xfer is finished also noticed in PR kern/43169 by Matthias Pfaller, but contrary to suggested fix done in way to keep the disk marked busy during the timeouts, as I think it's more correct
1.428.2.28	21-Jul-2017	jdolecek	KASSERT() that chaosed xfer actually ends up with error; might end up being returned as successful due to bugs in error recovery code
1.428.2.27	19-Jul-2017	jdolecek	defflag WD_CHAOS_MONKEY into opt_wd.h together with WD_SOFTBADSECT to set/unset this more easily
1.428.2.26	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.428.2.25	03-Jul-2017	jdolecek	reset xfer c_flags before retry, to clear flags like C_TIMEOU, or C_NCQ, so that retry, and no-NCQ downgrade logic actually works - drivers typically doesn't reset this field print number of retries to make it easier to spot the same xfer being retried several times in wddone(), hold the wd lock only when reading/changing wd softc structures, and not e.g. when calling malloc(), rnd_add_uint32() or ata_free_xfer(), which have their own locks; initially done to fix diagnostic assertion about held spin lock in kpause() within ata_reset_drive hook, but need to run that hook with AT_POLL anyway, since wddone() is typically invoked from interrupt context fix another interrupt context bug for WD_SOFTBADSECT - the malloc() needs to be called with M_NOWAIT
1.428.2.24	03-Jul-2017	jdolecek	introduce some code to test retry paths
1.428.2.23	24-Jun-2017	jdolecek	only limit the openings for I/O xfer if the drive actually supports NCQ; if it's non-NCQ drive, the tag is not going to be used, so we can use any xfer
1.428.2.22	23-Jun-2017	jdolecek	add debug printf if we fail to obtain xfer in wddump()
1.428.2.21	23-Jun-2017	jdolecek	change wd_standby() to remove the ata_channel_start() call - the purpose of that particular command is to make the drive idle fix locking bug in wddetach() exposed by calling ata_channel_start() in wd_standby() - move wd_standby() call out of the section protected by drive mutex, to avoid lock against itself should it need to get the lock change wd_flushcache() to only call ata_channel_start() when called from the ioctl; particularly, don't call it when suspending, closing, or on shutdown
1.428.2.20	23-Jun-2017	jdolecek	restart I/O processing after freeing xfer, i.e. now even after commands like cache flush or standby; the command handling no longer use on-stack xfer, hence use queue slot and compete with normal I/O for the xfers the restart give change to all drives attached to the same channel in round-robin fashion, for fair usage and to recover from situation when disk is idle due to all xfers being consumed by other drives make special concession for flush cache - ignore any new I/O requests on the particular disk while the flush cache is waiting for xfer, so that I/O queue won't starve the flush cache and the flush cache would be done ASAP tested on piixide(4), ahci(4), siisata(4)
1.428.2.19	20-Jun-2017	jdolecek	remove ata queue downsizing - every device, attached to the same channel, uses slots according to it's own limits wdc code changed to expect maximum one active xfer, and not check number of openings in the channel; this is to facilitate using wdc functions for e.g. handling of atapi commands for drivers which support both ATAPI and NCQ
1.428.2.18	19-Jun-2017	jdolecek	add ata_channel lock, use it to protect queue manipulation (only that for now); add ata_channel_detach() to destroy the locks change ata_get_xfer() so that it can wait for xfer, convert all on-stack xfer code to use the blocking variant fix siisata_reset_drive() to use polled reset and not try ata_activate_xfer(), convert drive probe code also over from slot0 XXX to ata_get_xfer() drive reset and PMP now works on siisata(4) too; changes tested also on piixide(4), ahci(4), mvsata(4)
1.428.2.17	19-Jun-2017	jdolecek	ata subsystem needs more work before wd(4) could be MPSAFE
1.428.2.16	16-Jun-2017	jdolecek	adjust reset channel and dump paths - channel reset now always kills active transfer, even on dump path, but now doesn't touch the queued waiting transfers; also kill_xfer hook is always called, so that HBA can free any private xfer resources and thus the dump request has chance to work - kill_xfer routines now always call ata_deactivate_xfer(); added KASSERT()s to ata_free_xfer() to expect deactivated xfer - when called during channel reset before dump, ata_kill_active() drops any queued waiting transfers without processing - do not (re)queue any transfers in wddone() when dumping - kill AT_RST_NOCMD flag This should also hopefully fix the 'polled command has been queued' panic as reported in: PR kern/11811 by John Hawkinson PR kern/47041 by Taylor R Campbell PR kern/51979 by Martin Husemann dump tested working with piixide(4) and ahci(4). mvsata(4) dump times out, but otherwise tested working, will be fixed separately. siisata(4) mechanically changed and not tested.
1.428.2.15	14-Jun-2017	jdolecek	put back reset of retry count to 0 when starting processing a buf do not use NCQ for retried transfers to speed up recovery, unless FUA is requested
1.428.2.14	24-Apr-2017	jdolecek	set NCQ priority field to 'high' for BPRIO_TIMECRITICAL transfers if drive supports it
1.428.2.13	24-Apr-2017	jdolecek	ata_free_xfer() must be called with lock held, same as ata_get_xfer()
1.428.2.12	24-Apr-2017	jdolecek	sync with HEAD, remove the nonfunctional Seagate 'mod15write' bug workaround
1.428.2.11	23-Apr-2017	jakllsch	Simplify wi_find() with container_of(). No need to search a list of wd_ioctls when the *bp is always going to be within the wd_ioctl structure. Correctly initialize wi->wi_bp.b_dev in the non-data case anyway. Fixes ATAIOCCOMMAND procesing on driver instances other than wd0.
1.428.2.10	22-Apr-2017	jakllsch	Avoid destroying uninitialized xfer on error in wdioctlstrategy().
1.428.2.9	19-Apr-2017	jdolecek	add ATA FUA support
1.428.2.8	19-Apr-2017	jdolecek	up to 65536 sectors can be transferred for LBA48 commands; adjust wdminphys() and wdstart1() accordingly
1.428.2.7	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.428.2.6	15-Apr-2017	jdolecek	tweak english
1.428.2.5	15-Apr-2017	jdolecek	probe for NCQ and WRITE DMA FUA EXT support
1.428.2.4	15-Apr-2017	jdolecek	pass also ata_command via ata_xfer, callers of ata_exec_command() is now responsible for allocation/disposal of the structure change code to allocate ata_xfer for commands on stack same way as previously the ata_command were, using c_slot 0; adjust asserts so that it would allow several xfers with same c_slot, as long as only one such transfer is active at a time
1.428.2.3	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.428.2.2	12-Apr-2017	jdolecek	more MP
1.428.2.1	12-Apr-2017	jdolecek	convert to mutex, mark MPSAFE
1.429.2.1	30-Apr-2017	pgoyette	And a few more DEVSW_MODULE_INIT for modular drivers
1.438.2.3	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.438.2.2	06-Sep-2018	pgoyette	Sync with HEAD Resolve a couple of conflicts (result of the uimin/uimax changes)
1.438.2.1	25-Jun-2018	pgoyette	Sync with HEAD
1.439.2.4	21-Apr-2020	martin	Sync with HEAD
1.439.2.3	13-Apr-2020	martin	Mostly merge changes from HEAD upto 20200411
1.439.2.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.439.2.1	10-Jun-2019	christos	Sync with HEAD
1.441.2.12	06-Oct-2018	jdolecek	actually, just make dump use the same queue skip as recovery, and remove the no longer necessary ata_queue_reset() call from wd(4) also for PR kern/47041
1.441.2.11	06-Oct-2018	jdolecek	remove AT_RST_EMERG, do the queue reset explicitly in wd(4) this should explicitly fix PR kern/47041 with sync during heavy disk activity, even thought it was actually already implicitly fixed by calling ata_thread_run() for drive reset in previous commit already, since the function already called ata_queue_reset() drop now unused ch_reset_flags and drive_reset_flags
1.441.2.10	06-Oct-2018	jdolecek	fix dump to also hold channel lock for the drive reset
1.441.2.9	04-Oct-2018	jdolecek	relax the NOERROR + C_CHAOS, just rerun the xfer instead of panic, to handle another variant of bad/missing AHCI error recovery, this time under Parallels
1.441.2.8	04-Oct-2018	jdolecek	don't add +1 to retry on error
1.441.2.7	03-Oct-2018	jdolecek	change channel reset and drive reset for all ATA controllers to always run via thread, and with channel lock held the whole time; the queue is frozen while reset is pending for this repurpose ata_reset_channel() into new ata_thread_run() also adjust some device printfs to not leak xfer pointer, and avoid aprint_* for non-autoconf messages
1.441.2.6	22-Sep-2018	jdolecek	remove explicit ata_channel_start() calls, it's no longer necessary now that ata_xfer's are allocated via pool and not really limited; replace by just a callout to restart the processing for rare cases where system runs out of memory
1.441.2.5	22-Sep-2018	jdolecek	fix use-after-free in wd(4) dump, detected by switch to the pool change code in wd_dumpblocks() to use it's own non-pool ata_xfer, which skips the deallocation step and thus keeps the contents when the I/O is finished
1.441.2.4	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.441.2.3	17-Sep-2018	jdolecek	convert from malloc()/free() to kmem_zalloc()/kmem_free()
1.441.2.2	01-Sep-2018	jdolecek	stop assuming c_slot is immutable, use the xfer pointer rather than c_slot in printfs()
1.441.2.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.452.2.3	30-Dec-2022	martin	Pull up following revision(s) (requested by tsutsui in ticket #1557): sys/dev/ic/ahcisata_core.c: revision 1.83 sys/dev/ic/ahcisata_core.c: revision 1.102 sys/dev/ata/ata.c: revision 1.164 sys/dev/ata/ata_wdc.c: revision 1.115 sys/dev/ata/ata_recovery.c: revision 1.4 sys/dev/ic/siisata.c: revision 1.42 sys/dev/ic/wdc.c: revision 1.308 sys/dev/ic/mvsata.c: revision 1.56 sys/dev/scsipi/atapi_wdc.c: revision 1.138 sys/dev/ic/siisata.c: revision 1.49 sys/dev/ata/atavar.h: revision 1.105 sys/dev/ata/wd.c: revision 1.460 sys/dev/ata/ata.c: revision 1.155 sys/dev/ata/wd.c: revision 1.462 sys/dev/ata/atavar.h: revision 1.109 sys/dev/ata/satapmp_subr.c: revision 1.16 sys/dev/ic/wdc.c: revision 1.299 sys/dev/ic/ahcisata_core.c: revision 1.93 sys/dev/ata/ata_wdc.c: revision 1.120 sys/dev/ic/wdcvar.h: revision 1.100 sys/dev/scsipi/atapi_wdc.c: revision 1.141 sys/dev/ic/mvsata.c: revision 1.61 sys/dev/usb/umass_isdata.c (apply patch) drop wd lock in wdstart1() before calling the ata_bio hook; when called from ata thread context, that can still need to sleep for wdc attachments in wdcwait() fix use-after-free for ata xfer on bio submission found by KASAN driver ata_bio hooks read parts of the xfer after ata_exec_xfer() call in order to determine return value, change so that the hook doesn't return any value - callers do not care already, as all I/O requests are asynchronous this problem was uncovered by recent change for wd(4) to not hold wd mutex during ata_bio call, the interrupt for the xfer might thus actually fire immediately adjust also ata_exec_command driver hooks similarily - remove all completion and waiting logic from drivers, upper layer ata code using AT_WAIT/AT_POLL changed to call ata_wait_cmd() itself PR kern/55169 by Nick Hudson Function declaration formating whitespace consistency. NFCI. PR kern/56403 Fix kernel freeze for wdc(4) variants with ATAC_CAP_NOIRQ: (1) Change ata_xfer_ops:c_poll from void to int function. When it returns ATAPOLL_AGAIN, let ata_xfer_start() iterate itself again. (2) Let wdc_ata_bio_poll() return ATAPOLL_AGAIN until ATA_ITSDONE is achieved. A similar change has been made for mvsata(4) (see mvsata_bio_poll()), and no functional changes for other devices. This is how the drivers worked before jdolecek-ncq branch was merged. Note that this changes are less likely to cause infinite recursion: (1) wdc_ata_bio_intr() called from wdc_ata_bio_poll() asserts ATA_ITSDONE in its error handling paths via wdc_ata_bio_done(). (2) Return value from c_start (= wdc_ata_bio_start()) is checked in ata_xfer_start(). Therefore, errors encountered in ata_xfer_ops:c_poll and c_start routines terminate the recursion for wdc(4). The situation is similar for mvsata(4). Still, there is a possibility where ata_xfer_start() takes long time to finish a normal operation. This can result in a delayed response for lower priority interrupts. But, I've never observed such a situation, even when heavy thrashing takes place for swap partition in wd(4). "Go ahead" by jdolecek@.
1.452.2.2	21-Mar-2020	martin	Pull up following revision(s) (requested by riastradh in ticket #795): sys/dev/dksubr.c: revision 1.112 sys/arch/xen/xen/xbd_xenbus.c: revision 1.95 sys/dev/scsipi/sd.c: revision 1.328 sys/dev/dkvar.h: revision 1.32 sys/dev/ld.c: revision 1.109 sys/dev/cgd.c: revision 1.120 sys/dev/raidframe/rf_netbsdkintf.c: revision 1.380 sys/dev/ata/wd.c: revision 1.458 Add a flag to dk_dump for virtual disk devices. If a disk is backed by a physical medium other than itself, such as cgd(4), then it passes DK_DUMP_RECURSIVE to disable the recursion detection for dk_dump. If, however, a device represents a physical medium on its own, such as wd(4), then it passes 0 instead. With this, I can now dump to dk on cgd on dk on wd.
1.452.2.1	21-Mar-2020	martin	Pull up following revision(s) (requested by riastradh in ticket #788): sys/sys/dkio.h: revision 1.26 sys/dev/dkwedge/dk.c: revision 1.100 sys/sys/disk.h: revision 1.75 external/cddl/osnet/dist/uts/common/fs/zfs/vdev_disk.c: revision 1.14 external/cddl/osnet/dist/uts/common/fs/zfs/vdev_disk.c: revision 1.15 sys/dev/cgd.c: revision 1.121 sys/dev/ata/wdvar.h: revision 1.50 sys/kern/subr_disk_open.c: revision 1.15 sys/dev/ata/wd.c: revision 1.459 New ioctl DIOCGSECTORALIGN returns sector alignment parameters. struct disk_sectoralign { /* First aligned sector number. / uint32_t dsa_firstaligned; / Number of sectors per aligned unit. */ uint32_t dsa_alignment; }; - Teach wd(4) to get it from ATA. - Teach cgd(4) to pass it through from the underlying disk. - Teach dk(4) to pass it through with adjustments. - Teach zpool (zfs) to take advantage of it. => XXX zpool doesn't seem to understand when the vdev's starting sector is misaligned. Missing: - ccd(4) and raidframe(4) support -- these should support _using_ DIOCGSECTORALIGN to decide where to start putting ccd or raid stripes on disk, and these should perhaps _implement_ DIOCGSECTORALIGN by reporting the stripe/interleave factor. - sd(4) support -- I don't know any obvious way to get it from SCSI, but if any SCSI wizards know better than I, please feel free to teach sd(4) about it! - any ld(4) attachments -- might be worth teaching the ld drivers for nvme and various raid controllers to get the aligned sector size There's some duplicate logic here for now. I'm doing it this way, rather than gathering the logic into a new disklabel_sectoralign function or something, so that this change is limited to adding a new ioctl, without any new kernel symbols, in order to make it easy to pull up to netbsd-9 without worrying about the module ABI. Make getdiskinfo() compatible with a DIOCGWEDGEINFO. dkw_parent is defined to hold the disk name as used by disk_find(), not a partition (i.e. no partition letter appended). Use utility functions to handle disk geometry.
1.453.2.2	25-Jan-2020	ad	Sync with head.
1.453.2.1	17-Jan-2020	ad	Sync with head.
1.460.2.1	20-Apr-2020	bouyer	Sync with HEAD
1.469.2.1	02-Aug-2025	perseant	Sync with HEAD
1.9	12-Oct-1998	bouyer	Merge bouyer-ide
1.8	15-Aug-1998	mycroft	Assign my copyrights to TNF.
1.7	26-Apr-1998	mycroft	Update copyright.
1.6	26-Apr-1998	mycroft	In LBA mode, always use the `total capacity' count, rather than calculating the capacity based on the c/h/s numbers. In fact, don't use the c/h/s numbers for much of anything. For ATA-4 drives or later, always use LBA mode, since it's now required. Collectively, this allows >8GB disks (like the 12GB Bigfoot) to work.
1.5	23-Apr-1998	bouyer	From Robert V. Baron: Compute the disk block addr at command queing time rather than exec time. This fix a bug which could lead to data corruption on disk: when a command was reexecuted after an error condition (from wdcunwedge), the partition offset was re-added to the block addr, leading to a transfert at the wrong disk block. This should fix the problem reported by some laptop's users, where the first disk read after a suspend/resume returned garbage.
1.4	14-Jan-1998	cgd	Various cleanups and bug fixes to the wdc/wd code: * Fix bug in wdc that would overflow ATAPI transfer length. * Improve wdc probe code so that 'wdc' is probed in if present even if there are no drives attached, and so that it works properly even if the only device is an ATAPI slave. * bus_space-ify. * split the ISA attachment from the wdc driver, and remove ISA dependencies from non-ISA files. * claim that wd and wdc are now machine-independent (probably not completely true, but mostly so; they at least work on arm32 and i386). * Various other minor fixups and cleanups, some of which were pointed out by Kazuki Sakamoto.
1.3	03-Sep-1997	bouyer	Add a map to wdc_softc to track the already known devices to prevent them to be probed several times. This fixes the "ATAPI CD probed as wd drive" problem. Thanks to Geoff Wing <mason@primenet.com.au> for testing this on his hardware.
1.2	27-Aug-1997	bouyer	branches: 1.2.2; Merge scsipi branch in the mainline. This add support for ATAPI devices (currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2 busses to which devices can attach (scsibus and atapibus). This needed to change some include files and structure names in the low level scsi drivers.
1.1	01-Jul-1997	bouyer	branches: 1.1.2; file wdlink.h was initially added on branch bouyer-scsipi.
1.1.2.1	01-Jul-1997	bouyer	Add the ide atapi driver from my original atapi driver, ported to scsipi.
1.2.2.2	04-Sep-1997	thorpej	Update marc-pcmcia branch from trunk.
1.2.2.1	27-Aug-1997	thorpej	Update marc-pcmcia branch from trunk.
1.52	13-Apr-2025	jakllsch	Add physical sector and alignment info to struct disk_geom and the geometry plist, and handle in partutil. Bump version for disk_geom addition. Collect DIOCGSECTORALIGN handling into one place.
1.51	28-Dec-2021	riastradh	branches: 1.51.10; wd(4): Fix bugs in softbadsect handling. - Don't copyout kernel virtual addresses (of SLIST entries) that userland won't use anyway. => The structure still has space for this pointer; it's just always null when userland gets it now. - Don't copyout under a lock. - Stop and return error if copyout fails (unless we've already copied some out). - Don't kmem_free under a lock. XXX Unclear whether anyone actually uses WD_SOFTBADSECT or why -- it's always been disabled by default. Maybe we should just remove it?
1.50	02-Mar-2020	riastradh	New ioctl DIOCGSECTORALIGN returns sector alignment parameters. struct disk_sectoralign { /* First aligned sector number. / uint32_t dsa_firstaligned; / Number of sectors per aligned unit. */ uint32_t dsa_alignment; }; - Teach wd(4) to get it from ATA. - Teach cgd(4) to pass it through from the underlying disk. - Teach dk(4) to pass it through with adjustments. - Teach zpool (zfs) to take advantage of it. => XXX zpool doesn't seem to understand when the vdev's starting sector is misaligned. Missing: - ccd(4) and raidframe(4) support -- these should support _using_ DIOCGSECTORALIGN to decide where to start putting ccd or raid stripes on disk, and these should perhaps _implement_ DIOCGSECTORALIGN by reporting the stripe/interleave factor. - sd(4) support -- I don't know any obvious way to get it from SCSI, but if any SCSI wizards know better than I, please feel free to teach sd(4) about it! - any ld(4) attachments -- might be worth teaching the ld drivers for nvme and various raid controllers to get the aligned sector size There's some duplicate logic here for now. I'm doing it this way, rather than gathering the logic into a new disklabel_sectoralign function or something, so that this change is limited to adding a new ioctl, without any new kernel symbols, in order to make it easy to pull up to netbsd-9 without worrying about the module ABI.
1.49	05-Apr-2019	bouyer	branches: 1.49.4; Implement a DIRTY flag (copied from sd(4)) so avoid flushing the cache if there has been no write. This avoids a (long) timeout on the flush cache command triggered by atactl sleep, when the device is open only by the atactl command itself. If a drive has no partition open and goes to sleep, the WDF_LOADED flag is clear, and the next open will issue wd_get_params() command. But to wake up the drive a reset is required, and wd_get_params() doens't issue a reset on timeout, so there's no way to wake up the disk. Add a retry after reset to wd_get_params(). Tested by Hauke Fath; fixes PR kern/49457
1.48	19-Mar-2019	mlelstv	Set disk model name as disk type. The information can be queried through drvctl(4).
1.47	22-Oct-2018	jdolecek	Merge jdolecek-ncqfixes branch - ata_xfer's are dynamicall allocated as needed using a pool, no longer limited to number of possible openings supported by controller; dump and recovery paths use dedicated pre-allocated storage - moved callouts and condvars from ata_xfer to queue or channel, so that ata_xfer does not need special initialization - slot allocation now done when xfer is being activated, uncoupled from memory allocation; active slots are no longer tracked by controller code - channel and drive reset is done always via the atabus thread, and now executes with channel locked the whole time - NCQ recovery moved to shared function, and run via the thread also - added some workarounds for buggy error recovery AHCI emulation in QEMU and Parallels designed to primarily fix kern/52614, but might also help with kern/47041 and kern/53183
1.46	03-Nov-2017	mlelstv	branches: 1.46.2; 1.46.4; 1.46.6; Add WDF_OPEN flag to really disallow opening of a disk that has been invalidated. Restore wdbiorestart function to actually retry the failed I/O request instead of just restarting the queue. Fix compilation without ATADEBUG.
1.45	01-Nov-2017	mlelstv	refactor wd and ataraid drivers to use common disk subroutines.
1.44	07-Oct-2017	jdolecek	Merge support for SATA NCQ (Native Command Queueing) from jdolecek-ncq branch ATA subsystem was changed to support several outstanding commands, and use NCQ xfers if supported by both the controller and the disk, including NCQ error recovery. Set NCQ high priority for BPRIO_TIMECRITICAL xfers if supported. Added FUA support. Done some work towards MP-safe, all ATA code tsleep()/wakeup() replaced by condvars, and switched most code from spl* to mutexes (separate wd(4) and ata channel lock). Introduced new option WD_CHAOS_MONKEY to facilitate testing of error handling, fixed several uncovered issues. Also fixed several problems with kernel dump to wd(4) disk. Tested with ahcisata(4), mvsata(4), siisata(4), piixide(4) on amd64, with and without port multiplier, both disk and ATAPI devices; other drivers and archs mechanically adjusted and compile-tested. NCQ is supported for ahcisata(4) and siisata(4) for any controller, for mvsata(4) only Gen IIe ones for now. Also enabled ATAPI support in mvsata(4). Thanks to Matt Thomas for initial ATA infrastructure patch, and Jonathan A.Kollasch for siisata(4) NCQ changes and general testing. Also fixes PR kern/43169 (wd(4)); and PR kern/11811, PR kern/47041, PR kern/51979 (kernel dump)
1.43	22-Jul-2016	jakllsch	branches: 1.43.4; Add ATA8-ACS Long Logical Sector Feature Set support to wd(4).
1.42	13-Apr-2015	riastradh	branches: 1.42.2; Convert sys/dev to use <sys/rndsource.h>.
1.41	10-Sep-2014	matt	branches: 1.41.2; g/u_int[0-9]*_t/ s/u_int/uint/g
1.40	02-Feb-2012	tls	branches: 1.40.6; Entropy-pool implementation move and cleanup. 1) Move core entropy-pool code and source/sink/sample management code to sys/kern from sys/dev. 2) Remove use of NRND as test for presence of entropy-pool code throughout source tree. 3) Remove use of RND_ENABLED in device drivers as microoptimization to avoid expensive operations on disabled entropy sources; make the rnd_add calls do this directly so all callers benefit. 4) Fix bug in recent rnd_add_data()/rnd_add_uint32() changes that might have lead to slight entropy overestimation for some sources. 5) Add new source types for environmental sensors, power sensors, VM system events, and skew between clocks, with a sample implementation for each. ok releng to go in before the branch due to the difficulty of later pullup (widespread #ifdef removal and moved files). Tested with release builds on amd64 and evbarm and live testing on amd64.
1.39	19-Nov-2011	tls	branches: 1.39.2; First step of random number subsystem rework described in <20111022023242.BA26F14A158@mail.netbsd.org>. This change includes the following: An initial cleanup and minor reorganization of the entropy pool code in sys/dev/rnd.c and sys/dev/rndpool.c. Several bugs are fixed. Some effort is made to accumulate entropy more quickly at boot time. A generic interface, "rndsink", is added, for stream generators to request that they be re-keyed with good quality entropy from the pool as soon as it is available. The arc4random()/arc4randbytes() implementation in libkern is adjusted to use the rndsink interface for rekeying, which helps address the problem of low-quality keys at boot time. An implementation of the FIPS 140-2 statistical tests for random number generator quality is provided (libkern/rngtest.c). This is based on Greg Rose's implementation from Qualcomm. A new random stream generator, nist_ctr_drbg, is provided. It is based on an implementation of the NIST SP800-90 CTR_DRBG by Henric Jungheim. This generator users AES in a modified counter mode to generate a backtracking-resistant random stream. An abstraction layer, "cprng", is provided for in-kernel consumers of randomness. The arc4random/arc4randbytes API is deprecated for in-kernel use. It is replaced by "cprng_strong". The current cprng_fast implementation wraps the existing arc4random implementation. The current cprng_strong implementation wraps the new CTR_DRBG implementation. Both interfaces are rekeyed from the entropy pool automatically at intervals justifiable from best current cryptographic practice. In some quick tests, cprng_fast() is about the same speed as the old arc4randbytes(), and cprng_strong() is about 20% faster than rnd_extract_data(). Performance is expected to improve. The AES code in src/crypto/rijndael is no longer an optional kernel component, as it is required by cprng_strong, which is not an optional kernel component. The entropy pool output is subjected to the rngtest tests at startup time; if it fails, the system will reboot. There is approximately a 3/10000 chance of a false positive from these tests. Entropy pool _input_ from hardware random numbers is subjected to the rngtest tests at attach time, as well as the FIPS continuous-output test, to detect bad or stuck hardware RNGs; if any are detected, they are detached, but the system continues to run. A problem with rndctl(8) is fixed -- datastructures with pointers in arrays are no longer passed to userspace (this was not a security problem, but rather a major issue for compat32). A new kernel will require a new rndctl. The sysctl kern.arandom() and kern.urandom() nodes are hooked up to the new generators, but the /dev/*random pseudodevices are not, yet. Manual pages for the new kernel interfaces are forthcoming.
1.38	17-Dec-2009	bouyer	branches: 1.38.12; ATA sepcs changed the LBA48 boundary from 0xfffffff to 0xffffffe between ATA6r1 and ATA6r3, which caused drives conform to ATA6r3 or later to reject LBA28 commands at sector 0xfffffff. Get the right idea from the LBA48 boundary from IDENTIFY words 60-61. Remove the WD_QUIRK_FORCE_LBA48 quirk, associated tables entries and autodetect code, it's not needed any more. Based on patch sent to teck-kern by Christoph Badura, use of words 60-61 instead of a constant for the LBA48 boundary by me.
1.37	19-Oct-2009	bouyer	Remove closes 3 & 4 from my licence. Lots of thanks to Soren Jacobsen for the booring work !
1.36	18-Mar-2008	cube	branches: 1.36.4; 1.36.14; 1.36.22; Split device_t and softc for ATA devices, as well as wd(4). Other cosmetic changes where appropriate.
1.35	28-Feb-2008	drochner	use pmf again to register the shutdown handler, instead of shutdownhook_establish()
1.34	22-Feb-2008	dogcow	As pointed out by Geoff Wing, 1.356 of wd.c needs sc_sdhook restored.
1.33	09-Dec-2007	jmcneill	branches: 1.33.6; 1.33.10; Merge jmcneill-pm branch.
1.32	11-Dec-2005	christos	branches: 1.32.44; 1.32.46; 1.32.56; 1.32.58; merge ktrace-lwp.
1.31	15-Oct-2005	yamt	- change the way to specify a bufq strategy. (by string rather than by number) - rather than embedding bufq_state in driver softc, have a pointer to the former. - move bufq related functions from kern/subr_disk.c to kern/subr_bufq.c. - rename method to strategy for consistency. - move some definitions which don't need to be exposed to the rest of kernel from sys/bufq.h to sys/bufq_impl.h. (is it better to move it to kern/ or somewhere?) - fix some obvious breakage in dev/qbus/ts.c. (not tested)
1.30	25-Jul-2005	drochner	remove unused (hopefully:-) softc fields for disk geometry
1.29	27-Feb-2005	perry	branches: 1.29.4; nuke trailing whitespace
1.28	25-Sep-2004	thorpej	branches: 1.28.4; 1.28.6; Add support for wedges to the ATA disk driver.
1.27	30-Aug-2004	drochner	make the software managed bad-sector list optional (kernel option WD_SOFTBADSECT)
1.26	14-Dec-2003	thorpej	Move the wddone() prototype to wd.c
1.25	14-Dec-2003	thorpej	Garbage-collect __ATA_DISK_PRIVATE; it is no longer needed.
1.24	14-Dec-2003	thorpej	Move struct ata_device from wdvar.h to atavar.h.
1.23	14-Dec-2003	thorpej	Move struct ata_bustype from wdvar.h to atavar.h.
1.22	14-Dec-2003	thorpej	Move struct ata_bio from wdvar.h to atavar.h.
1.21	07-Nov-2003	mycroft	Replace home-grown locks with lockmgr().
1.20	02-Nov-2003	wiz	Fix some typos. From Tom Cosgrove via jmc@openbsd.
1.19	08-Oct-2003	bouyer	Make the ATA mid-layer appears as atabus, as proposed in http://mail-index.netbsd.org/tech-kern/2003/09/25/0006.html This adds a device (atabus) between IDE controllers and wd or atapibus, to have each ATA channel show up in the device tree. Later there will be atabus devices in /dev, so that we can do IOCTL on them. Each atabus has its own kernel thread, to handle operations that needs polling, e.g. reset and others. Device probing on each bus it defered to the atabus thread creation. This allows to do the reset and basic device probes in parallel, which reduce boot time on systems with several pciide controllers.
1.18	05-Oct-2003	bouyer	Remove references to University of California from my copyright notices.
1.17	15-Apr-2003	darrenr	branches: 1.17.2; Implement (and document!) "dkctl <dev> badsector .." to `manage' the bad sector information in the kernel. Doing this uncovered some shortcomings that should have been pretty obvious with the code committed prior, addressing the major kludge with a new struct - disk_bacsecinfo to be passed into DIOCBSLIST.
1.16	15-Apr-2003	darrenr	Automatically build up a list of bad sectors from failed read operations and prempt read operations on matching regions with a failure rather than waiting for the device to return a failure. The I/O operation must have already failed MAXRETRIES times before being added to the list - this can generally take up to 12 seconds. List is made accessible to userspace via DIOCBSLIST and DIOCBSFLUSH.
1.15	03-Apr-2003	fvdl	Check RAW_PART against the media size instead of the disklabel. Add the media size in 512-byte sectors to the softc, to avoid some 64 bit computations. Bump the capacity stored in softcs for disks to 64 bits.
1.14	21-Mar-2003	thorpej	Work around a buggy PHY in some Seagate S-ATA drives which get confused by the way some S-ATA controllers packetize data when the sector count of the transfer mod 15 is 1.
1.13	27-Jan-2003	thorpej	Experimental support for RAID volumes configured by ATA "RAID" controllers. Such RAID controllers are actually just IDE controllers with a BIOS that can create RAID volumes and write the configuration info to config blocks on the disks. The BIOS can do I/O to these volumes, and the OS must understand the config blocks and implement RAID in software in order to be able to use these volumes. Only SPAN (simple concatenation) and RAID0 are supported at this time, and writing back config blocks is also not supported at this time. Currently, only the Promise configuration scheme is supported, although supporting the Highpoint scheme should not be too difficult. In any case, this is sufficient to use the Promise RAID0 volume (thus preserving the win2k AS installation) on this new Intel server I have. Thanks to Soren Schmidt for doing the work in FreeBSD; it made this task much easier. The config block parsing code is adapted from his work.
1.12	23-Apr-2002	bouyer	More copyright fixes, pointed out by Thomas. Thanks !
1.11	13-Jan-2002	christos	Add LBA48 support based on patches from Keisuke YOSHIDA. KNF a bit while I am there.
1.10	03-Dec-2001	bouyer	Update my copyrigth.
1.9	03-Dec-2001	bouyer	Allow wd to attach to something else than wdc/pciide (like USB<->ATA bridges): - move some functions from ata.c to ata_wdc.c or wdc.c. - add callbacks to struct ata_bustype so that wd.c doesn't call directly functions from the lower level driver.
1.8	02-Dec-2001	bouyer	Clean up attach of wd/atapibus: kill ata_atapi_attach. Change atapibus to use a struct scsipi_channel instead of ata_atapi_attach as attach arch. Create a ata_device, compatible with scsipi_channel, to attach wd.
1.7	08-Jul-2001	wiz	branches: 1.7.2; Correct various misspellings of 'transfer' and inflected forms.
1.6	15-May-2000	bouyer	branches: 1.6.6; Sync my copyrigth notice
1.5	01-Apr-2000	bouyer	- DMA code cleanup: pciide_dma_finish() doesn't stop/unload the current DMA op if an IRQ was not detected, unless the force flag was given. Use this to detect if the IRQ was for us (closer to shared IRQ for controllers which don't have their own IRQ handler in pciide.c) and to poll for DMA xfer. Also makes the timeout recovery code simpler. - ATAPI cleanup: don't call controller-specific functions from atapiconf.c (wdc_), so that it's possible to attach an atapibus to something else than a wdc/pciide (Hi Lennart :). Overload struct scsi_adapter with struct atapi_adapter, defined as struct scsi_adapter + atapi-specific callbacks. scsipi_link still points to an scsi_adapter, atapi code casts it to atapi_adapter if needed. Move atapi_softc to atapiconf.h so that it can be used by the underlying controller code (e.g. atapi_wdc.c). Add an atapi-specific callback atapi_probedev(), which probe a drive in a controller-specific way, allocate the sc_link and fills in the ataparams if needed. It then calls atapi_probedev() (from atapiconf.c) to do the generic initialisations and attach the device. - While I'm there merge and centralise the state definitions in atavar.h. It should now be possible to use a common ata/atapi routine to set the drive's modes (will do later).
1.4	20-Oct-1999	enami	Cancel active transfers on aic/wdc detach. Also makes LS-120 drive works for me again.
1.3	11-Nov-1998	bouyer	branches: 1.3.4; 1.3.10; 1.3.12; 1.3.14; - clearify the boot messages (features supported vs features used). Thanks to Havard Eidnes for his complains about this :) - fix some typo in comments - hoppefully better detection of drives reporting bogus PIO modes.
1.2	12-Oct-1998	bouyer	Merge bouyer-ide
1.1	04-Jun-1998	bouyer	branches: 1.1.2; file wdvar.h was initially added on branch bouyer-ide.
1.1.2.2	13-Aug-1998	bouyer	- sync with HEAD - better error detection and report for wd drives - better wddump() support - wdcwait() now takes a timout argument, so that we can honnor the timeout argument of scsipi cmds.
1.1.2.1	04-Jun-1998	bouyer	Commit changes to the IDE system in a branch. This allows a better separation between higth-level and low-level (i.e. registers read/write) and generalize the queue for all commands. This also add supports for IDE DMA.
1.3.14.1	27-Dec-1999	wrstuden	Pull up to last week's -current.
1.3.12.1	15-Nov-1999	fvdl	Sync with -current
1.3.10.2	20-Nov-2000	bouyer	Update thorpej_scsipi to -current as of a month ago A i386 GENERIC kernel compiles without the siop, ahc and bha drivers (will be updated later). i386 IDE/ATAPI and ncr work, as well as sparc/esp_sbus. alpha should work as well (untested yet). siop, ahc and bha will be updated once I've updated the branch to current -current, as well as machine-dependant code.
1.3.10.1	20-Oct-1999	thorpej	Sync w/ trunk.
1.3.4.1	07-Jul-2000	he	Apply patch (requested by bouyer): Add support for the following PCIIDE controllers: o AMD 756 o CMD PCI0648 and PCI0649 o Hightpoint HPT366 o OPTi 82c621 (and a few of its derivatives) o Promise Ultra/33 and Ultra/66 o Intel 82801 (ICH/ICH0) Also fix PR#10437 (detect more ATAPI devices).
1.6.6.4	20-Jun-2002	nathanw	Catch up to -current.
1.6.6.3	28-Feb-2002	nathanw	Catch up to -current.
1.6.6.2	08-Jan-2002	nathanw	Catch up to -current.
1.6.6.1	24-Aug-2001	nathanw	Catch up with -current.
1.7.2.3	23-Jun-2002	jdolecek	catch up with -current on kqueue branch
1.7.2.2	11-Feb-2002	jdolecek	Sync w/ -current.
1.7.2.1	10-Jan-2002	thorpej	Sync kqueue branch with -current.
1.17.2.7	10-Nov-2005	skrll	Sync with HEAD. Here we go again...
1.17.2.6	04-Mar-2005	skrll	Sync with HEAD. Hi Perry!
1.17.2.5	19-Oct-2004	skrll	Sync with HEAD
1.17.2.4	21-Sep-2004	skrll	Fix the sync with head I botched.
1.17.2.3	18-Sep-2004	skrll	Sync with HEAD.
1.17.2.2	03-Sep-2004	skrll	Sync with HEAD
1.17.2.1	03-Aug-2004	skrll	Sync with HEAD
1.28.6.1	19-Mar-2005	yamt	sync with head. xen and whitespace. xen part is not finished.
1.28.4.1	29-Apr-2005	kent	sync with -current
1.29.4.5	24-Mar-2008	yamt	sync with head.
1.29.4.4	17-Mar-2008	yamt	sync with head.
1.29.4.3	27-Feb-2008	yamt	sync with head.
1.29.4.2	21-Jan-2008	yamt	sync with head
1.29.4.1	21-Jun-2006	yamt	sync with head.
1.32.58.1	11-Dec-2007	yamt	sync with head.
1.32.56.1	26-Dec-2007	ad	Sync with head.
1.32.46.2	23-Mar-2008	matt	sync with HEAD
1.32.46.1	09-Jan-2008	matt	sync with HEAD
1.32.44.2	06-Nov-2007	joerg	Refactor PNP API: - Make suspend/resume directly a device functionality. It consists of three layers (class logic, device logic, bus logic), all of them being optional. This replaces D0/D3 transitions. - device_is_active returns true if the device was not disabled and was not suspended (even partially), device_is_enabled returns true if the device was enabled. - Change pnp_global_transition into pnp_system_suspend and pnp_system_resume. Before running any suspend/resume handlers, check that all currently attached devices support power management and bail out otherwise. The latter is not done for the shutdown/panic case. - Make the former bus-specific generic network handlers a class handler. - Make PNP message like volume up/down/toogle PNP events. Each device can register what events they are interested in and whether the handler should be global or not. - Introduce device_active API for devices to mark themselve in use from either the system or the device. Use this to implement the idle handling for audio and input devices. This is intended to replace most ad-hoc watchdogs as well. - Fix somes situations in which audio resume would lose mixer settings. - Make USB host controllers better deal with suspend in the light of shared interrupts. - Flush filesystem cache on suspend. - Flush disk caches on suspend. Put ATA disks into standby on suspend as well. - Adopt drivers to use the new PNP API. - Fix a critical bug in the generic cardbus layer that made D0->D3 break. - Fix ral(4) to set if_stop. - Convert cbb(4) to the new PNP API. - Apply the PCI Express SCI fix on resume again.
1.32.44.1	16-Oct-2007	joerg	Replace shutdown hook to flush cache with PNP style power handler to flush the disk cache also on suspend.
1.33.10.1	03-Apr-2008	mjf	Sync with HEAD.
1.33.6.1	24-Mar-2008	keiichi	sync with head.
1.36.22.1	21-Apr-2010	matt	sync to netbsd-5
1.36.14.1	10-Jan-2010	snj	Pull up following revision(s) (requested by dsl in ticket #1220): sys/dev/ata/wd.c: revision 1.380 sys/dev/ata/wdvar.h: revision 1.38 ATA sepcs changed the LBA48 boundary from 0xfffffff to 0xffffffe between ATA6r1 and ATA6r3, which caused drives conform to ATA6r3 or later to reject LBA28 commands at sector 0xfffffff. Get the right idea from the LBA48 boundary from IDENTIFY words 60-61. Remove the WD_QUIRK_FORCE_LBA48 quirk, associated tables entries and autodetect code, it's not needed any more. Based on patch sent to teck-kern by Christoph Badura, use of words 60-61 instead of a constant for the LBA48 boundary by me.
1.36.4.1	11-Mar-2010	yamt	sync with head
1.38.12.1	17-Apr-2012	yamt	sync with head
1.39.2.1	18-Feb-2012	mrg	merge to -current.
1.40.6.1	03-Dec-2017	jdolecek	update from HEAD
1.41.2.2	05-Oct-2016	skrll	Sync with HEAD
1.41.2.1	06-Jun-2015	skrll	Sync with HEAD
1.42.2.1	26-Jul-2016	pgoyette	Sync with HEAD
1.43.4.10	07-Oct-2017	jdolecek	make usage of NCQ 'high' priority for BPRIO_TIMECRITICAL xfers settable via sysctl, too
1.43.4.9	28-Sep-2017	jdolecek	add sysctls to control if NCQ is being used, and how many max tags; I have a drive which is significantly slower with NCQ than non-NCQ, and it's generally useful to have this easily overridable while here, also move the frequency settings for WD_CHAOS_MONKEY to a sysctl and make it per-drive
1.43.4.8	02-Sep-2017	jdolecek	include opt_wd.h instead of opt_wd_softbadsect.h, the option doesn't have private file any more
1.43.4.7	19-Jul-2017	jdolecek	tighen and expand error handling, mostly for NCQ use cases: - make retry timeout callout per xfer, i.e. retry separately - zero whole bio struct on retry to avoid more stale state - add a REQUEUE option, which doesn't bump retry count - add ata_read_log_ext_ncq() for NCQ recovery - adjust logic for activating xfers - allow next command only when it's for same drive, several concurrent are only supported when HBA and driver support FIS-based switching - add new ata_timeout() which handles race between callout_stop() and the invokation, add appropriate handling on deactivate/free paths - stop using ch_status/ch_error in non-wdc code; later it will be dropped completely
1.43.4.6	23-Jun-2017	jdolecek	restart I/O processing after freeing xfer, i.e. now even after commands like cache flush or standby; the command handling no longer use on-stack xfer, hence use queue slot and compete with normal I/O for the xfers the restart give change to all drives attached to the same channel in round-robin fashion, for fair usage and to recover from situation when disk is idle due to all xfers being consumed by other drives make special concession for flush cache - ignore any new I/O requests on the particular disk while the flush cache is waiting for xfer, so that I/O queue won't starve the flush cache and the flush cache would be done ASAP tested on piixide(4), ahci(4), siisata(4)
1.43.4.5	23-Apr-2017	jakllsch	Simplify wi_find() with container_of(). No need to search a list of wd_ioctls when the *bp is always going to be within the wd_ioctl structure. Correctly initialize wi->wi_bp.b_dev in the non-data case anyway. Fixes ATAIOCCOMMAND procesing on driver instances other than wd0.
1.43.4.4	19-Apr-2017	jdolecek	adjust ata code to support more than one active command, including the timeout handling, add support for NCQ commands move probe for NCQ and number of tags to middle layer, negotiate mutual support between drive and controller implement NCQ support in ahci(4)
1.43.4.3	15-Apr-2017	jdolecek	probe for NCQ and WRITE DMA FUA EXT support
1.43.4.2	15-Apr-2017	jdolecek	make ata_xfer's allocated as part of ata_queue and make it include ata_bio; they are pre-allocated on attach and ata_get_xfer() now never sleep, drop the pool modify wd(4) to file the bio requests using the xfers and hence make it possible to have more than one active I/O request in flight; ata_bio callback doesn't need to allocate any memory any more, require it to never return ATACMD_TRY_AGAIN move lp, badsect, multi from ata_bio to ata_drive_datas, as they are per-drive, not per transfer drop unused drv atac_claim_hw/atac_free_hw hooks, and also drop again ata_bio c_hwslot
1.43.4.1	12-Apr-2017	jdolecek	convert to mutex, mark MPSAFE
1.46.6.4	22-Sep-2018	jdolecek	remove explicit ata_channel_start() calls, it's no longer necessary now that ata_xfer's are allocated via pool and not really limited; replace by just a callout to restart the processing for rare cases where system runs out of memory
1.46.6.3	22-Sep-2018	jdolecek	fix use-after-free in wd(4) dump, detected by switch to the pool change code in wd_dumpblocks() to use it's own non-pool ata_xfer, which skips the deallocation step and thus keeps the contents when the I/O is finished
1.46.6.2	22-Sep-2018	jdolecek	separate ata_xfer slot allocation and the memory allocation, so that there can be more queued xfers than number of supported slots by controller, and use a pool instead of custom pre-allocation primarily to help PR kern/52614 remove no longer needed custom wd(4) logic for flush cache switch also wd(4) trim/suspend/setcache/wdioctlstrategy to sleep waiting for the memory, they are all called from process context and this avoids spurious failures
1.46.6.1	31-Aug-2018	jdolecek	refactor ata_xfer to be just dumb structure; move all callouts/condvars out retry callout to wd(4); reset callout and the active/cmd finish condvars to channel queue; change code using the condvars so it works if there are multiple waiters simplify the async wait code for cmds, replace ata_wait_xfer()/ata_wake_xfer() with ata_wait_cmd() fix the callout_invoking/ack race handling code for timeouts to actually have chance to work; change mvsata(4) to use generic timeout func towards resolution of kern/52614
1.46.4.2	08-Apr-2020	martin	Merge changes from current as of 20200406
1.46.4.1	10-Jun-2019	christos	Sync with HEAD
1.46.2.1	26-Nov-2018	pgoyette	Sync with HEAD, resolve a couple of conflicts
1.49.4.1	21-Mar-2020	martin	Pull up following revision(s) (requested by riastradh in ticket #788): sys/sys/dkio.h: revision 1.26 sys/dev/dkwedge/dk.c: revision 1.100 sys/sys/disk.h: revision 1.75 external/cddl/osnet/dist/uts/common/fs/zfs/vdev_disk.c: revision 1.14 external/cddl/osnet/dist/uts/common/fs/zfs/vdev_disk.c: revision 1.15 sys/dev/cgd.c: revision 1.121 sys/dev/ata/wdvar.h: revision 1.50 sys/kern/subr_disk_open.c: revision 1.15 sys/dev/ata/wd.c: revision 1.459 New ioctl DIOCGSECTORALIGN returns sector alignment parameters. struct disk_sectoralign { /* First aligned sector number. / uint32_t dsa_firstaligned; / Number of sectors per aligned unit. */ uint32_t dsa_alignment; }; - Teach wd(4) to get it from ATA. - Teach cgd(4) to pass it through from the underlying disk. - Teach dk(4) to pass it through with adjustments. - Teach zpool (zfs) to take advantage of it. => XXX zpool doesn't seem to understand when the vdev's starting sector is misaligned. Missing: - ccd(4) and raidframe(4) support -- these should support _using_ DIOCGSECTORALIGN to decide where to start putting ccd or raid stripes on disk, and these should perhaps _implement_ DIOCGSECTORALIGN by reporting the stripe/interleave factor. - sd(4) support -- I don't know any obvious way to get it from SCSI, but if any SCSI wizards know better than I, please feel free to teach sd(4) about it! - any ld(4) attachments -- might be worth teaching the ld drivers for nvme and various raid controllers to get the aligned sector size There's some duplicate logic here for now. I'm doing it this way, rather than gathering the logic into a new disklabel_sectoralign function or something, so that this change is limited to adding a new ioctl, without any new kernel symbols, in order to make it easy to pull up to netbsd-9 without worrying about the module ABI. Make getdiskinfo() compatible with a DIOCGWEDGEINFO. dkw_parent is defined to hold the disk name as used by disk_find(), not a partition (i.e. no partition letter appended). Use utility functions to handle disk geometry.
1.51.10.1	02-Aug-2025	perseant	Sync with HEAD

OpenGrok